Digital PDFs

AA-2762G-TC

July 1989

522 pages

Original

23MB

Document:	RSTS/E V9.7 System Managers Guide
Order Number:	AA-2762G-TC
Revision:	0
Pages:	522
Original Filename:	http://bitsavers.org/pdf/dec/pdp11/rsts_e/V09/2_System_Mangement/AA-2762G-TC_RSTS_E_V9.7_System_Managers_Guide_Jul89.pdf

OCR Text

RSTS/E System Manager's Guide
Order Number: M-2762G-TC

This manual describes how to manage a RSTS/E system.

Operating System and Version: RSTS/E V9.7
Software Version:

digital equipment corporation
maynard, massachusetts

RSTS/E V9.7

July ~989
The information in this document is subject to change without notice and should not be
construed as a commitment by Digital Equipment Corporation. Digital Equipment Corporation
assumes no responsibility for any errors that may appear in this document.
The software described in this document is furnishod under a license and may be used or
copied only in accordance with the terms of such license.
No responsibility is assumed for the use or reliability of software on equipment that is not
supplied by Digital Equipment Corporation or its affiliated companies.

© Digital Equipment Corporation 1975, 1985, 1989.
All Rights Reserved.
Printed in U.S.A.
The postpaid Reader's Comments forms at the end of this document request your critical
evaluation to assist in preparing future documentation.
DATATRIEVE
DEC
DECmail
DECmate
DECnet
DECsystem-10
DECsystem-20
DECtape
DECUS
DECwriter
DIBOL

FMS-11
LA
MASSBUS
PDP
P/OS

Professional
Q-Bus
Rainbow
ReGIS
RSTS

HSX
HT
UNIBUS
VAX
VMS
VT
Work Processor

IBM is a registered trademark of International Business Machines Corporation.

Contents
Pref,ace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xvii

Summary of Teclhnlcal Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xxi

Chalpter 1

Overview of RSTS/E System Management

1.1

Glossary of RSTS/E Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1-1

1.2

System Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1-2

1.3

Disk Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1 .3.1
Types of Disks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3.2
File Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3.3
Level 0 Disk Structure (Pre-V8.0) . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3.4
Level 1 Disk Structure (for V8.0 and V9.0) . . . . . . . . . . . . . . . . . . . .
1.3.5
Differences Between Level 0, Level 1.1, and Level 1.2 . . . . . . . . . . . .
1.3.6
Converting Disk Structures (DSKCVT) . . . . . . . . . . . . . . . . . . . . . . .
1.3.7
Disk Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1-3
1-3
1-4
1-5
1-6
1-7
1-7
1-10

1.4

System Operation Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1-11

1.5

Privilege . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1-12

1.6

Multiple Privileges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1-12

1.7

Classes of System Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.7.1
Account Management Activities . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.7.2
File Access Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1-12
1-12
1-13

1.8

Multiple Privilege Masks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1-13

Chslpter 2

Starting, Halting, and Automatic Recovery of RSTS/E

2.1

Starting up RSTS/E . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.1
Bootstrapping RSTS/E with a Hardware Bootstrap Loader ........ .
2.1.2
Bootstrapping RSTS/E after a System Halt . . . . . . . . . . . . . . . . . . . .
2.1.3
Starting limesharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2-1
2-1
2-2
2-2

2.2

Halting the RSTS/E System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2-3

2.3

Automatic Recovery and Restart Facilities . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.1
Catastrophic Errors and System Crashes . . . . . . . . . . . . . . . . . . . . .
2.3.2
Automatic Recovery from Catastrophic Errors ................. .

2-4
2-4
2-5

iii

2.3.3
2.3.4
2.3.5
2.3.6

Chapter 3

2-6
2-6
2-6
2-7

The System Startup Command Flies

3.1

Creating and Using Command Flies .............. " ............... .

3-1

3.2

System Startup Command File: START.COM ....... " ............... .
3.2.1
Mount All Disks ..................... , ............... .
Assign Systemwide Logical Names ........ , ............... .
3.2.2
Start Error Logging. . . . . . . . . . . . . . . . . . . " . . . . . . . . . . . . . . . .
3.2.3
Enable Caching (Optional) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.4
Set Default System Characteristics . . . . . . . . , . . . . . . . . . . . . . . . .
3.2.5
Install Run-lime Systems ....................... ; ...... .
3.2.6
Install Libraries ....................."................
3.2.7
Set Terminal Characteristics ............................. .
3.2.8
Define System Commands ............................. .
3.2.9
3.2.10
Set Printer Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Start PrintlBatch Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.11
Start OPSER-Based Spooling Package (Optional) .............. .
3.2.12
Enable LAT (Optional) ................................ .
3.2.13
Start Error Analysis .................................. .
3.2.14
Enable Logins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.15
Logout the Initial Startup Job. . . . . . . . . . . . . . . . . .. . ....... .
3.2.16

3-2
3-6
3-7
3-7
3-7
3-7
3-7

Chapter 4

Automatic Restart Mode Initialization . . . . . . . . . . . . . . . . . . . . . . . .
Restarting with a Different Monitor ..... . . . . . . . . . . . . . . . . . . . .
Restarting with a Different Swap Size . . . . . . . . . . . . . . . . . . . . . . .
Power-Fail Hardware Used by RSTS/E . . . . . . . . . . . . . . . . . . . . . .

3-8
3-8
3-8
3-8

3-9
3-9
3-9
3-9
3-9
3-9

Shutting Down the RSTS/E System

4.1

Performing System Shutdown: SHUTUP .......................... .
4.1.1
Running SHUTUP ................................... .

4-1
4-1

4.2

SHUTUP Phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.1
Set-Up Dialogue Phase ................................ .
4.2.2
Warning Message Phase ............................... .
4.2.3
DECnetlE Shutdown Phase (When Necessary) . . . . . . . . . . . . . . . . .
4.2.4
Initial Job Killing Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.5
PrintlBatch Shutdown Phase (When Necessary) . . . . . . . . . . . . . . . .
4.2.6
OPSER Shutdown Phase (When Necessary) ................. .
4.2.7
EVTLOG Shutdown Phase (When Necessary) ................ .
4.2.8
ERRCPY Shutdown Phase (When Necessary) . . . . . . . . . . . . . . . . .
4.2.9
Rnal Job Killing Phase (When Nec:essary) ................... .
4.2.10
EMT Logging Shutdown Phase (When Necessary) ............. .
4.2.11
Remove Run-lime Systems and Resident Libraries Phase ........ .
4.2.12
Log File Closing Phase . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.13
Swap File Removal Phase .....' ........................ .
4.2.14
Disk Dismount Phase .........' ........................ .
4.2.15
Rnal Shutdown Phase ....... " ........................ .
4.2.16
SHUTUP Operation Examples .. " ........................ .
4.2.16.1
SHUTUP Example: Large RSTS/E System .......... .
4.2.16.2
SHUTUP Example: Small RSTS/E System .......... .
4.2.16.3
SHUTUP Example: DECnetlE Shutdown ............ .
4.2.17
Notes on SHUTUP Operation .. " ........................ .

4-2
4-2
4-4
4-5
4-6
4-7
4-7
4-8
4-9
4-9
4-9
4-9
4-10
4-10
4-10
4-10
4-10
4-10
4-12
4-12
4-14

Ch4apter 5

Managing Quotas

5.1

Disk QlUotas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.1.1
Disk Quota Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.1 .2
logged-In Disk Quotas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.1.3
logged-Out Disk Quotas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5-1
5-2
5-2
5-2

5.2

Job Quotas ............................................... .
5.2.1
Detached-Job Quota. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.2.2
Job Quotas and login . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5-3
5-3
5-3

5.3

Send/Receive Quotas ........................................ .
5.3.1
RIB Quota . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.3.2
Message Quota . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5-4
5-4
5-4

5.4

Dynamic Region Limits ....................................... .

5-4

Managing Privileges
6.1

RSTS/E-Defined Privileges .................................... .
6.1.1
DATES Privilege ..................................... .
6.1.2
DEVICE Privilege ..................................... .
6.1.3
EXQTA Privilege . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.1.4
GACNT Privilege . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , .
6.1.5
GREAD Privilege ..................................... .
6.1.6
GWRITE Privilege .................................... .
6.1.7
HWCFG Privilege ..................................... .
6.1.8
HWCTl Privilege ..................................... .
INSTAL Privilege ..................................... .
6.1.9
6.1.10
JOBCTl Privilege .................................... .
6.1 .11
MOUNT Privilege . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
PBSCTl Privilege .................................... .
6.1.12
6.1.13
RDMEM Privilege ..................................... .
6.1.14
RDNFS Privilege ..................................... .
6.1 .15
SEND Privilege . . . . . " . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.1 .16
SETPAS Privilege . . . " . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.1 .17
SHUTUP privilege . . . " . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.1 .18
SWCFG Privilege .... " . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.1.19
SWCTl Privilege . . . . " . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.1.20
SYSIO Privilege ..... " . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.1.21
SYSMOD Privilege .................................... .
6.1.22
TMPPRV Privilege .................................... .
6.1.23
TUNE Privilege ...................................... .
6.1.24
WACNT Privilege .... , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.1.25
WREAD Privilege ..................................... .
6.1.26
WRTNFS Privilege ... " . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.1.27
WWRITE Privilege .................................... .

6.2

User-Defined Privileges ....... " ............................... .

6.3

Multiple Privileges and Jobs ... " ............................... .
6.3.1
Privilege Masks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.3.2
Job Creation ....... " . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
login ............ , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.3.3
6.3.4
logout ............................................ .
Spawned jobs ....................................... .
6.3.5

6-1
6-3
6-3
6-3
6-3
6-3
6-3
6-3
6-4
6-4
6-4
6-4
6-4
6-4
6-4

6-5
6-5
6-5
6-5
6-5
6-5
6-6
6-6
6-6
6-7
6-7
6-7
6-7

6-8
6-8
6-8
6-8
6-9
6-9
v

6.4

Chapter 7
7,,1

7.2

Running Programs Under Multiple Prlvllegeu •.......................
6.4.1
Running System Programs ............. " ............... .
6.4.2
Temporary Privileges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.4.3
Designating a Program as Privileged ....... " ............... .
6.4.4
Running a Privileged Program ........... " ............... .
6.4.5
Program exit ....................... " ............... .

Accounts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ................
'
.
7.1.1
Project-Programmer Number (PPN) . . . . . . . . . . . . . . . . . . . . . . . .
7.1.2
Account Access Privileges ................ , ............. .

7-1
7-2
7-2

Account Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , ............. .
Account Name . . . . . . . . . . . . . . . . . . . . . . . . " ............. .
Datemme/Keyboard of last login . . . . . . . . . . . . . . . . . . . . . . . . .
Datemme of last Password Chan~~e ...................... .
Password . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.2.5
Interactive Flag . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.2.6
Captive Account Flag . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.2.7
Network Flag. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.2.8
Dial-Up Flag . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.2.9
Expiration Date ............ " . . . . . . . . . . . . . . . . . . . . . . . . .
7.2.10
Authorized Privilege Mask ..... " . . . . . . . . . . . . . . . . . . . . . . . . .
7.2.11
Creation Date ............. " . . . . . . . . . . . . . . . . . . . . . . . . .
7.2.12
UFO Cluster Size ........... " . . . . . . . . . . . . . . . . . . . . . . . . .
7.2.13
Quotas .................. " . . . . . . . . . . . . . . . . . . . . . . . . .
7.2.14
System Data . . . . . . . . . . . . . . " . . . . . . . . . . . . . . . . . . . . . . . . .

7-2
7-3
7-3
7-3
7-3
7-4
7-4
7-4
7-4
7-4
7-4

DCl Command Descriptions .......... , . . . . . . . . . . . . . . . . . . . . . . . . .
CREATE/ACCOUNT Command . . . . . . . . . . . . . . . . . . . . . . . . . . .
DELETE/ACCOUNT Command . . . . . . . . . . . . . . . . . . . . . . . . . . .
SET ACCOUNT Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
SET PASSWORD Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.3.5
SHOW ACCOUNT Command . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7-5
7-5

7.3.1
7.3.2
7.3.3
7.3.4

Chapter 8

6-10
6-10
6-10
6-11

Managing Accounts

7.2.1
7.2.2
7.2.3
7.2.4

7.3

6-9
6-9

7-5
7-5
7-5
7-5

7-10
7-13
7-16
7-18

Managing Terminals

8.1

Understanding the SET TERMINAL Command

8-1

8.2

Terminal Line Speed Characteristics File: TERSPD.SYS ............... .

8-2

8.3

Setting Terminals Other Than Your Own . . . . . . . . . . . . . . . . . . . . . . . . . . . .

8-3

8.4

User-Defined Terminal Characteristics File: TERMAC.SYS .............. .

8-4

8.5

Automatic Setting of Terminal Characterlst'ics: TERDFL.SYS ............ .

8-4

8.6

Setting Permanent Terminal Characteristics .... " .... " .............. .

8-5

8.7

Using the INOBROADCAST and IBROADCJ~ST Command Qualifiers ....... .

8-5

8.8

Terminal Servers (lAT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.8.1
Enabling lAT .................. " . . . . . . . . . . . . . . . . . . . .

8-7

8--6

8.8.2

Isolation of LAT Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

8.9

Pseudo Keyboards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .' ........ .

8.10

Enabling Forms Management Support (FMS-11) . . . . . . . . . . . . . . . . . . . . . .

8.11

DCL Command Descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.11 .1
CREATElSERVICElLAT Com mand . . . . . . . . . . . . . . . .' ........ .
8.11.2
DELETElSERVICElLAT Command . . . . . . . . . . . . . . . . . . . . . . . . .
8.11.3
SET NODEILAT Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.11.4
SET SERVICEILAT Command . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.11.5
SET SYSTEM Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.11.6
SET TERMINAL Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.11.7
SHOW COUNTERS/LAT Command . . . . . . . . . . . . . . . . . . . . . . . .
8.11.8
SHOW NODEILAT Command . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.11.9
SHOW SERVICEILAT Command . . . . . . . . . . . . . . . . . . . . . . . . . .
8.11.10
SHOW SESSIONSILAT Command . . . . . . . . . . . . . . . . . . . . . . . . .
8.11.11
SHOW TERMINAL Command . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.11.12
SHOW TERMINAL_SERVERS/LAT Command . . . . . . . . . . . . . . . . .
8.11.13
START/LAT Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.11.14
STOP/LAT Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

888-1
8-1
8-1
8-1
8-1
8-2
8-2

Chapter 9

8-2
8-2
8-2
8-3
8-3
8-3

Managing Print/Batch Services

9.1

Managing the Print/Batch Services Package . . . . . . . . . . . . . . . . . . . . . . . . .
9.1.1
Queue Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.1 .1 .1
Queue Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.1 .1 .2
Queue Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.1 .1 .3
Default Queues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.1 .1 .4
Assigning Servers to Queues . . . . . . . . . . . . . . . . . . . . .
9.1 .1 .5
Queue Priorities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.1 .1 .6
Opening and Closing Queues . . . . . . . . . . . . . . . . . . . .
9.1.1.7
Starting and Stopping Queues . . . . . . . . . . . . . . . . . . . .
9.1 .1 .8
Processing Requests . . . . . . . . . . . . . . . . . . . . . . . . . .
9.1 .1 .9
Accessing Entries on a Queue . . . . . . . . . . . . . . . . . . . .
9.1 .1 .10
Job Limits for Queues . . . . . . . . . . . . . . . . . . . . . . . . .
9.1.2
Print Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.1.2.1
Device Ownership . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.1.2.2
Job Copies and File Copies . . . . . . . . . . . . . . . . . . . . . .
9.1.2.3
Header Pages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.1.2.4
Page Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.1.2.5
Control Characters and Escape Sequences .......... .
9.1.2.6
Eight-Bit Characters . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.1.3
Batch Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.1.3.1
Executing Command Files . . . . . . . . . . . . . . . . . . . . . . .
9.1.3.2
CPU and Time Limits . . . . . . . . . . . . . . . . . . . . . . . . . .
9.1.4
Managing Forms for PBS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.1 .4.1
Forms Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.1 .4.2
The Forms Definition Rle . . . . . . . . . . . . . . . . . . . . . . .
9.1 .4.3
Forms Definition Keywords . . . . . . . . . . . . . . . . . . . . . .
9.1 .4.4
Changing Forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

9.2

Multiple Jobs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.2.1
The Primary Job . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Secondary Jobs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.2.2

9.3

Del Command Descriptions ... " . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
ASSIGN/QUEUE Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.3.1

99999999999999999-1
9-1
9-1
9-1
9-1
9-1
9-1
9-1
9-1,
9-1
9-1

9.3.2
9.3.3
9.3.4
9.3.5
9.3.6
9.3.7
9.3.8
9.3.9
9.3.10
9.3.11
9.3.12
9.3.13
9.3.14
9.3.15
9.3.16
9.3.17

Chapter 10

9-21
9-22
9-23
9-24
9-25
9-29
9-31
9-32
9-35
9-37
9-39
9-40
9-42
9-44
9-45
9-46

Managing System Errors

10.1

Use of the Error Logging Programs: ERRINl' and ERRCPY ............. .
10.1.1
Error Logging Initialization: ERRINT . . . . . . . . . . . . . . . . . . . . . . . .
10.1 .2
Examples of ERRINT Dialogue . . . . . . . . . . . . . . . . . . . . . . . . . . .
10.1.3
Error Logging: ERRCPY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

10-1
10-2
10-3
10-3

10.2

Displaying Errors: ERR DIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10.2.1
Running ERRDIS ............ , . . . . . . . . . . . . . . . . . . . . . . . .
10.2.2
Help Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10.2.3
Summary Report .............' . . . . . . . . . . . . . . . . . . . . . . . .
10.2.4
Bad Block Report ............ " . . . . . . . . . . . . . . . . . . . . . . . .
10.2.5
Adding Bad Blocks to the Bad Bloc~~ File . . . . . . . . . . . . . . . . . . . . .
10.2.6
Full Report ................ , . . . . . . . . . . . . . . . . . . . . . . . .
10.2.6.1
User Description in Full Report . . . . . . . . . . . . . . . . . . . .
10.2.6.2
Disk Error Detailed Description . . . . . . . . . . . . . . . . . . . .
10.2.6.3
MSCP Variations on the Full Report ............... .
10.2.6.4
Nondisk Peripheral Device Error Detailed Description .... .
10.2.6.5
Nonperipheral Error Detailed Description ............ .
10.2.7
Zero File

10-4
10-4
10-6
10-6
10-7
10-8
10-9
10-10
10-11
10-14
10-17
10-19
10-20

10.3

Analyzing System Crashes: ANALYS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10.3.1
Running the ANALYS Program. . . . . . . . . . . . . . . . . . . . . . . . . . . .
10.3.2
ANALYS Output ............. " . . . . . . . . . . . . . . . . . . . . . . . .

10-21
10-21
10-23

Chapter 11

viii

CLOSE/QUEUE Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DEASSIGN/QUEUE Command . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DELETE/QUEUE Command ............... , ............ .
DELETE/SERVER Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
INITIALIZE/QUEUE Command .............. , ............ .
INITIALIZE/SERVER Command . . . . . . . . . . . . . . . . . . . . . . . . . . .
OPEN/QUEUE Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
SET QUEUE Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
SET SERVER Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
SHOW SERVER Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
START/QUEUE Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
START/QUEUE/MANAGER Command . . . . . . . . . . . . . . . . . . . . . .
START/SERVER Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
STOP/QUEUE Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
STOP/QUEUEIMANAGER Command . . . . . . . . . . . . . . . . . . . . . . .
STOP/SERVER Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Device Management and Caching Control

11.1

Principles of Device Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11-1

11 . 2

Optimizing Disk Directory Structure: REORDR . . . . . . . . . . . . . . . . . . . . . . .
11.2.1
Why Use REORDR? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11 .2.2
Dialogue Questions and Responses . . . . . . . . . . . . . . . . . . . . . . . .
11 .2.3
Reordering Your Disks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11 .2.4
Error and Processing Messages . . . . . . . . . . . . . . . . . . . . . . . . . . .
11.2.5
REORDR Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

'11-2
'11-2
'11-3
11-4
'11-5
'11-6

11 . 3

Caching Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11 .3.1
Size of the Cache . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11-7
11-8

11.3.2
11.3.3
11.3.4

Sequential and Random Caching Modes . . . . . . . . . . . . . . . . . . . . .
Random Mode Caching ............................... .
Sequential Mode Caching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11~

11.4

Caching Guidelines ........ " . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11-~

11.5

Virtual Disks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11-1~

11.6

DCl Command Descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
AllOCATE Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11.6.1
CREATElVIRTUAl_DISK Command ....................... .
11.6.2
DEAllOCATE Command .............................. .
11.6.3
11.6.4
DELETE/VIRTUAL_DISK Command ....................... .
11.6.5
SET CACHE Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11 .6.6
SET DEVICE Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11.6.7
SET PRINTER Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11.6.8
SHOW CACHE Command .............................. .
11.6.9
SHOW DEVICE Command ............................. .
11.6.10
SHOW DISKS Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11.6.11
SHOW PRINTER Command ............................ .

Chapter 12

11~
11-~

11-1~

11-1 ~
11-1-'
11-H
11-1i
11-H
11-2~
11-2~

11-2E
11-2S
11-29
11-30

Managing System Flies and Logical Names

12.1

Run-Time System Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
12.1.1
Adding and Removing Auxiliary Run-Time Systems ............. .
12.1.2
Associating a File with a Run-Time System .................. .

12-2
12-4
12-6

12.2

Resident Library Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
12.2.1
Resident Library DCl Commands ......................... .

12-7
12-7

12.3

System logical Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
12.3.1
Assigning logical Names .............................. .
12.3.2
Removing logical Names ...................... " ....... .
12.3.3
Listing System logical Names ................... " ....... .

12-8
12-9
12-10
12-10

12.4

Defining Concise Command language Commands ................... .
12.4.1
Adding a CCl Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
12.4.2
Listing Currently Defined CCl Commands ................... .
12.4.3
Removing a CCl Definition ............................. .

12-10
12-11
12-12
12-13

12.5

System File Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
12.5.1
Adding and Removing Swap Files ........................ .
12.5.2
Adding and Removing Overlay and Error Files ................ .
12.5.3
Using the DUMP/SYSTEM Command ...................... .

12-13
12-13
12-16
12-16

12.6

DCl Command Descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
ASSIGN/SYSTEM Command ............................ .
12.6.1
DEASSIGN/SYSTEM Command .......................... .
12.6.2
12.6.3
DEFINE/COMMAND/SYSTEM Command .................... .
DELETE/COMMAND/SYSTEM Command ................... .
12.6.4
12.6.5
DUMP/SYSTEM Command ............................. .
INSTAlUERROR_FllE Command ........................ .
12.6.6
INSTAlULIBRARY Command ........................... .
12.6.7
INSTAlUOVERLAY_FILE Command ....................... .
12.6.8
INSTALURUNTIME_SYSTEM Command .................... .
12.6.9
12.6.10
INSTALUSWAP_FILE Command

12-17
12-17
12-18
12-19
12-20
12-21
12-22
12-23
12-25
12-26
12-28

12.6.11
12.6.12
12.6.13
12.6.14
12.6.15
12.6.16
12.6.17
12.6.18
12.6.19
12.6.20
12.6.21
12.6.22
12.6.23
12.6.24

Chapter 13

12-30
12-31
12-32
12-33
12-34
12-35
12-37
12-38
12-39
12-40
12-41
12-42
12-43
12-44

Job Management

13.1

Controlling the Number of Logged-In Jobs . . . . . . . . . . . . . . . . . . . . . . . . . .

13-1

13.2

Broadcasting Messages to Terminals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

13-3

13.3

Controlling Jobs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

13-3

13.4

Removing Jobs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

13-4

13.5

Placing a Job on HOLD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

13-5

13.6

Setting Job Priority, Run Burst, and Maximum Size . . . . . . . . . . . . . . . . . . .

13-5

13.7

Controlling Keyboards and Remote Lines ........... , ............. .

13-7

13.8

DCL Command Descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13.8.1
ATTACH Command ............... , .................. .
13.8.2
BROADCAST Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13.8.3
DETACH Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13.8.4
FORCE Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13.8.5
HANGUP Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13.8.6
LOGIN Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13.8.7
REMOVE/JOB Command .............. " ............... .
13.8.8
SET JOB Command ................................... .
13.8.9
SET SYSTEM Command .............. " ............... .
13.8.10
SHOW JOB/PRIVILEGE Command . . . . . . . . . . . . . . . . . . . . . . . . .

13-7
13-8
13-9
13-11
13-12
13-13
13-14
13-16
13-17
13-19
13-20

Chapter 14

REMOVE/ERROR_FILE Command . . . . . . . . . . . . . . . . . . . . . . . . .
REMOVEILIBRARY Command . . . . . . . . . . . . . . . . . . . . . . . . . . . .
REMOVE/OVERLAY_FILE Command ....... " .............. .
REMOVE/RUNTIME_SYSTEM Command ................... .
REMOVE/SWAP_FILE Command . . . . . . . . . . . . . . . . . . . . . . . . . .
SET FILE Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
SHOW COMMAND/SYSTEM Command . . . . . . . . . . . . . . . . . . . . .
SHOW FILE/SYSTEM Command ............ " ............ .
SHOW LIBRARY Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
SHOW LOGICALS Command . . . . . . . . . . . . . . . . . . . . . . . . . . . .
SHOW LOGICAUSYSTEM Command . . . . . . . . . . . . . . . . . . . . . . .
SHOW RUNTIME_SYSTEM CommEmd ..................... .
UNLOAD/LIBRARY Command . . . . . . . . . . . . . . . . . . . . . . . . . . . .
UNLOAD/RUNTIME_SYSTEM Command .................... .

Miscellaneous System Management Functions

14.1

Loading and Unloading Resident SAT . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

14-1

14.2

Loading and Unloading Monitor Overlays . . . . . . . . . . . . . . . . . . . . . . . . . . .

14-2

14.3

Suspending System Operations .................................. .

14-2

14.4

Changing System Date and Time ....... " . . . . . . . . . . . . . . . . . . . . . . . . .

14-3

14.5

Recording System Activities: EMT Logging . . . . . . . . . . . . . . . . . . . . . . . . .
14.5.1
Programming for the EMT Logger . . . . . . . . . . . . . . . . . . . . . . . . . .

14-4
14-4

How EMT Logging Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Data Returned by EMT Logging . . . . . . . . . . . . . . . . . . . . . . . . . . .
14.5.3.1
Loggable EMTs Enabled or Disabled by SET SYSTEM .. .
EMT Logging and System Security . . . . . . . . . . . . . . . . . . . . . . . . .

14141414-

14.6

Monitoring System Status: SYSTAT ..................... " ........ .

14-

14.7

Dynamic Display of System Status: DISPLY ........................ .
14.7.1
Running the DISPLY Program . . . . . . . . . . . . . . . . . . " ........ .
14.7.2
Screen Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14.7.3
Job Status Statistics . . . . . . . . . . . . . . . . . . . . . . . . . " ........ .
14.7.4
Busy Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14.7.5
Disks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14.7.6
Message Receivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14.7."
Free Buffers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . " ........ .
14.7.8
Run-lime Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14.7.9
Resident Libraries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14.7:10
Memory Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1414-

14-1
14-1
14-1
14-1
14-1
14-1
14-1
14-2
14-2

14.8

Electronic Messaging Systems ................................. .

14-2

14.9

DCL Command Descriptions ................................... .
14.9."
LOAD/INDEX Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14.9.2
LOAD/OVERLAY Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
SET DATE Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14.9.3
SET SYSTEM Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14.9.4
SET TIME Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14.9.5
SHOW SYSTEM Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14.9.6
UNLOAD/INDEX Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14.9.7
UNLOAD/OVERLAY Command . . . . . . . . . . . . . . . . . . . . . . . . . . .
14.9.S

14-2
14-2
14-2
14-2
14-2
14-2

14.5.2
14.5.3
14.5.4

Chapter 15

1~
1~
1~

Managing Disks and Tapes

15.1

Working with Disks and Tapes ................................. .
15.1.1
Initializing Disks and Tapes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
15.1.2
Mounting and Dismounting Tapes and Disks . . . . . . . . . . . . . . . . . . .

151515-

15.2

Del Command Descriptions ................................... .
15.2.1
DISMOUNT Command for Disks . . . . . . . . . . . . . . . . . . . . . . . . . .
DISMOUNT Command for Tapes . . . . . . . . . . . . . . . . . . . . . . . . . .
15.2.2
INITIALIZE Command for Disks . . . . . . . . . . . . . . . . . . . . . . . . . . .
15.2.3
INITIALIZE Command for Tapes . . . . . . . . . . . . . . . . . . . . . . . . . . .
15.2.4
MOUNT Command for Disks . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
15.2.5
MOUNT Command for Tapes . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
15.2.6

15151515-

Chapter 16

15-1
15-1
15-1

Backing Up and Restoring Flies

16.1

Recovering Your System Disk (RECOVR) .......................... .

16-

16.2

DCl Command Descriptions .................................... .
16.2.1
BACKUP Command " . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
16.2.2
RESTORE Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

161616-'

Appendix A
A.1

Overview of Operator Services ...... . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.1.1
OPSER Program Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.1 .2
QUEMAN Program Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.1 .3
SPOOL Program Overview .................. , .......... .
A.1 .4
BATCH Program Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.1 .5
RJ2780 Program Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.1 .6
Controlling RESTOR with OPSER . . . . . . . . . . . . . . . . . . . . . . . . . .
A.1.7
Overview of OPSER Shutdown . . . . . . . . . . . . . . . . . . . . . . . . . . .

A-1
A-2
A-4
A-4
A-5

A-5
A-5
A-5

A-6

Operator Services Program: OPSER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.2.1
OPSER Operator Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.2.2
Message Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Valid Operator and Online Job Lists . . . . . . . . . . . . . . . . . . . . . . . .
A.2.3
A.2.4
Operator INTERRUPT Command . . . . . . . . . . . . . . . . . . . . . . . . . .
A.2.5
OPSER Startup Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.2.6
OPSER Action Under Various Startup Conditions .............. .

A-7
A-10
A-12
A-12
A-13
A-13

A.3

Queue Manager Program: QUEMAN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.3.1
QUEMAN Startup Commands and Switches .................. .
A.3.2
QUEMAN Interrupt Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.3.3
QUEMAN Startup Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.3.4
QUEMAN Action Under Various Startup Conditions ............. .
A.3.5
QUEMAN Consistency Checking . . . . . . . . . . . . . . . . . . . . . . . . . .

A-14
A-16
A-17
A-19
A-20
A-20

A.4

Line Printer Spooling Program: SPOOL . . . . . . . . . . . . . . . . . . . . . . . . . . . .
SPOOL Startup Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.4.1
Line Printer Spooling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.4.2
Keyboard Spooling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.4.3
Startup Error Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.4.4
SPOOL Interrupt Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.4.5
A.4.6
SPOOL Startup Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.4.6.1
Line Printer Startup with All Defaults ............... .
A.4.6.2
Line Printer Startup with Narrow Width ............. .
A.4.6.3
Keyboard Startup on an LA36 . . . . . . . . . . . . . . . . . . . .
A.4.6.4
Keyboard Startup on an LA 180 .................. .
Recovery from Line Printer Errors . . . . . . . . . . . . . . . . . . . . . . . . . .
A.4.7
Line Printer Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.4.8
Error Messages During User Output . . . . . . . . . . . . . . . . . . . . . . . .
A.4.9
Changing and Aligning Forms . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.4.10

A-21
A-22
A-27
A-28
A-28
A-30
A-31
A-31
A-31
A-32
A-33
A-34
A-35
A-36
A-36

A.5

Batch Processor Program: BATCH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.5.1
BATCH Startup Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.5.2
BATCH Interrupt Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.5.3
BATCH Startup Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.5.4
Operator Action Requests from BATCH . . . . . . . . . . . . . . . . . . . . . .

A-37
A-39
A-40
A-41
A-41

A.6

Operator Communication Program: PLEASE . . . . . . . . . . . . . . . . . . . . . . . .
A.6.1
Running and Terminating PLEASE . . . . . . . . . . . . . . . . . . . . . . . . .
A.6.2
OPSER Commands through PLEASE: . . . . . . . . . . . . . . . . . . . . . . .
A.6.3
PLEASE as a CCL Command . . . . . . . . . . . . . . . . . . . . . . . . . . . .

A-42
A-42
A-43
A-44

A.7

Terminating Operator Services and Spooling . . . . . . . . . . . . . . . . . . . . . . . .
A.7.1
OPSER Shutdown Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
OPSER Manual Shutdown Procedum . . . . . . . . . . . . . . . . . . . . . . .
A.7.2

A-44
A-44
A-44

A.2

xii

Managing the OPSER Spooling Package

A.S

RESTOR as an OPSER Controlled Program ........................ .

Appendix B

Number Conversion

Appendix C

Disk Device Sizes

Appendix D

General RSTS/E Messages

0.1

Appendix E

Special Characters Used In Error Messages ........................ .

A-4:

D-

SAVRES System Program

E.1

When to Use SAVRES ........................................ .

E-

E.2

Definitions of SAYRES Terms .................................. .

E-

E.3

Running SAYRES ........................................... .

E-:

E.4

SAYRES Switches .......................................... .

E-

E.S

SAYRES Dialogue ........................................... .
E.5.1
Device Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
E.5.2
Privileges Required . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
E.5.3
Checking the Input Vo~ume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
E.5.4
Checking the Output Volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
E.5.5
Restoring a RSTS/E Disk using the RESTORE Dialogue ......... .
E.5.6
Copying a RSTS/E Disk using the IMAGE Dialogue ............. .
E.5.7
IDENTIFY Dialogue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
E.5.8
Full Function Command Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

EE--

E-E--

EEE-1
E-1'
E-1:

E.6

SAYRES and Booting ........................................ .

E.7

Operator Interface During Processing ............................ .
E.7.1
Mounting and Dismounting Volumes . . . . . . . . . . . . . . . . . . . . . . . .
E. 7.2
Re-Accessing Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
E.7.3
Aborting SAYRES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

E-1'
E-21
E-2
E-2

E.S

SAYRES Summary Report ..................................... .
E.8.1
Summary Report Format and Example . . . . . . . . . . . . . . . . . . . . . .
E.8.2
Sum mary Report Run Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . .

E-2
E-2:
E-2:

E.g

SAYRES Error Handling ...................................... .
E.9.1
General SAYRES Error Messages . . . . . . . . . . . . . . . . . . . . . . . . .
E.9.2
Transfer Errors-Fatal and Nonfatal . . . . . . . . . . . . . . . . . . . . . . . . .

Appendix F

Device Testing Package (DEVTST)

F.1

Introduction to DEVTST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

F-1

F.2

DEVTST Operating Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

F-2

F.3

DEVTST Dialogue Errors . . . . . . . . . . . . . . . . . . . . . . . . . . , ........... .

F-4

F.4

CPEXER
F.4.1
F.4.2
F.4.3

CPEXER Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
CPEXER Dialogue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
CPEXER Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

F-5'
F-5
F-6

CPUTST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
F.5.1
CPUTST Dialogue ............ " . . . . . . . . . . . . . . . . . . . . . . .
F.5.2
CPUTST Errors .............. " . . . . . . . . . . . . . . . . . . . . . . .

F-7
F-7

DSKEXR
F.6.1
F.6.2

DSKEXR Dialogue ............ " . . . . . . . . . . . . . . . . . . . . . . .
DSKEXR Errors .............. " . . . . . . . . . . . . . . . . . . . . . . .

F-S
F-9
F-9

DSKSEK
F.7.1
F.7.2

DSKSEK Dialogue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DSKSEK Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

F-10
F-10
F-10

DXEXER
F.S.1
F.S.2

DXEXER Dialogue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DXEXER Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

F-11
F-11
F-11

KBEXER
F.9.1
F.9.2

KBEXER Dialogue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
KBEXER Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

F-12
F-12
F-13

LPEXER
F.10.1
F.10.2

LPEXER Dialogue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
LPEXER Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

F-13
F-13
F-13

MTEXER
F.11.1
F.11.2

MTEXER Dialogue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
MTEXER Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

F-13
F-14
F-14

1-1

MFOs and UFOs in Level 0 File Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . .

'1-5

F.5

F.6

F.7

F.8

F.9

F.10

F.11

F-4

F-8

Index

Figures

xiv

1-2

MFO, GFOs, and UFOs in the Level 1 File Structure . . . . . . . . . . . . . . . . . . . . .

1-6

9-1

One-to-One Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

9-4

9-2

One-to-Many Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

9-3

Many-to-One Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

9-4
9-5

11-1

Caching Mode Checks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11-10

11-2

Monitor Caching Checks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11-19

13-1

Priority Byte Format . . . . . . . . . . . . . . . . . . . . . . . . . . . " . . . . . . . . . . . . . . .

13-5

A-1

System Controlled Programs and Operator Interaction . . . . " . . . . . . . . . . . . . . .

A-3

E-1

Summary Report Format, , , , , , , , , , , , , , , , , , , , , , , , , . , , . , , , , , , , , , , ,

E-22

2-1

Initialization Option Summary, , , , , , , , , , , , , , , , , , , , , , , , , , . , , , , , , , . , ,

2-3

6-1

RSTS/E Privileges , , , , , , , , , , , , . , , , , , , , , , , , , , , , . , , , . , , , , . , , , , , ,

6-2

7-1

DCl Commands for Managing Accounts , , , , . , . , , , , . , , , , , . , , , , , , , , , , ,

7-1

8-1

RSTS/E Supported Controller Types , , , , , , , , , , , , , , , , , , , , , , , , . , , , , , , ,

8-2

9-1

9-1
9-12

10-1

DCl Commands for the Print/Batch Services Package , , , , , , , , , , . ,....... , .
PBS Batch Job Exit Status Messages , , , , , , , , , , , , , , , , , , , . , , , , , , , , , , ,
ERRDIS Dialogue Explanation , . , , . , . , , , , , , , . , , , , , , , , , , , , , , , , , , . , ,

10-2

User Description Data, , , , , , . , , , " , , , , , , , , , , , , , . , , . , , , . . . , , , , , , , ,

10-11

10-.3

Disk Error Detailed Description , . , , " , , , , , , , , , , , , , , , , . , , . , , , , . , , , . , ,

10-13

10-4

Nondisk Peripheral Device Format , , , . , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,

10-18

10-5

Nonperipheral Error Format , , , , , . , , , , , , , , , , , , , . , , . . , . . . , , , , , , , , , ,

10-20

10-6

ANALYS Program Dialogue , , , , , , , , , , , , , , , , , , , , , , , .............. .

10-22

10-7

System Crash Error Code . . . . . . . . . . . . . . . . . . . . . , , , , , , , , , , , , , , , , , ,

10-23

11-1

11-3

DCl Commands for Device Management and Caching Control, , , , , . , , , , , , , ,
REORDR Dialogue Questions , , , , , , , , , , , , , , , , , , , , , , , , , , . , , , , , , , , ,
REORDR Message and Response , , , , , , . , , , , , , , , , , , .. , , , , , , , , , , , . ,

11-6

12-1

DCl Commands for Managing System Files and logical Names, .. , . , , , , , , , ,

12-1

Tables

9-2

11-2

10-5

11-3

13-1

DCl Commands for Job Management, , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,

13-1

14-1

DCl Commands for Miscellaneous System Functions " " " " ' , . , " " " "

14-1

14-2
14-3

DISPlY Program Switches " " " " , " " , . , " " " " " " " " ' , ... .
DISPlY Program Commands ........... , .. , .. , .................. .

14-11

14-4

STATE Column Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

14-15

14-5

Busy Devices Status Abbreviations: WHY Column ... , " " " " " " ' , " "

14-16

14-6

14-17

14-7

Disk Status Abbreviations: COMMENT Column " " " " " , ' , " " " " ' "
Message Receivers Abbreviations, , . , . , , , , , , , , , , , , , , , , , , , . , , , , , , , , ,

14-8

Run-Time Systems and Resident Libraries Report Abbreviations , , , , ,. , , , , , , , ,

14-19

14-9

Memory Status Report Abbreviations. , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,

14-21

15-1

DCl Commands for Disk and Tape Handling , , , . , , , , , , , , , , , , . , , , , , , , , ,

15-1

15-3

Disk Size and Cluster Size " " , . " , " ' , ' , . , " " " " , .. , , . , " " "
The IPUBLlC, IPRIVATE, and I[NO]SHARE Qualifiers for the MOUNT Command, ,

15-8
15-16

16-1

IINITIALIZE and lOUT_lABEL Interaction During BACKUP , , , , , , , , , , , , , , , , ,

16-8

16-2

BACKUP Quick Reference Table ,.", .. , " " " " " " " , . , . " . , . , ' "
IINITIALIZE and lOUT_lABEL Interaction During RESTORE, , , , . , , , . , , , , , , ,

16-10
16-18

A-1

RESTORE Quick Reference Table ' " " " " " " ' , . , . , " ' , . , ' , . , ' , . , '
OPSER Commands, , , , . , , , , , , , , , , , , , , , , , , . , , , , , , , , , , , , , , , , , , ,

A-2

OPSER Message and Action Request Contents , , , , , , , , , , , , , , , , , , , , , , , ,

A-11

A-3

OPSER Online Job List , , " " ' , . , " " " " " " " " " " " " " " "

A-12

A-4

QUEMAN Startup Commands " " " " " , ' " . , " ' , . , " " " " " , . , '
QUEMAN Startup Switches, , , , , , . , , , , , , , , , , , , . , , , , , , , , , , , , , , , . , ,

A-16
A-17

A-8

QUEMAN Interrupt Commands, , , , . , , , , , , , , , , , , , , , , , , , , , . , , . , , , , , ,
NEXT Command Error Text , , , , , , ... , , , , , , , , , , , , , , , , , , , , , , , , , , , . ,
SPOOL. Startup Options . , , . , , , , , , . , , , , , , , , , , . , , , , , , , , , , , , , , . , , .

A-9

SPOOL. Line Printer Default Values, . . . . . . . . . . . . . . . . . . . . . , ... , . , , , ..

A-27

A-10

SPOOL. Keyboard Default Values , . , . , . , . , , .... , ...... , , . , , . , .. , . , ,

A-28

A-11

SPOOL. Syntax Error Messages. , , , .. , , . , , .. , . , ..... , .. , . , . , . , .. , .

A-29

15-2

16-3
16-4·

A-5
A-6
A-7

14-9

14-18

16-16
A-8

A-16
A-19
A-23

A-12

SPOOL Interrupt Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

A-30

A-13
A-14
A-15

SPOOL User Requested Output Error Message~3 . . . . . . . . . . . . . . . . . . . . . . .

A-36

BATCH Startup Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..

A-39

BATCH Interrupt Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

A-16

BATCH Device Type Designators ......... " . . . . . . . . . . . . . . . . . . . . . . . .

A-17

PLEASE Commands to OPSER .......... " . . . . . . . . . . . . . . . . . . . . . . . .
RESTOR Commands through OPSER ...... " . . . . . . . . . . . . . . . . . . . . . . . .
Bit Values in the PDP-11 Word. . . . . . . . . . . " . . . . . . . . . . . . . . . . . . . . . . . .

A-40
A-41
A-43
A-46

A-18

B-1

xvi

B-1

C-1

Disk Device Sizes . . . . . . . . . . . . . . . . . . . " . . . . . . . . . . . . . . . . . . . . . . . .

C--2

0-1

General RSTS/E Messages . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . .

0-2

E-1

SAYRES Functions .................. " . . . . . . . . . . . . . . . . . . . . . . . .

E-3

E-2

SAYRES Switches ................... '. . . . . . . . . . . . . . . . . . . . . . . . .

E-5

E-3

SAYRES Device Specification Switches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

E-6

E-4
E-5

SAYRES Input Volume Error Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

E-7

RESTORE Dialogue Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

E-10

E-6

IMAGE Dialogue Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

E-7

IDENTIFY Dialogue Question . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

E-8

Booting RSTS/E and SAVE Set Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

E-9

Summary Report Run Totals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

E-10

General SAYRES Error Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

E-11
F-1

SAYRES Nonfatal Transfer Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DEVTST Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

E-14
E-17
E-19
E-23
E-25
E-29

F-2

RSTS/E Device Designators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

F-3

DEVTST Error Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

F-4

F-2

Preface

Objectives
This guide explains how to manage a RSTSIE system on a day-to-day basis.
Th use the features described in this guide, RSTSIE requires you to have certain privileges. Only the RSTSIE system manager and users with the needed
privileges should have access to this guide.

Audience
This guide is primarily for use by system managers and operators. You do not
need to have a great deal of knowledge of and experience with RSTSIE to manag«
a RSTSIE system. However, to perform certain system management tasks, you
should understand how to use a text editor like EDT and be familiar with the
DIGITAL Command Language (DCL).
In addition, because not all customers install their own systems, the manual
includes information for DIGITAL field service engineers and software support
representatives. This information is clearly indicated.

D10cument Structure
This manual contains 16 chapters and six appendixes:
Chapter 1

Contains a glossary of RSTSIE terms and an overview of the
RSTSIE disk structures, system management concepts, multiple
privileges, account management, and file access.

Chapter 2

Describes procedures for starting, halting, and restarting RSTSIE.

Chapter 3

Describes what the system startup command file should include.

Chapter 4

Describes the phases the RSTSIE system goes through when shutting down.

Chapter 5

Explains the quotas that limit the amount of various system resources that users can have.

Chapter 6

Describes the privileges that you can assign to a user when you
create an account.

Chapter 7

Gives background information on managing RSTSIE accounts and
describes the DCL commands used to create and maintain the
accounts.

Chapter 8

Gives background information on managing terminals and terminal
servers, and describes the relevant DCL commands.

xvi

Chapter 9

Gives background information on the PrintlBatch Services (PBS)
Package and describes the DCL commands for using PBS.

Chapter 10

Describes the available programs and DCL commands for managing
system errors.

Chapter 11

Introduces the concepts of device management and caching control
on your RSTSIE system and describes the relevant DCL commands.

Chapter 12

Gives background infonnation on RSTSIE system files and logical
names and describes the relevant DCL commands.

Chapter 13

Gives background information on RSTSIE jobs and describes the
DCL commands for managing them.

Chapter 14

Describes miscellaneouEI DCL commands.

Chapter 15

Gives background infonnation on working with disks and tapes and
describes the DCL commands for managing them.

Chapter 16

Describes the DCL commands for backing up and restoring disk
files and accounts.

Appendix A

Describes the OPSER-based spooling package.

Appendix B

Lists the octal and decimal values of each bit in the PDP-11 word.

Appendix C

Lists the cluster size and total size of each disk that RSTSIE
supports.

Appendix D

Lists the general RSTSIE messages that you may encounter when
using RSTSIE.

Appendix E

Describes the SAVEIRESTORE system program for copying and
backing up disks.

AppendixF

Describes the Device Testing Package (DEVTST) used to verify the
reliability of RSTSIE supported hardware.

Related Documentation
As system manager, you are responsible for setting certain parameters that
affect the monitor and, consequently, the monitor directives. You also need to
make decisions related to installing and updating your RSTSIE system. For more
information on these topics, refer to:

RSTS / E System Installation and Update Guide
RSTS / E System Directives Manual
RSTS / E Programming Manual
For more information on RSTSIE documentation, consult the RSTS / E
Documentation Directory.
In addition to the RSTS/E documentation, you should have access to the following
manuals:

PDP-ll Peripherals Handbook
DECnet / E System Manager's Guide

xviii

Conventions
Symbols and other conventions represent keyboard characters or aid in the presentation of infonnation. This guide uses the following s~bols and conventions:
<RET>

Represents a carriage-returnlline-feed combination. Press the Return
key on your keyboard to produce this combination.
Represents a line-feed character. Press the Line Feed key on your
keyboard to produce this character.

Represents the Ctrl key and some other character, represented here
by x. For example, enter <Ctrl/U> by holding down the Ctrl key and
pressing the key labeled U. RSTS echoes or displays the character at
your terminal as AU.

color

In examples, black characters represent data produced by the compute1
Red characters indicate information that you enter.

[]

Square brackets show the optional parts of a command in format
statements. For example:
DIRECTORY [file-spec[, ... ]]
In this example, you can include one file specification ([file-spec]), or
more than one ([, ... ]). Square brackets also indicate the choice you hav,
in using a qu~difier. For example:
![NO]DELETE
This means you can enter either !DELETE or INODELETE.
A vertical bar indicates a choice. For example:
ILOWERCASE=[INPUT I OUTPUT]
means you can enter ILOWERCASE=INPUT or ILOWERCASE=OUTPl

Summary of Technical Changes
The RSTS / E System Manager's Guide for RSTSIE V9.7 contains the following
revisions and updates:
•

Changing the monitor when restarting the system (Chapter 2)

•

Changing the maximum size of the swap file when restarting the system
(Chapter 2)

•

Running SHUTUP from any local terminal (Chapter 4)

•

Running SHUTUP with the automatic restart feature (Chapter 4)

•

Managing dynamic region limits (Chapter 5)

•

Managing Local Area Transport (LAT) (Chapter 8)

•

Managing pseudo keyboards (Chapter 8)

•

Enabling the Forms Management System (FMS-11) (Chapter 8)

•

Managing the virtual disk (Chapter 11)

•

Fine-tuning the EMT logging program (Chapter 14)

•

Using electronic messaging systems (Chapter 14)

•

Using the revised SAVRES backup utility (Appendix E)

xxi

Chapter 1

Overview of RSTS/E System Managemen1
RSTSIE runs on a PDP-11 computer. It allows simultaneous, timeshared access
to PDP-11 hardware and to RSTSIE software components through either local
or remote asynchronous terminals. The RSTS / E Software Product Description,
which is part of the RSTSIE media kit, describes the hardware and software
RSTSIE supports. This guide describes the software a system manager uses to
control the operation of a RSTSIE system. The following sections introduce some
impOl'tant RSTSIE concepts.

1 . 1 Glossary of RSTS/E Terms
To understand how RSTSIE works, you need to know these terms:
CCL (Concise
Command
Language)

A concise way to run a RSTSIE system program. The CCL syntax
lets you run a program without the RUN command and, unlike the
RUN command, lets you place the entire command on one line. After
the program completes executing, control returns to your job keyboard monitor. The system manager chooses the CCL commands for
a particular RSTSIE system. In DCL, CCL's are defined using the
DEFINE/COMMAND/SYSTEM command (see Chapter 12).

DCL (Digital
Command
Language)

A set of commands available on many different DIGITAL systems.
These commands perform basic tasks like copying and printing files,
and running programs. On RSTSIE, the DCL run-time system manage
the DCL command environment. See the RSTS / E System User's Guid
for more information about DCL on RSTSIE. See the RSTS / E Quick
Reference Guide for a complete list of all DCL commands available on
RSTSIE.

Default
Keyboard
Monitor

The main keyboard monitor on a RSTSIE system. You enter the defau:
keyboard monitor (DCL) after you log in.

Job

The unit that RSTSIE uses to keep track of you (and other users)
during a terminal session. When you log in, the system creates a job
for you and assigns it a job number. The system uses your job numbeJ
to keep track of everything you do from the beginning to the end of
your terminal session.

Job Keyboard
Monitor

The keyboard monitor that manages a job. The job keyboard monitor:
the same as the default keyboard monitor, unless you decide to chang~
it with the SET JOBIKEYBOARD command. After you change the jo1:
keyboard monitor, you remain under its control until you log out or us
SET JOBIKEYBOARD again to change it.

Overview of RSTS/E System Management 1-

Keyboard
Monitor

The part of a run-time system which interprets keyboard commands.
When working in the Digital Command Language (DCL) environment,
for example, you enter commands that the DCL keyboard monitor
receives and interprets. Each RSTSIE keyboard monitor has an identifying prompt that it displays to indicate when it expects command
input. Common keyboard monitor prompts on RSTSIE are: dollar sign
($) for DCL; "Ready" for BASIC-PLUS; angle bracket (» for RSX; and
period (.) for RTll.

Monitor or
Executive

The master control system Hoftware that observes, supervises, controls
or verifies the operation of a computer system. Also, the collection
of routines that controls thEl operation of user and system programs,
schedules operations, allocates resources, and performs 1/0.

Operating
System

The collection of programs, including a monitor or executive and system
programs, that organizes a central processor and peripheral devices
into a working unit for the development. and execution of application
programs.

Program
Development

The process of writing, entering, translating, and debugging source
programs.

Public Structure

The set of all disks that are public. When you do not include a device
name in the file specification, the system by default accesses one of
the disks on the public structure. The device name SY: represents the
name for all disks in the public structure. Thus, if you do not have any
public disks other than the system disk, SYO: and SY: are equivalents.
If you have more than the system disk in the public structure, SY:
refers to the aggregate of all public disks. SYO: refers to the system
disk.

Run-Time
System

System software that manages part of the RSTSIE system. For example, the BASIC-PLUS run-time system manages the BASIC-PLUS programming environment. The run-time system controls the interaction
between a user or program and the monitor.

System Disk

The disk required by the RSTSIE monitor to start the system and to
allow the system to run properly under timesharing. RSTSIE assigns
device name SYO: to the system disk.

System Program

Any general-purpose program included in an operating system to
perform common functions. CUSP is an acronym for Commonly Used
System Program.

1.2 System Management
Management of RSTSIE begins with providing a properly tailored hardware
and software configuration, proceeds through initializing the software at system
installation time, and continues with the daily functioning of timesharing. To
perform these steps efficiently, you should be familiar with timesharing concepts
and practices or have a close working relationship with a senior programmer or
analyst experienced with timesharing systems.

To manage a system well, you should know the processing requirements of the
system and its capabilities and structure. Information in the RSTS / E System
Installation and Update Guide describes important aspects of hardware and
software options and provides memory requirements to assist in configuring a
RSTSIE system.
As the system manager, you should be particularly careful about initializing your
system for the first time. Because there are so many variables, no prepackaged
approach can give you a completely optimized RSTS/E system. The RSTS/E
System Installation and Update Guide provides detailed information about all the
initialization options.
1-2

Overview of RSTS/E System Management

To make sure that important information reaches the right persons within your

organization, you should make someone responsible for current documentation
of the system. This responsibility includes both locally generated and Digitalsupplied procedures and guides. The responsible person should make sure that
specified members of the staff receive the most current system information. In
too many cases, improper use of resources occurs when responsible individuals de
not have the latest information.

1,,3 Disk Organization
System managers should understand that the organization of files on disks is
important. Only with this knowledge can managers make the best use of all
system resources. The following section introduces the concepts of disk and file
structures and how to use them to best advantage.

1.3.1

Types of Disks
RSTS/E supports two types of disks as part of its logical disk structure: private
and public. The public disk structure consists of a system disk (the one that was
booted and contains the RSTSIE monitor) and additional public disk packs or disl
cartridges. The system disk must be physically on line and logically mounted
whenever the system is running; this way, users can gain access to the system
disk during timesharing operations. All other public disks that users may need t(
access must be physically on line and logically mounted.
The system disk contains:
•

Monitor code

•

Initialization code

•

The default keyboard monitor (DCL)

•

The command files for system startup and crash recovery

Some installations may also use the system disk for storing active user jobs that
are temporarily swapped out of memory.
Remaining space on the public disk structure is used for:
•

Auxiliary run-time systems

•

Optional system files

•

System programs

•

User programs and data files

You can use disk drives not devoted to the public structure for private disk
packs or disk cartridges. Unlike public disks, you can logically and physically
dismount and move private disks to other drives during timesharing operations.
In addition, private disks make it possible to restrict disk storage to a defined
set of users, which can be a distinct advantage when numerous projects are in
progress. The file structure of a private disk is the same as that of a public disk.
In addition to private and public disks, both of which any job on the system can
share and access, you can logically mount a disk for use by only one job. In this
case, only that job can access the disk. When the job terminates (for example, by
logging out), the disk is logically dismounted. See "MOUNT Command for Disks"
in Chapter 15.

Overview of RSTS/E System Management

1-3

1.3.2

File Structures
The RSTSIE file structures let the monitor access system and user data in an
organized way. RSTSIE currently supports three kinds of file structures:
•

Level O.O-is for RSTSIE disks created before VB.O. You can mount disks with
this structure as read-only, private disks. You cannot mount them as system
or public disks.

•

Level 1.1-is the structure created for VB.O. You can mount disks with this
structure as read/write, private disks. You cannot mount them as system or
public disks.

•

Level 1.2-is the structure created for V9.0 and later. You can mount disks
with this structure as read/write, system, public, or private disks.
NOTE

If you are upgrading a single disk RSTS/E system to V9.0, you can
mount a Level 0.0 or Level 1.1 disk as the system disk. See the
following section "Converting Disk Structures."
The following three sections describe the three kinds of structures. If you have
a RSTSIE V9 system with all new disks, you can skip the sections that describe
the disk structures before V9. Furthermore, if you have pre-V9 disks, DIGITAL
recommends that you use the new DSKCVT disk conversion program to convert
them to the latest disk structure (see the section "Converting Disk Structures
(DSKCVT)" later in this chapter).
To see which structure applies to disks on your system, type SHOW DISKS

as shown. The "Level" column indicates the RSTSIE disk structure level. For
example:
$ SHOW DISKS <RET>

Disk Structure:
Size
Dsk Open
DR1
66 131648
DR2
0 242576
DR3
33 500352
DR4
0 242572
DR5
0 500352

Free
Clu Err
29424 22%
4
0
33040 13%
8
0
57744 11%
8
0
17528
7%
4
0
76152 15%
8
0

1-4

Overview of RSTS/E System Management

Name
Level
BOWER
1.2
CRONIN 0.0
WOJTAS 1.1
MODNE
1.2
HOGAN
0.0

Comments
Pub, DLW
Pri, R-O, DLW
Pri, DLW
Pri, DLW
Pri, R-O, DLW

1.:1.3

Level 0 Disk Structure (Pre-V8.0)
Pre-VS.O users gained access to files on a RSTSIE system by a Master File
Directory (MFD) and by User File Directories (UFDs). Figure 1-1 shows these
structures.
Figure 1-1:

MFOs and UFOs In Level 0 File Structure

Files

UFDs
[1,2]
MFD
(Located in [1,1])
[43,98]
Contains pOinters
to files, as
well as to
UFDs, in a
"linked list"

[100,150]

[254,254]

ZK-8042-HC

As Figure 1-1 shows, each disk initialized for use on a RSTSIE system contains
an MFD, located in [1,1]. The system uses the MFD to catalog other accounts
on the disk. The MFD on the system disk lists the accounts that can be used to
log in to the system. On a private disk, the MFD contains entries of accounts
that can create files on that disk. Any user can access any file on any disk if the
protection code of the file pennits. Only users whose accounts are entered in the
MFD on a private disk can create files on that disk.
The system creates one UFD for each user account on a disk when the manager
sets up the account (or, optionally, when a user creates the first file for the
account). The UFD catalogs all program and data files under an account and
maintains accounting and access information for these files. The UFD contains
all retrieval information for the files because each file is pure data and has no
linkage or structural information.

Overview of RSTS/E System Management

1-5

1.3.4

Level 1 Disk Structure (for VS.O and V9.0)
In VB.O and V9.0, users access files on a USTS/E system through the use of three
structures:
•

A Master File Directory (MFD)

•

A Group File Directory (GFD)

•

A User File Directory (UFD)

Figure 1-2 shows how these structures are related.
Figure 1-2:

MFO, GFOs, and UFOs In the Level 1 File Structure

UFDs

Files

[1,1]
GFDs

[1, *]

MFD

[1,25]

[43,*]
[43,98]
Contains pOinters
to GFDs by
"direct access"
[100,60]
[100,*]

[100,210]

(254,.J
L------......I

~ (254~254J ~
..

ZK-8043-HC

As Figure 1-2 shows, the MFD contains pointers to the Group File Directories
(GFDs). The GFDs point to UFDs within each group. For example, a pointer in
the MFD could access the GFD for files in [43,*] directories, from which it could
access a specific UFD, such as [43,98].

1-6

Overview of RSTS/E System Management

1,,3.5

Differences Between Level 0, Level 1.1, and Level 1.2
The new disk structure (Level 1.2) differs from the previous disk structures (Leve

o and Level 1.1) in the following ways:
•

The account data on Level1.2 contains attribute blockettes for the new or
expanded attributes such as quotas, multiple privileges, and long passwords.

•

The MFD on Level 1.1 and Level 1.2 is not associated with an account
number.
On Level 0, the MFD was located in [1,1], and had to be at Block #1. Now,
Block #1 points to the location of the MFD, allowing the MFD to be placed
anywhere on the disk.

•

The MFD and GFDs on Level 1.1 and Level 1.2 do not store files.
On Level 0, the MFD could store files as well as point to accounts.

•

GFDs on Level 1.1 and Level 1.2 allow more direct access to accounts.
On Level 0, the MFD contained a linked list of pointers to all the UFDs.
UFDs were listed in the order in which they were created; the MFD was
searched in sequence until it located a UFD.
Now, the MFD contains direct pointers to GFDs, which in turn point directly
to any UFD. (However, the structure of the UFDs remains the same as in
Level 0.) Thus, DCL commands that display accounts (for example, SHOW
ACCOUNT) list the accounts in ascending order, rather than in order of
creation.

•

You can create up to 65,025 accounts with the Level 1.1 or Level 1.2 structure
With the Level 0 structure only 1735 accounts could be created and referenced
on a disk. Now, you can create accounts for 255 groups, each with 256 users.
Only the disk size restricts the nwnber of accounts.

The users on your system should not see the differences between the three
structures. Level 1 takes up a 'little more disk space. There are more pointers,
because each group is handled separately, but access to files is faster and the
system can have more accounts. The linked list structure of the MFD is replaced
by direct access to GFDs.

1.:3.6

Converting Disk Structures (DSKCVT)
The DSKCVT program perfornls the following functions:
•

Converts Level 0.0 disks to Level 1.2

•

Converts Level 1.1 disks to Level 1.2

•

Converts accounts with LOOKUP passwords to NOLOOKUP (optional)

DSKCVT converts disks directly; you need no second disk to perform the conversion. DIGITAL recommends that you first back up the data on the disks to
be converted, so that any hardware or software errors that may occur during
conveI'sion will not cause loss of data.
DSKCVT requires some work space if the pack being converted is Level 0.0
format. Make sure there are 32 blocks times the number of different group
nwnbers used on the pack. For example, if there are ten group numbers, the
pack should have 320 blocks of free space for conversion.

Overview of RSTS/E System Management

1-7

DSKCVT accepts either logically mounted or logically dismounted packs.
However, the pack should not be in use when you run DSKCVT. 'lb make
sure no one is using the pack, DSKCVT dismounts and remounts it at the start of
the conversion.
In converting Level 1.1 disks to Level 1.2, DSKCVT adds several new atrtibute
blocks to each account. Since some of the new attributes are only relevant to user
accounts, DSKCVT asks:
Convert accounts to NOUSER <no>?

If you answer yes to this question (the default is no), DSKCVT changes all
accounts to non-user accounts. Non-user accounts have no passwords or privileges
and can only serve to store files.
CAUTION

If you intend to move the accounts (via BACKUP and RESTORE) to
the system disk from a disk you are converting with DSKCVT, then do
not convert the accounts to NOUSER. Otherwise, you will not be able
to log into the accounts or create jobs in them.
DSKCVT converts accounts to Level 1.2 by adding the following new attribute
blocks to each account:
•

Disk quota/usage block

•

Privilege block

•

Nondisk block

In creating the privilege attribute block, DSKCVT uses the following rules:
•

For account [1,1], it assigns all privileges except EXQTA

•

For all other [1,*] accounts, it assigns all privileges except SYSMOD and
EXQTA

•

For accounts not in the [1,*] group, it assigns no privileges

This scheme assigns privileges that are ,compatible with previous versions of
RSTSIE. Once the conversion is complete, you can modify privileges as necessary
using the SET ACCOUNTIPRIVILEGE command. ]li'or example, if you want all
users to be able to change their password, type:
$ SET ACCOUNT/PRIVILEGE=SETPAS [*,*]

DSKCVT sets the quota values as follows:
•

Logged-Out Disk Quota: Quotas of 0 (unlimited in VB.O and earlier) are
changed to -1 (unlimited in V9.0 and later). Other values are retained.

•

Logged-In Disk Quota: Set to unlimited.

•

Job Quota: Set to unlimited.

•

Detached Job Quota: Set to unlimited for accounts [1,*], zero for others.

•

RIB Quota: Set to unlimited for accounts [1,*], 3 for others.

•

Message Quota: Set to unlimited for accounts [1,*], 12 for others.

These values are compatible with VB.O. Again, you can use the SET ACCOUNT
command to change them as needed.

1-8

Overview of RSTS/E System Management

The accound flags are set as follows:
•

NOEXPIRE (If the account's password was ?????? then the account is marked
EXPIRED.)

•

NOCAPrIVE

•

INTERACTIVE

•

DIALUP

•

NETWORK

•

PASSWORD_PROMPT

•

LOOKUP or NOLOOKUP (according to what you specified)
NOTE

Accounts with any ?s in their passwords convert properly but cannot be
logged into, since? is an invalid character in V9.0 and later. This is
compatible with the meaning of? in passwords in previous versions.
When converting a disk, you can decide whether to convert all account passwords to their hashed form (NOLOOKUP) or leave them in their readable form
(LOOKUP). If your site still uses applications that must look up passwords, do not
hash passwords on the converted disk. (When you no longer need LOOKUP passwords, you can use DSKCVT to convert all account passwords to NOLOOKUP on
an existing Level 1.2 disk.)
If you convert all accounts to NOUSER accounts, then DSKCVT will not create a
password, privilege, or non-disk quota block for all accounts on the disk.
You need all privileges to run DSKCVT. DSKCVT displays an error and aborts if
you do not have all privileges currently enabled. There are two ways to obtain all
privileges in order to use DSKCVT:
•

Log into an account that has all privileges. Such an account has all the power
and capabilities of the [1,1] account in V8.0, including the ability to poke
memory (SYSMOD privilege).

•

During system startup, answer no to the Proceed with system startup?
prompt. You will return to the DCL prompt and will be running in a job that
has all privileges. You can then run the DSKCVT program to convert disks
as necessary. Once you have finished using DSKCVT, you can reswne system
startup by typing the command:
$ @[O,I]START

The second method shows that you do not require any account on your system
that has all privileges. Whenever you need to perform an operation that requires
a privilege not assigned to any account, you can do so by interrupting the initial
system startup procedure.
The following is an example of the DSKCVT program:
$ RUN $DSKCVT
DSKCVT V9.3-03

RSTS V9.6-11

Disk to convert? DU2:

** CURRENT DISK ATTRIBUTES **
Disk:
Label:
Level:

DU2:
WORK
1.1

Overview of RSTS/E System Management

1-9

Convert accounts to NOUSER <NO>?
Convert passwords to NOLOOKUP <YES>?
Display accounts <YES>?
Proceed (YES or NO)? Y
Converting disk DU2: to level 1.2 .•.
Account [ 0,1 J converted
Account [ 1,1 J converted
Account [ 1,2 ] converted

Converting disk DU2: passwords to NOLOOKUP
Account [ 0,1 T is a NOUSER account
Account [ 1,1 J is a NOUSER account
Account [ 1,2 J is a NOUSER account

Disk

DU2: conversion completed on 19-Apr-90 at 11:04 AM

In the example, DU2: is the disk pack to convert to Level 1.2. When you type
DU2: in response to the prompt, DSKCVT displays the pack ID DU2:WORK. This
lets you verify that you are converting the right pack. If the pack was already
converted, DSKCVT displays an error message and then repeats the Disk to
convert: prompt. Otherwise, DSKCVT goes on with the dialogue.
When you answer Y to the Proceed (YES or NO)? prompt, DSKCVT begins the
conversion. DSKCVT processes the accounts in ascending order. When the pack
is converted, DSKCVT dismounts it and updates the pack label. The pack is now
in Level 1.2 format. However, the storage allocation table (SAT) does not yet
reflect the free disk space on the pack. 1rherefore, the system marks the pack
"dirty" until DSKCVT updates the SAT. 'rhen the system updates the pack label
to show that the pack is now rebuilt. If the pack was mounted when DSKCVT
started, the system mounts it once again, with the pack ID as its logical name.
DSKCVT will not lose data even if the program aborts because of lack of work
space, a system crash, or a disk hardware problem. If there is a problem dwing
the conversion, you need only correct the problem and rerun the conversion. In
such a case, you do not need to restore the pack from a backup pack.
If the conversion proceeded far enough t.hat DSKCVT was able to convert the
pack, you only need to mount it. Otherwise, DSKCVT restarts the conversion
from the beginning, after deleting the work files left over from the first attempt
at conversion.
When you convert a Level 0.0 (pre-V8.0) disk, DSKCVT makes two passes. In the
first pass, DSKCVT converts the Level 0.0 pack to Level 1.1. The second pass
converts the Level 1.1 pack to Level 1.2.

1.3.7

Disk Optimization
When a user tries to access a file on the public structure, the system searches the
directories on all disks that are part of the public structure. The search verifies
whether the file exists. You can avoid the overhead of searching more than one
disk by placing frequently accessed fileEl on a private disk. This technique is
valid for systems with a number of disks in the public structure, as well as for a
system whose public structure consists of only one disk. The ability to place and
preextend directories can also help you optimize disks (see Chapter 7).

1-10

Overview of RSTS/E System Management

It is sometimes an advantage to dedicate an entire private disk to a single large

production file. Such organization ensures an efficient directory structure and
minimizes overhead to access file data. When more than one file is on the same
private disk, it is best to dedicate a whole account to each production file. The
system then spends less time searching the directory for file information.
In general, you should keep volatile files and stable files in separate accounts to
avoid fragmentation. Where file entries in the directory are spread across the faCE
of the disk rather than being confined to a few sequential blocks, fragmentation
requires the disk head to move more than is optimally efficient. Preextending
directories can ensure that the directory is in sequential blocks (see Chapter 7).
REORDR can restructure the directories on your system to improve disk access
time (see Chapter 11). You can also optimize disk usage by separating frequently
accessed files from those you use less often.
In an environment where distinct data files must be accessed by the same
program, the optimal organization is to keep each file on a different private
disk. If a program must access more than one file on the same disk, overhead is
increased because of disk-head movement. A large percentage of time is spent,
therefore, in moving the head back and forth. However, if each file referenced
by the program exists on a distinct private disk, head movement is not required
whenever a program references another file. Head movement is restricted to
locating the data itself. Positioning of directories can also help reduce head
movement (see Chapter 7).

1.4 System Operation Concepts
Immediately after logging in to the system, a user's terminal is under the control
of DCL (the system's default keyboard monitor). The terminal is said to be at the
system command level because you can type a system command and the keyboard
monitor processes it accordingly. The terminal returns to the system command
level when a command or program completes executing, or when you enter a
CtrllC at the terminal. When the terminal is at the system command level, the
keyboard monitor examines each ASCII text line entered and determines whether
that line is a system command. System commands are executed upon entry as
described in the RSTS / E System User's Guide.
The user job area is initialized at log-in time and set to a size of at least 1K words
(K = 1024). The job area can grow in increments of 1K words to a maximum
size set by the system manager. The maximum size for any job image, including
memory used by the current run-time system of the job, is 64K words. This is the
maximum size that the PDP-11 Instruction and Data (I&D) architecture (PDP11/44,45,50,53,55,70,73,83,84) allows. For a PDP-11 without I&D architecture
(PDP-11/23+,24,34,35,40,60) the maximum. size is 32K words.
Under the RSTSIE system, jobs run one at a time. A job runs until it enters an
Input/Output (1/0) wait state or until it exhausts the time quantum assigned to it
either by the system or the system manager. When the current job stops running,
the scheduler runs the next job that is ready. Meanwhile, the interrupt-driven
1/0 device handlers are processing requested data transfers. Mter completing a
transfer, the scheduler marks the job that requested the transfer as ready to run
again and starts it from the point at which execution stopped.
RSTSIE tries to keep as many jobs in memory as possible. When a job requires
more Inemory than is available, the system may move some jobs out of memory
to temporary storage in one of up to four areas known as swap files. This
operation is called swapping. By convention, these files are named SWAPO.SYS,
SWAPl.SYS, SWAP.SYS, and SWAP3.SYS. When memory becomes available for
a job stored in a swap file, it is swapped back into memory. Jobs waiting for
Overview of RSTS/E System Management

1-11

keyboard input and jobs waiting for devic·e 1/0 completion are most likely stored
in the swap files, while jobs currently ruIllning or involved in disk or magnetic
tape data transfers are stored in memory.
As the system processes each job, it maintains accounting infonnat.ion in memory
for that job. When the job is logged off the system, the information is used to
update the accounting information stored on the system disk for that account.

1.5 Privilege
You must have a way to prevent general access to activities that can damage
the system. The multiple privileges featu.re gives you finite control over access
to activities. You can limit user access to just those activities suitable to each
job. Multiple privileges gives you a tool to enhance both system performance and
security.

1.6 Multiple Privileges
The multiple privileges feature groups similar system functions into sets and
defines a privilege to control access to each set of functions. A group of 35
privileges govern the entire set of RSTSIE system functions. The privileges
given to an account determine the range of functions available to the user. Some
privileges apply to very specific functions, while others control functions within
broader classes of system use (see Chapter 6).

1.7 Classes of System Activities
The following sections divide some computer activities into general classes of
system functions and discuss the privileges that control them. They describe:

1.7.1

•

Account Management Activities

•

File Access Activities

Account Management Activities
One general class of system activity is account management. A user accesses a
computer through an account. The individual accou.nt is a member of a group,
which contains all other accounts with the same project number. The group,
in turn, is a subset of a world, which contains all accounts on the system.
Account management activities include creating and deleting accounts, as well as
changing passwords, disk quotas, and expiration dates. The following privileges
control account management:

1-12

GACNT

Group Account Managemont-Grants account management privileges
within the user's group.

WACNT

World Account Management-Grants account management privileges
for all accounts.

SETPAS

Set Password-Allows changing one's own password.

Overview of RSTS/E System Management

1 . 7.2

File Access Activities
Another general class of system activity is file access. Users routinely access
files. Some files are created by the user and reside in the individual's account.
Other files reside in the accounts of other users or in system accounts. File acces
activities include creating, deXeting, renaming, reading, writing, and executing
files.
Both the protection code of the file and the privileges granted to the user can
affect whether the system grants or denies file access.
In the case where no privileges are granted to a user, protection codes control th4
types of operations that a user can perform on a file. The SET PROTECTION
command (or the !PROTECTION qualifier in the RSTS/E file specification) passel
a value to the system that sets bits in the protection code byte. Each bit specifieE
that the activity named by that bit is prohibited or allowed.
Certain privileges also govern file access activities. Some privileges override
protection codes completely. The following privileges grant a user the right to
perform certain file access activities, regardless of protection codes:
GREAD

Group Read-Read any file within the group. Also, execute any program within the group, if the program's executable bit is set.

WREAD

World Read-Read any file on the system. Also, execute any program
on the system, if the program's executable bit is set.

GWRITE

Group Write-·Create, delete, extend, modify, or rename any file within
the group.

WWRITE

World Write--Create, delete, extend, modify, or rename any file on the
system.

NOTE

In addition to WWRITE, you need SYSIO privilege to create, delete,
extend, modify, or rename any file in group zero ([0,*]).

1.8 Multiple Privilege Masks
You assign a certain set of privileges to each account. The system stores this set
of privileges in privilege masks. A privilege mask is a set of flag bits with one bit
corresponding to each privilege.
For each active job, RSTS/E keeps 3 masks:
•

Authorized mask-The set of privileges that you give to the account. Users
can use the SHOW ACCOUNTIFULL command to list the set of privileges
available to their own account.

•

Current mask-The set of privileges now in effect for the job. The system
always references this mask when it performs a privilege check. Users can
raise or lower their privileges (up to their authorized limit) with the SET
JOB!PRIVILEGE command.

•

Saved mask-The saved record of the current privileges when a job gains
temporary privileges (see the section "Temporary Privileges" in Chapter 6).

Overview of RSTS/E System Management

1-13

When a user attempts to perform an activity that is restricted by one or more
privileges, the system performs a privilege check. It examines the current mask
to determine if the requesting job has all the privileges required to perform the
activity. If the requesting job has insufficient privilege to perform the activity, the
system returns one of the following errors:

1-14

•

?Protection violation

•

?Illegal SYS( ) usage

•

?XXXX privilege required

Overview of RSTS/E System Management

Chapter :

Starting, Halting, and Automatic Recovery 0
RSTS/E

This chapter describes how to start and halt a RSTSIE system and how the
system recovers automatically from catastrophic system errors. This chapter
also describes how to create the startup conditions for a RSTSIE system. See
the RSTS / E System Installation and Update Guide for a description of all other
initialization code options and their functions.

2.1

Starting up RSTS/E
To start a RSTSIE system, you must first load into memory a program containinl

the initialization code (INIT.SYS) and then answer YES to the Start timesharing
<YES> prompt.
INIT.SYS is a large, stand-alone program that, once you transfer it from the
RSTSIE distribution kit, exists on the system disk. INIT.SYS can provide a
number of important system functions, such as:
•

Creating the RSTSIE file structure

•

Creating system files

•

Installing a RSTSIE monitor

•

Setting system defaults

•

Creating startup conditions for RSTSIE

INIT.8YS does some of these automatically; however, they can also be done
through the Option: prompt.
Before you can use the initialization code to start your RSTSIE system, you must
move the code into memory. This is called bootstrapping. The following sections
describe how to transfer the intialization code into memory.

2.1.1

Bootstrapping RSTS/E with a Hardware Bootstrap Loader
The procedures for bootstrapping RSTSIE depend on the type of hardware
bootstrap loader and the type of disk you use as the system device. You should
make sure:
•

The system disk is physically mounted on a disk unit.

•

The drive on which the system disk is mounted has the READY light on and
is in the WRITE ENABLE condition.

Starting, Halting, and Automatic Recovery of RSTS/E

2-1

•

All required units are running and READY.

•

The console terminal is on line.

Read the RSTS / E System Installation and Update Guide for the proper instructions on using the hardware bootstrap on your SYSteln.
Mter you use the bootstrap to load the initialization code, INIT.SYS marks the
successful completion of the bootstrap procedure by printing an identification
line that includes the RSTS version number, the installation name, the disk unit
bootstrapped, and the INIT version number, followed by the Current date prompt.
Enter the date and press Return. INIT di.splays the Current time prompt. Enter
the time and press Return. INIT displays the Start timesharing prompt. For
example:
RSTS V9.7 installation-name (DMO) INIT V9.7
Current date <DD-MMM-YY>? 15-Jan-90<RET>
Current time <HH:MM AM/PM>? 8:15 AM<RET>
Start timesharing? <Yes>

If the program does not print this information, make sure the console terminal
is on line and try the operation again. If the system prints an error message on
the console terminal, see the RSTS / E System Installation and Update Guide for
recovery procedures.
If you want to enable the automatic restart facility, set the CPU switch register
so bit 0 is ON. The automatic facility relnains enabled as long as the CPU switch
register remains set in this way. If your CPU does not have a switch register, the
automatic restart facility is always enabled.

2.1.2

Bootstrapping RSTS/E after a System I-Ialt
When a RSTS/E system halts as a result of a catastrophic error, RSTS/E displays
the halt address in the address lights. F'or central processors that do not
have console lights, for example the 11/2:3+, the 11144, and CPUs with Remote
Diagnostic Consoles (RDCs), the system prints the address on the console
terminal. It is essential that you record this address. The exact procedure for
examining memory locations depends on the type of processor and front panel of
your system. Your Digital Software SpeCJlalist can describe the procedure for your
system. Once you record this information, bootstrap your system again using the
procedures described in the previous section. See the section "Automatic Recovery
and Restart Facilities" for a description of catastrophic errors, system crashes,
and crash dumps.

2.1.3

Starting Timesharing
Mter you bootstrap the system disk, RSTS/E loads the initialization code
(INIT.SYS) into memory. INIT then prints an identification line followed by
the Start timesharing? <Yes> prompt on the console terminal. You can answer in
one of four ways:

2-2

•

Type YES or press Return to start the monitor named in the INIT logo.

•

Type START and press Return to start the monitor named in the INIT logo
and display a list of disabled devices.

Starting. Halting. and Automatic Recovery of RSTS/E

•

Type START monitor-name and press Return to start another monitor in [O,J
and display a list of disabled devices.

•

Type NO to have INIT ret.urn to the Option prompt.
The program is ready to a.ccept one of the options that Table 2-1 summarize!
See the RSTS / E System Installation and Update Guide for a complete
summary of these options .

Table 2-1:
Option

Initialization Option Summary
Abbreviation

Meaning

BOOT dev

Boots another device.

DATE

Resets the current date and time. Also, establishes
default date and time formats.

DEFAULT

Establishes or changes a monitor's startup defaults
such as memory usage.

DSKINT

Initializes and optionally formats a disk. It also
checks for bad blocks.

HARDWR

Specifies device controller characteristics.

HELP

Prints the Help text.

IDENT

Prints INIT.SYS creation date and a list of patches
for INIT.SYS and the installed monitor.

PATCH

Modifies RSTSIE system code to correct program
errors and to add new program features.

REFRESH

Creates and positions system swap files, adds bad
blocks to the bad block file, and. positions optional
system files on the system disk or any auxiliary
disk.

SAVRES

Reconstructs (RESTORE), or makes an image copy
(IMAGE) of a RSTSIE disk.

START sil

ST sil

Installs the specified Save Image Library (SIL),
assigns and checks defaults, and starts the RSTSIE
system. If you do not enter a name for the SIL,
INIT prompts you for one.

2.:2 Halting the RSTS/E System
A system crash or setting the CPU console HALTIENABLE switch to its HALT
position causes a RSTSIE system to halt. For the PDP-11145, 50, 55, and 70,
RSTSIE considers the processor to be in a halt state if both RUN and PAUSE
lights are off. For all other PDP-11 processors, RSTSIE considers the processor to
be in a halt state if the RUN light alone is off.
The SHUTUP program shuts down the RSTSIE system. During this operation,
SHUTUP makes sure all files are properly closed and that system accounting information is accurately updated. If you want to turn the power off, run SHUTUp,
wait for INIT's Start timesharing? prompt, and press the HALTIENABLE switch.
See the section "Performing System Shutdown: SHUTUP" in Chapter 4 for a
complete description of the SHUTUP program.

Starting, Halting, and Automatic Recovery of RSTS/E

2-3

Digital recommends that you do not halt HSTSIE during timesharing by moving
the HALTIENABLE switch on the CPU console to its HALT position. If you
do this during system shutdown, SHUTUP may not have completed cleanup operations; consequently, file data can become corrupted and accounting
information may be lost. The only way to recover fro:m such a halt and to salvage
file information is to do the following:
1.

Raise the HALTIENABLE switch back to its ENABLE position before you
take any other action

Press the CONT switch to return the RSTSIE system to the state it was in
before the halt

2.3 Automatic Recovery and Restart FClcilities
This section describes how catastrophic errors occur and how your system can
automatically recover from them. This section also points out the need to include
commands in the START.COM file to perform critical system operations when the
system restarts in automatic mode.

2.3.1

Catastrophic Errors and System Crashes
Catastrophic errors and system crashes result from unexpected error traps. (For
information on error traps, see the related processor handbook.) The processor
can cause unexpected error traps when it:
•

Refers to a nonexistent or nonresponding memory address (for example, bus
time-out trap)

•

Refers to an odd address with an instruction that requires a word address

•

Attempts to execute a reserved or nonexistent instruction

Also, if the monitor detects a condition which it knows should not normally occur,
the monitor can force a software crash. Catastrophic errors and system crashes
may occur as a result of three types of problems:
•

Privileged job programming errors
RSTSIE system software is designed to protect itself against programming
errors occurring in nonprivileged jobn. When it detects such an error, the
system aborts execution of the user request and returns a corresponding error
message to the user's program.
RSTSIE software is vulnerable to some types of errors caused by jobs with
certain privileges. The privileges tha.t require special caution are SYSIO,
SYSMOD, and WRTNFS. Even though most errors do not cause the system to
crash, you should take special care when programming with privileged system
function calls such as Poke memory SYS call (SYS -6). See the RSTS / E
Programming Manual.

•

Hardware malfunctions
Hardware malfunctions can cause the system to crash. If system crashes or
ca tastrophic errors occur that you cannot explain (particularly on systems
that have traditionally been functioning well), it is likely that a hardware
problem has occurred. You can diagnose hardware problems by examining
the output logged by the programs in the RSTSIE System Error Package. See
Chapter 10 for information on error logging.

2-4

Starting, Halting, and Automatic Recovery of RSTS/E

•

System software malfunctions
Although every attempt has been made to detect and eliminate system
software en'ors, it is possible that RSTSIE software can crash unexpectedly,
If a problem of this type is discovered (it should be reproducible in a defined
environment and under defined conditions), contact a Digital Software
Specialist. As problems of this type become known, Digital reports them in
the "RSTSIE Software Dispatch" and corrects them as soon as possible.

2.3.2

Automatic Recovery from Catastrophic Errors
The system takes action in one of two ways when an unexpected error trap
occurs:
•

A catastrophic error that affects only one user

•

A system crash for which some software or hardware problem may be
responsible

For a catastrophic error, the system determines which user was responsible for
the error. It flags the user's job with a special code; this causes the system to
completely reinitialize that user's job area when it is the user's turn to run. The
system prints an error message on the user's terminal, followed by the text:
??Program lost-sorry. The system resumes normal time-sharing operations.
NOTE

Some run-time systems may issue other error messages.
When the system detects a condition from which it cannot recover, it does the
following:
1. Writes an image of read/write memory and tables to the CRASH.SYS file.
2.

Bootstraps the system disk.

Loads the initialization.code into memory.

Rebuilds the system disk.

Executes an automatic restart. (This occurs if the CPU switch register has
bit 0 set and if at least one minute has passed since startup.)

If the CPU register does not have bit 0 set or if one minute has not passed since
startup, the system does not perform an automatic restart. The system will
only reboot; it will not restart. In other words, the automatic restart operation
restarts timesharing using commands in the START. COM file. In contrast, a
reboot merely loads the initialization code (INIT.SYS) into memory. INIT.SYS
then prints its Start timesharing? <Yes> prompt, to which you can answer by
typing any of the options in Table 2-1.

If the system restarts automatically, it bootstraps itself into memory from the
systeln disk. Mter the system is bootstrapped into memory, control passes to the
initialization routines. The system recognizes startup was through an automatic
restart rather than through normal startup. Consequently, the system initializes
itself in automatic restart mode.
If an error trap occurs within one minute of system startup or if two error traps
occur within the same minute, the system halts at address 54. During automatic
restart the system protects itself against an infinite loop of error traps and
automatic restarts caused by a repeating hardware malfunction.

Starting. Halting. and Automatic Recovery of RSTS/E

2-~

2.3.3

Automatic Restart Mode Initialization
When the system is initialized in automatic restart mode, control bypasses all
parts of the initialization code that call for operator intervention and initializes
the system using parameters specified at previous startup. The system:
1.

Logs job 1 in to the system on KBO: under account [1,2]

2. Invokes [O,I]SYSINI.COM
3. Executes [O,I]START.COM with the PI argument set to 'CRASH' rather than
'START'
4. Performs specific operations appropriate for an automatic restart

2.3.4

Restarting with a Different Monitor
You can, if you like, keep several different monitors on your system. When you
start up timesharing, the system promptB you for your choice of monitor:
Monitor name to start <default name>?

You can also pre-set the choice of monitor so that the new monitor starts working
at the next system shutdown and reboot. Use the SET SYSTEMIMONITOR_NAME
command to do this. For example:
$ SET SYSTEM/MONITOR_NAME=ALTRNT

This command causes the system to use the monitor stored in the file
ALTRNT.SIL at the next automatic restart. ALTRNT.SIL also becomes the
default name for the next time you start up timesharing manually. You need the
SYSIO privilege to use the SET SYSTE:MfIMONITOR_NAME command.
If you have the WREAD privilege and UEle the SHOW SYSTEM command, it
displays the name of the monitor if it is different from the current monitor name.

2.3.5

Restarting with a Different Swap Size
If your system has I&D space, you set the maximum size of jobs when
you install the system. You can change the maximum size with the SET
SYSTEM/SWAP_MAXIMUM command. For example:
$ SET SYSTEM/SWAP_MAXIMUM=40

This command sets the maximum job siz;e at 40K words. You can reset the
maximum size to any value between 32K and 64K. The change in size does not
take effect until the next time you start the system. You need the TUNE privilege
to use the SET SYSTEM/SWAP_MAXIMUM command,
If you have the TUNE privilege and use the SHOW SYSTEM command, it

displays the new value of the swap maxilmwn if it is different from the current
swap maximwn.

2-6

Starting, Halting, and Automatic Recovery of RSTS/E

2.3.6

Power-Fan Hardware Used by RSTS/E
When power fails, a PDP-11 processor traps through the vector at location 24.
When power is restored, you can configure the system to do one of three things:
•

Halt.

•

Boot a specific device.
If the hardware is configured to boot the system disk, INIT.SYS prints its
identification line and the Start timesharing? <Yes> prompt and waits for yOl
to enter one of the options from Table 2-1.

•

Trap again through the vector at location 24.
In this case, the system must have core memory or battery backup. If memo~
contents are still valid (which means the battery backup is not exhausted),
RSTSIE performs an automatic restart.

Starting, Halting, and Automatic Recovery of RSTS/E

2-i

Chapter 3

The System Startup Command Files
This chapter describes how you control system startup with the system startup
command file (START.COM). Startup occurs when you answer YES to the Start
timesharing? <Yes> prompt, use the START option of the initialization code, or
when the system restarts automatically.
At startup, the monitor executes the command file [0,1]SYSINI.COM which
in turn executes [O,1]START.COM. You can modify START.COM (but not
SYSINI.COM) to suit the requirements of your installation.
'Th control startup efficiently, you should understand the conditions at startup
time:
•

•

The monitor logically mounts only the system disk read/write

•

No output is made to any terminal (except the console)

•

The monitor logs the console terminal KBO: into the system under the system
library account [1,2]

•

The monitor assigns ALL privileges to this initial job regardless of the
privileges assigned to accmmt [1,2]

•

The monitor loads the DCL run-time system

•

The monitor starts execution of the command file _SYO:[O,1]SYSINI.COM at
the console

3 ..1 Creating and Using Command Files
The SYSINI.COM command file contains commands necessary to initialize your
RSTSIE system properly. Do not edit this file in any way; this command file is
reserved by DIGITAL and is replaced with every update of the system.
The START.COM file that comes with the RSTSIE distribution kit is a sample
only; it may not execute properly on your system without modification. Make
sure START.COM includes the necessary commands to initialize your installation.
Use a text editor on your system to modify the sample START.COM file supplied
with the RSTSIE kit.
NOTE

Be sure both command files are in account [0,1] on the system disk.

The System Startup Command Files

3-1

The startup control file should perform the following operations:
•

Mount public and private disks

•

Assign system-wide logicals

•

Run the ERRINT program to start en'or logging

•

Enable data caching (if selected at system installation)

•

Set default system characteristics

•

Install run-time systems

•

Install libraries

•

Set terminal characteristics

•

Define system commands

•

Set printer characteristics

•

Start PrintlBatch Services (PBS) package

•

Start OPSER-based spooling package (optional)

•

Enable Local Area Transport (LAT) (optional)

•

Start error analysis if needed

•

Enable logins

•

Logout the initial startup job

The order in which the operations are petformed is critical. For example,
mounting all public disks ensures the integrity of the public structure by making
available all files on the system. You can mount any private disks in a similar
manner. The various steps tailor the system according to your installation's
requirements.

3.2 System Startup Command File: START.COM
A complete system startup command file contains the required commands to
perform initialization for timesharing and to petfonn routine operations. The
following sample shows a startup command file:
$

$
$
$
$
$
$
$
$

START.COM
Copyright (c) 1988 by
Digital Equipment Corporation, Maynard, MA.
P1 - Type of startup (either START or CRASH)

*** Prompt to continue ***

$ if P1 .eqs. "" then P1 = "START"
Default is START
$ S = "system startup"
$ if P1 .eqs. "CRASH" then S = "crash restart"
$ GO = ""
$
inquire/nopunc/exit=END/time out=10 GO "Proceed with "S'? <YES> "
.eq. 1 then _goto START
$ =if f$instr(1,"YES",f$edit(GO:--2»
$END:
$ exit

3-2

The System Startup Command Rles

$S'l~ART:

$ __open/logfi.le/replace 'P1' .LOG
$ set noecho/warning
$ --set verify
$ --write 0 ""
$ =~rite 0 " Beginning RSTS/E ,'S' ... "
$ set noon
$ 1 *** Install Monitor Overlays ***
$

$ write 0 f$time () +"
Installing monitor overlays"
$ --run $UTLMGR. TSK
IlC;:ad/ overlay ATTRIBUTE
Iload/overlay DCL
Iload/overlay DELETE RENAME
Iload/overlay DIRECTORY
lc.ad/overlay SYSTEM CALLS
Ilc.ad/overlay TERMINAL
$ eod
$ 1 *** Install Virtual Disk ***
$ 1
$1 write 0 f$time()+"
Installing Virtual Disk"
$1 =create/vi.rtual 200
$
Create a 200 block virtual disk
$
This will use 50KW of memory
$
*** Mount Disks ***
$
$ s: = ""
$ if P1 _E3lQS. "CRASH" then S = "and rebuilding"
$ --write 0 f$time () +"
Mounting' , S' disks"
$ 1- (If this is a crash recovery, MOUNT will rebuild
$
the disks if necessary.)
$
examples:
$
$ mount./nowrite DR3:LIBRAR L:
$
$ mount DB2:LOOKUP
$
*** Assign (or Replace) Systemwide Logicals ***
$
$ write 0 f$time () +"
Assigning logical names"
$ T
examples:
$
$ assign/system
_SY: [2,9] FERN:
$
$ assign/system/replace _DB2:[0,3] ERROR$:
$

*** Start Error Logging ***

$ write 0 f$time () +"
Starting error logging"
$ -set job/priority=O
$ --if P1 .nes. "CRASH" then _goto CONT1
$ run ERROR$:ERRINT
100
YES
$ EOD
$ =~goto CONT2
$CONT1:
$ run ERROR$:ERRINT
100
NO
$ EOD
$CONT2:
$ set job/priority=-8

The System Startup Command Files

3-3

$
$
$

*** Install Auxiliary Swap Files ***
If you have additional auxiliary swap files
install, add them here.

(SWAPO or SWAP3) to

Note: SWAP1.SYS already installed if present.

examples:
$ write 0 f$time()+"
Installing auxiliary swap files"
$ -install/swapfile=O _SYO:[O,l]SWAPO.SYS
$ =install/swapfile=3 _SYO:[0,l]SWAP3.SYS

*** Enable Caching and Set Default System Characteristics ***

$
$

$ write 0 f$time () +"
Setting system characteristics"
$ -run $UTLMGR.TSK
s;t cache/file/enable/keep=15/clustersize=2
set system/date format=alphabetic
set system/time=format=am-pm
set system/label=dos
set system/density=1600
set system/name="RSTS/E V9.7"
set system/powerfail_delay=300
set system/hangup=delayed
set system/password-prompt=(network,dialup)
$ eod
$
$
$
$

*** Install Run-time Systems and Libraries ***
Note: Resident library CSPLIB already installed.

Installing run-time systems and libraries"
$ write 0 f$time () + "
$ -run $UTLMGR.TSK
i~stall/runtime system BASIC
install/runtime-system TECO$:TECO
install/library7noaddress=unrestricted EDT$:EDT
install/library/address RMS$:RMSRES
install/library/noaddress RMS$:RMSLBA
install/library/noaddress RMS$:RMSLBB
install/library/noaddress RMS$:RMSLBC
install/library/noaddress RMS$:RMSLBD
install/library/noaddress RMS$:RMSLBE
install/library/noaddress RMS$:RMSLBF
install/library/address RMS$:DAPRES
$ EOD
$ , *** Set Terminal Characteristics ***
$
$ write 0 f$time () +"
Setting terminal characteristics"
$ -load/overlay TERMINAL
$ -run $TERMGR.TSK
s;t terminal kbO:/permanent/device_type=LA34
, System console as hard copy
set terminal kbdO:/permanent/device_type=LA100
, Pseudo-keyboard 0 for batch
'set terminal kbfO:/permanent/noautobaud/speed=9600/device type=VT100
, DHO line 0 as VT100 at 9600 baud
'set terminal kbgO: /permanent/noautobaud/ speed=9600/ device._type=VT100
, DZO line 0 as VT100 at 9600 baud
'set terminal kbg8: /permanent/noautobaud/ speed=9600/device._type=VT240
, DZ1 line 0 as VT240 at 9600 baud
$ EOD
$ =unload/overlay TERMINAL

3-4

The System Startup Command Files

$ ! *** Defi.ne System Commands ***
$

$ write 0 f$time () +"
Defining system commands"
$ T Install system commands (CCLs)
$ run $UTLMGR.TSK
d;fine/command/system BCKRMS$:RMSBCK.TSK
!define/command/system BPC-REF
AUXLIB$:BPCREF.TSK /line number=CCL
define/command/system BYE$LOGOUT.TSK
/privIlege
define/command/system CNVRMS$:RMSCNV.TSK
define/command/system DI-RECTORY $DIRECT.TSK
/line_number=CCL/privilege
define/command/system DEFRMS$:RMSDEF.TSK
define/command/system DESRMS$:RMSDES.TSK
define/command/system DSPRMS$:RMSDSP.TSK
define/command/system EDTEDT$:EDT.TSK
!define/command/system FITAUXLIB$:FIT.TSK
/line_number=CCL/privilege
define/command/system IFLRMS$:RMSIFL.TSK
define/command/system LBR$LBR.TSK
define/command/system LIBR$LIBR.SAV
/line number=8208
/line-number=8208
define/command/system LIN-K
$LINK.SAV
define/command/system MACR-O
$MACRO.SAV
/line=number=84 16
define/command/system MAC$MAC.TSK
TECO$:TECO.TEC
!define/command/system MAK-E
/line number=3584
!define/command/system MU-NG
TECO$:TECO.TEC
/line=number=3584
define/command/system PAT$PAT.'I'SK
$PIP.SAV
/line number=8208/privilege
d'efine/command/system PIPd,efine/ command/ system RSTRMS$:R.MSRST.TSK
/privIlege
d,efine/ command/ system SW-ITCH
/line number=CCL/privilege
$SWITCH.TSK
/line-number=CCL/privilege
$SYSTA,T. TSK
d1efine/ command/ system SY-STAT
TECO$:TECO.TEC
! d1efine/command/system TE-CO
/line=number=3584
$TKB.TSK
dl9fine/ command/ system TKB$ EOD
$ ! *** Set Printer Characteristics ***
$

Setting printer characteristics"
$ __write 0 f$time () + "
examples:
$
$ set printer LPO:/device type=LP11
$
$ -set printer LP1:/device-type=LA180
$
$ =set printer LP2:/device=type=LNOl
$
$

*** Start Spoolers ***

$
$

write 0 f$time()+"

Starting spoolers"

T Start PBS spooling package

$ __start/ queue/manager
$
** OPTIONAL** Start OPSER (old) spooling package
$
$

$
$

goto CONT3

! If OPSER is desired, delete this line.

T Install related system commands

$ define/conwand/system PL-EASE OPSER$:PLEASE.TSK
$ =~define/conwand/system QU-EUE
OPSER$:QUE.TSK
$ run OPSER$:OPSER
LOC; OP SER$ : OP SER. LOG; ALL
CHJ!~ KBO:
MESS ALL
OPE:R KB *: [1, 2 ]
DE'l?ACH
$ run OPSER$:QUEMAN
DE~;ACH/PRI: O/RUN: 6
$ set job/priority=-8
$ --run OPSER$: SPOOL
LPci: /HEAD: 2/RUN: 9/WIDTH: 80
$ set job/priority=-8
$ --run OP SER$ : SPOOL
LPi: /HEAD: l/RUN: 9/WIDTH: 132
$ set job/priority=-8
$ --run OPSER$ : BATCH
BAO::/ERROR:FATAL/NOQUEUE/NODELETE

/privilege/line number=CCL
/privilege/line=number=CCL

The System Startup Command Files

3-5

$ set job/priority=-8
$ -run OPSER$:BATCH
BAl:/ERROR:FATAL/NOQUEUE/NODELETE
$ EOD
$CONT3:
$ ! *** Start ANALYS if in Crash Recovery ***
$

$ if Pl .nes. "CRASH" then _goto CONT4
$ -write 0 f$time()+"
Starting analysis"
$ =run ERROR$:ANALYS
[O,l]CRASH.SYS
ERROR$ : ANAL YS . DMP
ERROR$:ERRCRS.FIL/DET
$ EOD
$CONT4:
$ ! *** Starting DECnet/E V4.0 ***
$

$ ! wri te 0 f$time () +"
Starting DECnet/E V4. 0 1"
$ !- First define the DECnet related CCL command::1
$! define/command/system NCP DECNET$:NCP.TSK/priv/line=ccl
$! -define/command/system NFT DECNET$:NFT.TSK/priv/line=ccl
$! -define/command/system NET DECNET$:NET.TSK/priv/line=ccl
$! -define/command/system TLK DECNET$:TLK.TSK/priv/line=ccl
$ !- Now start the network
$! ncp set system
$! ncp set exec state on
$! ncp set cir una-O state on
$ ! *** Start up LAT ***
$ !
$! write 0 f$time () +"
Starting LAT"
$! -SET NODE/LAT/ENABLE=O
$! -SET NODE/LAT/IDENTIFICATION="RSTS/E V9.7 LAT"
$! -CREATE/SERVICE/LAT RSTS/IDENTIFICATION="Unaut.horized use prohibited"
$! =START/LAT XEO:
$ ! *** Enable Logins ***
$ !
$ set system/logins
$ -broadcast/all "RSTS/E is on the air ... "
$ =close/logfile
$ _set noverify
$ set echo
$ _logout/br
$ exit

N()TE
The startup control files must be in [0,1] on the system disk (_SYO:, the
disk that was bootstrapped).

3.2.1

Mount All Disks
To ensure the integrity of the public structure by making available all files on

the system, you should mount all public disks. You can mount any private disks
in a similar manner. If this is a restart from a system crash, MOUNT rebuilds
the disks as required. See Chapter 15 for more information on mounting and
initialization of disks.

3-6

The System Startup Command Files

3.2.2

Assign Systemwide Logical Names
You can assign a systemwide logical name to a specified device type, unit numbe
and PPN (optionally) with the ASSIGN/SYSTEM command. See Chapter 12 for
more information on the use of systemwide logical names.

3,,2.3

Start Error Logging
It is important that you run the ERRINT program on your system. This prograll

sets initial conditions for system error logging and runs the ERRCPY program
which transfers to a disk file all errors logged by the RSTSIE monitor. If ERRINI
is the first program to be run (as DIGITAL recommends), it will occupy job
number 1 on the system and thus be easily monitored. Chapter 10 describes
ERRINT and ERRCPY.
NOTE

If your system has more than 256K words of memory, you should

consider adding /LOCK to the first line after the RUN ERRINT
command lines to keep ERRINT resident at all times.

3,,2.4

Enable Caching (Optional)
The SET CACHEIFILEIENABLE command enables caching of data files whose
UFD entry or OPEN MODE specifies caching. The !KEEP qualifier lets you spec
ify the time that an unused cache cluster is kept in memory. The /CLUSTERSIZl
qualifier controls the number of contiguous data blocks that are copied from the
disk to the cache whenever file data is cached. For more information on data ane
directory caching, see the section "SET CACHE Command" in Chapter 11.

3.2.5

Set !Default System Characteristics
The SET SYSTEM command sets the default system characteristics. In the
sample START.COM, for exanlple:
•

The date format is alphabetic

•

The time format is AM_PM

•

The magnetic tape label is DOS

•

The magnetic tape density is 1600 bpi

•

The system name is EDERE

•

The power fail delay is 300 seconds

•

NETWORK and DIALUP users must enter the system password

See Chapter 14 for more information on setting system default characteristics.

3.2.6

Install Run-Time Systems
The INSTALLIRUNTIME_SYSTEM command installs auxiliary run-time systems (like BASIC) in memory. See the section "Run-Time System Control" in
Chapter 12 for more information.

The System Startup Command Files

3-'

3.2.7

Install Libraries
The INSTALLILIBRARY command installs resident libraries (like RMSRES) in
memory. See Chapter 12 for more information on resident library control.

3.2.8

Set Terminal Characteristics
At startup time, the system sets the characteristics of all keyboard lines (except
line number 0) to hard copy, 80 column output at the line speed of 300 baud.
Thus, you need to use the SET TERMINAL command to set the correct characteristics of your terminals. Chapter 8 contains detailed descriptions of the SET
TERMINAL command qualifiers.

3.2.9

Define System Commands
It is most convenient to define Concise Command Language (CCL) sys-

tem commands on the system at the start of timesharing using the
DEFINE/COMMAND/SYSTEM command. The abbreviations shown in the
definitions of the CCL commands reflect the way the RSTSIE documentation
describes the commands. As the system lmanager you can redefine the abbreviations. You should not alter the fully defined command because each program is
coded to recognize the fully expanded cODlmand.
Note that the abbreviation point should be the same as the one used by DCL
itself. For most DCL commands, the abbreviation point follows the second
character, but for some, the abbreviation point follows the third or fourth
character, to assure a unique abbreviation for each command.
Use caution when you define system commands in this way. You may make
features of other keyboard monitors unavailable. This may happen if, for
example, you define PRINT as a CCL cOlnmand. BASIC-PLUS users will then
not be able to use the immediate mode PRINT statement.
Two CCL commands duplicate commands the BASIC-PLUS keyboard monitor
recognizes: HELLO and BYE. The commands are shown here to point out what
BASIC-PLUS does by default. Other keyboard monitors may not recognize the
commands HELLO and BYE if they do not exist as eCL commands.
Each CCL definition requires one small buffer. The buffer will be allocated from
the FIP small buffer pool. If this buffer pool is full, however, the monitor allocates
the buffer for the CCL definition from the general buffer pool. See Chapter 12 for
more information on the DEFINE/COM]\IIAND/SYSTEM command.

3.2.10

Set Printer Characteristics
Just as you did for terminals, you need to set the characteristics of your printers.
Minimally, you need to specify the device type of each printer on your system.
See Chapter 11 for more information about the qualifiers for the SET PRINTER
commahd.

3-8

The System Startup Command Files

3.2.11

Start Print/Batch Services
The PrintlBatch Services package (PBS) provides printing services for line
print.ers or terminals, and batch processing. The package is started with a singll
command. Chapter 9 contains detailed descriptions of the DCL commands you
can use to manage printlbatch services.

3.2.12

Start OPSER-Based Spooling Package (Optional)
The OPSER-based spooling package, like PBS, provides print and batch services
RSTSIE offers this optional package for compatibility with previous versions of
RSTSIE. New applications should use PBS. For more information on OPSER, seE
Appendix A of this manual and the RSTS/E Utilities Reference Manual.

3,,2.13

Enable LAT (Optional)
A new sample startup file is installed in Versions 9.6 and later. If you already
have a [O,I]START.COM file, then the sample uses the version number as its
extension, for example [0,I]START.096. If you do not have a [O,I]START.COM
file, then the sample is installed on your system with that name. The sample hal
commands in it to enable LAT, but these are in the form of comments. If you wan
to enable LAT automatically on system startup, edit the [O,I]START.COM file anI
remove the exclamation points (!) from the LAT commands. The commands are:
$ ! *** Start up LAT ***
$ !
$! write 0 f$time()+"
Starting LAT"
$! -SET NODE/LAT/ENABLE=O
$! -SET NODE/LAT/IDENTIFICATION="RSTS/E V9.7 LAT"
$! -CREATE/SERVICE/LAT RSTS/IDENTIFICATION="Unauthorized use prohibite<
$! =START/LAT XEO:

3.2.14

Start Error Analysis
Run the ANALYS program immediately after a system crash to recover valuable
diagnostic information. See Chapter 10 for a description of the ANALYS program

3.2.15

Enable Logins
The SET SYSTEMILOGINS command should appear after the commands that
perform all required and routine operations. The SET SYSTEMILOGINS command lets users log in to the system. You can then notify users that timesharing
has started. See Chapter 13 for more information on job management.

3.:2.16

Logout the Initial Startup Job
You should always logout the initial startup job, since it runs with ALL privileges.
Failing to do so can create a security risk if the system console is left unattended.

The System Startup Command Files

3-9

Chapter 4

Shutting Down the RSTS/E System

4.1

Performing System Shutdown: SHUTUP
The shutdown procedures for the RSTS/E system are critically important. If
you are not careful in performing system shutdown, valuable user data can
be irretrievably lost. To understand shutdown procedures fully, you need to
understand other RSTS/E system procedures. Chapter 12 describes these
concepts such as controlling logins, removing system files and resident libraries,
sending messages to all logged-in users, and forcing commands to unresponsive
terminals.

4:1.1

Running SHUTUP
You use the SHUTUP program to shut the system down in an orderly manner.
SHUTUP is in the system library account $ ([1,2]). SHUTUP must be installed
as a privileged program (protection code <232». If not, the system displays the
following error message:
?Program must be privileged

To run SHUTUP, you must have SHUTUP pnvilege. If you do not, the system
displays the following error message:
?SHUTUP privilege required

SHUTUP can be run from any local terminal over a dial up line or over a LAT
line. SHUTUP cannot be run from a pseudo keyboard (for example, by a SET
HOST connection).

To verify that conditions are suitable for shutting down the RSTS/E system, use
any of the SHOW commands, such as:
•

SHOW JOBS

•

SHOW DISKS

•

SHOW DEVICES/ALLOCATED

By executing these SHOW commands, you can determine what jobs are active on
the system and what disks and other peripheral devices are in use.

Shutting Down the RSTS/E System

4-1

4.2 SHUTUP Phases
As SHUTUP runs, it proceeds through the following 14 phases:
1. Set-up dialogue
2.

Warning message

DECnetiE shutdown (if necessary)

Initial job killing

5. PrintlBatch Services (PBS) shutdown (if necessaJry)
6.

OPSER (optional spooling package) shutdown (if necessary)

EVTLOG shutdown (if necessary)

8. ERRCPY shutdown (if necessary)
9.

Final job killing (if necessary)

10. EMT logging shutdown (if necessary)
11. Unload and remove run-time systems and resident libraries
12. Log file closing
13. Swap file removal
14. Disk dismount
15. Final shutdown
The following sections describe these phases, the operator interactions required,
and the expected results of the shutdown operation. Possible results, which you
may not expect are also described. Several sample runs of SHUTUP show the
results of shutting down the system.

4.2.1

Set-Up Dialogue Phase
The set-up dialogue phase processes all questions relating to the selection of
options in running SHUTUP. It also checks for the presence of PBS or OPSER.
In response to questions that appear dUl:ing this phase, you can press the
circwnflex (A) key to return to the previous question. Otherwise, you should
termina te your responses by pressing the Return key.
If the PrintlBatch Services (PBS) package is present, SHUTUP prints the
following message:
Allow Print/Batch entries to

complet~e

<YES>?

If you type NO, SHUTUP aborts all PBS entries that are in progress. If you press
Return or type YES, SHUTUP lets all currently running PBS entries finish.

If OPSER is present and running, SHUTUP prints the following question:
Use OPSER for utilities shutdown <YES>?

4-2

Shutting Down the RSTS/E System

SHUTUP operates in either of two modes:
•

Without OPSER: In this mode, regardless of whether OPSER is running, the
SHUTUP program treats all jobs in the system alike during the shutdown
procedure (except itself, the PrintlBatch Services program, the DECnetlE
EVTLOG program, the EM:T logger program, and ERRCPY).

•

With OPSER: In this mode SHUTUP allows OPSER to shut down its controlled jobs before SHUTUP continues with the regular shutdown procedures.

You should be familiar with the OPSER program and how it controls various
utility programs (such as QUMRUN and the OPSER spooling package programs).
Press Return or type YES to have SHUTUP communicate with OPSER during
the shutdown procedures. If you type NO, SHUTUP prints another question to
confinn your response:
Are you sure you don't want to use OPSER <NO>?

Pressing Return or typing NO causes the program to repeat the previous question
so you can again decide to use or not to use OPSER during shut down. Otherwise,
the program ignores OPSER during the shutdown procedures.
If you choose to use the OPSER program, SHUTUP prints the following additional
question:
Allow utilities to reach logical end point <YES>?

If you press Return or type YES, the program directs OPSER to tell its online
jobs to shut down their operations at the next logical breakpoint in their job
streams. For a line printer spooler, for example, this breakpoint occurs between
queued jobs.
If you type NO, the program directs OPSER to tell its online jobs to abort and
stop operations immediately. You should abort in this way only in an emergency:
jobs being processed at the time of shutdown may be lost unless you have
previously halted queuing and spooling operations and requeued these jobs.
SHUTUP then asks you to specify the number of minutes you want to elapse
before shutdown occurs:
Minutes until system shutdown (0-99) <5>?

This waiting period is called the "warning message phase" and may be as short
as 0 minutes or as long as 99 minutes. Press Return to select the default value
of 5 minutes. If you decide to hring the system down immediately by typing 0,
SHUTUP does not ask the two questions about disabling logins and the DECnet/E
network.
When you specify a shutdown period greater than 0 minutes, SHUTUP allows you
to select the amount of time that will elapse before further logins are disabled:
Minutes until logins are disabled (0-99) <O>?

This question gives you the opportunity to select the amount of time during which
users can continue to log in to the system as shutdown occurs. Users can log in to
the system (as if the system were not shutting down) for the number of minutes
you specify. The ability to allow users this flexibility is useful when you specify a
long shutdown period.
There will be times when you do not want to allow any more users on the system.
In this case, accept the default response by pressing Return. Then, users can no
longer log in to the system; however, those already logged in remain unaffected.

Shutting Down the RSTS/E System

4-3

NOTE

SHUTUP disables alllogins at least one minute before the initial job
killing phase.
The last message in the set-up dialogue phase allows you to choose the number of
minutes that will elapse before new network activity is disabled:
Minutes until new network activity is disabled (0-99) <nn>?

The number of minutes SHUTUP selects for the default value <nn> depends on
your response to the system shutdown question. As in the previous question,
being able to specify when the network will shut down allows users time to
complete network tasks. If you choose to allow no further logical links to be
created, type 0, and then press Return. Those still using the network can
continue to work but no new users are allowed access to the network.
Shutdown with automatic RESTART <NO>'?

If you type YES, you will automatically restart the system after the shutdown.
The system does not restart from the ternlinal you are using. INIT.SYS uses the
console terminal (KBO:), no matter what terminal you use for SHUTUP. You must
determine whether INIT.SYS failed to start by looking at the console terminal.
Use automatic restart to change the monitor name, the swap maximum, LAT
support, or the maximum number of pseudo keyboards.

4.2.2

Warning Message Phase
At the beginning of the warning message phase, as well as in the beginning of
all other phases that SHUTUP may execute, a message appears on the system
console terminal (KBO:) in the following format:
hh:mm AM dd-mmm-yy
PM

########

########

SHUTUP prints:
•

The time that it issues the message in the format hh:mm {AM I PM}

•

The date that it issues the message in the format dd-mmm-yy

•

The title of the current phase

SHUTUP prints the following message on the system console terminal when the
time specified for disabling logins has elapsed:
Further LOGINs are now disabled

The RSTSIE shutdown procedure requires that logins are disabled before the
monitor can perform the final shutdown operation. During the following phases,
SHUTUP continuously checks that logins remain disabled. If, for any reason,
logins are not disabled, SHUTUP immediately aborts all operations with the
following message:
?Logins not disabled
??SHUTUP aborting - please try again later

You should determine the cause of the change, correct the problem, and run
SHUTUP again.
If your system has DECnetiE support, th(~ SHUTUP program prints the following
message when the time specified for disabling network activity expires:
Further network activity is now disabled

4-4

Shutting Down the RSTS/E System

The message tells you that no new logical links can be formed. Those links
that are already active, however, can remain active until the time to network
shutdown has elapsed. Before the network shuts down, the DECnetlE program
EVTLOG may print messages indicating the status of the network. The initial
EVTLOG message marks the beginning of the network shutdown phase. The
second message, issued later in the shutdown process, ~ndicates that the network
has been shut off completely. See the section "SHUTUP Example: DECnetIE
Shutdown" for an example of these messages.
During the warning message phase, the program issues warning messages indicating the system is shutting down. These messages appear on most keyboards
in the system, including pseudo keyboards. The following keyboards are the
exceptions:
OSC - The OPSER operator services console
KBn:- The terminal running shutup
The program issues warnings after an appropriate waiting period, defined by the
following formula:
waiting time = (total time left)/5 + 1 minutes
For an initial waiting period of 60 minutes, warning messages appear at 60,
47,37) 29, 23, 18, 14, 11, 8, 6, 4, 3, 2, and 1 minute before shutdown begins.
SHUTUP prints the following lnessage on all system terminals at each warning
time:
hh:mm AM dd-mmm-yy

System going down in n minutes, please finish up

At the same time, the following message appears on the terminal running
SHUTUP:
n minute message sent

When the time to shutdown has expired or if the initial waiting period is specified
as 0, the final warning message appears:
hh:mm AM dd-mmm-yy

**** FI~AL WARNING! ! ! !! System shutting down ****

At this point, the program enters the next phase. SHUTUP prints the time and
date for both the SYSTEM GOING DOWN and the FINAL WARNING messages.
4.~!.3

DECnet/E Shutdown Phase (When Necessary)
The tilme to DECnetIE shutdown begins as soon as you enter the number of
minutes that should elapse before new network activity is disabled:
Minutes until new network activity is disabled (0-99) <nn>?

The number of minutes, shown in angle brackets <nn>, is the default response.
It assmnes the same value you typed in response to the SHUTUP question:
Minutes until system shutdown (0-99) <nn>?

Rather than accept the default response (that is, to the network activity question)
and have new network activity disabled at the same time the system shuts
down, it is often better to disable new network activity before shutdown occurs.
Entering a value less than the number of minutes to shutdown establishes a
period before shutdown in which no new links can be created; it also allows time
for those already active to complete network tasks. If you want new network
activity disabled immediately, type 0 and press Return. When the time to disable

Shutting Down the RSTS/E System

4-5

network activity has elapsed (0 to 99 minutes), SHUTUP prints a message to
indicate no further network activity is allowed:
Further network activity disabled

SHUTUP disables new network activity by executing the equivalent of the
Network Control Program (NCP) command SET EXECUTOR STATE SHUT.
This prevents the creation of new logical links but does not affect existing logical
links. In reality, a user can create new network links until the network activity is
disabled. Mter that period ends, no new links can be created, but users already
on the network can continue working until the system shuts down. The network
remains up until there are no links active or until the system shuts down.
As soon as all logical links have been disconnected by their users, the network
state changes automatically to OFF. The state is immediately set to OFF if
no links existed at the time SHUTUP disabled new network activity. If the
DECnetIE event logging program EVTLOG is on and has event 2.0 enabled, it
logs both the transition to SHUT and the one to OFF. (See the DECnet / E System
Manager's Guide for a description of EVTLOG.)
SHUTUP never sends special messages indicating the approach of a DECnetlE
shutdown. Users can tell that SHUTUP is shutting down the system only from
SHUTUP messages to all terminals and from the fact that the network is in the
SHUT instead of the ON state.
If the network has still not completely shut down at the end of the warning
message phase (because active links remain), SHUTUP explicitly sets the
network to OFF. This action disconnects all remaining logical links. SHUTUP
then waits 15 seconds to allow affected jobs to complete network tasks and exit,
before it enters the job killing phase.
If you decide not to shut the system down (thus, not shut off the network) after
you have started SHUTUP, entering a CtrllC to stop the execution of SHUTUP
may reactivate the network. A CtrllC causes SHUTUP to print a message telling
you the state in which the network was left. If the network is not in the state you
want, you must use NCP to place the network and the event logger (EVTLOG) in
the desired state. See the DECnet/ E System Manager's Guide for a description of
NCP.

4.2.4

Initial Job Killing Phase
In the initial job killing phase, SHUTUP begins to clear the system of currently
active jobs. If SHUTUP finds there is only one job (itselO running on the system,
it skips this phase and then continues the shutdown procedure at the unload and
remove run-time systems and resident libraries phase.
If SHUTUP is using PBS in the next phase, during this phase the program
ignores all jobs currently associated with PBS (as indicated by entries in PBS's
on-line job table). The program also ignores any job whose primary keyboard is
a pseudo keyboard. Such a job is probably being controlled by one of the batch
servers.
SHUTUP further divides all other jobs active in the system into two classes:
attached and detached. This classification depends on whether they have an
attached primary keyboard (KB:). For all attached jobs, SHUTUP forces the
following text string to the keyboard input buffer:
AC

$_LOGOUT/BRIEF

For all detached jobs, SHUTUP issues the Kill job SYS call (SYS 8) to remove the
job.
4-6

Shutting Down the RSTS/E System

SHUTUP makes two passes through the current active job table during this
phase. In the first pass, the program terminates all active jobs not being ignored
At the end of the first pass, it establishes the number of attached jobs that were
found and forced to logout. SHUTUP then waits for a specified period to allow
the LOGOUT program to complete operations on each of the attached jobs.
Mter the waiting period (if any) expires, the program makes a second pass
through the active job table. During this pass, all jobs not being ignored are
removed with the kill job SYS call. At the end of the pass, the program checks
the table. If any jobs that should have been killed were not, SHUTUP aborts
operations with the following error message:
?SHUTUP failed in initial job killing phase
??SHUTUP aborting - please try again later

You should determine why SHUTUP could not kill all jobs, correct the problem,
and run SHUTUP again.
If SHUTUP successfully removes all jobs during the phase, or if at any time
during either pass the number of active jobs becomes one (SHUTUP only), the
next phase begins.

4.:2.5

Print/Batch Shutdown Phase (When Necessary)
The program skips the PrintlBatch Services (PBS) shutdown phase if PBS is not
running on the system. If PBS is not running, SHUTUP continues with the next
phase; no message is displayed.
If PBS is running, SHUTUP sends a message to QMAN telling QMAN to
shutdown PBS. QMAN acknowledges the shutdown command by sending back
a confirmation message. If SHUTUP does not receive the confirmation message
within two minutes after the offline request was sent, SHUTUP aborts the
shutdown operation and displays the following:
?No response from Print/Batch Services after 2 minutes

Otherwise, SHUTUP displays the number of entries currently processing in PBS
in the following format:
Print/Batch Services shutting down -- # jobs are completing.
Print/Batch Services shutting down -- # jobs are being terminated.

Then SHUTUP waits for PBSMGR to shutdown PBS. The maximum wait period
is one hour if jobs are allowed to complete and one minute if jobs are aborted. If
shutdown occurs in the specified time, SHUTUP displays the following message
and goes to the next phase:
Print/Batch Services shutdown complete at hh:mm am/pm

Otherwise, SHUTUP aborts and displays the following error message:
?Print/Batch Services shutdown taking too long
??SHUTUP aborting - please try again later

4.:~.6

OPSER Shutdown Phase (When Necessary)
The program skips the OPSER shutdown phase if:
•

OPSER is not running on the system

•

You elect not to use OPSER during the shutdown procedures

Shutting Down the RSTS/E System

4-7

If you choose to use OPSER, SHUTUP in1mediately detaches from the terminal
running SHUTUP. OPSER reattaches to KBO: for use during OPSER's shutdown
procedures. As the detach occurs, the following messages appear:
DETACHING ...
'OPSER' ATTACHING
-if:

SHUTUP sends a message to OPSER directing it to begin its shutdown procedures by selecting the appropriate mode of OPSER shutdown. SHUTUP then
waits 60 seconds to see if OPSER receives the message andlor is functioning properly. If OPSER does not respond properly within that time, SHUTUP attempts
to reattach to the terminal running SHUTUP. If the operation is successful, the
following message appears:
?OPSER not active
??SHUTUP aborting - please try again later

If the terminal is attached or assigned, SHUTUP waits for one second and tries
again. You should take whatever steps are necessary to free the terminal running
SHUTUP for use so SHUTUP can complete its activities.
If OPSER is shutting down its online programs in the immediate mode, it sends
each program the appropriate message to stop operations. OPSER pauses to
make sure all programs have completed their assignments. OPSER then closes
its files, prints the following message on the system console (KBO:), and stops
execution:
'OPSER' TERMINATING

SHUTUP allows OPSER 120 seconds to complete its shutdown in the immediate
shutdown mode. If unsuccessful, SHUTUP aborts its operations, signaling the
failure as follows:
?OPSER shutdown taking too long
??SHUTUP aborting - please try again later

If OPSER is shutting down in the logica.l end mode, it proceeds to shutdown
successive levels in its online job table. Appropriate system console messages
signal the end of each job (see "SHUTUP Operation Examples"). When all online
OPSER jobs are gone, OPSER closes its files and kills itself, freeing the system
console for SHUTUP use again. OPSER tells you when it finishes processing with
the message:
'OPSER' TERMINATING

SHUTUP allows 60 minutes in the logical end mode for OPSER to complete
operations. If the procedure is not successful, SHUTUP aborts operations with
the error message described previously. Otherwise, the shutdown of OPSERrelated jobs and activities is complete and the program enters the next phase.

4.2.7

EVTLOG Shutdown Phase (When Necessary)
The SHUTUP program enters the EVTLOG shutdown phase if it finds EVTLOG
in the message/receiver table. (DECnetJE does not need to be active for SHUTUP
to perform this phase.) SHUTUP sends EVTLOG a special message that causes
EVTLOG to log the shutdown occurrence on the system console, perform various
other shutdown operations, and kill itself.

4-8

Shutting Down the RSTS/E System

When SHUTUP detects that EVTLOG has completed its shutdown activities and
is no longer present, SHUTUP proceeds to the next phase. SHUTUP allows 60
seconds for EVTLOG to complete its tasks, and if failure occurs, SHUTUP abortE
operations with the following error message:
?EVTLOG failed to shutdown
??SHUTUP aborting - please try again later

See the DECnet / E System Manager's Guide for information about the EVTLOG
program.

4.2.8

ERRCPY Shutdown Phase (When Necessary)
The program enters the ERRCPY shutdown phase if the system error logging
utility ERRCPY is present. SHUTUP sends ERRCPY a special message which
causes ERRCPY to log the shutdown occurrence in the system error log, close thE
file, and kill itself.
When SHUTUP detects that ERRCPY has completed its shutdown and is no
longer present, SHUTUP proceeds to the next phase. SHUTUP allows 60 second:
for ERRCPY to complete its tasks. If failure occurs, SHUTUP aborts operations
with the following error message:
?ERRCPY failed to shutdown
??SHUTUP aborting - please try again later

4.2.9

Final Job Killing Phase (When Necessary)
If SHUTUP reaches this point, and is not the only job left running in the system
apart from the EMT logger, SHUTUP makes one last attempt at killing all
other remaining jobs. If SHU'rUp is successful, it enters the next phase. If not
successful, SHUTUP aborts operations with the following error message:
?SHUTUP failed in final job killing phase
??SHUTUP aborting - please try again later

4.2.10

EMT Logging Shutdown Phase (When Necessary)
The program enters the EMT logging shutdown phase if the EMT logging utility
is present. (See Chapter 14 for a description of the EMT logger.) SHUTUP sends
a shutdown message to the EMT logger, informing the EMT logger to stop. If th€
EMT logger does not shut down in 60 seconds, SHUTUP aborts the EMT logger
with the following warning message:
%EMT logger failed to shutdown - killing EMT logger job = n

In this message, n is the job number of the EMT logger.

4.2.11

Remove Run-Time Systems and Resident Libraries Phase
During the remove run-time system and resident libraries phase, the program
removes all run-time systems and resident libraries from the system tables.
SHUTUP does not attempt to remove the primary run-time system.

Shutting Down the RSTS/E System

4-~

4.2.12

Log File Closing Phase
During the log file closing phase, SHUTUP closes the log file, if it is open, and
logs out. It then logs in again and goes on to the next phase.

4.2.13

Swap File Removal Phase
During the swap file removal phase, SHUTUP makes sure all installed swap files
are removed.

4.2.14

Disk Dismount Phase
During the dismount phase, the program dismounts all mounted disks, both
private and public, from the system. The: system disk always remains mounted.
If you have opened a file using the DeL commands OPEN or OPENILOG, or if
you are executing SHUTUP from an indirect command file, and the file resides
on a disk other than the system disk, an unexpected error occurs when SHUTUP
tries to dismount the affected disk. SHUTUP aborts and you must close the file
before retrying the shutdown.

4.2.15

Final Shutdown Phase
SHUTUP enters the final shutdown phaBe to output the last status message,
clears the terminal buffers, and executes the special System shutdown SYS
call (SYS -16). The monitor transfers control to the system initialization code
(INIT.SYS) and bootstraps the system disk. The monitor loads the initialization
code and INIT prints the Start Timesharing <Yes>? prompt (unless you are using
automatic restart, in which case the system restarts).

4.2.16

SHUTUP Operation Examples
Each of the following sections contains an example of a SHUTUP operation.
These examples and the description of the SHUTUP program should help you to
shut down your system correctly and in orderly fashion.

4.2.16.1

SHUTUP Example: Large RSTS/E System

The system in the following example is being shut down in a leisurely fashion. This gives system and network users enough time to complete their tasks.
SHUTUP allows users to log in to the system during the entire shutdown phase;
logins are disabled at system shutdown, after 5 minutes. Those using the network
can log in until the 5-minute mark, after that time the system does not allow network access by users. However, those on the network before new network activity
is disabled can continue to work until the system shuts down completely.
$ run shutup

SHUTUP V9.7 RSTS V9.7 EDERE
10:11 AM 28-May-89
Type

4- to

fAf

########

Set-up Dialogue Phase

to any query to backup one question

Shutting Down the RSTS/E System

########

Allow Print/Batch Services jobs to complete <YES>? yes
Use OPSER for utilities shutdown <YES>? yes
Allow OPSER utilities to reach logical end point <YES>? yes
Minutes until system shutdown (0-99) <5>? 15
Minutes until logins are disabled (0-99) <O>? 5
Minutes until new network activity is disabled (0-99) <5>? 5
Shutdown with automatic RESTART <NO>? NO
10:12 AM 28-May-89 ########
Warning Message Phase
15 minute warning message sent
11 minute warning message sent
Further LOGINs are not disabled
Further network activity is now disabled
8 minute warning message sent
6 minute warning message sent
4 minute warning message sent
3 minute warning message sent
2 minute warning message sent
1 minute warning message sent

########

10:27 AM 28-May-89

########

DECNET Shutdown Phase

########

10:27 AM 28-May-89

########

Initial Job Killing Phase

########

10:27 AM 28-May-89 ########
Print/Batch Shutdown Phase
Print/Batch Services shutdown complete at 10:27 AM

########

10:27 AM 28-May-89
Detaching ...

########

OPSER Shutdown Phase

########

, OPSER' ATTACHING
#
JOB #6 ' BA1SPL' TAKEN OFFLINE
#
JOB #7 'BA2SPL' TAKEN OFFLINE
#
JOB #5 ' BAOSPL' TAKEN OFFLINE
#
JOB #4 'LPOSPL' TAKEN OFFLINE
#
JOB #3 , QUEMAN' TAKEN OFFLINE
#'OPSgR' TERMINATING
Re-att.aching ...
10:29 AM 28-May-89

########

EVTLOG Shutdown Phase

########

10:29 AM 28-May-89

######41'#

ERRCPY Shutdown Phase

########

10:29 AM 28-May-89

########

Remove RTS/RES LIB Phase

########

10:29 AM 28-May-89

########

Log File Closing Phase

########

10:29 AM 28-May-89

########

SWAP File Removal Phase

########

10:29 AM 28-May-89

########

Disk DISMOUNT Phase

########

10:29 AM 28-May-89

########

Final Shutdown Phase

########

Please wait for system to re-boot itself
RSTS V9.7 EDERE (DR1) INIT V9.7
28-May-89 10:30 AM
Start timesharing <Yes>?

Shutting Down the RSTS/E System

4-11

4.2.16.2

SHUTUP Example: Small RSTS/E System

The following example shows a system shutdown when OPSER is not present.
Except for the absence of the first OPSEH-related question J the example applies
when OPSER is present but not used during shutdown.
$ run shutup
SHUTUP V9.7 RSTS V9.7 EDERE

12:24 PM 22-May-89
Type

fAf

########

set-up Dialogue Phase

to any query to backup one question

Allow Print/Batch Services jobs to complete <YES>? yes
Minutes until system shutdown (0-99) <5>?
5
Minutes until logins are disabled (0-99) <O>? 1
Shutdown with automatic RESTART <NO>? NO
12:24 PM 22-May-89 ########
5 minute warning message sent
Further LOGINs are now disabled
3 minute warning message sent
2 minute warning message sent
1 minute warning message sent

Warning Message Phase

########

Initial Job Killing Phase

########

12:25 PM 22-May-89 ########
Print/Batch Shutdown Phase
Print/Batch Services shutdown complElte at 12: 25 PM

########

12:24 PM 22-May-89

########

12:25 PM 22-May-89

########

EV'I'LOG Shutdown Phase

#######:fI:

12:25 PM 22-May-89

########

ERRCPY Shutdown Phase

########

12:25 PM 22-May-89

########

Remove RTS/RES LIB Phase

########

12:25 PM 22-May-89

########

Log File Closing Phase

########

12:25 PM 22-May-89

########

SWAP File Removal Phase

########

12:25 PM 22-May-89

########

Disk DISMOUNT Phase

########

12:25 PM 22-May-89

########

Final Shutdown Phase

########

Please wait for system to re-boot itself
RSTS V9.7 EDERE (DR1) INIT V9.7
22-May-89 - 12:24 PM
Start timesharing <Yes>?

4.2.16.3

SHUTUP Example: OECnet/E Shutdown

This example shows SHUTUP when OPSER is used to shut down a spooling
system of five spoolers. It also illustraten how the network shuts down when a
system includes DECnetIE support and event logging by the DECnetIE EVTLOG
program has been enabled.
$ run shutup
SHUTUP V9.7 RSTS V9.7 EDERE

12:39 PM 22-May-89
Type

fAf

########

Set-up Dialogue Phase

########

to any query to backup one question

Allow Print/Batch Services jobs to complete <YES>? yes
Use fOPSER f for utilities shutdown <YES>? yes
Allow OPSER utilities to reach logical end point <YES>? yes
Minutes until system shutdown (0-99) <5>1 5
Minutes until logins are disabled (0-99) <O>? 0
Minutes until new network activity is disabled (0-99) <5>? 2
Shutdown with automatic RESTART <NO>? NO

4-12

Shutting Down the RSTS/E System

12:40 PM 22-May-89 ********
Further LOGINs are now disabled
5 minute warning message sent

Warning Message Phase

********

Further network activity is now disabled
3 minute warning message sent
Event type 2.0, Local node state change
Occurred 22-May-89 12:42:05.0 on node 135 (OTHG)
Reason for state change: Operator command
Old node state
On
New node state = Off
2 minute warning message sent
1 minute warning message sent
12:45 PM 22-May-89

********

DECNET Shutdown Phase

********

12:45 PM 22-May-89

********

Initial Job Killing Phase

********

12:45 PM 22-May-89 ********
Print/Batch Shutdown Phase
Print/Batch Services shutdown complete at 12:45 PM

********

12:45 PM 22-May-89
Detaching ...

********

OPSER Shutdown Phase

*****:fI:**

'OPSER' ATTACHING

JOB *6 ' BA1SPL' TAKEN OFFLINE
*
JOB *7 'BA2SPL' TAKEN OFFLINE
*
JOB *5 'BAOSPL' TAKEN OFFLINE
*
MESSAGE
949 : 22-May-89 12:45 PM JOB:3 DET QUMRUN[1,2]
BA1SPL (6) REQUESTED OFF-LINE -- TAKEN OFF-LINE
*
MESSAGE
950 : 22-May-89 12:45 PM JOB:3 DET QUMRUN[1,2]
BA2SPL (7) REQUESTED OFF-LINE -- TAKEN OFF-LINE
*
MESSAGE
951 : 22-May-89 12:45 PM JOB:3 DET QUMRUN[1,2]
BAOSPL (5) REQUESTED OFF-LINE -- TAKEN OFF-LINE
*
JOB *4 ' LPOSPL' TAKEN OFFLINE
*
MESSAGE
952 : 22-May-89 12:45 PM JOB:3 DET QUMRUN[1,2]
LPOSPL (4) REQUESTED OFF-LINE -- TAKEN OFF-LINE
*
JOB *3 , QUEMAN' TAKEN OFFLINE
*'OPSER' TERMINATING
Re-at taching ...
12:45 PM 22-May-89 ********
EVTLOG Shutdown Phase
Shutting down EVTLOG by operator request.

********

12:45 PM 22-May-89
12:45 PM 22-May-89
12:45 PM 22-May-89
12:45 AM 22-May-89
12:45 PM 22-May-89
12:45 PM 22-May-89
12:45 PM 22-May-89

********

ERRCPY Shutdown Phase

********

Final Job Killing Phase

********

Remove RTS/RES LIB Phase

********

Log File Closing Phase

********

SWAP File Removal Phase

********

Disk DISMOUNT Phase

********

Final Shutdown Phase

********

Shutting Down the RSTS/E System

4-13

Please wait for system to re-boot itself
RSTS V9.7 EDERE (DR1)

INIT V9.7

22-May-89 - 12:45 PM
Start timesharing <Yes>?

4.2.17

Notes on SHUTUP Operation
You can help the users on your system by establishing administrative procedures
that govern the hours the system is in operation. If you have a fixed shutdown
schedule, users can plan their work load and complete tasks in the allotted hours
of timesharing.
You can keep them informed by placing messages in the system NOTICE.TXT
file. The system displays the contents of NOTICE.TXT each time a user logs in
using a comma (,) in the PPN. This is a convenient way to make sure all users
know about important system operations.

4-14

Shutting Down the RSTS/E System

Chapter f

Managing Quotas

This chapter describes the quotas that limit the use of various system resources.
You can assign quotas to a user when you create an account. RSTSIE quotas fall
into four categories:
•

Disk quotas

•

Job quotas

•

SendlReceive quotas

•

System dynamic memory limits

The remainder of this chapter describes these quotas in detail. See Chapter 7 fOI
information on assigning or changing an account's quotas.

5.1

Disk Quotas
You can limit the amount of space available to individual users on public or
private disks by creating and maintaining quotas on those disks. Quotas are
maintained and enforced on a per-disk basis, except for the public structure
where it is maintained across the structure.
You identify PPNs and assign quotas to them. The system automatically
maintains usage counts during nonnal file activities.
Each account that exists on a Level 1.2 disk, except for the [0,1] account, has a
disk quota attribute block associated with it. The quota blocks have the followin!
information:
•

U sage--Actual number of blocks on the disk taken up by your files

•

Quotas-Maximum numbE~r of blocks on the disk that your files can take up
before the system issues an error message
NOTE

For disk levels 1.1 and earlier, RSTSIE used only 16 bits to store the
disk quotas. Thus, quotas could not be as large as the size of the
largest disk supported by RSTSIE.

Managing Quotas

5-1

Disk Quota Operations

5.1.1

For pre-V9.0 disk structures, RSTSIE could only enforce disk quota limits at the
time a user logged out. Thus, as a SysteDl manager, you could limit the amount
of permanent disk space a user could have, but you could not limit the amount of
temporary disk space.
On level 1.2 disks, RSTSIE maintains two types of disk related quotas: a loggedin quota and a logged-out quota. If at least one user is logged in to the file's
account, the system automatically checks the usage count against the logged-in
quota as users create, delete, and extend files. If no one is logged in to the file's
account, then RSTSIE checks the usage count against the logged-out quota. The
system also enforces the logged-out quota when a user logs out.
If extending the file causes the total disk usage to exceed the quota, RSTSIE
rejects the operation and returns a ?Quota exceeded error message. When
checking for a possible quota violation, RSTSIE takes into consideration the total
write request. So, a write request that extends a file by multiple disk blocks is
never only partially completed.

To perform the quota checks, the monitor keeps a record of the number of blocks
that it allocates to each account on each disk. For public disks, this record is kept
on a per disk basis, but, when quota checks are made, the counts are added up
across the public structure. The monitor keeps the usage information in memory
and writes it back to the disk when appropriate. If the system crashes (leaving
old information on a disk), the disk rebuild operation (CLEAN) recomputes the
usage and writes the correct data on the disk.

5.1.2

Logged-In Disk Quotas
Logged-in disk quotas are the number of disk blocks that the system lets a user
allocate while logged in. The number of blocks in use is the total number of
blocks allocated. The allocation of any file is its actual size in blocks, rounded up
to the nearest full cluster. The allowable values are 0 to 16,777,214 (2"24-2),
and UNLIMITED. Zero means that no disk storage is allowed. UNLIMITED is a
special value that means the monitor does not enforce the logged-in quota.
Logged-in quotas do not apply on:

5.1.3

•

Level 0 or Level 1.1 disk structures

•

Disks mounted using MOUNTINOQUOTA (although the usage is still kept up
to date)

•

Disks mounted using MOUNTINOWIUTE (because no file extends can take
place on such a disk)

•

DVO:, which is always mounted with INOQUOTA

Logged-Out Disk Quotas
Logged-out disk quotas are the number of disk blocks that the system lets a user
retain at logout time. The logged-out quota is also checked when a user tries to
create or extend a file in a logged-out account.
The allocation of any file is its actual size in blocks, rounded up to the nearest
full cluster. The number of blocks in use by any user is the total number of blocks
allocated for all of the user's files.

5-2

Managing Quotas

The allowable values are 0 to 16,777,214 (2"24-2), or UNLIMITED. Zero means
that no disk storage is allowed to remain allocated at logout time. UNLIMITED
is a special value that means the monitor does not enforce the logged-out quota.
When a job logs out, the monitor checks the logged-out disk quota of each disk
against the usage of that disk. For the public structure, the system disk quota is
checked against the total usage on the public structure. If the quota is exceeded
(and no other interactive jobs are logged in under the same account) then the
monitor rejects the logout request.
For level 0 and level 1.1 disk structures, the logged-out quotas remain the
same as pre-V9.0. The allowable values are 0 to 65,535, where zero means
UNLIMITED.

5,,2 Job Quotas
For pre-V9.0 disk structures, there was no restriction on the number of attached
jobs that a user was allowed t.o run. Thus, a user who had access to several
terminals could gain control of a large number of job slots. In V9.0, each account
has a job quota.
The job quota specifies the total number of jobs (attached or detached) that each
account can have active at any time. If the user attempts to exceed this quota,
the system rejects the request and returns the ?Quota exceeded error message.
If the user attempts to log in when the quota is reached, the system rejects the
log-in request (see "Job Quotas and Login" below).
The default job quota is UNLIMITED.

5.2.1

Detached-Job Quota
The detached job quota is an account attribute. It indicates the actual number
of jobs that are allowed to run detached under each account. If detaching a job
causes this quota to be exceeded, then the system rejects the request and displaYI
the following message:
?Quota exceeded

The quota check applies even if the user issuing the request is running under
some other account. The detached job quota does not affect the automatic
detaching operation that takes place when carrier is lost on a dial-up line
controlling a job. In that case the job is unconditionally detached, but the quota
is enforced when the user attempts to log back on (see the section "Job Quotas
and Login").
The detached job quota defaults to zero when you create an account.

5.:2.2

Job Quotas and Login
When a user attempts to log in, the system checks the current list of jobs under
the account being logged in to against the quotas for that account. If either the
detached job quota is exceeded or the job quota would be exceeded, the system
rejects the log-in request. The system returns the list of currently detached jobs
in the usual manner.

Managing Quotas

5-3

If a user attempts to log in to some account and the job quota has been reached,
the result depends on whether there currently are jobs running detached under
that account. If there are none, the system rejects the log-in request. If there are
detached jobs, the system lets the user attach to one of them.
If a user attempts to log in to some account and the detached job quota for that
account has been exceeded, the system rejects the request but the user can attach
to one of the detached jobs. Thus, the monitor ensures that one user cannot
create an unlimited number of detached jobs.

5.3 Send/Receive Quotas
You can also limit the nwnber of receiver ID blocks (RIBs) and the number of
messages that a job can have at any time.

5.3.1

RIB Quota
Specifies the maximwn nwnber of RIBs that a job running under an account
can set up for itself as a message receiver. The allowable range is 0 to 256 or
UNLIMITED. UNLIMITED is a special keyword that means no RIB quota is to
be enforced.

5.3.2

Message Quota
Specifies the maximwn total message quota that a job running under an account
can request when declaring itself as a receiver. This maximum total must be
shared by however many RIBs (local and network) the job wants to set up. The
allowable range is 0 to 65,535 or UNLIMITED. UNLIMITED is a special keyword
that means no message quota is to be enforced.

5.4 Dynamic Region Limits
Dynamic regions are areas in memory cr'eated by some application programs;
Fortran 77 virtual arrays are examples of dynamic regions. You must have the
INSTAL privilege to allocate dynamic regions. If users or applications do not
have INSTAL, they can still call on a limited pool of publicly available space. The
size of this pool is the dynamic region limi.t.
You can adjust the dynamic region limit with the DCL command SET
SYSTEMIDYNAMI C_REGI ON_LIMIT. For example:
$ SET SYSTEM!DYNAMIC_REGION_LIMIT=20

This command sets the dynamic region limit at 20K words. As long as this
command remains in effect, no user or application can take more than 20K words
of dynamic region without the INSTAL privilege. There may be less than 20K
words available at any given time, since some of the pool may be assigned to
another user or application.
You need the SWCFG privilege to use the SET SYSTEMIDYNAMIC_REGION_
LIMIT command.

5-4

Managing Quotas

Chapter 6

Managing Privileges
This chapter describes the privileges that you can assign to a user when creating
an account. Privileges restrict the performance of certain system activities to
certain users. These restrictions protect the integrity of the operating system's
performance and thus the integrity of service provided to users. You should grant
privileges to each user on the basis of two factors:
•

Whether the user has a legitimate need for the privilege

•

Whether the user has the skill and experience to use the privilege without
disrupting the system

There are two types of privileges: RSTSIE-defined and user-defined.

6.1

RSTS/E-Oefined Privileges
RSTSIE-defined privileges fall into several categories. The categories are based
on the level of impact the user has on the system:
•

User-Minimum privileges to use the system effectively

•

Group----Potential to affect users inside the same group

•

World-Potential to affect users outside the same group

•

System-Potential to affect normal system operation

•

File-Potential to read or write file data

•

All-Potential to control the system

Users can use the DeL command SET JOBIPRIVlLEGE to enable and disable
privileges for which they are authorized. Table 6-1 lists the privileges by categor)
and gives a brief definition.

Managing Privileges

6-1

The sections that follow Table 6-1 describe each privilege in alphabetical order.
For a function-by-function description of privilege requirements, see the RSTS / E
Programming Manual, the RSTS / E System Directives Manual, or the RSTS / E
System User's Guide.
Table 6-1:

RSTS/E Privileges

Category

Privilege

Activity Permitted

User

none

Most normal activity except those listed below

SETPAS

Change own password

Group

World

Files

GACNT

Manauement of all accounts in group

GREAD

Read access to all files in group

GWRITE

Write access to all files in group

EXQTA

Exceed quotas

RDMEM

Read physical memory

WAC NT

Management of all accounts

WREAD

Read access to all files

WWRITE

Write access to all files

DATES

Change file last access date

DEVICE

AccesB restricted devices

GREAD

Read access to all files in group

GWRITE

Write access to all files in group

INSTAL

Add/r~~move/list system files, run-time systems,

libraries

System

All

6-2

Managing Privileges

SYSIO

Set privilege bit (128) of a data file

WREAD

Read .access to all files

WWRITE

Write access to all files

DATES

Chan~~e system clock

HWCTL
JOBCTL
SWCTL

PerfoIID system hardware control functions
PerfOIID job control functions
Control logins and networking

HWCFG
SWCFG

Set dE~vice characteristics
Set system parameters

INSTAL

Add/rl~move/list ceL commands

MOUNT

Mount and dismount devices

PBSCTL

Control Print/Batch services

SEND

Send to a restricted receiver, BROADCAST
messages

SHUTUP

Shut down the system

TUNE

Alter priority, run burst, and swapping mode

RDNFS

Perform read-only non-file-structured 110

SYSIO

Write files in [0,*] (with WWRITE)

SYSMOD

Modify the system

TMPPRV

Set privilege bit of an executable program

WRTNFS

Perform read/write non-file-structured 110

6.1.1

DATES Prfivilege
The DATES privilege lets you change the system date or time. You also need thif
privilege to change a file's date of last access or modification.

6.1.2

DEVICE Privilege
The DEVICE privilege lets you allocate, read, or write data on a restricted device

6.1.3

EXQTA Privilege
The exceed quota privilege (EXQTA) lets the space taken by the user's files on a
given volume exceed any quota set for the user (as determined by the PPN) on
that volume. However, this privilege does not let users exceed their logged-out
quota when logging off the system.
This privilege also lets users specify a priority, page limit, CPU limit, or time
limit value with a PRINT or SUBMIT command that exceeds the maximum valu4
for the specified queue.
The primary use of EXQTA is by privileged programs that need to perform certaiI
operations regardless of user quotas. Normally, you would not assign EXQTA to
users. Instead, you should set the user's quotas to get the effect you want.

6.1.4

GACNT Privilege
The group accounting privilege (GACNT) lets a user perform account management functions, such as create/delete account and change passwords/quotas for
all accounts with the same project number (group) as the user. GACNT also lets
a user create batch jobs in any account in the user's group.

6.1.5

GREAD Privilege
The group read privilege (GREAD) allows the user's job read and execute access
to any file whose project number is the same as the user's.

6.1.6

GWRITE Privilege
The group write privilege (GWRITE) gives the user's job create, delete, rename,
and write access to all files that have the same project number as the user's.

6.1.7

HWCFG Privilege
The hardware configuration privilege (HWCFG) lets you set terminal characteris·
tics for any terminal, declare a device as restricted, create and delete the virtual
disk, and set line printer parameters.

Managing Privileges

6-:

6.1.8

HWCTL Privilege
The hardware control privilege (HWCTL) lets you perform system control
functions, for example seizing control of a device, or hanging up a modern line.

6.1.9

INSTAL Privilege
The install privilege (INSTAL) lets you add, remove, or list system files. You can
also add or remove run-time systems, resident libraries, system logical names, or
system commands (CCLs), and create dynamic regions above the dynamic region
limit.

6.1.10

JOBCTL Privilege
The job control privilege (JOBCTL) lets you read the status of any other job, kill
a job, or remove the receiver ID block (RIB) of another job. You can also create a
job even if logins are disabled.

6.1.11

MOUNT Privilege
The MOUNT privilege lets you mount shared disks, or use certain restricted
MOUNT command qualifiers.

6.1.12

PBSCTL Privilege
The PBSCTL privilege lets you control the PrintlBatch Services. For example,
you can start or stop servers, or change printer forms.

6.1.13

RDMEM Privilege
The RDMEM privilege lets you read mernory (PEEK) and perform read-only
mapping of physical memory by executing the .PLAS system directive.

6.1.14

RDNFS Privilege
The RDNFS privilege lets the user's job perform read-only non-file-structured I/O
operations.
NOTE

This privilege lets a job access data anywhere on the disk without the
benefit of any file structure. Grant:it only to users who need it. If this
privilege is given to unqualified users who do not need it, the operating
system and service to other users can be disrupted. Such disruptions
can include the exposure of confidential information to unauthorized
persons.

6-4

Managing Privileges

6. "1.15

SEND Privilege
The SEND privilege lets you send a message to a restricted receiver. You also
need this privilege to BROADCAST a message to a terminal.

6. "1.16

SETPAS Privilege
The SETPAS privilege lets users change their own password. This privilege
enables the SET PASSWORD command.

6.11.17

SHUlUP privilege
The SHUTUP privilege lets you run the SHUTUP program to shut down the
system.

6.11.18

SWCFG Privilege
The SWCFG privilege lets you perform system configuration functions such as:

6.11.19

•

Setting magnetic tape label defaults

•

Setting date/time presentation formats

•

Initializing queues and servers

•

Managing DECnet

SWCll Privilege
The software control privilege (SWCTL) lets you enable or disable logins. You
also need this privilege to perform operational network management functions
(for example, turning the network or individual circuits ON and OFF).

6.11.20

SVSIO Privilege
The SYSIO privilege lets the user's job:
•

Create, modify, rename, or delete files in [0,*]

•

Issue commands that write into files in [0,*] such as the DUMP/SYSTEM
command

•

Declare a network receiver that accepts inbound network connection requests
(a network server program)

•

Declare a receiver with a name that does not end with your job number

•

Set the privilege bit in the protection code of a data file (execute bit (64) not
set)

•

Force commands to a terminal

Managing Privileges

6-5

6.1.21

SYSMOD Privilege
The SYSMOD privilege lets you perform functions that could modify the system:
•

Write memory with POKE

•

Map physical memory read/write

•

Ma p the I/O page

•

Perform non-file-structured reads and writes of a mounted disk
NOTE

This privilege lets a job access data anywhere on the disk without the
benefit of any file structure. Grant it only to users who need it. If this
privilege is given to unqualified users who do not need it, the operating
system and service to other users can be disrupted. Such disruptions
can include the destruction of inforrnation on the system device, the
destruction of user data, and the exposure of confidential information
to unauthorized persons.

6.1.22

TMPPRV Privilege
The TMPPRV privilege lets you set the privilege bit (128) in the protection code
of a file with the execute bit (64) set. This privilege is not assigned to privileged
programs when they are executed. Thus programs with temporary privilege
cannot mark other programs as being privileged; a user must have TMPPRV
privilege to perform this function.
NOTE

Grant the TMMPRV privilege with care because users with it can
write and execute programs that perform system functions that are
controlled by privileges the user does not have.

6.1.23

TUNE Privilege
The TUNE privilege lets you:
•

Enable or disable disk caching

•

Enable or disable file caching

•

Lock your job in memory

•

Set the maximum size any job can occupy in physical memory (SWAP MAX)

•

Set the maximum time any job can run when compute bound (run burst)

•

Set the priority of any job

•

Change a batch server's priority and run burst

You can increase or decrease the priority of a job by executing the UU .PRI system
directive or the DCL command SET JOBIPRIORITY. With the same privilege, a
job can spawn another job with a priority other than its own. You can create such
a job by using an optional argument to the system directive UU.JOB.
The TUNE privilege also lets you control whether a job can be swapped out of
memory by executing the .SET system directive with the JFLOCK bit set.

6-6

Managing Privileges

Because of their significant effect on system performance, the BACKUP and
RESTORE commands also require TUNE privilege.
NOTE

Do not grant the TUNE privilege widely; if unqualified users have
lmrestricted ability to set job priorities, the fair and orderly scheduling
of jobs for execution can be disrupted.
Grant the TUNE privilege only to users who need to lock a job in
memory for performance reasons. ('IYpically, this will be a real-time
application.) If unqualified users have the unrestricted ability to lock
jobs in memory, physical memory may be held unnecessarily, thus
degrading system perfornlance.

6. "1.24

WACNT Privilege
The world accounting privilege (WACNT) lets you perform account management
functions, such as create/delete account and change password/quotas, for all
accounts. You can also use it to spawn jobs or run batch jobs in any account on
the system.
To affect or to examine other accounts inside its own project number, a job needs

only the GACNT privilege. But to affect or examine accounts outside its own
project number, a job needs the WACNT privilege.

6.11.25

WREAD Privilege
The world read privilege (WREAD) gives you read and execute access to any file.

6.1.26

WRTNFS Privilege
The WRTNFS privilege lets you perform read/write non-file-structured 110
operations.
NOTE

This privilege lets a job access data anywhere on the disk without the
benefit of any file structure. Grant it only to users who need it. If this
p:rivilege is given to unqualified users who do not need it, the operating
system and service to other users can be disrupted. Such disruptions
can include the destruction of information on the system device, the
destruction of user data, and the exposure of confidential information
to unauthorized persons.

6.1.27

WWRITE Privilege
The world write privilege (WWRITE) gives you create, delete, rename, and write
access to all files except those in [0,*]. For files in [0,*], you need to have SYSIO
privilege as well as WWRITE privilege.

Managing Privileges

6-7

6.2 User-Defined Privileges
The user-defined privileges (USERl-8) are eight privileges that have been
reserved for your own definition. Third-party and user software can test them at
the system program level to control various aspects of their application programs,
such as restricted functions or restricted access to certain files. RSTSIE does not
use these privileges for any access or privilege checks.

6.3 Multiple Privileges and Jobs
The following sections describe how the Jnonitor handles privilege information
during the life of a job. They describe:

•
•
•
•
•
6.3.1

Privilege Masks
Job Creation
Login
Logout
Spawned Jobs

Privilege Masks
A user's privileges are recorded in the job structure of the user's job in three
privilege masks: authorized, current, and saved. When a user logs in to the
system, the user's authorized and current privilege masks are initialized frorn
the job's account attributes. The saved privilege mask is a copy of the current
privilege mask upon entry to a privileged program. When temporary privileges
are dropped, the current privilege mask is restored from the saved privilege
mask.

6.3.2 Job Creation
At job creation, the monitor initializes both the current mask and the authorized
mask, giving them all privileges except SYSMOD.

6.3.3

Login
When a job logs in, the system looks up the authorized mask in the account
attributes. It copies this mask into the EJaved and authorized masks, ORs it
into the current mask, and sets the job s,tatus to indicate the job has temporary
privileges in effect.
If a program logs in, it now has all the privileges it originally had plus possibly
some new ones. When the program exits, the user has all authorized privileges
enabled; temporary privileges are dropped.
A user who logs in may not want all authorized privileges to be active at login. In
that case the user can use a LOGIN.COM file to initially drop some privileges.

6-8

Managing Privileges

6.:3.4

Logout
When a job logs out, the monitor clears the group-related privileges GACNT,
GREAD, and GWRITE in all three privilege masks. This is done because the job
is currently running with a project-programmer number (PPN) of [0,0] effectively
putting it in group zero. The monitor drops group privileges because the intent oj
these privileges is to allow access to the user's group, not group zero.
Apart from losing group privileges, a job neither gains nor loses any privileges as
a result of logging out.

6.:3.5

Spawned jobs
You can spawn a job using the UU.JOB system directive (see the RSTS / E
System Directives Manual). A spawned job can be either logged-in or loggedout. For jobs spawned logged-in, the monitor usually gives the spawned job
the set of authorized privileges for the account it logs in to. This is done before
the program, if any, is run. If the program is a privileged program, the usual
additional privilege processing takes place (see the section "Running a Privileged
Program").
Jobs spawned logged-out are given the privileges of the job executing the UU.JOE
directive.
Both logged-out and logged-in jobs spawned to an account other than the caller's
require accounting privilege (GACNTIWACNT). Spawning therefore lets users
with accounting privilege create jobs that have some other account's privileges,
possibly more than their own.

6.4 Running Programs Under Multiple Privileges
The following sections explain how the monitor handles privileges when running
programs. They describe:

6.4.1

•

Running System Programs

•

Temporary Privileges

•

Designating a Program as Privileged

•

Running a Privileged Program

•

Program Exit

Running System Programs
System programs provide tools for performing common tasks, such as managing
files and developing programs. Some programs are available to all users and
require no particular privileges to run. They have a protection code <104>. EDT
is an example of a nonprivileged system program.
Other system programs perform functions for privileged users. Most privileged
system programs have a protection code <104>, unless world file access is
required to successfully execute the program. In that case the protection code
is <124>. The difference between the nonprivileged program and the privileged
program that have the same protection code is that the latter verifies that a user
has the required privileges before proceeding.

Managing Privileges

6-~

Some privileged system programs like SHUTUP have a protection code <232>
but require the user to have a particular privilege to gain access. In the case
of SHUTUP, the user needs the SHUTUP privilege to run the program. Access
to privileged system programs depends on the user's set of privileges. When a
user attempts to run a privileged prograln, the program first performs a privilege
check to determine if the user is sufficiently privileged. If the user has the correct
set of privileges, the program grants access. See the individual description of
each system program to determine what privileges are required to gain access.

6.4.2

Temporary Privileges
A program that performs privileged functions for a nonprivileged user receives
the special designation of privileged program. When run, it gives the user
temporary privileges, that is, all privileges except SYSMOD and TMPPRV.
When the program exits, the monitor withdraws the extra privileges from the
user. Temporary privileges give the user access to privileged programs without
compromising system security.
Sometimes programs need to perform privileged functions for nonprivileged users.
For example, SHUTUP needs to PEEK at memory, an activity that requires
RDMEM privilege. Because it is a program, SHUTUP can PEEK in a controlled
manner. Therefore, it is not a security risk. On the other hand, unrestricted use
of the RDMEM privilege by a user can compromise system security.

6.4.3

Designating a Program as Privileged
A user designates an executable program as privileged by setting the privilege bit
(128) and the executable bit (64) in the protection code. Therefore, any program
with a protection code of <192> or higher is privileged. The normal protection
code associated with privileged executable programs is <232>, granting execute
access to all, but restricting read/write access to the owner.
For security purposes, there are two restrictions regarding privileged programs:

6.4.4

•

A user needs the TMPPRV privilege to be able to designate a program as
privileged.

•

A privileged program cannot reside on a disk that is mounted /NOSHARE.
This restriction prevents an outsider from acquiring privileges by bringing in
a privileged program on a private pack. To be able to mount a disk ISHARE,
a user needs the MOUNT privilege. Thus, it is important that users who
have the ability to mount disks ISHARE know the contents of the disk they
are mounting.

Running a Privileged Program
When a user runs a privileged program, the monitor copies the account's current privilege mask into the saved privilege mask, then it ORs all privileges
except SYSMOD and TMPPRV into the current privilege mask, giving the user
temporary privileges (all but SYSMOD and TMPPRV).
It is good programming practice to have privileged programs drop privileges on

entry, and raise them only when needed. This practice provides better protection
against programming errors damaging the system. It thereby improves system
security.

6-10

Managing Privileges

6.4.5

Program exit
Whenever a program exits or chains to another program, the monitor performs
the following privilege-related cleanup:
•

If temporary privileges are in effect, the monitor copies the saved mask bacl~
into the current mask. This cancels the temporary privileges.

•

The monitor cancels any third-party privilege check currently in effect (see
the RSTS / E Programming Manual, SYS call 31, or the RSTS / E System
Directives Manual, general directive UU.3PP).

•

If the job is currently logged out and does not have WACNT privilege, and

the program exits, then the monitor kills the job. Passing control from one
program to another program is possible without restriction when a job is
logged out, but other operations that exit the current program result in a
self-kill.
•

If the program being exited is a privileged program, then the monitor clears

the job's memory and sets the job size to the minimum size for the job's
default keyboard monitor.

Managing Privileges

Chapter

Managing Accounts

This chapter gives background infonnation on managing RSTS/E accounts and
describes the DCL commands you use to create and maintain accounts. Table 7-:
summarizes these commands.

7.1

Table 7-1:

DCl Commands for Managing Accounts

Command

Description

CREATE/ACCOUNT

Sets up a new account on a file-stru.ctured disk

DELETE/ACCOUNT

Deletes an account from a file-stru.ctured disk

SET ACCOUNT

Changes an account's attributes

SET PASSWORD

Changes an account's password

SHOW ACCOUNT

Displays account attributes

Accounts
An account is the structure by which RSTS/E recognizes a timesharing user, and
in which RSTSIE maintains usage and access data. Each account is identified by
a PPN, and can exist on any disk-public, or private. However, you can only log
in to accounts that are defined on the booted system disk.
The public disk structure can be one or more disks and collectively these disk(s)
are represented by the logical name, SY:. Within the public structure, the booted
disk is known as SYO:. You can explicitly create, delete, or modify accounts on
SYO: or any private disk, but not on other public disks. The system automatically
extends accounts on public disks as needed. The only command that lets you
specify a public disk in the device specification is SHOW ACCOUNT.
RSTS/E has two types of accounts:
•

User-Contains files and login attributes

•

Nonuser-Only contains files

The system decides which type account to create by the following rules:
•

Explicitly by qualifier: !USER or INOUSER

•

Implicitly by the presence of any qualifier that suggests a user account such
as lIN_QUOTA, !PRIVILEGES, ![NO]LOOKUP, or ![NO]NETWORK

•

Neither of the above: !USER is assumed if the disk is SYO:, INOUSER is
assumed if the disk is private

Managing Accounts

7-1

7.1.1

Project-Programmer Number (PPN)
RSTSIE identifies each account by a project-programmer number (PPN). You can
specify the project or programmer number as a single value or a range of values.
You assign the PPN when you create the account. The PPN cannot be changed.
Valid PPNs are: [0,1-254] and [1-254,0-25~U.
Note that [0,1-254] are legal PPNs for nonuser accounts only; they are not legal
PPNs for user accounts. That is, no job can log in under [0,*]. Account [0,1]
is created when the disk is initialized, and cannot be deleted. Also, [0,1-199]
are reserved for DIGITAL software, [0,200] is reserved for RSTS/E and layered
product installations, and [0,201-254] are reserved for user software packages.

7.1.2

Account Access Privileges
There are three categories of access to account infonnation. A different set of
privilege flags control each category. The eategories are:
•

Read information-Read accounting data and account attributes
None-You can read and display accounting information in the current
account
GACNT-You can read and display accounting information of any account
in your group
WACNT-You can read and display accounting information of any account

•

Set information-Write accounting data and attributes to, or create or delete
accounts
GACNT-You can write accounting data of any account in your group
WACNT-You can write accounting data of any account on the system

•

Set password-Write (change) the passwords
SETPAS-You can change your own password (using the SET
PASSWORD command only)
GACNT-You can change the password of any account in your group
WACNT-You can change the pasElword of any account on the system

The privileges are inclusive: that is, if you have sufficient privilege to perfonn a
given operation on any account in your group, you can also perform that operation
on your own account. Similarly, if you have sufficient privilege to perform a given
operation on any account on the system, you can perform that operation on any
account in your group (including your own account).

7.2 Account Data
The following sections describe the data am account has to manage timesharing
access and usage. Unless the section notes otherwise, each account element is:

7-2

•

Initialized by CREATE/ACCOUNT

•

Changed by SET ACCOUNT

•

Displayed by SHOW ACCOUNT

•

Deleted by DELETE/ACCOUNT

Managing Accounts

7.:2.1

Account Name
The account name can be 1 to 13 characters, chosen from the ASCII character
set whose decimal values are 9, 32 through 126, and 161 through 254, which
includes all keyboard characters. This field is displayed by the SHOW ACCOUNT
command, by PBS on job and file header pages, and by the BROADCAST
command on message headers.

7.:~.2

Date/Time/Keyboard of Last Login
The date and time that the account was most recently logged in to, and the
keyboard at which that login occurred. The monitor maintains these data. You
cannot change these fields with account management commands or monitor
directives.
The keyboard number is useful only for hardwired terminals. Since LAT terminals are created as they are needed, their keyboard numbers are not useful
identifiers. If you use OPSER logging, you can match up LAT keyboard numbers
and login times to determine the server and port used.

7.:!.3

Date/Time of Last Password Change
The date and time that the account's password was most recently changed. The
monitor maintains these data. You cannot change these fields with account
management commands or monitor directives.

7 .~!.4

Password
Normally, the system stores an account's password on the disk in such a way that
it cannot be looked up. You select this form of storage with the INOLOOKUP
qualifier. If you need compatibility with earlier versions of RSTSIE, the /LOOKUP
qualifier directs the system to store an account's password on the disk in such a
way that the UU.RAD directive can be used to look it up (see the RSTS / E System
Directives Manual). No DIGITAL-supplied program displays passwords.
For a INOLOOKUP account, the password can be 6 to 13 characters long, chosen
from the ASCII character set whose decimal values are 9, 32 to 62, 64 to 126, and
161 to 254, which includes all keyboard characters. For a /LOOKUP account, the
password must consist of six alphanumeric characters.
A user can change an account's password using the SET PASSWORD command if
the user has SETPAS privilege.

An account can also be set to the special state INOPASSWORD_PROMPT. When
this is so, LOGIN permits access to the account without requiring (or prompting
for) a password. The password still exists; however, it is not used until the special
state is cleared by setting /PASSWORD_PROMPT.

Managing Accounts

7-3

7.2.5

Interactive Flag
If set, the interactive flag lets a user log in under an account. If the interactive

flag is not set, the system does not allow a job to be logged in under this account's
PPN, except by the create job directive UU.JOB (Sb~ the RSTS / E System
Directives Manual). No logins (by use of the UU.LIN sy~tem directive), interactive
or otherwise, are allowed.
NOTE

Batch jobs are still allowed even if the interactive flag is not set.

7.2.6

Captive Account Flag
The captive account flag controls the way in which a job under a given account
exits. If the captive flag is set, the systel1n expects the job to remain totally under
the control of a predetermined procedure, such as a login command procedure or
a menu processor. Any attempt to exit to the keyboard monitor signals either an
error or some devious action. The monitor converts such an exit request into a
self-kill directive and removes the job.

7.2.7

Network Flag
The network flag, if set, lets you initiate a job under a given account over a
network connection. This flag allows or disallows any network activity, for
example, a remote COPY or SET HOST.

7.2.8

Dial-Up Flag
The dial-up flag, if set, lets you initiate a job under a given account over dial-up
lines.

7.2.9

Expiration Date
The expiration date specified by this field is the last date on which you can
initiate a job under the associated account. The system can be instructed
(!EXPIRE without a date argument) never to allow initiation of a job under the
PPN of this account. The system does not delete files after the expiration date.
That is left to the system manager.
You can disable this field using /NOEXPIRE. The /NOEXPIRE qualifier lets an
account remain usable indefinitely.

7.2.10

Authorized Privilege Mask
The authorized privilege mask specifies the maximwn set of privileges that a
job running under this account can exercise (other than by running privileged
programs).
When a job logs in, all authorized privilel~es for the account are copied to the job's
current privilege mask. You can use comlmands in your LOGIN.COM file to tum
off some privileges at log in.

7-4

Managing Accounts

7 .:~.11

Creation Date
The date that the account was created. The system assigns the creation date
when you create the account. The creation date cannot be changed.

7 .:~.12

UFO Cluster Size
UFD cluster size is the number of 512-byte blocks that each cluster allocated to
the account's directory contains. You assign the UFD cluster size of a directory
when you create an account. The UFD cluster size cannot be changed.

An account can have a maximum of seven clusters; therefore, the UFD clustersize
affects the maximum number of files that can be stored in the account. This
maximum value is further affected by the size of each file, its cluster size and
attributes.
7.:~.13

Quotas
You can assign quotas on disk usage, job slots, and message receiver slots when
you create an account. See Chapter 5 for a discussion of these quotas.

7 .:~.14

System Data
The system account [0,1] on the system disk, is used to support certain system
parameters such as the system password.

7.:3 Del Command Descriptions
The following sections describe the commands which you can use to manage
accounts. All parameters can be specified on the command line.
7.~S.1

CREATE/ACCOUNT Command

The CREATE/ACCOUNT comrnand sets up a new account on a file-structl ',ted
disk. CREATE/ACCOUNT requires GACNT or WACNT privileges.

Managing Accounts

7-5

Format
CREATE/ACCOUNT

dev:[p,pnj

Command Qualifiers
I[NO]CAPTIVE
ICLUSTER_SIZE=ufd-cluster-size
IOETACHEO_JOB_QUOTA=detached-job-quota
I[NO]OIALUP
I[NO]EXPIRE=date-spec
IIN_QUOTA=logged-in-quota
I[NO]INTERACTIVE
IJOB_ QUOTA=job-quota
I[NO] LOG
I[NO]LOOKUP
1M ESSAGE_QUOTA=message-quota
INAME="account name"
I[NO]NETWORK
lOUT_QUOTA=logged-out-quota
I[NO]PASSWORO_PROMPT
IPOSITION=position
IPRIVILEGES=(priv, ... )
IRI B_ QUOTA=rib-quota
IS IZE=ufd-size
ITEMPLATE=dev:(p,pn]
I[NO]USER
Prompts
Account:
Password:
Password again, for verification:

Defaults
INOCAPTIVE
none
IOETACHED_JOB_QUOTA==O
IDIALUP
INOEXPIRE
IIN_QUOTA=UNLIMITED
See discussion
IJOB_QUOTA=UNLIMITED
IlOG
INOlOOKUP
IMESSAGE_QUOTA=12
none
INETWORK
lOUT_QUOTA=UNLIMITED
IPASSWORD_PROMPT
IPOSITION=INDEX
IPRIVllEGES=SETPAS
IRIB_QUOTA=3
ISIZE=1
none
See discussion
Defaults
Optionally, for user accounts
Optionally, for user accounts

Command Parameters
dev:[p,pn]
Device specification of a mounted device and the PPN for the account to create.
You must specify either the device specification or PPN. The default value for dev:
is _SYO:. The default PPN is your PPN. You can have wildcards and ranges in
the PPN specification. The allowable values for the PPN depend on the privileges
you have.

Command Qualifiers
I[NO]CAPTIVE
Indicates whether the account is to be a captive account. The default is

INOCAPTIVE.

7-6

Managing Accounts

ICLUSTER_SIZE=ufd-cluster-slze

The number of 512-byte blocks that each cluster in this accounes UFD contains.
Acceptable values are 1, 2, 4, 8, or 16. Disks with a PCS of 32 or 64 require
a cluster size of 16. The value cannot be less than the pack cluster size of the
specified disk.
IDETACHED_JOB_QUOTA=detached-job-quota

See Chapter 5. The default is O.
I[NO]DIALUP

Indicates whether logins from dial-up terminal lines are allowed. The default is
/DIALUP.
I[NO]EXPIRE=date-spec
INOEXPIRE means that the account is never to expire.

/EXPIRE=date-spec means that the account expires at the end of the date you
specify.
/EXPIRE without a date specification means that the account is to be marked as
"already expired," that is, unavailable for any job to run or be created under its
PPN. The default is INOEXPIRE.
IIN_QUOTA=logged-ln-quota

See Chapter 5. The allowable values are 0 to 16,777,214 and UNLIMITED.
Zero means that no disk storage is allowed. UNLIMITED is a special keyword
that means the monitor does not enforce the logged-in quota. The default is
UNLIMITED.
I[NO]INTERACTIVE

This qualifier controls the way in which you can create a job under this account.
If you set INOINTERACTIVE, the system does not allow jobs to be logged in
under this account's PPN except by the create job directive UU.JOB (see the
RSTS / E System Directives Manual),
The default is !INTERACTIVE.
IJOB_ QUOTA=job-quota

See Chapter 5. The default is UNLIMITED.
I[NO]LOG

If !LOG is in effect, the CREATE function logs the creation of the new user or
nonuser account with one of the following messages:
User account dev: [p,pn] created at <device cluster number>
Non-user account dev: [p,pn] created at <device cluster number>

The default is !LOG.
I[NO]LOOKUP
INOLOOKUP means that RSTSIE stores the password in such a way that it
cannot be looked up. The passwords must be six to fourteen characters long.

!LOOKUP means that RSTSIE stores the password in a manner that follows
the rules for RSTSIE versions earlier than V9.0. The passwords must be six
characters long.
The default is /NOLOOKUP.

Managing Accounts

7-7

IMESSAGE_QUOTA=message-quota

See Chapter 5. The allowable values can be from 0 to 65,535 or UNLIMITED.
The UNLIMITED keyword means that the monitor does not enforce the message
quota.
The default is 12.
INAME="account name"

Can be 1 to 13 characters, chosen from the ASCII character set whose decimal values are 9, 32 to 126, and 161 to 254, which includes all the keyboard
characters.
I[NO]NETWORK

Indicates whether logins over network connections are allowed.
The default is !NETWORK.
lOUT_ QUOTA=logged-out-quota
See Chapter 5. The allowable values are 0 to 16,777,214 and UNLIMITED. Zero
means that no disk storage is allowed to remain at logout time. UNLIMITED is a
special keyword that means the monitor does not enforce the logged-out quota.

The default is UNLIMITED.
I[NO]PASSWORD_PROMPT
!NOPASSWORD_PROMPr means that LOGIN is not to require or prompt for a
password for a user to gain access to the specified account (that is, specifying the
[p,pn] in response to User: is sufficient to log a user in to the account).

The default is !PASSWORD_PROMPT.
IPOSITION=posltlon

The starting position of the account's UFD expressed in device cluster numbers
(DCNs). Legal values are 1 through the lnaximum DCN for the specified disk.
MIDDLE is a special keyword that means create the UFD for this account as
near as possible to the middle of the disk. INDEX]s a keyword value that
means create the UFD as near as possiblle to the storage allocation table (file
[O,l]SATT.SYS).
The default is INDEX.
/PRIVILEGE=(prlv, ... )
The possible values for each priv are: NONE, ALL, a privilege keyword, or a
privilege keyword with a NO prefix. ALL means all privileges that the current
user has enabled.

RSTS/E scans the list from left to right, with each successive term adding or
taking away privileges. For example, (ALL,NOWWRITE) means all privileges
except WWRITE.
If you specify ITEMPLATE=dev:[p,pn], RSTS/E treats the template account's privileges as though they appeared to the left of all the entries in the !PRIVILEGE
qualifier.

IRI B_ QUOTA=rlb-quota

See Chapter 5. The allowable values can be from 0 to 256 or UNLIMITED. The
UNLIMITED keyword means that the monitor does not enforce the RIB quota.
The default is 3.

7-8

Managing Accounts

ISIZE=ufd-slze
The initial size of the new account's UFD expressed in clusters. Acceptable values
are 0 - 7.

The default is 1.
ITEMPLATE=dev:[p,pn]
If present, this qualifier indicates that the specified account's parameters are to
be used as defaults for the new account's parameters. You must specify either the
device specification or the PPN.

The default value for dev: is _SYO:. The default PPN is your PPN.
NOTE

The specified account's password, size, and position are not part of the
defaults for the new account.
I[NO]USER
Indicates the type of account to create. IUSER means create an account with
login attributes and file storage capabilities. /NOUSER means create an account
for file storage only.

The default is /NOUSER.

Managing Accounts

7-9

7.3.2

DELETE/ACCOUNT Command

The DELETE/ACCOUNT command removes an account from a file-structured
disk. DELETE/ACCOUNT requires WACNT or GACNT privilege and WWRITE
or GWRITE privilege. You also need SYSIO privilege to delete a [0,*] account.

Format
DELETE/ACCOUNT

dev:[p,pn]

Command Qualifiers
/[NO] lOG
/[NO]PURGE[=QUERY]
/[NO]QUERY
/[NO]RESET[=QUERY]
/[NO]USER

Defaults
flOG
See discussion
See discussion
See discussion
See discussion

Prompts
Account:

Command Parameters
dev:[p,pn]
Device specification of a mounted disk, and the PPN of the account to delete from
the specified disk. You can omit the device specification (the default is _SYO:);
you must specify the PPN. The PPN field allows wildcards and ranges. Unless
you specify /NO QUERY, the system requests confirmation that you really want to
delete the account:
Really delete account dey: [p,pn] <NO> ?

Command Qualifiers
I[NO]lOG

If /LOG is in effect, the DELETE function logs the deletion of the account:
Account dev:[p,pn] deleted

The default is /LOG.
I[NO]PURGE[=QUERY]
This qualifier controls the actions to be taken if the account contains files.
When you specify /PURGE, DELETE/ACCOUNT deletes any files which it finds,
and then deletes the account. If you specify the optional argument (=QUERY),
RSTSIE displays a warning message if files exist for the specified account and
then requests permission to proceed.

When you specify /NOPURGE, DELETE/ACCOUNT displays a warning message
if files exist in the account, skips the aecount, and goes to the next account if a
wildcard account was specified:
%Account dey: [p,pn] has n blocks of disk allocated
Account dev:[p,pn] not deleted

7-10

Managing Accounts

If you do not specify /PURGE, INOPURGE, or if /PURGE=QUERY is specified,
"DELETE/ACCOUNT proceeds if no files exist. Otherwise, before proceeding,
RSTSIE displays a prompt indicating that there are files in the account:
Proceed (yiN) <N> ?

Deleting files (either by /PURGE or by answering YES to the prompt) requires
GWRITE or WWRITE privilege. If any file deleted by DELETE/ACCOUNT has
protected data (protection code = 128), the system writes over the file with zeros
before deletion.
I[NO]QUERY

Controls the action taken if the account has nonzero accounting data or contains
files.
/QUERY indicates that RSTSIE displays a prompt before deleting any account.
INOQUERY indicates that RSTSIE only displays a message if the account has
nonzero accounting data or contains files.
The default is INOQUERY.
I[NO]RESET[=QUERY]

Controls the action taken if the specified account has nonzero accounting data
(CPU time, device time, connect time, or kilo-core-ticks). If you specify !RESET,
RSTSIE deletes the account even if it has nonzero accounting data.
If you specify the optional argument (=QUERY), RSTSIE displays a warning
message that the specified account has nonzero accounting data and asks for
permission to proceed.
If you specify INORESET, DELETE/ACCOUNT displays a warning message if
the account has nonzero accounting data, skips the account, and goes to the next
account if a wildcard account was specified:
%Account dev: [p,pn] has nonzero accounting statistics
Account dev: [p,pn] not deleted

If you do not specify !RESET, INORESET, or if !RESET=QUERY is specified,
the delete account operation proceeds if the accounting data is zero. Otherwise,
RSTSIE displays a warning message indicating that the specified account has
nonzero data and asks for pennission to proceed.
I[NO]USER
You can restrict the search for a given account to a specific type (user or nonuser)
by using this qualifier. The default is to search all accounts.

Managing Accounts

7-11

Examples
1.

$ DELETE/ACCOUNT/PURGE/RESET/NOQUERY

This command allows you to delete an account regardless of its current state.
2.

$ DELETE/ACCOUNT/NOPURGE/NORESET

This command prevents you frOll}. deleting an account that has files or nonzero
accounting data.
3.

$ DELETE/ACCOUNT/QUERY

This command allows you to be prompted before deleting an account.
4.

DELETE/ACCOUNT

This command allows you to be prompted only for an account that has files or
nonzero accounting data.

7-12

Managing Accounts

7.3.3

SET ACCOUNT Command
The SET ACCOUNT command modifies account parameters on a file-structured
disk. SET ACCOUNT requires GACNT or WACNT privilege.

Format
SET ACCOUNT

dev:[p,pn]

Command Qualifiers

Defaults

I[NO]CAPTIVE
IDETACHED_JOB_QUOTA=detached-job-quota
I[NO]DIAlUP
I[NO] EXP IRE[ =date-spec]
liN _ QUOTA=logged-in-quota
I[NO]INTERACTIVE
IJOB __QUOTA=job-quota
I[NO]lOG
I[NO]lOOKUP
IMESSAGE_QUOTA=message-quota
INAME="account name"
I[NO]NETWORK
lOUT _ QUOTA=logged-out-quota
I[NO]PASSWORD_PROMPT
IPRIVllEGES=(priv, ... )
IR IB__QUOTA=rib-quota
I[NO]USER

none
none
none
none
none
none
none
IlOG
none
none
none
none
none
none
none
none
none

Prompts
Account:

Command Parameters
dev:[p,pn]
Device specification of a mounted disk and the PPN of the account to modify
on the specified disk. You can omit either the device specification or the PPN,
but not both. The default device specification is _SYO:. The default PPN is the
user's PPN. You can have wildcards and ranges in the PPN specification. RSTS/E
uses all qualifiers that apply to a specific account and displays the following
informational message:
Account dev: [ppn] modified

If no qualifiers apply to a selected account, the system displays the following
warning message:
%Account dev: [ppn] not modified

Managing Accounts

7-13

Command Qualifiers
I[NO]CAPTIVE

Indicates whether the account is to be a captive account.
IDETACHED_JOB_QUOTA=detached-job·,quota
See Chapter 5.
I[NO]DIALUP

Indicates whether logins from dial-up tenninallines are allowed.
I[NO]EXPIRE=date-spec

INOEXPIRE means that the account is never to expire.
/EXPIRE=date-spec means that the account is to expire at the end of the day
specified.
/EXPIRE without a date specification means that the account is to be marked as
unconditionally expired, that is, unavailable for any job to run under its PPN.
IIN_QUOTA=logged-in-quota

See Chapter 5.
I[NO]INTERACTIVE

This qualifier controls the way in which you can create a job under this account.
If you set INOINTERACTIVE, the systenl does not allow jobs to be logged in

under this account's PPN except by the create job directive UU.JOB (see RSTS / E
System Directives Manual).
IJOB _ QUOTA=job-quota

See Chapter 5.
I[NO]LOG

If !LOG is in effect, the SET ACCOUNT function logs the modification of the
account:
Account dev: [p,pn] modified

If INOLOG is in effect, only warning messages are logged.

The default is /LOG.
I[NO]LOOKUP

INOLOOKUP means that RSTSIE stores the password in such a way that it
cannot be looked up. /LOOKUP means that RSTSIE stores the password in at
manner that follows the pre-V9.0 rules.
IMESSAGE_QUOTA=message-quota
See Chapter 5.
INAME="account name"

Can be 1 to 13 characters, chosen from the ASCII character set whose decimal
values are 9, 32 to 126, and 161 to 254.
I[NO]NETWORK

Indicates whether logins over network connections are allowed.
lOUT_QUOTA=logged-out-quota

See Chapter 5.

7-14

Managing Accounts

I[NO]PASSWORD_PROMPT

If you specify INOPASSWORD__ PROMPT, the account is set to not require a
password for logging in. To resume password prompting, specify IPASSWORD_
PROMPT; the password is the same as it was before INOPASSWORD_PROMPT
was specified.
IPRIVILEG E=(priv, ... )

This qualifier modifies the privileges currently assigned to the account. A priv
specification occurring in the positive form adds the associated privilege to
the account's authorized privilege mask; a NOpriv specification removes the
associated privilege. Note that the change affects only future logins; jobs already
logged in are unaffected.
A positive specification (priv) is effective only if you are authorized to have that
privilege. However, any user with WACNT or GACNT (as appropriate) can
remove any privilege.
IRIB_QUOTA=rlb-quota
See Chapter 5.
I[NO]USER
You can restrict the search for a given account to a specific type (user or nonuser)
with this qualifier.

The default is to search all accounts.

Managing Accounts

7-15

7.3.4

SET PASSWORD Command
The SET PASSWORD command lets you change the password of the specified
user account. When you issue this command, the system responds with:
New password:

You then enter the desired new password. If the entry is invalid or a null string
(Return only), the system gives the error message ?Invalid password, and returns
to the New password prompt. The systenl accepts an oversize password string;
however, it ignores the excess characters and displays a warning message.
Mter you enter a new password, the system prompts:
New password again, for verification:

You then reenter the same password. If the passwords do not match, the system
displays the error message ?Verification Dlismatch, does not change the password,
and returns to command level. If the passwords match, the system displays the
/LOG message, if any, and returns to command level.
Users need SETPAS (or either GACNT or WACNT) privilege to change their
own password. Users need GACNT or WACNT privilege to change another
account's password. If you only have SETPAS, the system prompts you for the old
password:
Old password:

You must enter the correct current password for your account. A mismatch
results in the error message:
?Old password validation error.

Format
SET PASSWORD

[de v:[p, pm11

Command Qualifiers
/[N 0] lOG
/SYSTEM

Defaults
flOG
none

Prompts
See description

Command Parameters
dev:[p,pn]
Device specification of a mounted disk, and the PPN of the account for which to
change the password. You can omit the device specification, the PPN, or both.
The default device specification is _SYO:. The default PPN is the user's PPN. If a
nonuser account is specified, the system displays a warning message:
% [ppn] is not a user account

7-16

Managing Accounts

You can have wildcards and ranges in the PPN specification. If a wildcard or
range specification includes a nonuser account, RSTSIE skips the account and
does not display a message.
NOTE

Users who only have SETPAS privilege cannot specify a dev:ppn
parameter; in that case, RSTSIE rejects the command and displays the
following error message:
?Need xACNT privilege to access dev: [ppnJ

Command Qualifiers
I[NO]LOG
If you specify /LOG, successful setting of a new password yields:
Password set for account dev: [p,pn]

The default is /LOG.
ISYSTEM
If you specify /SYSTEM, the new password replaces the old system password.
You must have WACNT privilege to change the system password.

Managing Accounts

7-1 ~

7.3.5

SHOW ACCOUNT Command
The SHOW ACCOUNT command displays account attributes. No privilege is
required to display your own account. You need GACNT or WACNT privilege to
display other accounts.
The following are samples of the output from the SHOW ACCOUNT command
with different qualifiers.

$ SHOW

ACCOUNT/BRIEF [*,194]

Account

Name

Allocation

[1,194]
[2,194]
[3,194]
[4,194]
[10,194]

Blade Runner
JT Sysgen
JT PIP
JT ACTMGR

6016
5080
4336
3280
752

IDNCLP Last Login
IDN
I
IDN
IDN

P KB26:
P KB32:
P KB32:
P

Expires

0:3-Dec-89 01:24 PM 15-AUG-90
12-Nov-89 02:56 PM
06-Nov-89 03:06 PM

5 accounts matching _SYO: [*,194]
Total allocation:
19464

NOTE

If an asterisk (*) character appears after the value in the allocation
field, it means the account is over quota.
$ SHOW ACCOUNT/FULL
-

SYO: [1, 214 ]

"Blade Runner"

Created 09-May-90

Privileges:

GACNT WACNT
HWCFG HWCTL
USER2 USER3
SHUTUP SWCFG

GREAD GWRITE WREAD WWRITE DATES DEVICE EXQTA
INSTAL JOBCTL MOUNT PBSCTL RDMEM USERO USER1
USER4 USERS USER6 USER7 RDNFS SEND
SETPAS
SWCTL SYSIO SYSMOD TUNE
WRTNFS

Attributes:

INTERACTIVE
NOLOOKUP

DIALUP
NETWORK
PASSWORD PROMPT

Quotas:

Disk usage
- Logged out:
Job limits
Detached:
Send/Receive RIB:

Accounting:

CPU Time:
Device Time:
Connect Time:

unlimited
unlimited
256

58:23:24.9
214:35
903:25

NOCAPTlVE
EXPIRES: 15-Aug-90
Logged in:
Total:
Message:

unlimited

Kilo-Core-Ticks:
UFD Clustersize:
Blocks allocated:

52377442
16
6000

unlimited

Last Password Change on 28-Nov-89 at 03:21 PM
Last login on KB26: on 03-Dec-89 at 01:24 PM

NOTE

If an asterisk (*) appears after the value in any field in the quotas
section, it means the account is over quota.

7-18

Managing Accounts

$ SHOW ACCOUNT/ACCOUNTING_DATA [*,194]

Account

Name

[1,194]
[2,194]
[3,194]
[~I, 194]
[10,194]

Blade Runner
JT Sysgen
JT PIP
JT ACTMGR

Allocation

KCT

CPU-time

Connect

Device

6016
5080
4336
3280
752

52377442
2215
859
797

58:23:24.9
00:00:17.5
00:00:08.6
00:00:09.7

903:25
00:15
00:17
00:02

214:35
00:00
00:00
00:00

5 accounts matching _SYO: [*,194]
Total allocation:
19464

NOTE

If an asterisk (*) character appears after the value in the allocation
-field, it means the account is over quota.

Format
SHOW ACCOUNT

dev:[p,pnj

Command Qualifiers

Defaults

IACCOUNTING_DATA
IAll
IBRIEF
IFUlL
IOUTPUT=filespec
I[NO]RESET
I[NO]USER

See discussion
See discussion
IBRIEF
none
none
INORESET
See discussion

Prompts
none

Command Parameters
dev:[p,pn]
Device specification of a mounted disk, and a PPN. The PPN can be a wildcard
or range specification (for appropriately privileged users). You can omit both the
device specification and the PPN. The default device specification is _SYO:. The
default PPN is the user's PPN.

Command Qualifiers
I ACCOUNTING_DATA

Produces a one-line display of accounting data for the specified accounts. This
qualifier conflicts with !BRIEF and /FULL. If you specify them in combination,
the rightmost qualifier overrides the others and execution proceeds without an
error.
IAll

Produces a display of all accounts on the specified disk. If an account is specified
along with IALL, the account specification overrides fALL.

Managing Accounts

7-H

IBRIEF
Produces a display that uses only one line per account. You cannot use !BRIEF in
combination with !RESET. When shown the display, the letters IDNCLP indicate
that the account has the following attributes:
I
D
N
C
L
P

Interactive
Dial-up
Network
Captive
Lookup password
Password prompt

IFULL
Produces a complete display for the accounts specified.
IOUTPUT=filespec
If specified, the display is written to the file you specify.
I[NO]USER
Specifies which accounts are to be included in the display. IUSER means only
display accounts with login attributes. fNOUSER means only display accounts
without login attributes.

The default is to display all accounts.
I[NO]RESET
Indicates whether the accounting data is reset after being displayed. The !RESET
qualifier requires GACNT or WACNT privilege. You. cannot use !RESET in
combina tion with !BRIEF. If neither IACCOUNTING_DATA or /FULL is specified,
IACCOUNTING_DATA is assumed.

The default is INORESET.

7-20

Managing Accounts

Chapter f

Managing Terminals

This chapter gives background information on managing terminals, either
on your system or connected by Local Area Transport (LAT). The RSTS/E
operating system supports a variety of terminals. The RSTS/E operating system
automatically sets the default characteristics of all interfaces to certain standard
values:
•

Produces hard-copy output.

•

Prints data up to 80 columns wide

•

Sets autobaud detection on multiplexed lines

You must use the SET TERMINAL command to set the characteristics of the
terminals that do not have these standard default values.

8.1

Understanding the SET TERMINAL Command
The SET TERMINAL command sets characteristics for terminals attached locally
to a RSTS/E system or for terminals connected by dial-up lines. Users can use
the SET TERMINAL command to set characteristics for their own terminals.
Only users with HWCFG privilege can use the SET TERMINAL command to
define characteristics for other terminals.
The IAUTOBAUD qualifier lets the system detect and set the baud rate of any
multiplexed terminal when the user logs in. As system manager, you can include
SET TERMINALINOAUTOIPERM commands in the START.COM file for all
multiplexed terminals you want to manually set the baud rate.
There are two methods for setting the characteristics of dial-up lines:
•

Nonprivileged user sets characteristics. All dial-up lines start out with the
standard default settings. Consequently, the user of a dial-up line must log
in to the system at the standard default settings, and then use the SET
TERMINAL command to set the characteristics of the terminal. Note that
the terminal characteristics revert to the standard settings when the user
logs out.

•

A user with HWCFG privilege sets permanent characteristics. You can issue
the SET TERMINALIPERMANENT command to set the permanent characteristics. This causes the system to automatically use the characteristics you
set, instead of the standard values, each time a user logs in on a particular
dial-up line. Note that the permanent characteristics remain in effect for
the current time-sharing session, unless you issue a new SET TERMINAL
cOlnmand.

Managing Terminals

8-1

8.2 Terminal Line Speed Characteristics File: TERSPD.SYS
You need to create a special tenninalline speed characteristics file if you have
these two circumstances:
•

You decide to restrict the valid line speeds at certain tenninals to a subset of
the line speeds allowed by the tenninal inteIface.

•

You install an interface that has been modified so nonstandard line speeds
are substituted for one or more standard interface line speeds.

The following guidelines can help you prepare a tenninalline speed characteristics file:
1.

Gather the infonnation on the restrictions or modifications and the keyboard
numbers of the tenninals affected.

Create a file named TERSPD.SYS with a text editor.

Include all of the infonnation from step 1 in the file.

Place the file in the system library account.
NOTE

The IAUTOBAUD qualifier does not. honor the restricted speeds in
TERSPD.SYS.
You can use the file TERSPD .SYS to allow only certain speeds on certain
terminals. The format for each line of the file is the terminal device designator,
a colon, and a list of speeds to allow for that terminal; each speed separated by a
comma. For example:
KBF2: 300,1200, ...

The terminal device designator can be KBnn:, TTnn:, or KBcnn:. The form KBnn:
or TTnn: produces a result that may change as new hardware is added. In the
form (KBcnn:), c indicates the hardware controller and nn is a subline on that
type controller. Table S-llists controller types that RSTS/E supports.
DIGITAL recommends that you use the controller form to specify terminal devices
in the START.COM, TERSPD.SYS, and TERDFL.SYS files. In this way, if you
change the number of pseudo keyboards~ you do not have to change those files.
The SHOW DEVICE command displays both forms of the device designator for
each keyboard.

8-2

Table 8-1:

RSTS/E Supported Controller Types

Code

Controller Type

DLll C,D type single line interface

DLIIE type single line modem interface

Pseudo keyboards

DH multiplexor

DZlDZVIDZQ multiplexor

DHVIDHUIDHQlDHF/CXF/CXAlCXB multiplexor

Managing Terminals

You can specify up to 16 speeds for each terminal. You can specify a range of
terminals on one line as low-end terminal device designator, an optional colon,
dash, high-end terminal device designator, mandatory colon, and then the list of
speeds.
For example:
KBGO-KBH8: 300,1200, ...

When you specify a terminal number without any speeds after it, no speeds are
allowed and the speed cannot be changed. You can use blank lines, comment
lines (lines beginning with an exclamation mark), and comments at the end of
lines to document the file.
An example of a TERSPD.SYS file follows:
T E R S PD. S Y S
This file is used by the system manager to define
those speeds which are allowed on each terminal.
The format is as follows:
KBn:list of speeds separated by commas.
KBFO:
!No speeds allowed for KBFO
KBF1:2400,9600
!KBF1 can have either 2400 or 9600
KBF2:-KBF4:9600
!KBF2 to KBF4 can have 9600 only
KBF5: 9600,EXTA, EXTB
!KBF5 can use 9600 & external speeds
KBF6:0,50,75,110,150,200,300,600,1200,1800,2400,4800,9600
KBF7:0,50,75,110,150,200,300,600,1200,1800,2400,4800,9600,EXTA,EXTB
KBF8-KBF15:0,50,75,110,150,200,300,600,1200,1800,2400,9600
KBGO-KBH8:50,300,1200,2400,4800,9600
KBH9:0,50,75,110,150,200,300,600,1200,1800,2400,9600
KBH10:-KBH15:300,600,1200 !KBH10 to KBH15 are dial-up lines

NOTE

The DLIIE-type interface and the individual local interfaces (KLll
and DLIIA through DLIID) do not have programmable baud rates.
For this reason, you cannot execute commands to change baud rates on
a keyboard line having any of those interfaces (KBAnn, KBBnn, and
KBCnn).

8.3 Setting Terminals Other Than Your Own
If you have HWCFG privilege, you can set the characteristics of other terminals
in the RSTSIE system with the SET TERMINAL command. For example, if you
want to set the characteristics of a VT100 terminal at KB32, type:
$ SET TERMINAL/VT100 KB32:

Mter you type the keyboard number and press the Return key, RSTSIE immediately sets the characteristics of the line to those of a VT100 alphanumeric display
terminal. DCL prints the dollar sign ($) prompt again to let you enter another
command.
You can also change specific characteristics of a terminal. For example, to limit
the line length of the terminal at KB40, type:
$ SET TERMINAL/WIDTH=60 KB40:
$

Managing Terminals

8-3

Each time 60 characters print on KB40, the system performs a carriage return
and line feed.
Use the !PERMANENT qualifier to specify characteristics on a logged-out
terminal. For example, you can use !PERMANENT to set an initial WIDTH
characteristic for a pseudo keyboard:
$ SET TERMINAL/WIDTH=70/PERMANENT KHDO:

8.4 User-Defined Terminal Characteristics File: TERMAC.SYS
Each line of TERMAC.SYS contains the unique name for a terminal type and
the command qualifiers that define the characteristics of that terminal type. The
qualifiers are explained in the description of the SET TERMINAL command at
the end of this chapter.
The ITYPE=n qualifier assigns a unique 8-bit binary value to each terminal type.
The values 0 to 128. are reserved for use by DIGITAL. The values 129. to 255.
are available for user-defined terminal types. The SHOW TERMINAL command
uses this value to locate a terminal's cha:racteristics.
An exatnple of a TERMAC.SYS file follows:
T E R MAC . S Y S
VT125X/DEV=VT125/WIDTH=80/SPEED=(300,9600)/TYPE=201
GIGI/DEV=VKIOO/SETUP=GREEN/TYPE=202
! VKIOO W/GREEN BACKGROUND
DIABLO/DEV=LA34/SPEED=9600/WID=80/SETUP=DIABLO/TYPE=203
LA34X/DEV=LA34/TAB/SPEED=9600/TYPE=204
VT52X/DEV=VT52/SPEED=1200/TYPE=205
FOO/DEV=BAR/WIDTH=132/TYPE=206
BAR/DEV=LA120/SETUP=NARROW/WIDTH=80/TYPE=207

For example, to define an LA120 that has only 80-column paper, place the
following line in the file:
MYLA120/DEVICE=LA120/WIDTH=80/TYPE=208

Then, use the following command to setKB19: to these characteristics:
$ SET TERMINAL/DEVICE=MYLA120/PERMANENT KB19:

You can initialize your terminals to the ]mown terminal types and their default
characteristics through the use of the TE:RMAC.SYS file. You can use blank lines,
comment lines (lines beginning with an exclamation mark), and comments at the
end of lines to document the file.

8.5 Automatic Setting of Terminal Characteristics: TERDFL.SYS
The file TERDFL.SYS establishes default terminal characteristics for nonstandard keyboards. The file is used whenever the user executes the SET
TERMINAUINQUIRE command and the system was unable to determine the
terminal type. The system looks up the terminal number in the file and, if found,
sets the characteristics that are on that line. If the terminal number is not found,
the system displays an error message.
For most DIGITAL terminals, SET TERMINALIINQUIRE can determine the
type. The only supported exception is the LA36. The same may apply to older,
unsupported terminals or terminals made by other manufacturers.

8-4

Managing Terminals

If you want the SET TERMINALIINQUIRE command to work for terminals whose type cannot be determined; enter them in the TERDFL.SYS
file. (It is not necessary to put any other supported terminals in the file.) If
SET 'rERMINALIINQUIRE can determine the type, it ignores any entry in
TERDFL.SYS.
An example of a TERDFL.SYS file follows:
T E R D FL. S Y S
KBH10:-KBH15:/DEV=LA36

The format for each line of the file is the terminal device designator, a colon, and
then the command qualifiers that define the characteristics of the keyboard. A
range of terminals can be specified on one line as the low-end terminal device
designator, an optional colon, a dash, the high-end terminal device designator, a
mandatory colon, and then the qualifiers.
You can use blank lines, comment lines (lines beginning with an exclamation
mark), and comments at the end of lines to document the file.

8.16 Setting Permanent Terminal Characteristics
The /PERMANENT qualifier lets you set the permanent characteristics of a
terminal. For example, to set the characteristics of the dial-up line on keyboard
14 for the current time-sharing session, you log in to an account that has HWCFG
privilege and type the command:
$ SET TERMINAL/PERMANENT KB14:/DEV=LA36

When you include the !PERMANENT switch with the SET TERMINAL command,
the command takes effect immediately. If you do not have HWCFG privilege or
the characteristics of the line conflict with the command, RSTSIE prints an error
message and reprints the DCL prompt. The characteristics set by executing a
SET TERMINAL!PERMANENT command are in effect until the next system shut
down or until you change them.. If you set characteristics without !PERMANENT,
they revert to the standard characteristics on logout.
Use the !PERMANENT qualifier with all SET TERMINAL commands in
START. COM.

8.'4' Using the INOBROADCAST and IBROADCAST Command
Qualifiers
The INOBROADCAST qualifier prevents your terminal or any terminal on the
system from receiving broadcasts sent by the BROADCAST command. Use this
when:
•

You need to print a document on a letter-quality printer and do not want
messages corrupting the printout

•

You are using a text editor in screen mode at a dial-up tenninal set at a low
baud rate. Repainting the screen each time someone sends a message can be
time consuming, especially at 300 baud.

Managing Terminals

8-5

Use the !BROADCAST qualifier to allow your tenninal to receive messages again.
Use the /NOBROADCAST qualifier to keep messages from printing on your own
tenninal. For example:
$ SET TERMINAL/NOBROADCAST
$ SHOW TERMINAL/BRIEF
Terminal: KB29:
Device_Type: VT52

Break
CRFill=O
No Hostsync
Scope
Width=80

No Broadcast
No Delimiter
Interactive
Speed=9600

Control=(R,T,C)
No Eight_bit
Lowercase
Tab

No Control=X
No Form
Resume=Control C
TTSync

The SHOW TERMINAL command lets you verify that the command worked and
that the rest of the tenninal settings are set properly. If you decide to reverse
this condition, use the !BROADCAST qualifier. For example:
$ SET TERMINAL/BROADCAST
$ SHOW TERMINAL/BRIEF
Terminal: KB29:
Device_Type: VT52

Break
CRFill=O
No Hostsync
Scope
Width=80
$

Broadcast
No Delimiter
Interactive
Speed=9600

Con'trol= (R, T, C)
No Eight_bit
Lowercase
Tab

No Control=X
No Form
Resume=Control C
TTSync

Your tenninal can now receive BROADCAST messages.
You can also use the /NOBROADCAST a.nd !BROADCAST qualifiers on other
tenninals on your system. Use the SET TERMINAL command as follows:
$ SET TERMINAL KB25:/NOBROADCAST/PERM
$ SHOW TERMINAL/BRIEF KB25:
Terminal: KB25:
Device_Type: VT100

Break
CRFill=O
Hostsync
Scope
Width=80
$

No Broadcast
No Delimiter
Interactive
Speed=9600

Control=(R,T,C)
No Eight_bit
Lowercase
Tab

No Control=X
No Form
Resume=Control C
TTSync

8.8 Terminal Servers (LAT)
RSTSIE includes support that lets systems with DECnet/E connect tenninals
directly over the Ethernet from tenninal servers. These tenninal connections are
called Local Area Transport (LAT). LAT lets users access your system directly
from tenninals connected to the following tenninal servers:
•

DIGITAL Ethernet Tenninal Server (DECSA)

•

DECserver 100

•

DECserver 200

•

DECserver 500

•

DECserver 550

•

VAXmate systems

•

IBM -PC systems running DECnet-DOS

RSTSIE cannot access printers through tenninal servers.
8-6

Managing Terminals

The DECSA and DECserver tenninal servers require that software be loaded
into them from a host system on the Ethernet. Since RSTS/E cannot load thls
software, it must be loaded from another system. The systems currently available
to load this software are:

•
•
•
•
•
•
•
8.8:.1

VAXlVMS
ULTRIX-32
ULTRIX-32m
DECsystem-lO
DECSYSTEM-20
RSX-IIM-PLUS
MicrolRSX

Enab~ing LAT

Th support LAT terminal servers, you must have DECnetiE and the Ethernet

hardware. If you have these, RSTS/E automatically makes LAT available.
LAT support requires about 2K words of memory. If you are not using LAT
on your Ethernet, use the following command to prevent LAT support in your
system:
$

SET SYSTEM/NOLAT

Mter entering thls command, you must shut the system down and restart it to
change the state of LAT. Th restore LAT support, use the SET SYSTEMILAT
command, followed by another shutdown and restart.
Use the SHOW SYSTEM command to detennine the current state of LAT and
the state that will take effect after the next reboot (if different from the current
state).
Use the SHOW TERMINAL cornmand to list the LAT server and port name for
those connections coming from lLAT tenninal servers.
Use the following DCL commands to set parameters for the LAT software:
Command

Description

CREATE/SERVICEILAT

Creates a LAT service

DELETE/SERVICEILAT

Deletes a LAT service

SET NODEILAT

Sets LAT node characteristics

SET SERVICEILAT

Sets LAT service characteristics

SHOW COUNTERSILAT

Shows LAT related counters

SHOW NODEILAT

Shows LAT node characteristics

SHOW SERVICEILAT

Shows LAT service characteristics

SHOW SESSIONS

Shows infonnation about LAT sessions

SHOW TERMINAL_SERVERSILAT

Shows tenninal servers known to LAT

STARTILAT

Starts LAT on an Ethernet device

STOPILAT

Stops LAT on an Ethernet device

Each of these commands is described more fully at the end of this chapter.

Managing Terminals

8-7

8.8.2

Isolation of LAT Problems
1b isolate LAT problems, use the DCL commands SHOW COUNTERSILAT and
SHOW COUNTERSILATIDEVICE, and the DECnetIE Network Control Program
(NCP) utility. 1b invoke NCP, type:
$ RUN DECNET$:NCP

You should then get the NCP> prompt. The following NCP commands are
particularly useful in isolating LAT problems:
NCP> SHOW LINE dev COUNTERS
NCP> LOOP CIRCUIT dev PHYSICAL ADDRESS ethernet-address

Note that dev is the name of the device being used (UNA-O or QNA-O) and
ethernet-address is the Ethernet address of the terminal server under question.
See the DECnet / E System Manager's Guide for more information on these
commands.
For more information on LAT activity, consult the console terminal. The LOGIN
and LOGOUT commands automatically send the server and port names for LAT
terminals to the OPSER program, which relays the names to the console. (If
OPSER is not running, LOGIN and LOGOUT send the names to KBO:.)
The LOGIN command also sends the server and port names to [O,1]LOGIN.COM,
which you can modify to respond to the information as you see fit. LOGIN.COM
automatically passes the names to the gTOUp and user LOGIN.COM files in
Parameter P5, and executes a SET TERMINALIINQUIRE command for the LAT
terminal.

8.9 Pseudo Keyboards
The monitor dynamically creates pseudo keyboards as they are needed and
eliminates them when they are closed. Because of this, you have no chance to
preset any of the terminal characteristics for these pseudo keyboards. The only
time pseudo keyboards are not available is when all 128 keyboards are in use or
when the system is very low on buffer space.
Some older applications require static pseudo keyboards instead of dynamic
pseudo keyboards. OPSER BATCH requires static keyboards. 1b create these
static pseudo keyboards, use the following command:
$ SET SYSTEM/PSEUDO_KEYBOARD=n

where n is the number of static pseudo keyboards needed. You can preset terminal characteristics for these pseudo keyboards with the SET TERMINALIPERM
command.
For these commands to have any effect, you must shut the system down and
restart.
Use the SHOW SYSTEM command to find out the number of pseudo keyboards
available the next time the system is relbooted.

8-8

Managing Terminals

8 ..10 Enabling Forms Management Support (FMS-11)
1b enable or disable FMS, use the command:
$ SET SYSTEM/[NO]FMS

Mter you issue the SET SYSTEM command, you must shut the system down and
restart before FMS is available (or becomes unavailable).
Use the SHOW SYSTEM comnland to find out the current state of FMS and the
state that will occur the next ti.me the system is rebooted.

8. -11

Del Command Descriptions
The following sections explain the commands available for managing terminals:
SET TERMINAL, SHOW TERMINAL, and the LAT-related commands.

8.11.1

CREATE/SERVICE/LAT Command

Creates a service which is then offered by the LAT host node. You need the
SWCFG privilege to use the CREATE/SERVICEILAT command. Example:
$ CREATE/SERVICE/LAT EDERE

CREATE/SERVICE/LAT
Command Qualifiers
IIDENTIFICATION="service_id"
I[NO]l.OG
ISTATIC_RATING=rating

Defaults
none
INOLOG
ISTATIC_RATING=255

Prompts
Service name:

Command Parameters
service_name
Gives a name for the service bei.ng created. Service names may have as many
as 16 characters, including the multinational alphanumeric characters and the
characters period (.), underscore L), hyphen (-), and dollar sign ($). Typically, the
service name is the same as the DECnetIE node name.

Command Qualifiers
IIDENTIFICATION="servlce_ld"
Specifies a description for the service. This description is used to further identify
the service being offered. It may be up to 64 characters in length.

Managing Terminals

8-9

I[NO]LOG

Specifies whether or not a confirmation message is displayed indicating that the
requested action was taken. The default is INOLOG.
/STATIC_RATING=ratlng

Specifies the static rating associated with the service. The valid range is 0 to 255.
The default is /STATIC_RATING=255.

8-10

Managing Terminals

8.11.2

DELETE/SERVICE/LAT Command
Deletes a service currently offered by the LAT host node. You need the SWCFG
privilege to use the DELETE/SERVICEILAT command. For example:
$ DELETE/ SERVICE/LAT EDERE~

Format
DELETE/SERVICE/LAT
Command Qualifiers
/[NO]LOG

service_name
Defaults
/NOLOG

Prompts
Service name:

Command Parameters
service_name
Specifies the service being deleted. Service names may have as many as 16 characters, including the multinational alphanumeric characters and the characters
period (.), underscore (_), hyphen (-), and dollar sign ($).

Command Qualifiers
/[NO]lOG
Specifies whether or not a confirmation message is displayed indicating that the
requested action was taken. The default is INOLOG.

Managing Terminals

8-11

8.11.3

SET NODE/LAT Command

Sets or changes the initial LAT host node characteristics. You must have the
SWCFG privilege to use the SET NODEILAT command. Example:
$ SET NODE/LAT/IDENTIFICATION="Welcome to LAT on EDERE"

Format
SET NODE/LAT
Command Qualifiers
IOISABLE={group_list I ALL}
IENABLE={group_list I ALL}
IIOENTIFICATION="node_id"
I[NO]LOG
IMULTICAST_TIMER=seconds

Defaults
none
none
none
INOLOG
IMULTICAST_TIMER=60

Prompts
none

Command Qualifiers
IDISABLE=group_list

Removes the specified group codes from the list of group codes associated with the
LAT host node. The group_list can include multiple codes, separated by commas.
For example, !DISABLE=(0,1,4,10) disables group codes 0,1,4, and 10. Use the
keyword ALL to indicate codes 0 to 255. Specify the !DISABLE qualifier no more
than once on a command line. No error is returned if a specified group code is not
enabled.
IENABLE=group_"st
Adds the specified group codes to the list of group codes associated with the LAT
host node. The group_list can include multiple codes, separated by commas.
For example, !ENABLE=(5,6,8,10) enables group codes 5,6,8, and 10. Use the
keyword ALL for codes 0-255. Specify the !ENABLE qualifier no more than once
on a command line. No error is returned if a specified group code is already
enabled.
IIDENTIFICATION="node_ld"

Specifies a description for the LAT host node. You can use this description as
an announcement string or to further identify the node. It may be up to 64
characters.
I[NO]LOG
Specifies whether or not a confirmation nlessage is displayed indicating that the
requested action was taken. The default :is INOLOG.
IMULTICAST_ TIMER=seconds

Specifies the interval, in seconds, between the multicast messages sent for service
announcements. The minimum value is 10 seconds, the maximum value is 255
seconds. The default is 60 seconds.

8-12

Managing Terminals

8.11.4

SET SEBVICE/LAT Command

Modifies the information associated with a particular service being offered by th
node. You need the SWCFG privilege to use the SET SERVICEILAT command.
Exatnple:
$ SET SERVICE/LAT/STATIC_RATING=40 EDERE

Format
SET SERVICE/LAT

service_name

Command Qualifiers
IIDENTIFICATION="service_id"
/[NO]LOG
/STATIC_RATING=rating

Defaults
none
INOLOG
none

Prompts
Service name:

Command Parameters
service_name
Identifies the service being modified. Service names may have as many as
16 characters, including the multinational alphanumeric characters and the
characters period (.), underscore L), hyphen (-), and dollar sign ($). Typically, th.
service name is the same as the node name.

Command Qualifiers
/IDENITIFICATION="service_ld"
Modifies the description for the service. You can use this description to further
identify the service being offered.
/[NO]LOG
Specifies whether or not a confirmation message is displayed indicating the
requested action was taken. The default is INOLOG.
/STATIC_ RATING=rating
Modifies the static rating associated with the service.

Managing Terminals

8-13

8.11.5

SET SYSTEM Command
The SET SYSTEM command sets the default system characteristics. You need
one or more privileges to use this command (see the discussion for each command
qualifier).
See Chapter 14 for additional SET SYSTI~M command qualifiers.

Format
SET SYSTEM
Command Qualifiers

Defaults

/[NO]FMS
/[NO]LAT
/PSEUDO_KEYBOARDS=n

none
none
/PSEUDO_KEYBOARDS=4

Command Qualifiers
/[NO]FMS

Enables FMS the next time the system restarts. The /NOFMS qualifier disables
FMS. You need the SWCFG privilege to use this qualifier.
/[NO]LAT

Enables LAT support the next time the system restarts. The /NOLAT qualifier
disables LAT support, saving 2K words of memory. You need the SWCFG
privilege to use this qualifier.
IPSEUDO_KEYBOARDS=n

Specifies the maximum nwnber, n, of static psuedo keyboards available on the
system the next time the system restarts. n can range from 1 to 127. You need
the SWCFG privilege to use this qualifier.

8-14

Managing Terminals

8.'11.6

SET TERMINAL Command

Sets characteristics for terminals attached locally to a RSTS/E system or for terminals connected by dialup lines. Users can use the SET TERMINAL command
to set characteristics for their own terminals. You need HWCFG privilege to
define characteristics for other terminals. Example:
$ SET TERMINAL/ADVANCED_VIDEO KB04

F()rmat
SET TERMINAL

[KB[c]nn[:]]

Command Qualifiers
I[NO] 132_COLUMNS
I[NO]ADVANCED_VIDEO
I[NO]ALT_MODE
I[NO]ANSI
I[NO]ANSWERBACK
I[NO]AUTOBAUD
I[NO]BREAK
I[NO]BROADCAST
IBUFFER_QUOTA=n
I[NO]CONTROL=[(option[,option, ... ])]
I[NO]CRFILL[=n]
I[NO]DELIMITER[=c]
I[DEVICE_TYPE=]terminal type
I[NO]DIALUP
I[NO]EIGHT_BIT
I[NO]ESCAPE_SEQUENCE
I[NO]FORM_FEED
IHARDCOPY
I[NO]HOST_SYNC
IINQUIRE
I[NO]INTERACTIVE
I[NO]KATAKANA
I[NO]LOADABLE_CHARACTERS
I[NO]LOCAL_ECHO
ILOWERCASE[=IN PUT I OUTPUT]
I[NO]PARITY[=option]
IPERMANENT
I[NO]PRINTER_PORT
I[NO]REGIS
IRESET
I[NO]RESTRICT
IRESUME=[ANY I CONTROL_C]
ISCOPE
ISELECT_ERASE
ISETLJP=filename
I[NO]SIXEL
ISPEED=(input[,output])
I[NO]TAB

Defaults
none
none
none
none
none
none
none
none
IBUFFER_QUOTA=180
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
See discussion
none
none
none
none
none
none
none
none
Managing Terminals

8-15

I[NO]TTSYNC
/TYPE=n
I[NO]UP_ARROW
IUPPERCASE[=INPUT IOUTPUT]
I[NO]USER_DEFINED_KEYS
IWIDTH=n

none
none
none
none
none
none

Prompts
none

Command Parameters
[KB[c]nn[:]]
Indicates the terminal to set the characteristics for. If you do not specify a
terminal, RSTSIE sets the the characteristics of the terminal that issued the
command. You must have HWCFG privilege to set the characteristics for another
user's terminal.

Command Qualifiers
I[NO]132_COLUMNS

If you specify 1132_COLUMNS, the terminal displays single-width characters
in an array of 24 lines by 132 columns. If you specify IN0132_COLUMNS, the
terminal displays single-width characterEl in an array of 24 lines by 80 columns.
I[NO]ADVANCED_VIDEO
If you specify IADVANCED_VIDEO, it means the terminal has the advanced

video option (AVO). The advanced features include character attributes (blinking, bolding, ... , and so on), screen memory, and ROM sockets. If you specify
INOADVANCED_VIDEO, it means the terminal does not support AVO.
I[NO]ALT_MODE
If you specify IALT_MODE, the system treats ASCII 027, 125, and 126 as

ESCAPE. If you specify INOALT_MODE, the system treats only ASCII 027 code
as ESCAPE. See /LOWERCASE for more information on the use of this qualifier.
I[NO]ANSI

If you specify IANSI, it means the terminal supports ANSI escape sequences.
If you specify INOANSI, it means the tenninal does not support ANSI escape
sequences.
I[NO]ANSWERBACK

If you specify IANSWERBACK, the terminal sends verifying identification to
electronic messaging services such as Telex or TWX. Users on dialup lines must
run the LOGIN program by pressing Return. IANSWERBACK cannot be used
with IAUTOBAUD. If you specify INOANSWERBACK, the terminal does not send
the identification message. The terminal can be used for normal sessions on the
system in either case.
I[NO]AUTOBAUD
If you specify IAUTOBAUD, the system monitors the terminal line to detect

and set baud rates without requiring the user to manually set or determine the
terminal's speed. If you specify INOAUT 0 BAUD , the system does not set the
terminal line's speed without manual intervention.

8-16

Managing Terminals

I[NO]BREAK

If you specify /BREAK, the system translates the BREAK key as a CtrllC
character (ASCII 3). If you specify INOBREAK, the system translates the
BREAK key as a null character (ASCII 0).
I[NO]BROADCAST

The /BROADCAST qualifier enables the system broadcast capability for the
terminal. That is, RSTSIE displays messages from the SEND system function call
on the terminal. The /NOBROADCAST qualifier disables the system broadcast
capability for the terminal.
IBUFFER_QUOTA=n

Indicates the number of input characters that you can buffer before the system
issues an XOFF character to your terminal to stop transmission. The buffer
quota value is a multiple of 30. The allowable values range from 180 to 7650. By
default, the system uses a quota of 180 characters. The system uses one small
buffer for each 30 characters. ~rhus, the default buffer quota of 180 characters
uses a maximum of six small buffers for terminal input. Changing a terminal's
input buffer quota is useful for high-speed input terminals., However, if there are
not enough small buffers available, a terminal may not be able to reach its full
quota. You need HWCFG privilege to use the /BUFFER_QUOTA qualifier.
I[NO]CONTROL=[(optlon[,optlon, ... ])]

If you specify ICONTROL=C, the system traps a CtrllC typed by the user and
handles it accordingly. If you specify INOCONTROL=C, the system ignores a
CtrllC typed by the user.
If you specify ICONTROL=R, the system enables the CtrllR retype facility. If you
specify /NOCONTROL=R, the system disables the CtrllR retype facility.
If you specify ICONTROL=T, the system enables the CtrllT job status display
facility. If you specify INOCONTROL=T, the system disables the CtrllT job status
display facility.
If you specify ICONTROL=X, the system traps a CtrllX typed by the user and
clears the user's entire type-ahead buffer. If you specify /NOCONTROL=X, the
system interprets a CtrllX typed by the user as data.
If you want to set or reset more than one option on a single command
line, you can combine them in a list. For example: ICONTROL=(R,T,C) or
/NOCONTROL=(C,X).
I[NO]CRFILL[=n]

If you specify ICRFILL, the system sets the carriage return fill factor to n, where
n is between 0 and 6. If the argument is not specified or is zero, or if /NOCRFILL
is specified, the system does not generate any fill characters.
I[NO]DELIMITER[=c]

If the argument is a character within quotation marks (for example
/DELIMITER="\ "), the character within quotation marks becomes the private
delimiter.
If the argument is a number (for example /DELIMITER=9), the private delimiter
is the ASCII character whose decimal value is the number. This allows for
any ASCII character whose decimal value is between 1 and 127 to be a private
delimiter.
If you specify INODELIMITER., the system removes the private delimiter that
was previously set.

Managing Terminals

8-17

I[DEVICE_TYPE=]termlnal type

Sets the default characteristics for the terminal type specified or, if it is a name
of a user-defined macro, sets the characteristics defined in the macro. This
command recognizes the following terminal types:
LA12
LA34
LA36
LA38

LA50 LA180S VT100 VT125 VT240
LA75 VK100 VT101 VT131 VT320
LA100 VT52 VT102 VT132 VT330
LA120 VT55 VT105 VT220 VT340

See the section "User-Defined Terminal Characteristics: TERMAC.SYS" for a
description and example of the user-defined macro file.
I[NO]DIALUP

Indicates if dialup control is enabled for the specified terminal.
I[NO]EIGHT_BIT
If you specify !EIGHT_BIT, it means the terminal supports eight bit characters.

If you specify INOEIGHT_BIT, it means the terminal does not support eight bit
characters.
I[NO]ESCAPE_SEQUENCE
If you specify !ESCAPE_SEQUENCE, the system treats an ESC character (value

27) as an indication of an incoming escape sequence. The system does not
echo the character and or any of the characters in the sequence. If you specify
INOESCAPE_SEQUENCE, the system treats an ESC character (value 27) as a
line terminator and echoes it as a $ character.
I[NO]FORM_FEED

The /FORM_FEED qualifier indicates that the hardware has form feed and vertical tab capabilities. The system transmits form feed and vertical tab characters
without translation. The INOFORM_FEED qualifier indicates that the hardware
does not have form feed and vertical tab capabilities. The system transmits four
line feed characters in place of a form feed or vertical tab character.
IHARDCOPY
If you specify /HARDCOPY, it means the terminal is not a cathode ray terminal

(CRT) display device. The system echoes a DEL character (RUBOUT) by printing
a backslash (\) and the last character typed and removes the last character typed
from the terminal input buffer. Later DEL characters cause the next to last
characters to be sequentially printed and removed from the terminal input buffer
until a character other than DEL is received. As a result, the system echoes
another \ to delimit the erased characters and then echoes the correct character.
If you specify !HARDCOPY and I[NO]TTSYNC is not specified, INOTTSYNC is
set by default.
I[NO]HOSTSYNC

Specify /HOSTSYNC if the terminal has special hardware that lets the computer
interrupt transmission of characters frOlln the terminal by sending the terminal
an XOFF character (value 19). Similarly, the computer instructs the terminal
to resume transmission of characters by sending the terminal an XON character
(value 17). The terminal hardware must respond to XOFF and XON characters
by stopping and resuming transmission.
Specify INOHOSTSYNC if the terminal does not have the hardware required for
the HOSTSYNC feature.

8-18

Managing Terminals

IINQUIRE

Queries the terminal by sending an ANSI ESCAPE identifying sequence, then
determines the terminal type and sets the appropriate terminal characteristics.
If the. terminal does not respond within five seconds, the system searches the file
TERDFL.SYS to see if there are default characteristics for this terminal. If an
entry its found, the system sets the specified characteristics. If not, the system
displays an error message and the terminal characteristics are unchanged.
NOTE

This qualifier causes the terminal's type-ahead buffer to be cleared.
LA36 terminals do not support this feature.
See the section "Automatic Setting of Terminal Characteristics: TERDFL.SYS"
for more information on using /INQUIRE with terminals that do not respond to
the identifying sequence.
I[NO]INTERACTIVE

Indicates that you can use the terminal to log in to the system. If you specify
fNOINTERACTIVE, the system ignores any characters input from the terminal
while it is not assigned to another job or not opened. Use this attribute for KB
devices set up as printing devices.
I[NO]KATAKANA

Indicates that the terminal supports the Katakana character set.
I[NO]L.OADABlE_CHARACTERS

Indicates that the terminal supports dynamically redefinable character sets (for
example, changing fonts).
I[NO]L.OCAl_ECHO
If you specify /LOCAL_ECHO, the system does not echo characters it receives

from the terminal. This is only used for a terminal that does its own local echo.
If you specify fNOLOCAL_ECHO, characters generated are sent only to the

computer. Therefore, the system echoes each character it receives so the charactm
is displayed locally and then translates certain characters to perform the proper
action. For example, the system echoes a CR character as a carriage return and
line feed sequence.
IlOWERCASE[=INPUT I OUTPUT]

The ILOWERCASE=INPUT qualifier indicates that the system enables lowercase
input from the terminal. The ILOWERCASE=OUTPUT qualifier indicates that
the system enables lowercase output to the terminal.
If you specify /LOWERCASE with no argument, the system enables lowercase
input and output. If you specify /LOWERCASE or /LOWERCASE=INPUT and
I[NO]ALTMODE is not specified, the system sets fNOALTMODE by default.
I[NO]PARITY=EVEN 1000

If you specify fNOPARITY, the system ignores the parity bit on characters it
receives and treats the parity bit on characters it transmits to the terminal as if
the parity bit were a data bit.
If you specify /PARITY=EVEN, the system sends characters to the terminal with
the parity bit properly set for even parity but ignores the parity bit on characters
it receives.
If you. specify /PARITY=ODD, the system sends characters to the terminal with

the parity bit properly set for odd parity but ignores the parity bit on characters
it receives.
Managing Terminals

8-H

IPERMANENT
If you specify !PERMANENT, the other qualifiers on the same command line

become the terminal's permanent characteristics. Permanent characteristics are
the characteristics of the terminal when it is not in use. The characteristics of
the terminal revert to the permanent characteristics when the user logs out.
You can use !PERMANENT to specify any characteristics of any terminal,
including dialup controlled terminals. You must use !PERMANENT with
IANSWERBACK, IAUTOBAUD, /DIALUP, and !INTERACTIVE. You must
have HWCFG privilege to specify this qualifier.
I[NO]PRINTER_PORT

Indicates that the terminal has a printer port.
I[NO]REGIS

Indicates that the terminal supports the Remote Graphic Instruction Set (ReGIS).
IRESET

Resets the terminal's characteristics to the permanent characteristics.
I[NO]RESTRICT
If you specify !RESTRICT, it means that only users with DEVICE privilege can
access the specified terminal. If you specify INORESTRICT, it means that any
users can access the specified terminal.

This qualifier only reflects OPENs and ASSIGNs of the device: not LOGINs.
IRESUME=[ANY I CONTROL_C]

Defines the XONIXOFF processing. The argument ANY enables type out and
echo when any character is typed after XOFF. ANY is valid only on LA36 and
VT52 terminals. The argument CONTROL_C enables type out and echo only
when you type XON or CtrllC after XOFF.
ISCOPE
Specify ISCOPE if the terminal is a CRT display terminal that has the following

characteristics:
•

Conforms to synchronization as described Wlder the flTSYNC command.

•

System echoes a DEL character (RUBOUT) as backspace, space, and
backspace sequence.

If you specify ISCOPE, and I[NO]TTSYNC is not specified, /TTSYNC is set by
default.
I[NO]SELECT_ERASE

Indicates if the terminal supports the selectively erasable character attribute.
If the terminal supports this feature, the terminal recognizes two forms of the
erase in line and erase in display commands. In one form all characters within
the set parameter are erased regardless of logical attributes. The second form
changes all characters designated with the selectively erasable character attribute
into spaces.
ISETUP=filespec

Sends the specified file's data to the ternlinal, in binary mode. If you specify a
file name only, the system assumes the £tIe is in the user's account on the system
disk with a file type of .ESC. The user tllust have access to the file or an error
results. Use this command to initialize a terminal for which you can make special
software settings.

8-20

Managing Terminals

I[NO]SIXEL

Indicates if the terminal supports Sixel Graphics. Sixel Graphics transfer binary
graphic images between the host and the terminal or the terminal and a printer.
ISPEED=(input[,output)]

Specifies the speed at which the terminal sends and receives data.
You can specify the speed as a single value to set the input and output baud
rates to the same speed. To specify a split speed for input and output, specify
/SPE}l~D=(n,m), where the values nand m indicate the input (terminal to host)
and output (host to terminal) baud rates, respectively.
NOTE

The /SPEED qualifier works on CXA, CXB, CXF, DH, DHF, DHQ, DHV,
DHU, DZ, DZQ, and DZV terminals. Split speed does not work on DZ,
DZQ, and DZV terminals.
You can use the keywords EXTA and EXTB to specify a baud rate associated witb
the DH11 External Clock A and External Clock B baud rates, respectively.
See the section "Terminal Line Speed Characteristics File: TERSPD .SYS " for a
description and example of the speed file.
I[NO]TAB

/TAB indicates the terminal hardware has tab capability. The system transmits
TAB characters without translation.

INOTAB indicates the terminal hardware does not have tab capability. To move
to the next tab stop, the system transmits the correct number of space characters
instead of transmitting a TAB character.
I[NO]TTSYNC

If you specify ITTSYNC, the terminal obeys the following synchronization
standard: if the terminal sends an XOFF character (equivalent to the CtrllS
combination), the computer interrupts transmission until the terminal sends
either an XON character (equivalent to the CtrllQ combination) or a CtrllC
combina tion.
If you specify INOTTSYNC, XON and XOFF characters sent by the terminal hav€
no special meaning.
ITYPE=n

Specifies the value of the terminal's type code. This type code coincides with
a part.icular terminal type and. the SHOW TERMINAL command uses it to
display the terminal type. This qualifier is best used in the TERMAC.SYS file to
associate a terminal with a user-defined macro.
Allowable values are 0 to 255. The values 0 to 128 are reserved for DIGITAL use.
I[NO]lJP_ARROW

If you specify /UP_ARROW, the system echoes a control and graphic character combination as the 1\ character (value 94) followed by the proper graphic
character. For example, CtrllE prints out as I\E.
If you specify /NOUP_ARROW, the system echoes the control and graphic
character combination as is.
IUPPERCASE[=INPUT I OUTPUT]

The /uPPERCASE=INPUT qualifier indicates that the system disables lowercase
input from the terminal. The IUPPERCASE=OUTPUT qualifier indicates that
the system disables lowercase output to the terminal.
Managing Terminals

8-21

If you specify !UPPERCASE with no arglunent, the system disables lowercase
input and output. If you specify !UPPERCASE or !UPPERCASE=INPUT and
I[NO]ALTMODE is not specified, the system sets IALTMODE by default.
/[NO]USER_DEFINED _KEYS

Indicates if the terminal supports User Defined Keys (UDKs). User Defined Keys
let you save a full command string and nlvoke it with a single key.
/WIDTH=n

Sets the width of the print line for the terminal to n. The width can be between
1 and 254. As a result, the system autOlnatically generates a carriage return/line
feed combination if n printing characters have been printed or echoed without
a carriage return/line feed combination and another printing character is to be
transmi tted.
If the terminal executing the program is in the VT100/200 family and the print
line width is changed, then the appropriate ESC sequences are sent to the
terminal to change the width of the screen accordingly. For example, if the width
changes from 80 or less to 81 or more, then the width of the screen is set to 132;
if the width changes from 81 or more to 80 or less, the width of the screen is set
to 80.

8-22

Managing Terminals

8.'11.7

SHOW COUNTERS/LAT Command

Displays information about counters relevant to managing LAT on the host node.
You need the SWCTL privilege to use the SHOW COUNTERSILAT command.
There are three types of counter information: LAT host node counters, terminal
server counters, and Ethernet portal counters. If you do not specify one of these,
the system displays information on node counters as the default. For example:
$ SHOW COUNTERS/LAT

LAT Host Node Counters
Seconds since last zeroed
Messages Transmitted
Bytes Transmitted
Messages Received
Bytes Received
Messages Retransmitted
Messages Received Out of Sequence
Invalid Messages Received
Invalid Slots Received
Illegal Messages Received
Illegal Slots Received
Resource Allocation Errors
Transmit Errors
Circuit Timeouts

:
:
:
:

882294
614796
13103390
579669
26923552
847
0
8
0
0
0
0
0
1

Format
SHOW COUNTERS/LAT

[server_name]

Command Qualifiers

Defaults

IDEVICE
IOUTPUT=output_file
INODE
/TERMINAL_SERVERS
IZERO

none
none
none
none
none

Prompts
none

Command Parameters
[server_name]
Specifies the terminal server for which counter information displays. Server
names may have as many as 16 characters, including the multinational alphanumeric characters and the characters period (.), underscore (_), hyphen (-), and
dollar sign ($). If you want to examine a terminal server, use the ITERMINAL_
SERVERS qualifier. If you do not use that qualifier or use the INODE qualifier
instead, the SHOW COUNTERSILAT command displays information on the host
node counters.

Managing Terminals

8-23

Command Qualifiers
IDEVICE

Displays the Ethernet device counters. These device counters pertain only to
the portal which Ethernet has open for LAT service. The device parameter is
optional. If you leave it out, you get information on counters for all Ethernet
portals open for LAT service.
IOUTPUT=output_file

Sends the display of the requested counters to the output file specified.
INODE

Displays the counters for the LAT host node. !NODE is the default if neither
!DEVICE nor !TERMINAL_SERVERS is specified.
ITERMINAL_SERVERS

Displays the counters for the terminal server(s) known to the LAT host node.
If a server name parameter is specified, only the counters for that server are
displayed. If no parameter is specified, then counters for all servers known to the
LAT host node are displayed.
IZERO

Zeroes the counters after they are displa.yed. This qualifier acts only on the
counter(s) that are displayed.

8-24

Managing Terminals

8.11.8

SHOW NODE/LAT Command
Displays information about the current LAT host node characteristics. The
information displayed includes the LAT version number, the node name and
identification, the multicast tinIer, and the groups to which the LAT host node is
associated. For example:
$ SHOW NODE/LAT

Node Name:
Identification:
Multicast Timer:
Group Codes:

EDERE
RSTS V9.7-08 Software at its finest
60
141 175 210 211 216

LAT is enabled on line XEO:

FClrmat
SHOW NODE/LAT
Command Qualifiers

Defaults

/OUTPUT =output_file

none

Prompts
none

Command Qualifiers
/OUTPUT =output_file
Requests that the output of this display be put into a file, and not be displayed at
the terminal.

Managing Terminals

8-25

8.11.9

SHOW SERVICE/LAT Command
Displays infonnation about the service(s) offered by the LAT host node. If you
specify a service name, only that service displays. Otherwise, all services offered
by the LAT host will be displays. For example:
$ SHOW SERVICE/LAT

Service:
EDERE
Identification: Unauthorized use prohibited
Static rating: 255

Format
SHOW SERVICES/LAT
Command Qualifiers
IOUTPUT=output_file

[service_name]
Defaults
none

Prompts
none

Command Parameters
[service_name]
Specifies the service for which infonnation is displayed. Service names may have
as many as 16 characters, including the multinational alphanumeric characters
and the characters period (.), underscore (_), hyphen (-), and dollar sign ($).
Typically, the service name is the same as the node name. If you do not specify a
service, the SHOW SERVICEILAT command displays all the services offered by
the LAT host.

Command Qualifiers
/OUTPUT =output_file
Requests that the output of this display be put into a file, and not be displayed at
the terminal.

8-26

Managing Terminals

8,,11.10

SHOW SESSIONS/LAT Command

Displays information about the sessions which are currently running LAT on
the host node. The information includes the RSTSIE keyboard number that the
user is logged in to, the terminal server that the user's terminal is physically
connected to, and the name of the service the user is currently running. For
example:
$ SHOW SESSIONS/LAT
KB43
Port: TWA30
KB46
Port: WTA4
KB48
Port: PORT 7
Port: PORT- 53
KB50
KB47
Port: PORT 68
KB44
Port: TWA62
KB45
Port: TWA63
Port: WTA6
KB54
KB55
Port: WTA9
KB49
Port: TWA87
KB51
Port: PORT 88
KB52
Port: PORT 3

Server:
Server:
Server:
Server:
Server:
Server:
Server:
Server:
Server:
Server:
Server:
Server:

BEANEY
JTSVAX
M12M12
M12M12
M12M12
BRAIDE
BRAIDE
SHERMN
SHERMN
BOTTLE
M12Mll
M12M26

Service:
Service:
Service:
Service:
Service:
Service:
Service:
Service:
Service:
Service:
Service:
Service:

ED ERE
EDERE
EDERE
EDERE
ED ERE
ED ERE
ED ERE
EDERE
EDERE
EDERE
ED ERE
EDERE

F~Drmat

SHOW SESSIONS/LAT
Command Qualifiers
IDEVICE=Ethernet_device:
lLAT
IOUTPUT=output_file

[server_name]
Defaults
none
none
none

Prompts
none

Command Parameters
[server_name]
Specifies the server for which information is displayed. Server names may have
as many as 16 characters, including. the multinational alphanumeric characters
and the characters period (.), underscore (_), hyphen (-), and dollar sign ($). If
you do not specify a server, the SHOW SESSIONSILAT command displays all th€
servers operating on the LAT host.

Command Qualifiers
IDEVICE=device:
Restricts the search to those users who are currently running LAT on the
particular Ethernet device specified.

Managing Terminals

8-27

flAT
Displays LAT sessions rather than other Elessions. ILAT is the default; SHOW
SESSIONS gives you the same display as SHOW SESSIONSILAT.
IOUTPUT=output_file
Sends the output of this display to a file instead of displaying it at the terminal.

8-28

Managing Terminals

8.11.11

SHOW TERMINAL Command
Displays the current characteristics of a specific terminal. You can change each
of these characteristics with a corresponding option of the SET TERMINAL
command. The following examples show the output format for the SHOW
TERMINAL command:

$ SHOW TERMINAL/BRIEF
Terminal: KB26:

Device_Type: VT220

Broadcast
No Delimiter
Interactive
Speed=9600

No Break
CRFill=O
Hostsync
Scope
Width=80

$ SHOW TERMINAL/FULL
Terminal: KB26:

132 Columns
No Answerback

Control=(R,T,C)
Eight_bit
Lowercase
Tab

No Control=X
No Form
Resume=Control C
TTSync

Device_Type: VT220
Advanced Video

No Alt Mode

ANSI

No Autobaud

No Break

Broadcast

No Control=X
Eight bit
Interactive
Lowercase
Restricted
No Sixel
Up_Arrow

CRFill=O
No Escape Sequence
No Katakana
No Parity
Resume=Control C
Speed=9600
User_Defined_Keys

Control=(R,T,C)
Buffer_Quota=180
No Delimiter
No Dialup
Hostsync
No Form
Loadable Characters No Local Echo
No ReGis
Printer Port
Select Erase
Scope
Tab
TTSync
Width=80
$

Fc)rmat
SHOW TERMINAL

[KB[c]nn[:]]

Command Qualifiers

IAll
IBRIEF
IFUlL
I[NO]PERMANENT

Defaults
Current Terminal

IBRIEF
IBRIEF
INOPERMANENT

Prompts
none

Command Parameters
[KB[c]nn[:]]
Indicates the terminal described in the display. When you do not specify a
terminal, the system displays the characteristics of the terminal executing the
command. You must have HWCFG privilege to specify another user's terminal.

Managing Terminals

8-29

Command Qualifiers
fAll
Displays the characteristics of all the tertninals configured on the system. If a
terminal is specified along with fALL, the terminal specification overrides fALL.
fBRIEF
If you specify !BRIEF, or neither !BRIEF nor /FULL, then a subset of the full
display is shown.
fFUll
If you specify /FULL, the full display is shown.

f[NO]PERMANENT
If you specify !PERMANENT, the system displays the permanent characteristics
of the terminal. If you specify INOPERMANENT, the system displays the current

characteristics of the terminal.

8-30

Managing Terminals

8,,11.12

SHOW TERMINAL_SERVERS/LAT Command

Displays infonnation about the tenninal server(s) known to the LAT host node.
The server name is optional; if you omit it, infonnation displays about all server!
known to the LAT. For example:
$ SHOW TERMINAL SERVERS/LAT BEANEY
Server: BEANEY
Active sessions:
1
Ethernet address:
AA-00-00-40-DO-OA
Keep alive timer:
20 seconds
Maximum buffer size:
1518 bytes
LAT Protocol:
V5.0

SHOW TERMINAL_SERVERS/LAT
Command Qualifiers
/OUTPUT=output_file

[server_name]
Defaults
none

Prompts
none

Command Parameters
[server_name]
Specifies the server for which infonnation is displayed. Server names may have
as many as 16 characters, including the multinational alphanumeric characters
and the characters period (.), underscore (_), hyphen (-), and dollar sign ($). If
you do not specify a server, the SHOW TERMINAL_SERVERSILAT command
displays all the servers operating on the LAT host.

Command Qualifiers
IOUTPUT=output_file
Requests that the output of this display be put into a file, and not be displayed a1
the tenninal.

Managing Terminals

8-31

8.11.13

START/LAT Command

The STARTILAT command enables LAT service. You can start LAT on a particular Ethernet device by specifying the device on the command line. If you specify
no device, LAT starts on all the Ethernet devices found. You must issue a SET
NODEILAT command before issuing the first STARTILAT command. You must
have the SWCTL privilege to use the STARTILAT command. For example:
$ START/LAT XEO:

Format
START/LAT

EtherneLdevice:

Command Qualifiers
I[NO]lOG

Defaults
IlOG

Prompts
Ethernet device:

Command Parameters
Ethernet_device:
Specifies the Ethernet portal to use for LAT. The device name is usually XEO: or
XHO:.

Command Qualifiers
I[NO]lOG
Specifies whether or not a confirmation message is displayed indicating whether
or not LAT has started up. The default is ILOG.

8-32

Managing Terminals

8,,11.14

STOP/LAT Command

The STOPILAT command disables LAT service on a particular Ethernet device
or on all Ethernet devices. If no Ethernet device parameter is specified, LAT is
disabled on all Ethernet devices. Jobs detach if they are currently running on
Ethernet devices that get disabled. You need the SWCTL privilege to use the
STOPILAT command. For example:
$ STOP/LAT XEO:

Format
STOP/LAT

[Ethernet_device:]

Command Qualifiers
f[NO]lOG

Defaults
flOG

Prompts
none

Command Parameters
[Ethernet_device :]
Specifies the Ethernet portal on which LAT is disabled. If you do not specify a
portal, the STOPILAT command disables LAT on all portals. Typically, there is
only one LAT portal, XEO: or XHO:.

Command Qualifiers
I[NO]LOG
Specifies whether or not a confirmation message is displayed indicating whether
or not LAT has stopped. The default is fLOG.

Managing Terminals

8-3:

Chapter 9

Managing Print/Batch Services
This chapter gives background information on the PrintlBatch Services (PBS)
package. PBS contains routines to help you manage your RSTSIE V9.0 and later
print and batch servers.
The chapter then describes the DCL commands for using PBS. Table 9-1
summarizes the DCL commands.
Table 9-1:

DCl Commands for the Print/Batch Services Package

Command

Description
Package-Related

START/QUEUEIMANAGER

Starts the PrintlBatch Services package

STOP/QUEUEIMANAGER

Shuts down the PrintlBatch Services package
Queue-Related

ASSIGN/QUEUE

Assigns a queue to a server

CLOSE/QUEUE

Closes a queue, preventing it from accepting additions
entries

DEASSIGN/QUEUE

Deassigns a queue from a server, or all servers

DELETE/QUEUE

Deletes a queue

INITIALIZE/QUEUE

Defines a print or batch queue

OPEN/QUEUE

Opens a queue, allowing entries to be placed on it

SET QUEUE

Modifies the attributes of one or more queues

SHOW QUEUE

Displays the attributes and status of one or more
queues

START/QUEUE

Starts a queue, allowing entries to be routed to server
for processing

STOP/QUEUE

Stops a queue, preventing entries from being routed t
servers for processing
Server-Related

DELETE/SERVER

Deletes a print or batch server

INITIALIZE/SERVER

Defines a print or batch server

SET SERVER

Modifies the attributes of one or more servers

(continued on next pagE:

Managing Print/Batch Services

Table 9-1 (Cont.):

DCl Commands for the Print/Batch Services Package
Server-Itelated

SHOW SERVER

Displays the attributes and status of one or more
servers

START/SERVER

Starts a print or batch server that was stopped

STOP/SERVER

Stops a print or batch server immediately or when a
specified event occurs

Entry-Related

9.1

DELETEIENTRY

Deletes one or more print or batch entries selected. by
entry-spec or entry-number

Queues a request for printing one or more files

SET ENTRY

Modifies the attributes of one or more print or batch
entries

SHOW ENTRY

Displays the attributes and status of one or more print
or batch entries

SUBMIT

Queues a request for batch processing one or more
command files

Managing the Print/Batch Services Package
The PrintlBatch Services (PBS) package provides several important services for
RSTSIE users and system managers. The major components of PBS are:
•

Queue Manager

•

Prin t Servers

•

Batch Servers

•

User Interface

The Queue Manager accepts all user requests, controls all print and batch
queues, schedules jobs to run, and createl3 new jobs as needed to process requests.
Print servers process print requests on selected printers. Both line printers (LP)
and terminals (KB) can be handled by print servers.
Batch servers control execution of submitted command procedures, using the DeL
command file processor (see RSTS / E Guide to Writing Command Procedures).
You can establish multiple batch servers to process several command procedures
at the same time.
The user interface is how an application program communicates with PBS. It
lets the user program issue requests for print and batch services. User Request
Packets (see RSTS / E Programming Manual) provide this link between PBS and
the user application programs.

9-2

Managing Print/Batch Services

9.1.1

Queue Manager
The Queue Manager module is an integral part of PBS. The term queue refers
to a list of pending print or batch requests. The term entry describes a request
on a queue. Normally, PBS adds new entries to the end or tail of a queue. PBS
processes an entry when it reaches the front or head of the queue.
PBS creates an entry on a queue in response to a user request for print or batch
services. The entry remains on a queue until its requested job is completed, or
until it is explicitly deleted by an operator or user. PBS assigns a priority to
every queue entry which establishes its position on the queue relative to other
entries. For example, PBS inserts a high-priority entry in the queue ahead of
existing lower priority entries instead of at the end.
The system file PBS$:PBS.SYS contains all the queue definitions. You define a
new queue with the INITIALIZE/QUEUE command. Once you define a queue
it remains in the file until it is explicitly deleted. You delete a queue with the
DELETE/QUEUE command.

9.1.1.1

Queue Types

PBS u.ses two types of queues:
•

Print-The PRINT command creates entries in a print queue

•

Batch-The SUBMIT command creates entries in a batch queue

You can define several queues for each type, but each queue can only have a
single type associated with it. Note that once you define a queue as a print or
batch queue, you cannot change its type.
9.1.1.2

Queue Names

Every queue must have a unique name. This name can consist of 1 to 9 characters from the set A-Z, 0-9, $, and _, and must begin with an alphanumeric
character. PBS always forces lowercase characters in queue names to uppercase. You need a trailing colon when you include the queue name in an entry
specification. Otherwise, the colon is optional.
Queue names must be unique, regardless of type. For example, if you define a
print queue named SYSTEM, then you cannot also define a batch queue with the
name SYSTEM.
9.1.1.3

Default Queues

You can define one default print queue and one default batch queue. When
a user issues a PRINT or SUBMIT command and does not specify a queue
name, PBS places the request on the appropriate default queue. You can change
default queues at any time by using the SET QUEUE command. As distributed,
PBS defines the default print queue SYS$PRINT and the default batch queue
SYS$BATCH.
You do not have to define default queues. However, if you do not, a user must
specify the queue on which to place each print or batch request.

Managing Print/Batch Services

9-3

9.1.1.4

Assigning Servers to Queues
Once you define a queue, users can submit requests to be placed on that queue.
You must also indicate which servers are to process requests on that queue. You
can define servers with the INITIALIZE/SERVER command.
In the simplest configuration, you assign each queue to a single server, and you
do not assign two queues to the same server. Figure 9-1 shows this configuration.

Figure 9-1:

One-to-One Configuration

Queue LQ

I ~ ~erver

LPO:

Queue BQ

I ~ ~erver

BAO:

The one-to-one configuration is useful when users need to have their print requests printed on a particular printer or their batch request processed by a
specific batch server. In such cases, you can choose the queue name to match its
assigned server's name, such as print queue LPO: or batch queue BAO:.
It may be useful to assign more than one server to a queue, an option which is
useful when multiple devices with the same characteristics exist. For example, at
a site with two letter-quality printers (for example, KB11: and KB12:), you could
define one print queue named LQ and assign it to servers controlling each device.
Figure 9-2 shows this configuration.

Figure 9-2:

One-to-Many Configuration

~[serverKBll: I

Queue LQ

[server KB12J

The one-to-many configuration is useful when one class of devices can process
similar jobs. When a user places a print request on the queue LQ, the first
available server assigned to that queue processes the request. To balance the use
of servers, PBS uses a round-robin method to determine which available server
to use next. That is, in deciding which server to use next, the most recently used
server is the last one PBS examines.
You can also assign more than one queue to the same server. For example,
suppose a batch queue named BAO is currently assigned to batch server BAO:,
while batch queue BA1 is assigned to batch server BA1:. If you notice that queue
BAO has many entries waiting for proceBsing while queue BA1 is empty, you can
assign queue BAO to both servers. Figure 9-3 shows this configuration.

9-4

Managing Print/Batch Services

Figure 9-3:

Many-to-One Configuration

I Queue BAO ,

Queue BA1

---.-,

I ~I

Server BAO:

Server BA1:

Note that in Figure 9-3 entries on queue BAI are processed only by server BAl:,
while entries on queue BAO are processed by either server.
The combination of a one-to-many and a many-to-one configuration is a manyto-many configuration. That is, you can assign any set of queues to any set of
servers. Thus, you have complete control over allocating resources (servers) for
processing queue entries.
You can use the SHOW QUEUEIFULL command to display the servers assigned
to a queue. You can use the SHOW SERVERIFULL command to display the
queues assigned to a server. The ASSIGN/QUEUE and DEASSIGN/QUEUE
commands let you assign and deassign each queue's servers.
9.1.1.5

QueUie Priorities

You assign each queue a maximum priority and default priority in the range 1
tQ 255. The maximum priority of a queue defines the maximum priority a user
can specify when issuing a request for that queue. The default priority of a queuE
defines the priority to assign to a request issued with no explicit priority specified.
These two parameters let you set up queues so PBS processes all requests in one
queue ahead of all requests in another queue.
Users with EXQTA privilege can submit PRINT or BATCH requests with a priority greater than the queue's maximum. See the section "Processing Requests" for
more information on processing order.
9.1.1.6

Opening and Closing Queues

Normally, a queue accepts entries even if no servers are available to process the
requests. At times, you may want to prevent PBS from placing any additional
requests on a queue. This is called closing a queue; existing entries continue
moving toward the front of the queue (if there are servers available to process
them), but the queue does not accept additional entries. You can close a queue
with the CLOSE/QUEUE command.
If a user issues a PRINT or SUBMIT command for a closed queue, PBS retums an error and rejects the request. You can open a closed queue with the
OPEN/QUEUE command.
9.1.1.7

Starting and Stopping Queues

Entries move toward the head of a queue, and PBS processes them when they
reach the queue's head. You can stop a queue with the STOP/QUEUE command
to prevent any entries on the queue from being processed. Users can place
additional entries on a stopped queue, but PBS does not process them. Any
entries currently in process when you stop a queue are not affected. You can
restart a stopped queue with the START/QUEUE command.

Managing Print/Batch Services 9-5

Note the distinction between stopping a queue and stopping a server assigned to
a queue. When you stop a queue, PBS does not process any pending entries on
that queue. When you stop a server, PBS does not process any pending entries
on any queue assigned to that server. The two cases are similar in the one-to-one
configuration. However, the two cases are quite different in a many-to-many
configuration.
For example, if you assign both queues A and B to the same server, stopping
queue A does not cause that server to become idle, since processing can still
continue from queue B. However, if you stop the server, then processing stops on
both queues.

9.1.1.8

Processing Requests
In the one-to-one or one-to-many configurations, PBS always processes the entry
at the head of the queue. However, when you assign multiple queues to the same
server, there is no implied order to processing entries; all of the entries at the
front of each queue are, by definition, eligible for servicing. PBS resolves any
conflicts that occur by checking the priority and age of the entries.
When several entries are waiting for processing, PBS chooses the one with the
highest priority. If several entries have the same priority, then PBS selects the
one with the earliest entry date and time.
Before PBS can process an entry, it must be in the :READY state. This means
that the entry:
•

Is not in an AFTER state. An entry can be given an 'after' date and time
value, which prevents the entry from being processed until the specified date
and time has passed.

•

Is not in a HOLD state. An entry can also be prevented from being processed
until an operator or the entry's owner explicitly releases it.

Once an entry is in the READY state, PBS processes it as soon as all of the
following conditions are met:

•

PBS cannot process any other READY entries of the same type but with a
higher priority

•

PBS cannot process any other READY entries of the same type and the same
priority but with an earlier entry date and time

•

A server assigned to the entry's queu.e is available to process the entry. That
is, the server is not currently processing any other entry.

•

For print entries, a server's controlled device is not currently allocated or
opened by any other job on the syste1TI.

•

For print entries, at least one device is available with the entry's required
forms installed on it.

•

For batch entries, at least one pseudo keyboard of the proper type (static or
dynamic, as appropriate to the server) is currently available to the server
(that is, not allocated to or opened by another job on the system).

When all of the above conditions are met, PBS starts a print or batch job on
behalf of the entry. PBS changes the entry's state to STARTED, and records the
name of the server processing the job in the entry. Use the SHOW ENTRYIFULL
command to display this information when an entry is in process.
Since PBS always processes print entrie!3 and batch entries by different servers,
entries of a different type never compete with each other for processing.
9-6

Managing Print/Batch Services

9.1' .1.9

Accessing Entries on a Queue

Users must also be able to locate entries in a queue. PBS provides two methods
for accessing entries in queues:
•

Entry number-When PBS adds an entry to a queue as a result of the PRIN'I
or SUBMIT command, PBS assigns a unique entry number to it. Entry
numbers are unique across all queues, regardless of type. Thus, each entry
has a specific number assigned to it, whether it is in a print queue or a batch
queue. PBS uses this number to record the entry in its queue file.

•

Entry specification-You can also access an entry by identifying some or all of
the following attributes:
The name of the entry's queue
The PPN of the entry's owner
The name of the entry
These attributes are permanently established when you create the entry; they
cannot be modified once PBS adds the entry to its queue.
Several commands (for example, DELETEIENTRY and SHOW ENTRY) let
you identify an entry or group of entries by specifying an entry specification
parameter with the command. The format of an entry specification is:
queue-name: [PPN]entry-name

Unlike entry numbers, entry specifications do not uniquely identify an entry;
several entries with the same attributes can exist at the same time. Therefore,
be careful when using entry specifications with commands to modify or delete
queue entries, since doing so can cause multiple entries to be modified or deleted
inadvertently.
9.1" 1.10

Job Limits for Queues

Each queue has maximum and default limits associated with it. For print jobs,
you can define page limits; for batch jobs, you can define both CPU and elapsed
time limits. The servers assigned to the queue enforce these limits.
This arrangement lets you set up queues that do not accept requests requiring
many resources. For example, you could assign a low maximum page limit to
a queue assigned to a server controlling a slow-speed printer, thus restricting
requests to relatively small print jobs. Similarly, a queue assigned to a batch
server running jobs at low priority could accept requests that include large CPU
or elapsed time limits.
Users with EXQTA privilege can submit print or batch requests with job limits
that exceed the maximum limits defined for the queues on which their requests
are placed.

9.1.2

Print Serveit's
Each print server must have a unique name. PBS always names print servers
after the devices they control (for example, server LPO: controls device _LPO:,
server KB45: controls device _KB45:, and so on).

Managing Print/Batch Services

9-7

Some commands permit wildcard print server names. For example:
•

*:-Means all servers

•

LP*:-Means all LP-type servers

•

KB*:-Means all KB-type servers

PBS does not allow the question mark character (ASCII 63) in print server
names.
PBS associates each entry on a print queue with a form name. This name
identifies the form that you must install on a printer before the server controlling
that device can start the job. If you issue a PRINT command and do not specify
a form name, PBS assigns the form set up as the default for the queue on which
the request is placed.
9.1.2.1

Device Ownership

A print or batch server can retain ownership of its controlled device and prevent
other jobs on the system from accessing the device. Use the /NOSHAREABLE
qualifier with the INITIALIZE/SERVER and SET SERVER commands to indicate
that the server's device should be permanently allocated to PBS. A device so
allocated will be deallocated under any of the following conditions:
•

An operator changes the server's NOSHAREABLE setting to SHAREABLE

•

The device's server is deleted

•

PBS is shut down

When PBS restarts after a shut down, it attempts to allocate any device whose
server is marked as NOSHAREABLE. PBS continues to try allocating any device
that is in use by another job on the system. PBS places such a server in the
DEVICE-WAIT state indicating that the Berver is trying to gain access to the
device.
If you initialize a server with NOSHAREABLE and its device is not available,

PBS displays a warning message. PBS allocates the device as soon as it becomes
available.
/SHAREABLE indicates that the device can be shared by other jobs on the
system, and PBS allocates it only when the server starts processing a request.
For a batch server, /SHARABLE also means the server uses a dynamic pseudo
keyboard, while /NOSHARABLE means a. static pseudo keyboard. The default is
/SHAREABLE.
9.1.2.2

Job Copies and File Copies

Each print job consists of one or more files to be printed. Qualifiers are available
with the PRINT command to specify the number of job or file copies to be printed.
The /JOB_COUNT=n command qualifier specifies the number of copies of the job
to be printed. The /COPIES=n qualifier specifies the number of copies of a file to
be printed. The default for both qualifiern is one copy.
If a print job consists of a single file, then the two qualifiers produce the same result except for the header pages (see the next section "Header Pages"). However,
if a print job consists of more than one file, then the qualifiers produce different
results.

9-8

Managing Print/Batch Services

For example, if you issue the following command:
$ PRINT/COPIES=2 FILE1.LST,FILE2.LST

the order of the job output is:
Job header (1 of 1)
File header (1 of 2)
FILE1.LST
File end
File header (2 of 2)
FILE1.LST
File end
File header (1 of 2)
FILE2.LST
File end
File header (2 of 2)
FILE2.LST
File end

But, if you issue the following command:
$ PRINT/JOB_COUNT=2 FILE1.LST,FILE2.LST

the order of the job output is:
Job header (1 of 2)
File header (1 of 1)
FILE1.LST
File end
File header (1 of 1)
FILE2.LST
File end
Job header (2 of 2)
File header (1 of 1)
FILE1.LST
File end
File header (1 of 1)
FILE2"LST
File end

9.1.2.3

Header Pages

You can print special job and file header pages to identify the start of a job or a
file within a stack of listings. Header pages include large block letters to help you
locate the beginning of job and file copies. You can control the number of job or
file header pages printed by the JOB_PAGES=n and FLAG_PAGES=n attributes
of the printer form.
Header pages consist of fields printed in single-height and double-height block
letters and other information printed in six repeated lines. Three lines are
printed at the top and bottom of each header page. These lines help you find the
start of each job in a large listing.
The job header pages contain the following information:
•

Owner's account name-This field is in single-height block letters

•

Entry name-This field is in double-height block letters

•

Owner's PPN-This field is in single-height block letters

•

The three lines at the top and bottom of each job header page contain the
following:
Current date
Current time

Managing Print/Batch Services 9-9

Entry entry-spec
Copy nnn of mmm
Printer device-name
Form form-name
Installation-name
The file header pages contain the following information:
•

Device name and PPN-This field is in single-height block letters

•

Filename-This field is in double-height block letters

•

File type-This field is in double-height block letters

•

Owner's PPN-This field is in single-height block letters

•

The three lines at the top and bottoml of each file header page contain the
following:
Current date
Current time
File filespec
Copy nnn of mmm
Printer device-name
Form form-name
Installation-name

9.1.2.4

Page Limits

You can limit each print job to a specified maximum number of pages. The page
limit can be:
•

User specified

•

The default for the queue

•

Unlimited

Print servers count the number of pages they print. When a print job exceeds its
page limit, PBS prints an error on the last page of the job listing and terminates
the job. Note that header pages do not count toward a job page limit.
When the user specifies the page limit, the value can not exceed the maximum
limit for the queue. If the user does not specify the page limit, PBS uses the
default limit of the queue.
9.1.2.5

Control Characters and Escape Sequences

Some printers, such as the LNOl, LN03, and LA2xx, can understand escape
sequences. In some cases, the action taken in response to a particular escape
sequence is device dependent. Therefore, you must route files that contain escape
sequences to the correct server. This is best done by providing queues that assign
jobs only to single servers or to servers that all serve the same type of printer.

9-10

Managing Print/Batch Services

The /CONTROLS qualifier, provided with the INITIALIZE/SERVER and SET
SERVER commands, indicates how the server should process control characters
and escape sequences:
•

ICONTROLS-Indicates that the server sends control characters and escape
sequences to its controlled device. The action taken is dependent on the
device's handling of such characters. The print server does not update the
fonn's current horizontal and vertical position when sending such characters
to the printer.

•

ICONTROLS=UP_ARROW-Indicates that the server translates control
characters to an 'up arrow' format, consisting of a circumflex (A.) character
followed by a printable character that identifies the control character, and
that the server translates the ESC character to the dollar sign ($), causing
PBS to display all of the characters in the escape sequence. This is the
default.

•

INOCONTROLS-Indicates that the server does not process control characters or escape sequences. PBS discards any such characters.
NOTE

Control characters and escape sequences can cause a print server to
lose track of line and page positioning. This is because print servers
do not know what action is taken by the printer in response to such
characters. Normally, form feed characters correct positioning within
a file unless the server is doing simulated form feed handling. In such
cases, it may be necessary manually to adjust the printer alignment to
correct positioning errors introduced by control characters and escape
sequences.
DIGITAL recommends that servers processing control characters always use
forms with the NOSIMULATE attribute, to prevent positioning errors between
jobs.
9.1.2.6

Eight-Bit Characters

PBS passes eight-bit characters unchanged to printers designated as eight-bit
devices with the SET PRINTERIEIGHT_BIT or SET TERMINALIEIGHT_BIT
commands.
For seven-bit devices, PBS converts eight-bit control characters to their seven-bit
equivalents. For example, <CSI> becomes <ESC>[. PBS does not handle eight-bii
printable characters; they are truncated to seven bits as they pass to seven-bit
devices.

9. "1.3

Batch Servers
Batch servers are named BAn:, where n is a unit number in the range 0-31.
Some commands permit wildcard batch server names, for example:
•

*:-Means all servers

•

BA*:-Means all batch servers

PBS does not allow the question mark character (ASCII 63) in batch server
names.

Managing Print/Batch Services

9-11

9.1.3.1

Executing Command Flies

Batch servers use the DCL command file processor on pseudo keyboards to create
the batch job environment. Batch servers execute a batch job by creating a
temporary command file. This file:
•

Opens a log file (unless the you specify /NOLOG_FILE with the SUBMIT
command)

•

Ini tializes error handling

•

Executes LOGIN.COM files

•

Turns VERIFY on

•

Executes the individual user command files

PBS passes the parameters you specify with the SUBMIT command only to the
first command file in the batch job.
Because batch servers create their own command procedure to execute the user's
command files, the number of levels of command nesting that PBS allows is
one less than the levels available when executing command procedures at the
interactive level (see RSTS / E Guide to liVriting Com,mand Procedures).
The exit status that PBS reports in the batch log file is the status of the most
severe error occurring in the job, not just the last one. The format of the exit
status display is:
Exit status = <message>

Table 9-2 lists the possible exit status messages.
Table 9-2:

PBS Batch Job Exit Status Messages

Message

Meaning

Entry deleted

Job was deleted during processing

Error

An error message occurred in the job, but no severe errors
occurred

9.1.3.2

Exceeded CPU limit

Job exceeded the CPU time limit

Exceeded time limit

Job exceeded the time limit

Severe error

A severe error message occurred in the job

Success

No errors or warnings encountered

Unknown

Job logged out or was killed (exit status is unknown)

Warning

A warning message occurred in the job, but no errors occurred

CPU and Time Limits

Batch servers also monitor their batch jobs to enforce CPU and time limits and
determine when the job is done. An active batch server examines its controlled
job's CPU and time usage approximately every 30 seconds. If either value exceeds
the allowable limit for the batch job, PBS terminates the job immediately. If you
specify a log file with the SUBMIT command, PBS includes an error message in
the log describing the reason for the job termination.

9-12

Managing PrintlBatch Services

~t.1.4

Managing Forms for PBS
Printer forms are the physical paper you install in a line printer. In some cases:
forms are standard stock paper as normally used for printing program listings. ]
other cases, forms are customized, preprinted paper, such as checks or invoices.
PBS identifies each form by name. A form name consists of 1-9 characters from
the set {A-Z, 0-9, $, _}. PBS converts lowercase characters to uppercase.
When you define a print server, you can specify the name for the form currently
installed on the server's print device. Each form definition includes the form
handling characteristics for the printer to which it applies. A Forms Definition
File (FDF), maintains all the form characteristics for each server in PBS (see th.
next section).
When you issue a print request, you can specify the form name required using
the IFORMS=form-name qualifier. If you do not include a form name, PBS
assigns the queue's default form name to the print request. PBS verifies that th.
specified form name is in the FDF. If not, PBS returns an error and rejects the '
print request.
Before PBS can start a print request, a server assigned to the request's queue
must be available with the required form installed on its printer. If not, the
request remains READY until you change forms or until you assign a new serve]
with the correct form name to the request's queue.
You can define the same form name differently for different printers. For
example, NORMAL can have different characteristics for the printer LPO:
from those it has for LP1:. Generally, the physical dimensions of a form page
(length and width) shou1d rernain the same from one printer to another, but the
attributes that describe how a printer shou1d handle the form may be different,
depending on the device.

9."1.4.1

Forms Attributes

PBS defines the following attributes for printer forms:
•

Form name--A unique 1 to 9 character name that identifies the form.

•

Server name--The name of the server to which the form definition applies.
This can refer to a single server or multiple servers of the same type (LP or
KB).

•

Length-The physical length of a form page, measured in lines of print. This
value can be in the range 1 to 255.

•

Width-The width of a form page, measured in characters printed across the
page. The value can be in the range 1 to 255.

•

Job pages-The number of header pages to print at the start of each new
print job. The value can be in the range 0 to 127.

•

Flag pages-The number of flag pages to print at the start of each new file
within a print job. The value can be in the range 0 to 127.

•

Form-feed handling-Depending on the form page length and the hardware
capability, the printer mayor may not be able to correctly handle FF characters that position the form page at the top of the next page. If not, then the
server must simu1ate the form feed action by outputting the proper number of
line feeds to reach the top of the next page.

Managing Print/Batch Services

9-13

9.1.4.2

•

Bottom margin-The number of lines to skip at the bottom of each page. The
value must be less than the form page length.

•

Trailer pages-Determines whether or not a single trailer page prints per file.

The Forms Definition File

PBS provides a Forms Definition File (FDF) to maintain all printer form definitions. This file is a standard ASCII strealln file, allowing anyone with write access
to add, delete, and update the form definitions using a standard text editor. FDF
defines form attributes by using keywords (for example, LENGTH=66), to simplify
definitions further.
PBS treats lines beginning with an exclamation point (I) as comment lines. Thus,
you can include descriptions in the definitions file.
PBS also supports continuation lines. If a definition ends with a hyphen (-),
the definition is continued on the next line similar to the DCL command line
continuation feature (see the RSTS / E System User's Guide).
NOTE

If a line to be continued contains a comment, the hyphen must precede
the comment (see the next section "Forms Definition Keywords").

Blank lines are ignored. So, you can include blank lines where you
want to separate definitions for readability.
The name of the Forms Definition File supplied on the RSTS/E distribution kit is
PBS$:FORMS.SYS. As distributed, this file contains two definitions for the form
named NORMAL: one is for LP-class printers and one is for KB-class printers.
The file has a protection code of <40>, allowing any user on the system to read it;
however, only users with WWRITE privil'ege can write to it.
The default print queue SYS$PRINT is initially set up with NORMAL as its default form name. Mter you install PBS, you can change any of these parameters.
You can add additional definitions to the file as needed.
The order of form definitions in the FDF is significant. If you initialize a print
queue and you do not include the !FORMS qualifier, PBS defaults to the first form
definition. Similarly, if you define a print server and do not include the !FORMS
qualifier, PBS defaults to the first form definition that applies to that server.
PBS accesses the FDF for the following reasons:
•

When you include a !FORMS qualifier with the PRINT command, PBS
searches the FDF for the specified fonn name.

•

When you issue a INITIALIZE/SERVER or SET SERVER command that
includes a !FORMS qualifier for a print server, PBS searches the FDF for the
definition of the specified form name.

•

When you issue a INITIALIZE/QUEUE or SET QUEUE command that
includes a !FORMS qualifier for a print queue, PBS searches the FDF for the
specified form name.

If the form name is not found, then PBS returns an error and rejects the request.

9-14

Managing Print/Batch Services

9.11.4.3

Forms Definition Keywords

A fonns definition consists of:
•

Form name

•

Server name

•

One or more keywords that describe the form attributes

You can use spaces or tabs to separate the keywords within a form definition.
Any standard RSTSIE line terminator (CR, LF, FF) terminates each definition.
For example, the definition of the default form named NORMAL for LP-class
servers is:
, form NORMAL for all LP servers
NORMAL= LP*:LENGTH=6666 lines per page
WIDTH=132132 columns per line
JOB P.A.GES=22 job header pages
FLAG PAGES=22 flag header pages
BOTTOM MARGIN=66 blank lines at the bottom of $ach page
trailer pages
TRAILER PAGESNOSIMULATE
hardware top-of-form

If a forms definition applies to a particular server, or class of servers, the server
name follows the form name and is separated from it by an equal sign. The
format of the server name in a forms definition is:
form-name[ =server-name[:]]
PBS does not permit embedded spaces or tabs between the form name and the
server name.
The server-name identifies the server(s) to which this definition applies. If the
definition only applies to a single server, then specify the explicit server name. If
the definition applies to a class of servers (all LP or all KB devices), then specify
form-name=_LP*: or form-name=_KB*: (wildcard unit number). If the definition
applies to all print servers, then specify form-name=*: or omit this field from the
definition. The default is all servers.
NOTE

If a server belongs to more than one definition for the same form, place
definitions that apply to specific servers before any that apply to a
general class of servers. PBS searches the FDF sequentially until it
finds the first occurrence of the requested form name with a matching
server name.

You can abbreviate all forms definition keywords to their first two characters.
Note that the form name always appears first in the definition and has no
associated keyword. All other fields in the definition can be in any order. If a
keyword is not included in a form definition, then PBS uses the keyword's default
value.

Managing Print/Batch Services

9-15

You can include the following keywords in a forms definition:
LENGTH=n

Indicates the number of lines per printed page for the specified form. The value n
can range from 1 to 255. The default length is 66; this is standard spacing for an
8-1/2 by 11-inch form with vertical spacing of 6 lines per inch.
NOTE

Some devices that print on single-sheet forms cannot print on all
physical lines of the form. For exarrtple, the LNOl prints a maximum
of 60 lines per page. In such cases, the form length should be set to the
device's maximum print lines; not the form's physical size.
WIDTH=n

Indicates the maximum number of characters that can appear in a printed line
for the specified form. The value n can range from 1 to 255. The default width is
132. The WIDTH value determines where a printed line will be truncated (if the
file is being printed with the /TRUNCATl~ qualifier of the DCL PRINT command
in effect) or wrapped around to a new line (if the file is being printed with the
fNOTRUNCATE qualifier in effect).
[NO]SIMULATE

Determines whether PBS should, on encountering a FF character or reaching
the end of a page, send a FF character directly to the device (NOSIMULATE), or
simulate the form feed by transmitting the proper number of line-feed characters
(SIMULATE). The default is SIMULATE.
JOB_PAGES=n

Defines the number of job header pages to be printed at the beginning of each
job listing. The value n can range from 0 to 127. Job header pages display
information about the job (job name, PPN, and so forth), printed in large block
letters. Job header pages make it easy to separate job listings. The default is
JOB_PAGES=1.
FLAG_PAGES=n

Defines the number of file header pages to be printed at the beginning of each
file listing. The value of n can range frOlTI 0 to 127. File header pages display
information about the file (file name, PPN, and so forth), printed in large block
letters. File header pages make it easy to separate printouts of files within a job.
The default is FLAG_PAGES=1.
NOTE

You can suppress the printing of file header pages by including the
fNOFLAG_PAGES qualifier with the file to be printed.
BOTTOM_MARGIN=n

Defines the number of lines to skip at the bottom of each page of the form. The
bottom margin value cannot exceed the form's length. The default is

9-16

Managing Print/Batch Services

BOTTOM_MARGIN=6. Printers that use single-sheet paper (for example, the
LN01) automatically end a page and start a new page when it reaches its maximum. print line. Forms definitions for these devices should normally specify
BOTTOM_MARGIN=O.
NOTE

You can suppress the printing of the bottom margin by including the
INOFEED qualifier with the file to be printed. This has no effect when
BOTTOM_MARGIN=O.
TRAILER_PAGES
Determines whether or not a single trailer page prints after each file.

9:1.4.4

Changing Forms
Whenever you need to change forms on a printer under the control of PBS,
perform the following steps:
1.

Use the STOP/SERVERlJOB_END command to stop the server at the end of
any current print job.

Use the SHOW SERVER command to confirm that the server is stopped and
is not processing any print job.

Install the desired form in the printer.

Use the SET SERVERIFORMS=form-name command to assign a new form
name to the server. Use the SHOW SERVERIFULL command to inspect the
new form attributes (LENGTH, WIDTH, and so on) as needed.

Use the START/SERVER command to restart the server. When you issue
this command, PBS routes any jobs that are waiting for the new form to be
installed to the server.

Note that you can issue a SET SERVERIFORMS=form-name command without
stopping the server, although DIGITAL does not recommended this unless the
new form definition applies to the same physical form currently installed in the
printer. Any print job in progress at the time its print server is assigned a new
form name is not affected. The form name last assigned to the server determine I
the next job to be processed by the server. Use the SHOW SERVER command to
display the name of the form most recently assigned.
If you modify the FDF to change the definition of a form name which is currentl:
assigned to a server, you must perform a form change (using the same form namE
to cause the new form definition to be read from the FDF.

9.2 Multiple Jobs
PBS consists of a single primary job and, optionally, one or more secondary jobs,
depending on the nmnber and use of concurrent print or batch servers. PBS is
designed to provide a compact and efficient package across all of its jobs. This
section provides some background information on how the PBS package makes
use of multiple jobs.

Managing Print/Batch Services

9-1

9.2.1

The Primary Job
Whenever you issue the START/QUEUE/MANAGER command, PBS creates its
primary job. This job runs detached. It contains the queue manager module and
is always present when the package is running.
The primary job contains two print and two batch server slots. A slot refers to
a sub-job in the PBS job that provides print or batch services. If the number of
slots available to support the active print or batch servers in the primary job is
insufficient, PBS creates secondary jobs.

9.2.2

Secondary Jobs
PBS creates secondary jobs when there are no print or batch server slots available
in the primary job to process a waiting request. PBS attempts to activate only as
many servers as necessary, depending on the number of waiting requests ready to
run, available devices, and so on. Each secondary job contains three print server
slots and four batch server slots.
When PBS starts, only the primary job is running. When a pending request is
ready to start, the queue manager first determines if a slot is available on which
to activate a server for the request. If the primary job contains no available
server slots of the correct type (print or batch), then PBS automatically creates a
secondary job.
Unlike the primary job, a secondary job remains present only while one or more
of its server slots is allocated to a server. Once all the server slots in a secondary
job become free (that is, there are no active servers in the job) PBS automatically
removes the job from the system. It is possible for several PBS secondary jobs to
be present on the system at the same tiIne, depending on the number of active
servers that you require at any time.
NOTE

PBS secondary jobs contain only print and batch server slots; they do
not contain a queue manager module.
Based on the configuration of PBS, the primary job is sufficient to handle two
concurrent print jobs and two concurrent batch jobs. If more than two print or
two batch servers are active at the same time, then PBS creates at least one
secondary job. Each secondary job allows for three more print jobs and four more
batch jobs to be started. Thus, if your printing volume causes six print devices
to be active concurrently, you need a total of three RSTSIE jobs (one primary and
two secondary) to run PBS.
PBS uses an "idle job timer" to determine when to remove a secondary job from
the system. PBS removes a secondary job when all of its servers have been
idle for five minutes. This time is sufficient for PBS to start a server's next job
without having to recreate a secondary job.
PBS creates secondary jobs as needed, as long as job slots are available on the
system. If none are available, then no secondary job is created, and PBS waits
until either a job slot becomes available, or an existing server slot becomes free.
You can limit the number of secondary jobs required by controlling the number
of servers that can process requests. You can do this by deassigning or stopping
one or more servers. Idle servers that are stopped or not assigned to any queue
cannot process a request. Thus, they do not require a server slot.

9-18

Managing Print/Batch Services

9,.3 DCl Command Descriptions
The following sections contain descriptions of the PBS commands that require
special privileges. See the RSTS / E System User's Guide for descriptions of
PRINT, SUBMIT, DELETEIENTRY, SET ENTRY, SHOW ENTRY, and SHOW
QUEUE commands.

Managing Print/Batch Services

9-19

9.3.1

ASSIGN/QUEUE Command

The ASSIGN/QUEUE command assigns a server to a queue. Server assignments are cumulative; any server assignEld to a queue is added to the set of
assigned servers. A server named BAn: can only be assigned to a batch queue.
Likewise, a server named LPn: or KBn: ean only be assigned to a print queue.
ASSIGN/QUEUE command requires PBSCTL privilege.

Format
ASSIGN/QUEUE

queue-name[:] server-name[:]

Command Qualifiers
none

Defaults
none

Prompts
Queue: queue-name[:]
Server: server-name[:]

Command Parameters
queue-name[ :]
Specifies the name of the queue to be assigned. This command does not pennit
wildcard queue names.
server-name[ :]
Specifies the server to assign to the queue. This command does not permit
wildcard server names.

9-20

Managing PrintlBatch Services

9.~1.2

CLOSE/QUEUE Command

This command closes a queue; that is, it prevents any additional requests from
being placed on the queue. If the specified queue is not open, RSTSIE ignores this
command. CLOSE/QUEUE requires PBSCTL privilege.

Format
CLOSE/QUEUE

queue-name[:]

Command Qualifiers
none

Defaults
none

Promp'ls
Queue: queue-name[:]

Command Parameters
queue-name[ :]
Specifies the name of the queue to be closed. This command does not permit
wildcard queue names.

Managing Print/Batch Services

9-21

9.3.3

DEASSIGN/QUEUE Command

The DEASSIGN/QUEUE command deassigns a server from a queue. Any job
already started on a deassigned server if!! not affected. However, no additional
jobs are started on the server from the queue. DEASSIGN/QUEUE requires
PBSCTL privilege.

Format
DEASSIGN/QUEUE

queue-name[:] server-namef.1

Command Qualifiers
/All

Defaults
See Discussion

Prompts
Queue: queue-name[:]
Server: server-name[:]

Command Parameters
queue-name[:]
Specifies the name of the queue to be deassigned. This command does not permit
wildcard queue names.
server-name[:]
Specifies the name of the server to be deassigned. This parameter overrides the
/ALL qualifier. This command does not permit wildcard server names.

Command Qualifiers
fAll
Indicates that all servers currently assigned to the queue should be deassigned.
If you specify /ALL, PBS does not prompt for a server name. If you specify a
server name, PBS ignores /ALL.

9-22

Managing Print/Batch Services

9.:3.4

DELETE/QUEUE Command
The DELETE/QUEUE command deletes a queue. If you issue DELETE/QUEUE
for a queue that has one or more entries, RSTS/E marks the queue for deletion
and closes it. PBS deletes the queue as soon as it becomes empty. You can
still display queues marked for deletion using the SHOW QUEUE command.
DELETE/QUEUE automatically deassigns all servers from the deleted queue.
DELETE/QUEUE requires SWCFG privilege.

F()rmat
DELETE/QUEUE

queue-name[:]

Command Qualifiers
none

Defaults
none

Prompts
Queue: queue-name[:]

Comrnand Parameters
queue,·name[:]
Specifi.es the name of the queue to be deleted. This command does not permit
wildcard queue names.

Managing Print/Batch Services

9-23

9.3.5

DELETE/SERVER Command

This command deletes a print or batch se·rver. If the server you specify is cur··
rently processing a job when you issue the command, RSTSIE marks it for deletion. PBS deletes the server when it beco:mes idle. DELETE/SERVER automatically deassigns the deleted server from alll assigned queues. DELETE/SERVER
requires SWCFG privilege.

Format
DELETE/SERVER

server-name[:]

Command Qualifiers
none

Defaults
none

Prompts
Server: server-name[:]

Command Parameters
server-name[:]
Specifies the name of the server to be deleted. This command does not permit
wildcard server names.

9-24

Managing PrintlBatch Services

9.3,,6

INITIALIZE/QUEUE Command

][NITIALIZE/QUEUE defines a print or batch queue and establishes its initial
properties. Use the SHOW QUEUE command to display the queues that are
currently defined. See the RSTS / E System User's Guide for details on the
SHOW QUEUE command.
When you initialize a queue, you also define its characteristics. Print and batch
queues have different characteristics. You can modify any of a queue's characteristics later by using the SET QUEUE command.
Note that certain qualifiers for this command apply only to print queues, while
others apply only to batch queues. An error results if you include both types of
qualifiers in the same command. PBS uses the following criteria to determine
whether a queue should be initialized as a print queue or a batch queue:
,t
,t

If you give a !PRINT or !BATCH qualifier, it determines the queue's type.
If you do not give either !PRINT or !BATCH, but you specify a print queue
only qualifier (for example, !PAGE_LIMIT) or a batch queue only qualifier (for
exalmple, /CPU_LIMIT), that qualifier determines the queue's type.
If you do not meet either of the previous conditions, the queue's type defaults
to a print queue.

PBS retains queue information in the queue file when the package is shut down.
The system uses this information to automatically reinitialize queues when it
restarts the package. INITIALIZE/QUEUE requires SWCFG privilege.

Format
INITIALIZE/QUEUE

queue-name[:]

Command Qualifiers

Defaults

IBATCH
/CLOSE
ICPU_LlMIT=(MAXIMUM=a,DEFAULT =b)
IDEFAULT
IFORMS[=form-name]
IPAGE_LlMIT=(MAXIMUM=a,DEFAULT=b)
IPRINT
IPRIORITY=(MAXIMUM=a,DEFAULT=b)
IPRIVILEGES=("priv[, ... ]")
!TIME_LIMIT=(MAXIMUM=a,DEFAULT=b)

See Discussion
See Discussion
See Discussion
See Discussion
See Discussion
See Discussion
See Discussion
See Discussion
IPRIVILEGES=NONE
See Discussion

Prompts
Queue: queue-name[:]

Managing Print/Batch Services 9-25

Command Parameters
queue-name[ :]

Specifies the name of the queue to be initialized. This command does not permit
wildcard queue names.

Command Qualifiers
IBATCH

Indicates that the queue is a batch queue, and only accepts entries with the
SUBMIT command.
ICLOSE

Indicates that the queue should initially be closed, preventing any PRINT or
SUBMIT entries from being entered on the queue. By default, PBS initializes the
queue as open.
ICPU_LlMIT=(MAXIMUM=a,DEFAULT=b)
ICPU_LlMIT=n

Batch queues only. Indicates the maximum and default CPU time limits (in
minutes) for each command file processed in a batch job started from this queue.
The maximum CPU limit is the largest value that you can specify when issuing
a request for the queue. The default CPU limit is the value that PBS assigns
to a request that had no limit specified. The keyword UNLIMITED indicates
that PBS does not enforce a CPU limit. If you do not specify the MAXIMUM
argument, then PBS uses MAXlMUM=UNLIMITED. If you do not specify the
DEFAULT argwnent, then PBS uses the maximum CPU limit as the default
limit.
If you use the /CPU_LIMIT=n form, then PBS uses the value you specify for both
the maximum and default limits. If you specify a numeric argument, it must be
in the range 1 to 65,535. The default CPU limit cannot exceed the maximum
CPU limit.
If you issue a SUBMIT command without an explicit CPU limit, then PBS uses
the queue default CPU limit. When you issue a SUBMIT command, you can
specify a CPU limit up to, but not exceed.ing the queue maximum CPU limit.
IDEFAULT

Indicates that the queue should be marked as the default queue for its type (print
or batch). You can define only one print queue and one batch queue as a default
queue.
If you specify /DEFAULT, then the queu.e becomes the default queue for its type,
and any queue currently defined as the default queue of the same type no longer
serves as the default queue. By default, PBS does not mark the queue as a
default queue.
IFORMS=form-name

Print queues only. Indicates if a default form name is defined for the queue.
If you specify /FORMS=form-name, then PBS defines the name you specify as
the default form name for the queue. If you issue a PRINT command without
a /FORMS qualifier, then the queue's default form name is the first form name
found in the FDF.

9-26

Managing Print/Batch Services

IPAGE_LlMIT=(MAXIMUM=a,DEFAULT=b)
IPAGE_LlMIT=n

Print queues only. Indicates the maximum and default page limits for each file
printed as part of a print job started from this queue. The maximum page limit
is the Xargest value that you can speeify when issuing a PRINT request for this
queue. The default page limit is the value assigned to a request for this queue
tha t had no limit specified.
The keyword UNLIMITED indicates that no page limit is enforced. If you do not
specify the MAXIMUM argument, then PBS uses MAXIMUM=UNLIMITED. If
you do not specify the DEFAULT argument, then PBS uses the maximum page
limit as the default limit.
If you use the IPAGE_LIMIT=n form~ then PBS uses the value you specify for
both the maximum and default limits. If you specify a numeric argument, it must
be in the range 1 to 4,294,967,295. The default page limit can not exceed the
maximum page limit.
If you issue a PRINT command without an explicit page limit, then PBS uses the
queue default page limit. When you issue a PRINT command, you can specify a
page limit up to, but not exceeding, the maximum page limit of the queue.
IPRINT

Indicates that the queue is a print queue and only accepts entries from the
PRINT command.
IPRIORITV=(MAXIMUM=a,DEFAULT=l»
IPRIORITY=n

Defines the maximum and default priorities for the queue, in the range 1 to 255.
If you do not specify the MAXIMUM argument, then PBS uses 255. If you do
not specify the DEFAULT argument, then PBS uses the maximum value as the
default priority. If you use the IPRIOJRITY=n form, then PBS uses the value you
specify for both the maximum and default priority. The default priority cannot
exceed the maximum priority.
/":RIVILEG ES=(prlv, ...)

Specifies the set of privileges required for a user to place entries on the queue.
A user who issues a PRINT or SUBMIT command must have all of the queue's
privileges for the request to be accepted. By default, a user does not need any
privileges to place an entry on the queue.
The argument can be any of the RSTS/E privilege keywords, plus the keywords
ALL and NONE. The !PRIVILEGES=.ALL qualifier assigns all defined privileges
to the queue; !PRIVILEGES=NONE assigns no privileges to the queue. Note that
you can specify any privilege name with !PRIVILEGES, even those that you do
not possess.
The argument can also be USERn, where n is an integer between 1 and 8. You
can use these USERn privileges to identify particular users or groups of users
and to restrict the queue's usage to those users.
ITIME_LlMIT=(MAXIMUM=a,DEFAULT::b)
ITIME_LlMIT=n

Batch queues only. Indicates the maximum and default elapsed time limits (in
minutes) for each command file processed in a batch job started from this queue.
l'he maximum time limit is the largest value that you can specify when issuing a
request for this queue. The default tilne limit is the value that PBS assigns to a
request that had no limit specified.

Managing Print/Batch Services 9-27

The keyword UNLIMITED indicates that PBS does not enforce a time
limit. If you do not specify the MAXIMUM argument, then PBS uses
MAXIMUM=UNLIMITED. If you do not specify the DEFAULT argument,
then PBS uses the maximum time limit as the default limit. If you use the
ITIME_LIMIT=n form, then PBS uses the value you specify for both the maximum and default limits. If you specify a numeric argument, it must be in the
range 1 to 65,535. The default time limit cannot exceed the maximum time limit
of the queue.
If you issue a SUBMIT command without an explicit time limit, then PBS uses
the queue default time limit. When you issue a SUBMIT command, you can
specify a time limit up to, but not exceeding, the maximum time limit of the
queue.

9-28

Managing Print/Batch Services

9.3.7

INITIALIZE/SERVER Command

The INITIALIZE/SERVER command defines a print or batch server. Use the
SHOW SERVER command to display the servers that are currently defined. See
the description of the SHOW SERv]~R command in this chapter for details.
When you initialize a server, you also define its characteristics. Print and batch
servers have different characteristics. You can modify any of a server's character·
istics later by using the SET SERVE~R command.
PBS retains the server information in the queue file when the package is shut
down. The system uses this server information to initialize servers automatically
when you restart the package. This eliminates the need to reinitialize servers
each time you start PBS. INITIALIZE/SERVER requires SWCFG privilege.

Format
INITIALIZE/SERVER

server-namef:]

Command Qualifiers
I[NO]CONTROLS[=UP~RROW]
IFORMS[=form-name]
IPRIORITY=n
IRUN_BURST=n
I[NO]SHAREABLE

Defaults
See Discussion
See Discussion
See Discussion
See Discussion
ISHAREABLE

Prompts
Server: server-name[:]

Prompts
Command Parameters
server-name[:]
Specifies the name of the server to be initialized. This command does not permit
wildcard server names.

Command Qualifiers
I[NO]CONTROLS[=UP_ARROW]
Print servers only. Indicates how print servers should process control characters
and escape sequences. The default is /CONTROLS=UP_ARROW (see the section
"Control Characters and Escape Sequences").

You must have PBSCTL privilege to specify this qualifier.
IFORMS[=form-name]
Print servers only. Identifies the form currently installed on the printer. The
server only prints requests with thiB form name. PBS displays an error message
if the form name does not exist in the FDF.

If you omit this qualifier, then PBS initializes the printer with the first form
definition found in the FDF that applies to the specified print server. You can
la ter issue a SET SERVER co:mmand to assign a different form to the server.

Managing Print/Batch Services

9-2!

IPRIORITY=n

Batch servers only. Indicates the priority at which to run command procedures
controlled by this server. RSTSIE permits any value in the range -120 to +120;
however, batch jobs cannot exceed the priority of the PBS primary job. RSTSIE
rounds the priority down to a multiple of 8. The default is -8.
You must have TUNE privilege to specify this qualifier.
IRUN_BURST=n

Batch servers only. Indicates the run burst to assign to any command procedures
controlled by this server. RSTSIE permits any value in the range 1 to 127. The
default is 6.
You must have the TUNE privilege to specify this qualifier.
I[NO]SHAREABLE

Specifies if the server's device should be shared with other jobs on the system.
The default is /SHAREABLE.
If you specify INOSHAREABLE, RSTSIE immediately allocates the device to
the PBS job, thereby preventing other jobs on the system from gaining access
to the device. For batch servers, PBS allocates the next available static pseudo
keyboard to the server. The device remains allocated to the PBS job until you
delete its server (with the DELETE/SERVER command), or you shut down PBS
(with the STOP/QUEUEIMANAGER comlnand). You get a warning message if
you specify INOSHAREABLE for a device currently allocated to another job on
the system; PBS allocates the device when it becomes avail~ble.
If you specify /SHAREABLE, PBS allocates the device only when the server is
active. The device is available to other jobs when the server is idle. The default
is /SHAREABLE.
For batch servers, /SHAREABLE means the server uses a dynamic pseudo
keyboard with terminal characteristics that can not be set; INOSHAREABLE
means the server uses a static pseudo keyboard, with terminal characteristics
that can be set. DIGITAL recommends that you use /SHAREABLE unless you
need specific terminal characteristics

9-30

Managing Print/Batch Services

9.3.8

OPEN/QUEUE Command
The OPEN/QUEUE command opens a queue for additional requests after it has
been closed. If the specified queue is already open, RSTSIE ignores this command
OPEN/QUEUE requires PBSCTL privilege.

FClrmat
OPEN/QUEUE

queue-name[:]

Command Qualifiers

Defaults

none

Prompts
Queue: queue-name[:]

Command Parameters
queue-name[ :]
Specifies the name of the queu.e to be opened. This command does not permit
wildcard queue names.

Managing Print/Batch Services

9-31

9.3.9

SET QUEUE Command
The SET QUEUE command modifies the characteristics of one or more queues.
Changing a queue's characteristics does not affect entries already on the queue.
SET QUEUE requires PBSCTL privilege.

Format
SET QUEUE

queue-name[:]

Command Qualifiers

Defaults

IALL
IBATCH
ICPU_LlMIT=(MAXIMUM=a,DEFAULT =b)
I[NO]DEFAULT
IFORMS[=form-name]
IPAGE_LlMIT=(MAXIMUM=a,DEFAULT=b)
IPRINT
IPRIORITY=(MAXIMUM=a,DEFAULT=b)
IPRIVILEGES=("priv[, ... ]")
/TIME_LIMIT =(MAXIMUM=a,DEFAULT=b)

See Discussion
See Discussion
Attributes are unchanged
Attributes are unchanged
Attributes are unchanged
Attributes are unchanged
See Discussion
Attributes are unchanged
Attributes are unchanged
Attributes are unchanged

Prompts
Queue: queue-name[:]

Command Parameters
queue-name[:]
Specifies the name of the queue to be modified. This command permits wildcard
queue names. The rules for selecting queues when using wildcards are:
•

If you specify !PRINT or a print queue only qualifier, then PBS only selects
print queues.

•

If you specify !BATCH or a batch queue only qualifier, then PBS only selects
batch queues.

•

If neither of the above rules are true, then PBS selects any queue.

Command Qualifiers
fAll

Indicates that all queues should be modified. If you specify fALL, PBS omits the
queue name prompt and uses the queue name *. If you specify a queue name,
PBS ignores this qualifier.
fBATCH

Indicates that only batch queues should be modified. When you use fALL with
this qualifier, PBS modifies all batch queues.

9-32

Managing PrintIBatch Services

ICPU_LlMIT=(MAXIMUM=a,DEFAULT =b)
ICPU_LlMIT=n

Batch queues only. Indicates the maximum and default CPU time limits (in
minutes) for each command file processed in a batch job started from this queue.
The tnaximum CPU limit is the largest value that you can specify when issuing ~
request for the queue. The default CPU limit is the value that PBS assigns to a
request that had no limit specified.
The keyword UNLIMITED indicates that PBS does not enforce a CPU limit.
If you do not specify the MAXIMUM argument, then PBS does not change the
maxi:mum CPU limit. If you do not specify the DEFAULT argument, then PBS
does not change the default CPU litnit. If you use the fCPU_LIMIT=n form, ther
PBS use the value you specify for both the maximum and default limits. If you
specify a numeric argument, it must be in the range 1 to 65,535. The default
CPU limit cannot exceed the maxitnum CPU limit.
If you do not specify this qualifier, then PBS does not change the queue's curren1
default and maximum values.
I[NO]DEFAULT

Indicates whether the queue should serve as the default queue for its type. If
you specify /DEFAULT, then PBS marks the queue as the default queue for its
type, and any queue already defined as the default queue is no longer marked
as being the default queue. If you Elpecify INODEFAULT, then the queue is no
longer marked as being the default queue for its type. You then have no default
queue for that type. PBS does not return an error if you specify INODEFAULT
for a queue that is not currently marked as the default queue.
If you do not specify this qualifier, then any current default queue is not affected
IFORMS=form-name

Print queues only. Indicates if a default form name is defined for the queue. If
you specify /FORMS=form-name, then the name you specify becomes the default
for the queue and replaces any previous default form.
If you do not specify this qualifier, then the queue's current default form name if
not changed.
IPAGE_LlMIT=(MAXIMUM=a,DEFAULT=b)
IPAGE_LlMIT=n

Print queues only. Indicates the maximum and default page limit for each file
printed as part of a print job started from this queue. The maximum page limit
is the largest value that you can specify when issuing a request for this queue.
The default page limit is the value that PBS assigns to a request for this queue
that had no limit specified.
The keyword UNLIMITED indicates that no page limit is enforced. If you do no'
specify the MAXIMUM argument, then PBS does not change the maximum pagE
limit. If you do not specify the DEFAULT argument, then PBS does not change
the default page limit. If you. use the IPAGE_LIMIT=n form, then PBS uses
the value you specify for both the Dlaximum and default limits. If you specify a
numeric argument, it must be in the range 1 to 4,294,967,295. The default page
limit can not exceed the maximum page limit.
If you do not specify this qualifier, then the queue's current default and maximUl
values are not changed.
IPRINT

Indicates that only print queues should be modified. When you use fALL with
this qualifier, PBS modifies all print queues.

Managing Print/Batch Services

9-3

IPRIORITY=(MAXIMUM=a,DEFAULT=b)
IPRIORITY=n

Defines the maximum and default priorities for the queue, in the range 1 to 255.
If you do not specify the MAXIMUM argument, then PBS does not change the
current maximum priority. If you do not fJpecify the DEFAULT argument, then
PBS does not change the current default priority. If you use the IPRIORITY=n
fonn, then PBS uses the value you specify for both the maximum and default
priorities. The default priority cannot exceed the maximum priority.
If you do not specify this qualifier, then the queue's current default and maximum
values are not changed.
IPRIVILEG ES=(priv, ... )

Specifies the set of privileges required for a user to place entries on the queue.
A user who issues a PRINT or SUBMIT command must have all of the queue's
privileges for the request to be accepted. PBS accepts any of the RSTSIE
keywords, plus the keywords ALL and NONE.
The IPRIVILEGES=ALL qualifier assigns all defined privileges to the queue;
IPRIVILEGES=NONE assigns no privileges to the queue. Note that you can
specify any privilege name with IPRIVILI~GES, even those that you do not
possess.
The argument can also be USERn, where n is an integer between 1 and 8. You
can use these USERn privileges to identify particular users or groups of users
and to restrict the queues usage to those users.
If you specify this qualifier, PBS replaces the queue's current privileges with
those in the argument list. If you do not specify this qualifier, the queue's current

privileges are unchanged.
ITIME_LlMIT=(MAXIMUM=a,DEFAULT=b)
ITIME_LlMIT=n

Batch queues only. Indicates the maximum and default elapsed time limits (in
minutes) for each command file processed in a batch job started from this queue.
The maximum time limit is the largest value that a user can specify when issuing
a request for this queue. The default time limit is the value PBS assigns to a
request that had no limit specified.
The keyword UNLIMITED indicates that PBS does not enforce a time limit.
If you do not specify the MAXIMUM argument, then PBS does not change the
current maximum time limit. If you do not specify the DEFAULT argument,
then PBS does not change the current default time limit. If you use the ITIME_
LIMIT=n fonn, then PBS use the value you specify for both the maximum and
default limits. If you specify a numeric argument, it must be in the range 1 to
65,535. The default time limit cannot exceed the maximum time limit.
If you do not specify this qualifier, then the queue's current default and maximum

values (if any) are not changed.

9-34

Managing Print/Batch Services

9.:3.10

SElf" SERVER Command
The SET SERVER command changes the characteristics of one or more servers.
Use Uris command to change forms for print servers. SET SERVER requires
PBSCTL privilege.

FClrmat
SET SERVER

server-name{.j

Command Qualifiers
IAll
IBATCH
I[NO]CONTROLS[=UP_ARROW]
IFORMS[=form-name]
IPRINT
IPRIORITY[=n]
IRUN_BURST[=n]
I[NO]SHAREABLE

Defaults
See Discussion
See Discussion
Attributes are unchanged
Attributes are unchanged
See Discussion
Attributes are unchanged
Attributes are unchanged
Attributes are unchanged

Prompts
Server: server-name[:]

Command Parameters
server-name[:]
Specifies the name of the server to be modified. This command permits wildcard
server names.

Command Qualifiers
fAll

Indicates that all servers lhould be modified. If you specify fALL, PBS omits the
server name prompt. If you specify a server name, PBS ignores this qualifier.
fBATCH

Indicates that only batch servers should be modified. When you use fALL with
this qualifier, PBS modifies all batch. servers.
f[NO]CONTROlS[=UP_ARROW]
Print servers only. Indicates how print servers should process control characters
and escape sequences. If you do not specify this qualifier, then PBS does not
change the current CONTROLS setting. You must have the PBSCTL privilege to
use this qualifier.
fFORMS=form-name
Print servers only. Indicates the nalne of the form installed on the printer. The
server only prints requests with the same form name. PBS returns an error if th~
form name you specify for the device does not exist in the FDF.

If you do not specify this qualifier, then the server's current form is not changed.

Managing Print/Batch Services 9-35

IPRINT

Indicates that only print servers are to be modified. When you use fALL with this
qualifier, PBS modifies all print servers.
IPRIORITY[=n]

Batch servers only. Indicates the priority at which to run command procedures
controlled by this server. This value cannot exceed the priority of the PBS
primary job. PBS accepts any value in the range -120 to +120. PBS rounds the
value down to a multiple of 8.
If you do not specify an argument with !PRIORITY, the server's priority is set
to -8. If you do not specify this qualifier, then the batch server's priority is not
changed. You must have the TUNE privilege to use this qualifier.
IRUN_BURST[=n]

Batch servers only. Indicates the run bUrl3t to assign to any command procedures
controlled by this server. PBS accepts any value in the range 1 to 127.
If you do not specify an argument with /RUN_BURST, the server's run burst is
set to 6. If you do not specify this qualifier, then the batch server's run burst is
not changed. You must have the TUNE p:rivilege to use this qualifier.
I[NO]SHAREABLE

Indicates if the device should be made shareable with other jobs on the system.
If you specify INOSHAREABLE, then PBS immediately allocates the device,

preventing other jobs from accessing the device. The device is deallocated
whenever you delete the server (with the DELETE/SERVER command) or
shut down PBS (with the STOPfQUEUE/:MANAGER command). If you specify
INOSHAREABLE and the device is not currently available, PBS displays a
warning message and allocates the device as soon as it becomes available.
/SHAREABLE indicates that the device can be shared by other jobs on the system
and PBS only allocates it when the server is active.
For batch servers, fSHAREABLE means the server uses a dynamic pseudo
keyboard with terminal characteristics that can not be set; INOSHAREABLE
means the server uses a static pseudo keyboard, with terminal characteristics
that can be set. DIGITAL recommends that you use /SHAREABLE unless you
need specific terminal characteristics
If you do not specify this qualifier, then the [NO]SHAREABLE characteristic is

not changed.

9-36

Managing Print/Batch Services

9.3,.11

SHOW SERVER Command

The SHOW SERVER command displays the status of one or more servers in PBS.
The display includes information about the server's characteristics, its current
status, and the queues assigned to the server. The SHOW SERVER command
requires no privileges.

Format
SHOW SERVER

[server-name[:]]

Command Qualifiers
IAll
IBATCH

Defaults
IAlL

IBRIEF
IFUll

IBRIEF
IBRIEF

IPRINT

none

Prompts

none

Command Parameters
server·,name[:]
Specifies the name of the server to bt~ displayed. If you do not specify a server
name, all servers are displayed. This command does not permit wildcard server
names.

Command Qualifiers
IAll
Specifies that all servers are to be dif3played. fALL is the default if no server
name is specified. If a server name is specified, this qualifier is ignored.
IBATCH
Specifies that only batch servers are to be displayed.
IBRIEF
Specifies a brief display. It is the default. The brief display includes the following
information:

•

On-line or Off-line-Print servers only. Indicates the current status of the
server's print device. Note that PBS does not detect an off-line device until
it attempts to print on it; all devices are assumed to be on-line when PBS
first starts. KB-type printers are considered off-line when an XOFF (CtrIlS)
characters causes the device to stall for an extended period of time.

•

Active-Indicates that the server is currently processing a request.

•

Idle--Indicates that the server is not currently processing a request.

•

Stopped-Indicates that the server is currently stopped and will not accept
any new requests. The server mayor may not be active.

Managing Print/Batch Services 9-37

•

Job-wait-Indicates that PBS is unable to create a secondary job because no
job slots are currently available. PBS will keep trying to create the second job
until it succeeds.

•

Device-wait-Indicates that PBS is unable to allocate a device because the
device is not available. For print servers, device-wait means that the printer
controlled by the server is allocated to or opened by another job. For batch
servers, device-wait means that no pseudo-keyboards are currently available.
PBS will keep trying to allocate the device until it succeeds.

•

Shareable or Non-shareable-For print servers, shareable means that the
print device is available to other jobs on the system when it is not in use
by the server. Non-shareable means that the print device is permanently
allocated to the server.
For batch servers, shareable means t.he server uses a dynamic pseudo
keyboard, with non-specifiable characteristics. Non-shareable means that
the server uses a static pseudo keyboard that therefore does not change
cha.racteristics; rather, it uses the characteristics specified.

•

Form name-Print servers only. Indicates the name of the forms currently
installed, or "No forms" if no forms are installed.

IFULL
Specifies a full display. The full display includes all of the information in the lbrief
display, and also includes the following information:

•

Current entry-If the server is activE!, PBS displays the entry number and
entry spec of the current entry.

•

Assigned queues-PBS displays a list of the queues which can route requests
to the server.

IPRINT
Specifies that only printer servers are to be displayed.

9-38

Managing Print/Batch Services

9.:3.12

START/QUEUE Command

The START/QUEUE command starts a queue that was stopped with the
STOP/QUEUE command. Once you start a queue, PBS processes waiting entries on the queue. If you already started the specified queue, PBS ignores this
command. START/QUEUE requires PBSCTL privilege.

Format
START/QUEUE

queue-name[:]

Command Qualifiers
none

Defaults
none

Prompts
Queue: queue-name[:]

Command Parameters
queue-name[:]

Specifies the name of the queu.e to be started. This command does not permit
wildcard queue names.

Managing PrinV8atch Services

9-39

9.3.13

START/QUEUE/MANAGER Command

The START/QUEUEIMANAGER command starts PBS. Following this procedure,
it:
1. Creates the PBS primary job as a detached job running in the user's account.
RSTSIE initializes the job's priority and run burst according to the package
configuration parameters.
2.

Compresses the PBS system file by removing any records that are marked
as deleted. (This step reduces the overall size of the file.) During this step,
if RSTS/E finds an entry with status STARTED, it retains the entry, placing
it on HOLD. This can happen if RSTSIE restarts the package after a system
crash.

3. Loads all queue definitions into memory from the PBS system file.
4.

Loads all server definitions into memory from the PBS system file. If a print
server device is not a valid device on the system, PBS marks the server
DISABLED.

5. Assigns servers to queues based on the assignment information maintained in
the PBS system file.
6. Attempts to permanently allocate a device for any server marked as
NOSHAREABLE in its server's definition. If the allocation fails because
t.he device is currently in use, PBS marks the server DEVICE-WAIT and
attempts to allocate the device periodically until the allocation succeeds.
7.

Confirms startup has successfully cOlIlpleted by displaying a message.

The start-up process can take from a few seconds to several minutes, depending on the number of pending requests iIll the queue. Most users include the
START/QUEUEIMANAGER command in their system start-up command file
so the time required for PBS start-up is added to normal system start-up time.
START/QUEUEIMANAGER requires PBSCTL privilege.

Format
START/QUEUE/MANAGER
Command Qualifiers

Defaults

fPRIORITY=n
fRUN_BURST =n

fPRIORITY=-8
fRUN_BURST=6

Prompts
none

9-40

Managing Print/Batch Services

Command Qualifiers
fPRIORITY=n

Specifies the priority of the detached PBS job and any secondary jobs that PBS
creates. Value must be in the range -120 to +120; with any value specified
rounded down to a multiple of 8. The default is -8.
fRUN_BURST =n

Specifies the run burst of the detached PBS job and any secondary jobs that PBS
creates. Value must be in the range 1 to 127. The default is 6.

Managing Print/Batch Services

9-41

9.3.14

START/SERVER Command

This command restarts a server that wan stopped with the STOP/SERVER
command. If the server has a stalled job when you issue the command, the
command either resumes the job or terminates it, depending on the qualifiers you
use.
For print servers, the job can be resumed at a specified page of the current
file copy. PBS ignores this command if the specified server is already started.
For batch servers, PBS releases the controlled batch job from its held state.
START/SERVER requires PBSCTL privilege.

Format
START/SERVER

server-name[:]

Command Qualifiers
IBACKSPACE[=n]
IFORWARDSPACE[=n]
IN EXT_JOB
IPAGE=n
IRESTART
/TOP _OF _FILE

Default~;

See Discussion
See Discussion
See Discussion
See Discussion
See Discussion
See Discussion

Prompts
Server: server-name[:]

Command Parameters
server-name[:]
Specifies the name of the server to be started. This command does not permit
wildcard server names.

Command Qualifiers
IBACKSPACE[=n]
Print servers only. Indicates that printing should resume n pages back in the
current file listing. If you do not include an argument, then printing resumes one
page before the current page. If you specify an argument that extends beyond the
beginning of the file, then printing resumes at the beginning of the current file
copy.

PBS ignores this qualifier if the specified server is idle.
IFORWARDSPACE[=n]
Print servers only. Indicates that printing should resume n pages forward in the
current file listing. If you do not include an argument, printing resumes one page
after the current page. If you specify an argument that extends beyond the end
of the current file, printing resumes at the next file copy, if any.

PBS ignores this qualifier if the specified server is idle.

9-42

Managing Print/Batch Services

INEXT_JOB
Indicates that the current job should be aborted and printing or batch processing
should resume at the next job, if any.

PBS ignores this qualifier if the specified server is idle.
IPAGE=n
Print servers only. Indicates that printing should resume at page n in the curren1
file listing. If you specify an argument that extends beyond the end of the file,
printing resumes at the next file copy, if any. A value of 1 indicates the first page
of the file; PBS does not include header pages in the count.

PBS ignores this qualifier if the specified server is idle.
IRESTART
Print servers only. Indicates that printing should resume at the beginning of the
current job copy. If you submit the job with a /JOB_COUNT qualifier argument
larger than one, PBS only restarts the current job copy.

PBS ignores this qualifier if the specified server is idle.
ITOP_.OF_FILE

Print servers only. Indicates that printing should resume at the beginning of the
current file copy. If the file being printed was submitted with a /COPIES qualifier
argument larger than one, PBS only restarts the current file copy.
PBS ignores this qualifier if the specified server is idle.

Managing Print/Batch Services

9-43

9.3.15

STOP/QUEUE Command

The STOP/QUEUE command stops a queue; that is, PBS prevents any pending
requests on the queue from being processed. Any requests that are currently
being processed are not affected. STOP/qUEUE requires PBSCTL privilege.
PBS ignores this command if the specified queue is already stopped.

Format
STOP/QUEUE

queue-name[:]

Command Qualifiers
none

Defaults
none

Prompts
Queue: queue-name[:]

Command Parameters
queue-name[:]
Specifies the name of the queue to be stopped. This command does not permit
wildcard queue names.

9-44

Managing Print/Batch Services

91.3.16

STOP/QUEUE/MANAGER Command

The STOP/QUEUEIMANAGER command lets you shut down PBS, either immediately (causing all currently active jobs to be terminated) or when all servers
become idle. STOP/QUEUEIMANAGER requires PBSCTL privilege.
You ean shut down PBS in two ways:
•

As part of normal system shutdown. The SHUTUP program includes a phas
for automatically shutting down PBS, either immediately or when all print 0
batch jobs currently in progress complete.

•

Manually, using the STOP/QUEUEIMANAGER command.

Format
STOP/QUEUE/MANAGER
Command Qualifiers

Defaults

/[NO]ABORT

/NOABORT

Prompts
none

Command Qualifiers
I[NO]ABORT

Indicates whether any print or batch jobs currently running should be aborted
before shutting down PBS. Use IABORT to shut down PBS immediately; use
INOABORT to let all currently running print or batch jobs complete. The default
is INOABORT.

Managing Print/Batch Services

9-45

9.3.17

STOP/SERVER Command
This command stops a print or batch server. The action taken depends on
whether the server is currently active or not. For queues that are assigned to
only one server, stopping the server is equivalent to stopping the queue. For
queues that are assigned to more than one server, stopping one server does
not prevent requests from being routed to other active servers. STOP/SERVER
requires PBSCTL privilege.
If you issue a STOP/SERVER command for a server that is not currently processing a request, no additional requests are routed to that server. However, if
you issue a STOP/SERVER command for a server that is currently processing a
request, the following action is taken depending on the server:
•

Batch servers-If you specify the /JOlB_END qualifier, PBS lets the current
batch job complete before stopping thl3 server. Otherwise, the server suspends
the batch job by assigning it a priority of -128. The server then waits until
PBS issues a START/SERVER command to either resume or terminate the
batch job.

•

Print servers-If you specify the !PAGE_END, !FILE_END or /JOB_END
qualifier, PBS does not stop the server until it completes the current page,
file copy, or job copy. If the copy is the final copy of the job, PBS lets the job
finish before it stops the server. If you do not specify any of the qualifiers,
PBS stops the server immediately.
When PBS directs the server to stop, the server stops sending data to the
printer. However, system buffers continue to be printed on the device until
they become empty. This is necessary to prevent PBS from losing proper
page positioning on the printer. The server then waits until you issue a
START/SERVER command to either resume or terminate the print job.

Format
STOP/SERVER

server-name[:]

Command Qualifiers
/FILE_END
/JOB_END
/PAGE_END

Defaults
See Discussion
See Discussion
See Discussion

Prompts
Server: server-name[:]

Command Parameters
server-name[:]

Identifies the server to be stopped. This command does not permit wildcard
server names.

9-46

Managing Print/Batch Services

Command Qualifiers
/FILE_IEND

Print servers only. Indicates that printing should stop after the current file copy
is finished printing.
Note that this qualifier conflicts with /JOB_END and !PAGE_END.
/JOB_END

For print servers, indicates that printing should stop after the device has finished
printing the current job. For batch servers, indicates that the server should stop
when the current batch job completes.
Note that this qualifier conflicts with !FILE_END and !PAGE_END.
IPAGE._END

Print servers only. Indicates that printing should stop at the end of the current
page.
Note that this qualifier conflicts with !FILE_END and /JOB_END.

Managing Print/Batch Services

9-47

Chapter 10

Managing System Errors

The RSTSIE monitor automati.cally logs both hardware and system level software
errors. To gain the full advantages of this error detection capability, you must USE
the programs in the System Error Package. These programs fall into three main
categories:
•

Extraction and retention of system errors

•

Extraction and retention of system errors not yet retrieved at the time of a
system crash

•

Compilation and formatting of saved system errors

When the monitor detects an error, routines save critical error-related data and
send .a message to the ERRCPY program. The system invokes ERRCPY, which
then retrieves the saved data, performs minimal error message processing, and
stores the error information in a specially formatted disk file (ERRLOG.FIL).
Because RSTSIE limits the number of messages that may be queued to ERRCPY,
it should be run at a higher priority than most jobs to prevent the loss of valuabl.
diagnostic information. A separate program, ERRINT, initializes and validates
the elTor logging file to minimize the size of ERRCPY. Normally, you run ERRIN1
at the beginning of system start-up to start ERRCPY.
When a system crash occurs and the system crash dump facility is enabled,
the monitor preserves the contents of certain critical parts of the system in the
system file [0,1] CRASH.SYS. Run the ANALYS program immediately following
the recovery from a system crash to extract and format key information from the
crash file. Normally, this is done by the system start-up command file. One of
the ftmctions ANALYS performs is the creation of a separate error logging file (b~
default, ERRCRS.FIL) containing errors that were not processed by ERRCPY at
the time of the crash and an ERRDIS report for the errors.
NOTE

You must have SWCTL privilege to run any com.ponent of the System
Error Package.

10.1

Use of the Error Logging Programs: ERRINT and ERRCPY
The system program ERRCPY reads error-related information stored in the
monitor part of memory and writes it to a special disk file called ERRLOG.FIL.
The error file initialization and validation program (ERRINT) starts the ERRCP~
program. ERRINT performs various file checking or generation functions and
chains to ERRCPY. You must make sure the commands are in the START.COM
file to start ERRINT (thus, ERRCPY) and that ERRCPY is active during
time-sharing operations.
Managing System Errors

10-·

10.1.1

Error Logging Initialization: ERRINT
When the RSTSIE system starts up, DCL executes commands in the START. COM
command file. The following command starts ERRINT:
$ RUN ERROR$:$ERRINT

When the system executes the previous command, the ERRINT program tries to
locate the error file ERRLOG.FIL, which resides in ERROR$:. If it finds the file,
ERRINT checks certain critical control information in the file. If this information
is invalid, ERRINT renames the file ERRLOG.TMp, generates and initializes a
new ERRLOG.FIL, and issues an informational message:
Error File was found to be Invalid
Error File was generated and Invalid File renamed ERRLOG.TMP

If ERRINT does not find ERRLOG.FIL, it generates and initializes the file.
Two other commands, corresponding to responses to ERRINT questions, must
also be present in the command file. The two ERRINT dialogue questions for
which you must include responses are:
Change size to < 100 >?
Utilize crash file output (YES/NO) <NO>?

Once ERRINT validates or creates ERRLOG.FIL, it prints a message telling you
what percentage of the file has been usedl. For example:
ERRLOG file is 8% full

ERRINT also prints a question that includes (in angle brackets) the maximum
number of blocks allowed for the file:
Change size to < 100 >?

You can accept the current size of the file (originally 100 blocks) or change the
maximum size if you want. You can also specify an optional switch, /LOCK. If
you include /LOCK in your response, ERRCPY runs nonswappable. The default
is that ERRCPY can be swapped out.
NOTE

When an error occurs, RSTSIE writes it to a disk file. If ERRINT has
been swapped out to this same disk, RSTSIE does not log the error.
However, if you lock ERRINT in memory, RSTSIE does not lose these
errors. This is the only advantage of locking ERRINT in memory.
The disadvantage of locking ERRINT in memory is that the system
may hang, particularly if the memOlry is fragmented. The probability
of the system hanging for this reason is a function of free memory
available at any given time. Because the free memory is, in part,
a function of total available memory, the probability of your system
hanging for this reason decreases with increased total memory.
The program then asks if you want to append the contents of the special error
crash file ERRCRS.FIL to the contents of ERRLOG.FIL in the question:
Utilize crash file output (YES/NO) <NO>?

If you answer YES, ERRINT tries to find. the file and add its contents to the
end of the main error logging file. If it cannot find the error crash file, ERRINT
ignores the command and continues.

10-2

Managing System Errors

Mter it finishes processing all of its functions, ERRINT detaches and chains to
ERRCPY. At this point, the tenninal on which ERRINT started is free for other
use.
The ERRINT program must be located in the error library account ERROR$. 'Ib
run ERRINT from this account, type RUN ERROR$:ERRINT and press Return.
You can run ERRINT only if ERRCPY is not currently running. If you start up
ERRINT while ERRCPY is running, then ERRINT exits after displaying the
following message:
?Error logger is already running

Note that the version of ERRCPY that was running in a detached state on the
system continues to run.

10,,1.2

Examples of ERRINT Dialogue
An example of the ERRINT dialogue follows:
$ run error$:errint<RET>
ERRINT V9.7 RSTS V9.7 EDERE
ERRLOG file is 8% full
Change size to < 100 >? 120<RET>
Utilize crash file output (YES/NO) <NO>? NO<RET>
Detaching

In this example, a valid error log file exists, has a maximum length of 100 blocks,
and is 8% full. If you change the maximum size to 120 blocks but decide not to
add the special ANALYS error log file to the ERRINT error log file, the program
detaches and chains to ERRCPY.
In the next example, ERRINT finds an invalid error log file, renames the invalid
file to ERRLOG.TMP, creates a new file, leaves the maximum size at 100
blocks, and attaches the special error log file ERRCRS.FIL to the error log file
ERRLOG.FIL. Note that the 2% full message indicates a file containing only
control infonnation; this is the state of a newly generated error log file:
$ run error$:errint<RET>
ERRINT V9.7 RSTS V9.7 EDERE
Error file was found to be invalid
New file was generated and invalid file renamed ERRLOG.TMP
ERRLOG file is 2% full
Change size to < 100 >?<RET>
Utilize crash file output (YES/NO) <NO>? YES<RET>
Detaching

10,,1.3

Error Logging: ERRCPY
You can start the ERRCPY program only by running ERRINT. ERRCPY runs
detached, processes incoming error messages, and writes each message as a
variable-length record to a nons panned sequential file (ERRLOG.FIL). When its
message queue is empty, ERRCPY enters the receiver sleep state and may be
swapped out.
The monitor queues messages to ERRCPY regardless of the presence or absence
of ERRCPY. (If ERRCPY is not running, the messages are simply left waiting,
up to a limit of 40.) The first message the monitor queues on start-up is always
a "power fail/start up" message. On receiving a special error message from the
SHUTUP program, ERRCPY logs the message, closes the error file, and kills
itself.

Managing System Errors 10-3

10.2 Displaying Errors: ERRDIS
The error display program ERRDIS provides you with a convenient method of
displaying previously logged errors.
ERRDIS has four functions:
•

Provides a summary report by error type and unit number (where applicable)
of all errors logged to the error file

•

Provides a detailed report of one or all error types logged between any two
user-selected date and time pairs

•

Zeros some or all of the contents of the error file following the generation of a
report

•

Provides a list of potentially bad disk blocks

ERRDIS consists of two modules: ERRDIS and ERRDET. The first module performs the first and third functionEJ. To perform the other functions,
ERRDIS chains to ERRDET; control returns to ERRDIS when ERRDET finishes
processing.

10.2.1

Running ERRDIS
You can run the ERRDIS program by typing:
$ RUN ERROR$:ERRDIS

Mter ERRDIS prints header information, you select optional modes of operation
by means of an interactive dialogue. In the dialogue description shown in
Table 10-1, prompts are numbered for reference. An answer enclosed in angle
brackets
indicates the default response; select the default by pressing Return
in response to a program prompt. You can abbreviate most responses with two
characters. Optional characters are enclosed in square brackets []; for example,
HE[LP].

«»

10-4

Managing System Errors

Table 10-1:

ERR DIS Dialogue Explanation

Step

Prompt and Description

ERRDIS V9.7 RSTS V9.7 EDERE
ERRDIS prints an identification line.

Input File <ERROR$:ERRLOG.FIL>?
Enter a file specification for ERRDIS to process. The file must be formatted
as an error logging file.

Output to <KB:ERRDIS.OUT>?
Enter a file specification to which ERRDIS should send the report.

He[lp], Ba[d Blocks], Su[mmary], Fu[ll] Report, or Ze[ro] File? <Summary>?
Select one of five possible reports:
•
•

•

He[lp]. Prints a help file, followed by a list of mnemonics corresponding
to all possible error types.
Ba[d Blocks]. Outputs a report of possible bad blocks detected from the
disk errors logged in the input file entered in Step 2. The next dialogue
step is 11.
Su[mmary]. Outputs a report of the number of errors detected for each
error type and for each unit number within a specific error type. The
next dialogue step is 10.
Fu[ll]. Outputs a report of the detailed contents of each selected error
record in the error file. Selection of a specific record depends on the
answers to steps 5 through 9 in the ERRDIS dialogue.
Ze[ro]. Zeroes the error file, either completely or by error type. The
next dialogue step is 12.

Specific Error Type <All>?
Enter a two-character mnemonic identifying the specific error type that you
want ERRDIS to process. The default answer selects all error types. Note
that if you select All, you may also attach the INOTAPE switch if you want
magnetic tape errors to be omitted from the error listing.

Starting Date <First Error>?
Enter the date of the earliest error that you want ERRDIS to process. The
format is dd-mmm-yy or yy.mm.dd. The default answer is the date of the
first error in the error log file. If you press Return to select the default, the
next step is 8.

Starting Time <First Error>?
Enter the time of the earliest error that you want ERRDIS to process. The
format is hh:mm, either in 24-hour time or with AMlPM afterwards. The
default is the time of the first error in the error log file on the selected date.

Ending Date <Last Error>?
Enter the date of the latest error that you want ERRDIS to process. The
format is dd-mmm-yy or yy.mm.dd. The default is the date of the last error
in the error log file. If you select the default, then the next dialogue step is
11.

Ending Time <Last Error>?
Enter the time of the latest error that you want ERRDIS to process. The
format is hh:mm, either in 24-hour time or with AMlPM afterwards. The
default is the time of the last error in the error log file on the selected
ending date. The next dialogue step is 11.

(continued on next page)

Managing System Errors

10-5

Table 10-1 (Cont.):

ERRDIS Dialogue Explanation

Step

Prompt and Description

List Bad Blocks (YeslNo) <No>?
Press Return (or type No) if you do not want ERRDIS to generate a list of
possible bad blocks following the summary report. Type Yes if you want a
list of bad blocks.

Zero Error File upon completion (YeslNo) <No>?
Type Yes to have ERRDIS zero the previously specified error log file after
it successfully generates the desired report. (You must type all three
characters of Yes to zero the file.)

Zero Which Error Type <AI!>?
Press Return or type All to have ERRDIS zero all error types in the file.
Otherwise, enter the error type or types you want zeroed.

After you finish the dialogue, ERRDIS:
•

Processes the selected error records

•

Fonnats and generates the selected report

•

Zeros the error log file (if requested in step 11 of Table 10-1)

•

Returns to step 2 of Table 10-1.

At that point, you can enter CtrllZ in response to any question to return the
tenninal to the keyboard monitor prompt.

10.2.2 Help Report
The Help report prints information on the use of the ERRDIS program and then
prints a list of the error mnemonics other programs use. After you run ERRDIS
and select the Help report, the program prints the Help text.

10.2.3 Summary Report
A summary report at the end of this section supplies general information on
all errors logged. The first two columns list the two-character error mnemonic
and full error description. The TOTAL HEC/LOG column lists the total errors
received by the ERRCPY program and the total it stored in the error file. These
two totals may differ due to the limits on:
•

The number of errors that may be logged

•

The size of the error file

The UNIT NUMBERS column lists the total number of errors generated by each
device on a controller.
The CONTROLLER ERRORS column indicates how many of the errors logged under TOTAL REC/LOG are controller errors, as opposed to unit errors. (Controller
errors apply to the controller itself, while unit errors apply to a specific unit
being controlled.) If N/A appears under CONTROLLER ERRORS, that means
that ERRDIS is not keeping track of CONTROLLER ERRORS for that particular
ERROR CODE-DESCRIPTION.

10-6

Managing System Errors

The asterisk (*) next to 239/100 indicates that MM, under the ERROR CODEDESCRIPTION column, has exceeded the maximum number of errors (100) that
the EITor Package logs for that particular error type.
Mter you run ERRDIS and select the SUMMARY report, the program prints the
summary as follows.
$ RUN ERROR$:ERRDIS <RET>
ERHDIS V9.7 RSTS V9.7
EDERE
Input File <ERROR$:ERRLOG.FIL>? <RET>
Output to <KB:ERRDIS.OUT>? <RET>
Hel[lp], Ba[d Blocks], Su[mmary], Fu[ll] Report or Ze[ro] File <Summary>? <RET>
List Bad Blocks (Yes/No) <No>? <RET>
Zel:o Error File upon completion (Yes/No) <No>? <RET>
EmRDIS Summary Report taken on 29-May-89, 10:50 AM
Input File: ERROR$:ERRLOG.FIL Output File: KB:ERRDIS.OUT
Reported Date/Time Range:
13-May-89, 12:21:22 AM through 28-May-89, 12:35:47 AM
ERROR
CODE-DESCRIPTION

TOTAL
REC/LOG

PF Power Fail/Start-up 1/1
DU MSCP Disks
7/7
3/3
DB RH 11 /RP 04 -·05 - 06
DL RL01/RL02
3/3
* 239/100
MM RH11/TU16/TE16
SH MSG from SHUTUP
1/1

UNIT NUMBERS
2 3 4 5 6

1
49 7

CONTROLLER
ERRORS

3
2
5

N/A
3
N/A
N/A
N/A
N/A

Total of 115 Errors Logged out of 254 Received
34 out of 100 Blocks have been used in ERROR$:ERRLOG.FIL

********************************************************************
Input File <ERROR$:ERRLOG.FIL>? <Ctrl/Z>
$

10.2.4

Bad Block Report
The ERRDET program reports possible bad blocks that were detected by the
monitor by checking the error logging file for such entries. The criteria for a
possible bad block depends on the type of disk.
Each bad block report includes:
•

The logical block number

•

The logical name or pack identification of the disk

•

The physical device name of the disk

Using the logical block number that ERRDIS prints, you can add the possible
bad block to the bad block file with the BADS suboption of REFRESH. See the
RSTS / E System Installation and Update Guide for information on adding bad
blocks.

Managing System Errors 10-7

The following example shows a list of bad blocks that the ERRDIS program
prints:
$ RUN ERROR$:ERRDIS <RET>
ERRDIS V9.7 RSTS V9.7
EDERE
Input File <ERROR$:ERRLOG.FIL>? <RET>
Output to <KB:ERRDIS.OUT>? <RET>
Help, Bad Blocks, Summary, Full Report or Ze[ro] File <Summary>? BAD <RET>
Zero Error File upon completion (Yes/No) <No>? <RET>
ERRDIS Bad Block Report taken on 29-May-89, 10:51 AM
Input File: ERROR$:ERRLOG.FIL Output File: KB:ERRDIS .. OUT
Requested Date/Time Range:
First Error through Last Error
ERROR$:ERRLOG.FIL will not be zeroed upon completion
List of Possible Bad Blocks:

---------------------------RAIDER

Logical Name
Physical Name
Logical Block Number

DM1:
3401

Logical Name
Physical Name
Logical Block Number

OFTHE
DR4:
237968

Logical Name
Physical Name
Logical Block Number

LOST
DRS:
432952

Logical Name
Physical Name
Logical Block Number

ARK
DM2:
8

Total Number of Possible Bad Blocks: 4
********************************************************************
Input File <ERROR$:ERRLOG.FIL>? AZ

NOTE

Use caution before adding possible had blocks (that ERRDIS reports) to
the bad block file. Instead of corrupt disk surfaces, hardware controller
or disk drive problems may have caused the errors to occur.

10.2.5

Adding Bad Blocks to the Bad Block File
The ERRDIS program prints a list of potential bad blocks. If you decide to add a
block to the bad block file BADB.SYS on a disk, use the BADS suboption of the
REFRESH option in INIT.SYS (see the RSTS / E System Installation and Update
Guide). In preparation for adding bad blocks, which requires you to bring the
system down, use the PIP system program to copy the data in a bad file to a new
file. As your system discovers bad blocks, PIP lets you collect this infonnation in
a single file. You can then use this file to add the bad blocks to BADB.SYS when
you nonnally take your system down.
Follow these steps to use PIP for this pUlC'pose:
1. Specify the /GO switch (along with other necessary switches, such as /CL, and
so forth) to ignore possible ?Data error on device errors. Rename the file that
contains the bad block, but do not dellete it:
$ RUN $PIP
*FILE.TMP=FILE.OLD/GO
*FILE.BAD=FILE.OLD/RE
*FILE.OLD=FILE.TMP/R

10-8

Managing System Errors

2. When it is convenient, shut down the system and use the BADS suboption
of REFRESH to add any bad blocks to the disk's bad block file. Each new
bad block is now allocated to two files ([O,l]BADB.SYS and FILE.BAD in the
pJ'evious example).
3. The BADS suboption automatically rebuilds the allocation tables. It then tellE
you the file has a bad block and allows you to delete the file (FILE.BAD in
the previous example). The RSTS / E Installation and Update Guide describef
this operation.
When BADS deletes the file, it frees the blocks in the file that are not bad.
You must modify the previous procedure if the bad block is in a file directory. In
this case, use the BACKUP program to copy all accessible files in the account
(or disk) to an archive medimrt. When you add the bad block and clean the disk,
you can delete the account whose directory contains the bad block. Mter starting
timesharing, use BACKUP to restore the files from the archive medium.
The bad block report does not apply to MSCP disks; bad blocks on those devices
are recovered automatically.

101.2.6 Full Report
Each error displayed includes:
•

A title line containing the error code mnemonic

•

The error description

•

A sequence number

•

The date and time the error occurred

The monitor maintains the sequence numbers, starting at one, from start-up
to shutdown. A crash and subsequent automatic restart does not reset these
numbers.
There are four basic categories for individual errors displayed within the full
error report:
•

Nonperipheral errors such as processor traps and memory parity

•

Di sk errors

•

Nondisk peripheral device errors such as magnetic tape errors

•

Other errors such as the shutdown message received from SHUTUP or missed
errors

The program automatically checks for and reports possible bad blocks for disk
errors included in the scan. The following sections contain partial listings of a
FULL report.
When the monitor detects a hardware or software error, it sends information
about the error to the ERRCPY program. ERRCPY can accept and file error
infonnation until the number of queued errors reaches 40. The monitor logs all
errors it detects beyond this point as "missed errors." Only after the queue begins
to empty and the number of errors queued to ERRCPY falls below the maximum
of 40 can the monitor begin to send the normal error information. This means
that while the queue is full, the monitor reports all subsequent errors as "missed
errors," rather than sending the usual error information to ERRCPY.

Managing System Errors

10-9

The monitor also reports missed errors if it does not have a sufficient number of
general small buffers available to store elTor information. When the number of
general small buffers drops below 75, the monitor reports any detected error to
ERRCPY as a "missed error."
10.2.6.1

User Description In Full Report

All errors except shutdown, missed errorl3, and some MSCP controller errors
usually include a user description. The name fields in the user description relate
to a specific job whose number appears in the first field (Job number). (Under
certain conditions, a job number of 0 is possible. In this case, the user description
is not listed.)
The following example shows a complete error report for a tape device; the report
includes both a user description and a detailed description:
$ RUN ERROR$:ERRDIS <RET>
ERRDIS V9.7 RSTS V9.7 EDERE::
Input File <ERROR$:ERRLOG.FIL>? <RET>
Output to <KB:ERRDIS.OUT>? <RET>
Help, Bad Blocks, Summary, Full Report or Ze[ro] l?ile <Summary>? FULL <RET>
Specific Error Type <All>? <RET>
Starting Date <First Error>? 13-May-89 <RET>
Ending Date <Last Error>? <RET>
Zero Error File upon completion (Yes/No) <No>? <RET>
ERRDIS Full Report (All Types) taken on 29-May-89, 02:17 PM
Input File: ERROR$:ERRLOG.FIL
Output File: KB:ERRDIS.OUT
Requested Date/Time Range:
13-May-89 (First Error) through Last Error
ERROR$:ERRLOG.FIL will not be Zeroed upon completion

********************************************************************
MM

RH11/TU16/TE16 Seq #92 Occurred on 28-May-89 i3.t 02: 11: 44 PM

User Description:
Job Number
KB Number
Account
Program Name
User Job Physical Addr.
User Job Size
Control Parameters
RTS Name
RTS Physical Address

18
2

[1,170]
PASIKO
02310000
28K
200
RT11
01714000

Detailed Description:
Timeout Indicator
Physical Name

000
MMO:

DDB

017600
045471
000000
000056

CSR Address:

176700

10-10

Managing System Errors

173444
005171
147240
002004

104252
001630
000000
127311

000011
177400
001007

Contents of Registers:

---------------------145270

MT'CS1
MT'WC
MTBA
MTFC
MTCS2
MTDS
MTER
MTAS
MTCK
MTMR
MTDT
MTSN
MTTC
MTBAE
MTCS3

000400
132444
000000
000100
150660
100000
000001
000000
000000
000000
000000
000000
000016
002000

TRE Xfer Err

PE Mode
COR/CRC

T077 TM03

Table 10-2 swnmarizes the user description data.
Table 10-2:

User Description Data

Heading

Meaning

Job Number

For disk errors, the number of the job that requested
the 110; for nondisk peripheral errors, the job owning
the device; for nonperipheral errors, the job running a1
the time of the error.

KB Number

The keyboard number of the job, followed by DET if
the job is detached.

Account

The PPN of the user who created the job.

Program Name

The name of the program running in the job virtual
address space (low segment).

User Job Physical Address

The current physical memory address of the job.

User Job Size

The size of the user program running in the low
segment of the job virtual address space.

Control Parameters 1

Monitor control information.

RTS Name

The name of the run-time system mapped into the
highest segment of the user's virtual address space.

RTS Physical Address

The address of the location in physical memory where
the run-time system resides.

1A detailed description of this entry is beyond the scope of this manual. The entry is intended for
use by DIGITAL personnel.

10.2.6.2

Disk Error Detailed Description

A FULL report for a disk error includes three sections:
•

User description

•

Detailed description

•

Contents of registers

Table 10-3 describes the information found in the detailed description portion of
a disk error report. Before you refer to the table, study the following example.

Managing System Errors

10-11

The example shows what a single complete disk error looks like. If the entire
Full Report of DB type errors were reproduced here, you would see two more
identically structured error messages following this one. As you can tell from the
sample in the section "Summary Report," there are three DB errors under TOrrAL
RECILOG. The example follows:
$ RUN ERROR$:ERRDIS <RET>
ERRDIS V9.7 RSTS V9.7 EDERE
Input File <ERROR$:ERRLOG.FIL>? <RET>
Output to <KB:ERRDIS.OUT>? <RET>
Help, Bad Blocks, Summary, Full Report or Zarro] File <Summary>? FULL <RET>
Specific Error Type <All>? DB <RET>
Starting Date <First Error>? 13-May-89 <RET>
Starting Time <First Error>? <RET>
Ending Date <Last Error>? <RET>
Zero Error File upon completion (Yes/No) <No>? <RET>
ERRDIS Full Report (DB only) taken on 29-May-89, 11:00 AM
Input File: ERROR$:ERRLOG.FIL
Output File: KB:ERRDIS.OUT
Requested Date/Time Range:
13-May-89 (First Error) through Last Error
ERROR$:ERRLOG.FIL will not be Zeroed upon completion

********************************************************************
DB

RH11/RP04-05-06 Seq #81 Occurred on 28-May-89 at 07:12:14 AM

User Description:
Job Number
KB Number
Account
Program Name
User Job Physical Addr.
User Job Size
Control Parameters
RTS Name
RTS Physical Address

17
33 (Det)
[1,100]
LOGOUT
02744000
11K
200
... RSX
00000000

Detailed Description:
I/O Status
Timeout Indicator
Offset Position
Overlapped Seek Ind.
Unit Size in DC's
Device Cluster Size
Pack Cluster Size
Logical Name
Physical Name
Logical Block Number

000
000
Previous
377
121510
8
8
D
DB2:
19272

DSQ

004500
045501
000000
000056

173442
006171
146300
002004

104252
064540
000000
127310

000011
174000
001000
001003

WCB

100000
146274
113570
005166

004442
000000
004550
005413

000002
000000
004551
005414

000021
001310
005165
005415

FCB

147040
064511
001300
000011

070560
077430
113570
000226

000544
037404
001270
000010

076725
000001
113570
146300

CSR Address:

176700

10-12

Managing System Errors

[1,100]
TEMP33.TMP

Contents of Registers:
RPCS1
RP'WC
RPBA
RPDA
RPCS2
RPDS
RPgR1
RPJ\.S
RP:LA
RP1~R

RPlDT
RP;SN
RPOF
RPDC
RPCC
RPJ!!:R2
RPgR3
RP1!!:Cl
RPE:C2

145270
000400
000000
000000
000000
000000
000000
000000
000000
000000
000000
000000
000000
000000
000000
000000
000000
000000
000000

TRE Xfer Err

Not Ready

Table 10-3 describes the infonnation found in the detailed description portion of
the previous disk error report. The program prints an additional field if the error
signalled a possible bad block.
Table 10-3:

Disk Error Detailed Description

Heading

Meaning

I/O Status

See following Timeout Indicator.

Timeout Indicator

Timeout and 110 Status are considered as a pair. One of the
following states can exist:
Timeout Indicator

110 Status

Meaning
Idle

SEEK in progress

Waiting for ReadiWrite

Timeout

where x equals any nonzero value.
Offset Position

The contents of Offset Position for disks having offset capabilities:

= cente:rline position
20 = +x offset
220 = -x offset
40 = +2x offset
240 = -2x offset
60 = +3x offset
260 = -3x offset
100 = offset unknown
where x equals the number of micro inches of offset (a drive
dependent parameter).
Overla pped Seek
Ind.

377 = overlap seek driver
000 = non-overlap seek driver
You determine which of these two values appears during system
installation.

Unit Size in DC's

Size (in octal) of the disk expressed as the number of device clusters.

(continued on next page)

Managing System Errors

10-13

Table 10-3 (Cont.):

Disk Error Detailed Description

Heading

Meaning

Device Cluster Size

Device cluster size for this drive.

Pack Cluster Size

Pack cluster size for this disk.

Logical Name

The pack identification if no system wide logical name was given at
mount time or the syst1em logical name entered at mount time.

Physical Name

The physical name and unit number.

Logical Block
Number
DSQ 1

The block number where the error occurred.

WCB 1

The WeB (Window Control Block) is present for disk errors that
resulted from a file request by the user job.

FCB 1

The File Control Block..

eSR Address

The eSR address is the base address of the set of registers in the
110 page belonging to the device that caused the error.

Contents of
Registers

This section contains the name of each readable device register,
its contents and, possibly, an abbreviated description of various
error bits that were fOlmd to be set to one in the register. ERRDIS
stores a maximum of 21 error bits for each error type. The data
file ERRDAT.FIL contains the register mnemonics, error bit descriptions, and error code descriptions for all error types. ERRDIS
requires this file to run.

The disk request queuEI entry block.

See the PDP-ll Peripherals Handbook for a detailed description of
registers and bit descriptions. The register mnemonics displayed by
ERRDIS correspond to those given in the manual.
lA detailed description of this entry is beyond the scope of this manual. The entry is intended for
use by DIGITAL maintenance personnel.

10.2.6.3

MSCP Variations on the Full Report

For systems that support MSCP type disk hardware, (for example, the RA80),
there are two variations on the Full Report:
•

If there is a disk error, then the Full Report looks just like the normal Full
Report, except that it has two extra sections called the MSCP Description
and the Driver Related Data (see Example #1). The Full Report includes the
following five sections:
User Description
Detailed Description
Contents of Registers
MSCP Description
Driver-related Data

•

If there is a controller error, then the Full Report shows an abbreviated error

message. The Full Report does not have a User Description, and its Detailed
Description is not as large as in the normal Full Report (see Example #2).
The Full Report includes the following four sections:
Detailed Description
Contents of Registers

10-14

Managing System Errors

MSCP Description
Driver-related Data
If you refer back to the sample report in the section "Summary Report," you will
notice that the first example that follows is the error that was logged under UNIT
NUMBER 1. The second sample error message that follows is one of the three
errors logged under CONTROLLER ERRORS.

Example #1:
$ RUN ERROR$:ERRDIS <RET>
ERRDIS V9.7 RSTS V9.7 EDERE
Input File <O,3]ERRLOG.FIL>? <RET>
Out~put to <KB:ERRDIS.OUT>? <RET>
Help, Bad Blocks, Summary, Full Report or Ze[ro] File <Summary>? FULL <RET>
Specific Error Type <All>? DU <RET>
Starting Date <First Error>? <RET>
Ending Date <Last Error>? <RET>
Zel:O Error File upon completion (Yes/No) <No>? <RET>
ERRDIS Full Report (DU only) taken on 29-May-89, 02:17 PM
Input File: ERROR$:ERRLOG.FIL
Output File: KB:ERRDIS.OUT
RequestE~d Date/Time Range:
First Error through Last Error
ERROR$:E:RRLOG.FIL will not be Zeroed upon completion

********************************************************************
DU

MSCP Disks Seq

46 Occurred on 25-May-89 at 07:04:02 PM

User Description:
Job Number
KB Number
Account
Program Name
USE~r Job Physical Addr.
USE~r Job Size
Control Parameters
RTS Name
RTS Physical Address

12
16
[1,217]
PACKID
01610000
7K
000
BAS4F
01514000

Detailed Description:
I/O Status
Timeout Indicator
Offset Position
Overlapped Seek Ind.
Unit Size in DC's
Device Cluster Size
Pack Cluster Size
L09ical Name
Physical Name
L09ical Block Number

000
377
000
000
000000

o
o
DU7:
200806
000000
010147
000010
000003

CSR Address:

177430
000367
000000
004000

103664
023000
000000
000000

001424
000400
075576
003777

160414

Contents of Register:
SA

000000

Controller On Line

MSCP Description:
Packet Type

End Message

Managing System Errors

10-15

MSCP Envelope

000060

000001

MSCP Packet

074130
000203
000000
020424
000007
000001

140416
000043
000000
001003
000000
003400

Command Reference Number
Drive Number
Message Flags
End Flags
Endcode
Status/Event Code
First Bad Block

074130 140416
000007
000
000
GET UNIT s~rATUS command
203
000043
No vol mntd or r/s switch out
543437593.

000007
000736
001624
031431
000037
000020

000000
000000
000000
020144
000001
002001

Driver Related Data:
Internal Control Status Word 000300
Internal Unit Status Word
000000
Driver Error Code
000007
BBR Flag Word
000000

GUS failed; unit not functional

Example #2:
$ RUN ERROR$:ERRDIS <RET>
ERRDIS V9.7 RSTS V9.7 EDERE
Input File <ERROR$:ERRLOG.FIL>? <RET>
Output to <KB:ERRDIS.OUT>? <RET>
Help, Bad Blocks, Summary, Full Report or Ze[ro] File <Summary>? FULL <RET>
Specific Error Type <All>? DU <RET>
Starting Date <First Error>? <RET>
Ending Date <Last Error>? <RET>
Zero Error File upon completion (Yes/No) <No>? <RET>
ERRDIS Full Report (DU only) taken on 29-May-89, 02:19 PM
Input File: ERROR$:ERRLOG.FIL
Output File: KB:ERRDIS.OUT
Requested Date/Time Range:
First Error through Last Error
ERROR$:ERRLOG.FIL will not be Zeroed upon completion

********************************************************************
DU

MSCP Disks Seq

2 Occurred on 21-May-89 at 12:59:04 PM

Detailed Description:
I/O Status
Timeout Indicator

000
377

CSR Address:

172150

Contents of Register:
SA

000000

Controller On Line

MSCP Description:
MSCP Packet Type

Last Fail Packet

MSCP Envelope

000030

000020

MSCP Packet

000000
000400
000000

000000
000012
000406

Controller Class
Controller Model
Controller Hardware Version
Controller Software Version
UDA Internal Error Code

001
006
000
004
000013

Mass Storage
UDA50-A

10-16

Managing System Errors

000000
000000
000004

000000
000000
000013

ER.MST - Bus Master Error

Driver Related Data:
Internal Control Status Word 000102
Internal Unit Status Word
000000
Driver Error Code
000000
BBR Flag Word.
000000

Driver did not detect error

********************************************************************
10. 2.6.4

Nondisk Peripheral Device Error Detailed Description

Table 10-4 lists the fields found in the detailed description portion of the nondisk
peripheral device error report. Before you refer to the table, study the following
example. The example shows a single complete tape error. As you can tell from
the example in the section "Summary Report," this is one of the 239 tape errors
that were received. This particular error was one of the 100 logged, and also one
of the seven logged under UNIT NUMBER 1. The entire Full Report contains 99
more tape error messages. The example follows:
$ RUN ERROR$:ERRDIS <RET>
ERRDIS V9.7 RSTS V9.7 EDERE
Input File <ERROR$~ERRLOG.FIL>? <RET>
Out.put to <KB:ERRDIS.OUT>? <RET>
Help, Bad Blocks, Summary, Full Report or Ze[ro] File <Summary>? FULL <RET>
Sp~:wific Error Type <All>? MM <RET>
Starting Date <First Error>? <RET>
Ending Date <Last Error>? <RET>
Zero Error File upon completion (Yes/No) <No>? <RET>
ERRDIS Full Report (MM only) taken on 29-May-89, 02:17 PM
Input File: ERROR$:ERRLOG.FIL
Output File: KB:ERRDIS.OUT
Requested Date/Time Range:
First Error through Last Error
ERROR$:ERRLOG"FIL will not be zeroed upon completion

**'k************************************************** ***************
MM

RH11/TU16/TE16 Seq #44 Occurred on 28-May-89 at 05:51:08 AM

User Description:
Job Number
KB Number
Account
Program Name
User Job Physical Addr.
US~3r Job SiZE:!
Control Parameters
RTS Name
RTS Physical Address

11
33 (Det)
[1,100]
BACKUP
02610000
15K
200
... RSX
00000000

Detailed Description:
Timeout Indicator
Ph~{sical NamE:!

000000
MM1:

DDB

000016
000001
105536
000361
000000

CSR Address:

172440

000426
000000
120540
001002
007314

002102
000000
010004
010760
004400

100001
000000
000021
000053

Managing System Errors

10-17

Contents of Registers:
MTCS1
MTWC
MTBA
MTFC
MTCS2
MTDS
MTER
MTAS
MTCK
MTMR
MTDT
MTSN
MTTC

145260
000000
010004
000000
000100
150760
100000
000001
000010
001100
142011
060023
102301

TRE Xfer Err

PE Mode
COR/CRC

TU16/TE16

Table 10-4 describes the information found in the detailed description portion of
the previous nondisk peripheral device error report example.
Table 10-4:

Nondlsk Peripheral Device Format

Heading

Meaning

Timeout Indicator

Nonzero indicates that an expected response was not received
from the device in the allotted time.

Physical Name

The physical name and unit number of the device.

DDBl

The DDB (Device Data Block) contains various parameters
required by the system to control the device.

DDB Extension l

Supplementary DDB information.

CSR Address

The CSR address is the base address of the set of registers in
the 110 page belonging to the device that caused the error.

Contents of Registers

This section contains the name of each readable device
register, its contents and, possibly, an abbreviated description
of various error bits that were found to be set to one in the
register. ERRDIS stores a maximum of 21 error bits for each
error type. The data file ERRDAT.FIL contains the register
mnemonics, error bit descriptions, and error code descriptions
for all error typ«~s. ERRDIS requires this file to run.
See the PDP-ll Peripherals Handbook for a detailed description of registers and bit descriptions. The register
mnemonics displayed by ERRDIS correspond to those given
in the manual.

1 A detailed description of this entry is beyond the scope of this manual. The entry is intended for
use by DIGITAL maintenance personnel.

10-18

Managing System Errors

10.:t6.S

Nonperlpheral Error Detailed Description
See the related processor handbook for detailed descriptions of nonperipheral
error fields. Table 10-5 gives an abbreviated description of these items. Before
you refer to the table, study the following example. This example shows a single
complete error message extracted from the ERRLOG.FIL using ERRDIS to
reproduce only the PF type errors:

$ RUN ERROR$:ERRDIS <RET>
ERRDIS V9.7 RSTS V9.7 EDERE
Input File <ERROR$:ERRLOG.FIL>? <RET>
Output to <KB:ERRDIS.OUT>? <RET>
Help, Bad Blocks, Summary, Full Report 01: Ze [ro] File <Summary>? FULL <RET>
Specific Error Type <All>? PF <RET>
Starting Date <First Error>? <RET>
Ending Date <Last Error>? <RET>
Zero Error File upon completion (Yes/No) <No>? <RET>
ERRDIS Full Report (PF only) taken on 29-May-89, 02:21 PM
Input File: ERROR$:ERRLOG.FIL
Output File: KB:ERRDIS.OUT
Requested Date/Time Range:
First Error through Last Error
ERROR$:ERRLOG.FIL will not be zeroed upon completion

********************************************************************
PF

Power Fail/Start-up Seq #1 Occurred on 13-May-89 at 02:15:01 PM

Detailed Description:
RO
R1
R2
R3
R4
R5
Virtual PC
Physical PC
Processor Status
Stack Pointer
(SP)
(SP+2 )
(PC'-6 )
(PC'-4 )
(PC'-2 )
(PC)
CPU ID
CPU ERR

000000
004200
000000
000002
004100
000000
000077
00000077
000000
002074
000002
000000
034156
004356
034156
004357
-1

000000

Managing System Errors

10-19

Table 10-5 describes the infonnation found in the detailed description portion of
the previous example.
Table 10-5:

Nonperipheral Error Format

Heading

Meaning

RO-R5

Processor registers 0 through 5. Each register is listed on a
separate line.

Virtual PC

The address (within the user virtual address space) of the
instruction beinl~ executed at the time of the error.

Physical PC

The address (in the physical memory address space) of the
instruction being executed at the time of the error.

Processor Status

The processor status word.

Stack Pointer

The address of the current top of stack.

(SP)
(SP+2)

The contents of the top two words in the stack.

(PC-6)
(PC-4)
(PC-2)
(PC)

The contents of the three words preceding the current
program PC and the contents of the current PC.

CPUID

If it exists, the <!ontents of the CPU identification register.

CPU ERR

If it exists, the contents of the CPU error register.

MEDX

If it exists, the CPU has Maintenance ExaminelDeposit
Instruction.

Contents of Registers

Certain nonperipheral errors include contents of key registers. The format is identical to the contents of registers for
disk and nondisk peripheral errors.

NOTE

The PF error code is used for recording both start-up and power fail
recovery. The error log for a nonnal start-up has a sequence nwnber of
1. For a power fail restart, the sequence number is greater than 1.

10.2.7

Zero File
The Zero File option of ERRDIS lets you clear any or all types of errors from
the ERRLOG.FIL file without generating a report. Mter you select this option,
ERRDIS prompts you for the error type to zero. The default is to zero all error
types. Enter the two-character mnemonic for the error type you wish to clear, or
type Return to accept the default. The following example shows how to clear all
DB errors:

$ RUN ERROR$:ERRDIS
ERRDIS V9.0 RSTS V9.7 EDERE
Input File <ERROR$:ERRLOG.FIL>?
Output to <KB:ERRDIS.OUT>?
He[lp], Ba[d Blocks], Su[mmary], Fu[ll] Report or Ze[ro] File <Summary>? ZERO
Zero Which Error Type <All>? DB
Input File <ERROR$:ERRLOG.FIL>?

10-20

Managing System Errors

10.3 Analyzing System Crashes: ANALYS
Time-sharing operations halt when a crash occurs on a RSTS/E system. The
system then writes an image of read/write memory, tables, and XBUF to the
CRASH.SYS file in account [0,1]. The system bootstraps the system disk, loads
the initialization code into memory, and executes an automatic restart if the CPU
switch register has bit 0 set.
Unless you run the ANALYS program to save crash dump information, the next
system crash causes the CRASH.SYS file to be overwritten and the information
in the file to be lost. Normally, this is done as part of system start-up. See
Chapter 2 for a description of automatic system recovery procedures.
The crash analysis program ANALYS consists of four modules:
•

ANALYS

•

ANALY1

•

ANALY2

•

ANALY3

For simplicity, the documentation refers to only one program, ANALYS, which
functionally includes the four modules.

10.3.1

Running the ANALVS Progrann
Run the ANALYS system program by typing the command:
$ RUN ERROR$:ANALYS<RET>

Mter you press Return, the ANALYS program prints an identification line and
three dialogue questions:
ANALYS V9.7 RSTS V9.7 EDERE
INPUT <[O,l]CRASH.SYS>? <RET>
OUTPUT <ANALYS.DMP<188»? <RET>
Crash error log filename<ERROR$:ERRCRS.FIL>? <RET>
$

The ANALYS program uses the following switches:
•

!DET tells ANALYS to detach.

•

I[NO]DUMP tells ANALYS whether or not to include a memory dump in its
output. The default is !DUMP.

•

/NARROW tells ANALYS whether to use 80 or 132 columns in its output. The
default is 132.

•

/NOSTB tells ANALYS not to include the symbol table in its output.

•

/NOXBUFDUMP tells ANALYS not to include the octal dump of XBUF. It
also implies a !DUMP, so you do not need to include !DUMP with it.

•

/SIL lets Digital analyze the crash and .SIL files submitted with a Software
Performance Report (SPR).

Managing System Errors

10-21

Table 10-6 explains how to answer the ANALYS program questions.
Table 10-6:

ANALYS Program Dialogue
Question and Response

INPUT <[O,l]CRASH.SYS>?
Asks for the name of the file to be analyzed. Type the file specification of
the file you want ANALYS to analyze, which by default is CRASH.SYS in
account [0,1]. Press Return to accept the default response. However, you
may want to keep copies of different crash files.
ANALYS uses the currently installed monitor SIL file to extract symbolic
references. If the crash file being analyzed is not associated with the
currently installed monitor, you must append the ISIL: switch with the
name of the related monitor SIL. This switch allows DIGITAL to analyze
crash and .SIL files submitted with a Software Perfonnance Report (SPR).

OUTPUT <ANALYS.DMP<l88»?
Requests a disk file or a device designator for the output medium. You
can select the default output file ANALYS.DMP by pressing Return. Enter
another file specification if you do not want ANALYS to place the output
in the default account. ANALYS always prints an annotated version of a
memory dump and automatically prints a memory dump in 132-column
format unless you:
1.
2.
3.
4.

Request output to your tenninal (for example, KB:ANALYS.DMP).
Append the /NODUMP switch to omit the memory dump.
Append the /NOXBUFDUMP switch to omit the octal dump of XBUF.
Append the /NARROW switch to a file specification (or type INARROW
and press Return in response to the OUTPUT question if you want
ANALYS.DMP to be the output to be the output file) for SO-column
output.
5. Append the INOSTB switch to omit the symbol table. Do not use this
switch when submitting an SPR.
In any case, you get the annotated output in SO-column fonnat, but with
fewer memory locations printed on each line. If you do not want a dump of
memory, append the /NODUMP switch to your response.
Crash error log filename<[O,3]ERRCRS.FIL>?

Asks for the file specification of the file you want to hold certain error
infonnation from the crash file. Press Return to have ANALYS write this
infonnation to the ERRCRS.FIL in the error package account [0,3].
Note that the IDET qualifier causes ANALYS to detach.
ANALYS retrieves error information saved at the time of a system crash
but not written to the system Hrror logging file ERRLOG.FIL. The ERRDIS
program accepts the file you specify as input and produces an error log
report from its contents. To retain a single continuous error logging file, use
the ERRINT option.

Normally, ANALYS takes more than 30 :minutes to run. When the output stops,
the program automatically terminates and returns to the keyboard monitor
prompt.

10-22

Managing System Errors

10.3.2

ANALYS Output
Output of the ANALYS system program supplies valuable hardware and software
information that a software specialist can use to determine possible causes of
system crashes. It includes:
•

A report similar to SYSTAT

•

A memory dump of the critical contents of memory

•

A listing of all monitor symbols

The report also contains DECnetIE information if you configured your system
with DECnetlE. See the RSTS / E Utilities Reference Manual for an explanation oi
SYSTAT.
Table 10-7 shows the error codes that ANALYS reports in the crash dump data.
Table 10-7:

System Crash Error Code

Error Code (Octal)

Meaning

-1(177776)

Power fail error.

-2(177776)

Jump to O.

-3(177775)

This code is returned when RSTSIE does a crash dump as a
result of halting the CPU and continuing from location 52
(octal).

-4(177774)

This code is returned when RSTSIE detects an unrecoverable
problem and forces a software crash. This can be caused by
either a RSTSIE software problem or faulty harware.

Trap to 4.

Trap to 10.

Trap to 250 (Memory management violation).

Kernel SP Stack overflow.

Trap to 114 (Parity memory error).

To obtain crash information printouts automatically, include the proper commands
in the START.COM file. The commands run ANALYS to preserve the crash
information and run ERRDIS to create a report. See Chapter 3 for a description
of the commands you should include in the START. COM file.

Managing System Errors

10-23

Chapter 11

Device Management and Caching Control
This chapter introduces the concepts of device management and caching control
on your RSTS!E system. Later, it explains DCL commands available to you in
more detail. Table 11-1 summarizes these DCL commands.
Table 11-1:

DCl Commands for Device Management and Caching Control

Command

Meaning
Device Management

ALLOCATE

Allocates a device and optionally assigns it a logical name

DEALl.OCATE

Deallocates the specified device

SET DEVICE

Sets the characteristics for the specified device

SET PRINTER

Set the characteristics of the specified line printer

SHOW DEVICE

Displays information about the specified device

SHOW DISKS

Displays the status of all mounted disks

SHOW PRINTER

Displays the characteristics of the specified printer

Caching Control

11i.1

SET CACHE

Enables data and directory caching

SHOW CACHE

Displays the current cache settings

CREATE~TUAL_DISK

Reserves a section of memory for use as a virtual disk

DELErrE~RTUAL_DISK

Frees the seciton of memory reserved for the virtual disk

Principles of Device Management
Use the SET DEVICE command to set device attributes. You need HWCTL,
HWCFG, and DEVICE privileges to use the SET DEVICE command.
You can disable a specific device for the current time-sharing session with the
/DISABLE qualifier to the SET DEVICE command. To disable a device for
subsequent sessions, use the SET DEVICE command in the START.COM file. To
reenable a disabled device, use the !ENABLE qualifier.
The IRESTRICT qualifier limits the use of the specified device to users with
DEVICE privilege. The INORESTRICT qualifier lets any user allocate or open
the specified device.

Device Management and Caching Control

11-1

11.2 Optimizing Disk Directory Structure: REORDR
The REORDR program can restructure th.e disk directories on your system to
improve disk access time. This process begins once you run REORDR and answer
the set of questions in its dialogue. The following sections discuss:

11.2.1

•

Operations that REORDR performs

•

REORDR dialogue

•

Restructuring the public disk structure

•

Example of the entire process

Why Use REORDR?
The system catalogs RSTSIE files on disk in user file directories (UFD). As you
create, delete, and extend files, the file directory entries become scattered across
the surface of the disk. This scattering increases both the time to open files and
the time to retrieve data blocks.
The REORDR program can perform three operations to restructure disk directories. Each of these operations can improve the performance of your system. Mter
you run the program and answer a sequence of dialogue questions, REORDR
performs the following:
1.

Places the list of file names in as few physical disk blocks as possible. This
reduces the nwnber of directory acceBses the system needs to make for a file
open operation.

Attempts to place all directory infonnation concerning the physical location
of file data (retrieval pointers) for a file into contiguous disk blocks. This
reduces the nwnber of disk directory accesses required to perform disk read
and write operations.

Optionally sorts the file name list in one of four ways, depending on how you
answer the dialogue questions. This reduces the nwnber of directory accesses
the system needs to make for a file open operation.

When RSTS/E creates a new file in an account, it adds the file name at the end
of the current directory list. That is, RSTSIE places files last in the directory if,
during disk initialization, you typed NO to the DSKINT question NEW FILES
FIRST (see the RSTS / E System Installation and Update Guide). Answering
"Yes," while not efficient under most situations, may be more efficient for your
installation. You should have new files first only if users on your system do not
create many new files and their accounts are not large. Otherwise, because of the
way RSTSIE retrieves user directory information, it is more efficient to have new
files last. With the REORDR program, you can choose to organize files by access
date or creation date.
Note that DCL opens all work files in DCLWRK$:, or in your own account if
DCLWRK$: does not exist. In either case, the account with this file in it cannot
be reordered.

11-2

Device Management and Caching Control

11.2.2

Di~.logue Questions and Responses

The REORDR program prints a set of questions after you type RUN
AUXLIB$:REORDR:
$ RUN AUXLIB$:REORDR
REORDR V9.0 RSTS V9.0 EDERE
Sort Directory(s)

(YES/NO) <NO>? YES

Order by CRE[ATION] or ACC[ESS] Date<CRE>? <RET>
In FOR [WARD] or REV[ERSE] Order<FOR>? <RET>
Device and UFD Specification(s)? DB1:[*,*],DB2: [*,*]

Table 11-2 summarizes these questions and their responses.
Table 11-2:

REORDR Dialogue Questions
Question and Response

Sort Directory(s) (YESINO) <NO>?
Type NO or press Return to keep the existing order of the account(s) you ar,
processing. Either response causes REORDR to skip the next two questions
Type YES to have REORDR sort the account(s) you designate. REORDR
asks the next two questions to determine how you want the account(s)
sorted.

Order by CRE[ATION] OR ACC[ESS] Date<CRE>?
Type CRE or press Return to have REORDR sort the account(s) by creation
date. Type ACC to have REORDR use ACCESS date to sort the account(s).
Depending on how the disk was initialized, the access date can have one of
two meanings. It can mean:
•
When the file was last accessed
•
When it was last modified
Note that the abbreviation DLW in the comments part of the SHOW DISKS
command or the DISPLY program shows that a disk's access date is set to
the date of last modification.

In FOR[WARD] or REV[ERSE] Order<FOR>?
Type FOR or press Return if you want REORDER to sort oldest files first.
Type REV to have REORDR create a sorted file with newest files first.

(continued on next page)

Device Management and Caching Control

11-3

Table 11-2 (Cont.):

REORDR Dialogue Questions
Question and Response

Device and UFD Specification(s)?

Enter the device mnemonic(s) and unit number(s) of the device(s) you want
REORDR to process. Include the account(s) you are reordering with the
device designator. For example, the specification DBO:[1,210],DBl:[*,210]
is a valid response. Although SY: (indicating the entire public structure) is
not a valid device, SYO: (indicating the system disk, the device that was
bootstrapped) is acceptable for the device response. An asterisk (*) is a
valid character in either or both of the project-programmer account fields.
It tells REORDR to process all project or all programmer numbers on the
device. You can also include mUltiple specifications if you separate them
with commas (,). The account designators are:
•

•
•
•

[* ,*]

Process all user file directolies on the disk
[p,*]
Process all user file directories with project number p
[*,pn]
Process all user file directories with programmer number pn
[p,pn]
Process only the user file directory with project number p and programmer number pn

NOTE

For disks initialized prior to VB.O,
REORDR can not order account [1,1], even
though it may be implicit in an account
designation.

11.2.3

Reordering Your Disks
1b reorder all disks in the public disk structure, you must include the device
designator for each disk and the account numbers of the files you want reordered:
Device and UFD Specification(s)?DBO:[*,*], DB1: [*,*]

REORDR orders all the directories in the UFD on RP06 unit 0 and then all the
directories in the UFD on RP06 unit 1. Make sure you logically mount the disks
with write access before running the program.
1b prevent damage to a directory, REORDR requires that all files in a UFD be
closed during the entire operation.

11-4

Device Management and Caching Control

NOTE

REORDR verifies that no file is open when it starts to process the
UFD. However, it is unable to detect file opening, creation, or deletion
between the initial check and the completion of the UFD processing.
REORDR can damage the file structure of a disk if a file creation takes
place during reordering. For this reason, follow either of these steps as
a precaution:
II

Disable logins with the SET SYSTEMINOLOGINS command. Make
sure that no other jobs (including spoolers and detached jobs) are
active on the system.
Dismount the disk to be reordered and remount it using the
/NOSHARE qualifier.

The only safe way to reorder a UFD on the public structure is to disable
logins and to make sure t.hat no other user is currently logged in. You
should reorder a UFD on a private disk only when you are sure that no
other job will access the account(s) being reordered on that disk. If you
follow the second step, this is no problem.

11.2.4 Error and Processing Messages
During the reordering of a UFD, the REORDR program holds the reordered UFD
in a temporary file. At the conclusion of the ordering process, ·REORDR copies
the temporary file back to the original UFD. As REORDR processes each UFD, it
generates either a message indicating that. the reorder worked or one that tells
why an attempted reorder did not work. The messages REORDR prints in the
attempt to reorder an account are:
Directory dev:[m,n]
Directory dev:[m,n]
Directory dev:[m,n]
Directory dev: [m,n]
Directory dev:[m,n]
Directory dev: [m,n]
I <-----format----> I

cannot be Reordered
does not exist
has been Reordered
is not Reordered - File filename.type is OPEN
is not Reordered - No write access to UFD
is Null
I <-----·-----------text---------------------> I

The general format and the example of the format contain a device designator
and an account number:
Directory dev:[m,n] I <-----text-------> I
Directory DR3:[1,223] has been Reordered

Device Management and Caching Control

11-5

Table 11-3 explains each message.
Table 11-3:

REORDR Message and Response
REORDR Message and Response

cannot be Reordered
Account [0,1] contains open files during timesharing and thus REORDR
cannot reorder them.
does not exist
REORDR cannot find the account on the device.
has been Reordered
REORDR has successfully reordered the UFD.
is not Reordered. File filename.typ is OPEN
A user is accessing the UFD and thus REORDER leaves the UFD intact.
is not Reordered· No write access to UFD
REORDER encounters a UFD to which it cannot gain write access. It prints
the error message and continues.
is Null
No UFD exists for the account. If REORDR encounters an account without
a UFD, it prints the IS NULL error message.

Fatal errors cause REORDR to print a Inessage of the form:
?REORDR Fatal Error [code]

I <----me:3sage-----> I AT LINE nnnnnn

All fatal REORDR errors include the ?REORDR Fatal Error prefix, a program
error code, a message, and a program line nwnber. Code represents the nwnber
of the program error that caused the fatal condition to occur. (DIGITAL uses the
code nwnber to determine the reason for the fatal error.) REORDR prints a brief
message to help identify the error. Finally, the error message includes the line
nwnber at which the program came to an end. Submit a Software Performance
Report (SPR) to DIGITAL whenever REORDR encounters a fatal error.

11.2.5

REORDR Example
The following example shows the REORDR dialogue and the messages that result
from the reordering process:
$ RUN AUXLIB$:REORDR
REORDR V9.0 RSTS V9.0 EDERE
Sort Directory(s)

(YES/NO) <NO>? YES

Order by CRE[ATION] or ACC[ESS] Date<CRE>? CREATION
In FOR [WARD] or REV[ERSE] Order<FOR>? REVERSE

11-6

Device Management and Caching Control

Device and UFD Specification(s)? DB1: [*,*],DB2: [*,*]
Directory DB1:[0,1] cannot be Reordered
Directory DB1:[1,1] cannot be Reordered
Directory DB1:[1,3] has been Reordered
Directory DB1:[1,4] has been Reordered
Directory DB2:[1,8] has been Reordered
Directory DB1:[1,13] has been Reordered
Directory DB2:[1,44] is Null
Directory DB2:[1,77] is Null
Directory DB1:[1,200] has been Reordered
Directory DB1:[1,201] has been Reordered
Directory DB1:[1,202] has been Reordered
Directory DB2:[2,227] has been Reordered
Directory DB2:[2,240] has been Reordered
Directory DB2: [26,12] has been Reordered
Directory DB2:[100,100] has been Reordered
Directory DB1: [120,50] has been Reordered
Directory DB2:[120,54] is Null
Directory DB2: [210,203] has been Reordered
Directory DB2:[232,15] has been Reordered
Directory DB2:[232,13] has been Reordered
Directory DB2:[251,O] has been Reordered
Sort Directory(s)

(YES/NO) <NO>? <CTRL/Z>

1 ~1.3 Caching Control
Data caching stores blocks from a user file for direct memory access. Directory
caching stores the Master File Directory (MFD), the Group File Directory (GFD),
the User File Directory (UFD), monitor overlay code, and other frequently
accessed system files and directories.
When a user job executes a read request, the RSTSIE monitor performs a disk
access and transfers the requested block(s) of data and directory information fron
the disk to the user job's buffer. With data and directory caching, the monitor
stores the most recently read data blocks in an area of memory called the cache.
If a user job executes a read request for either data and/or directory information
in the cache, the monitor copies the requested information directly from the
cache into the job's buffer. Because the system can thus retrieve data blocks from
memory instead of disk, physical disk access is decreased. This can result in
improved 1/0 throughput and faster response time.
Data caching is only useful for read operations. Every write operation causes an
actual write to the disk. In a write operation that modifies existing cached data,
the data is updated, both in the cache and on the disk.
Both data and directory caching are enabled during system installation. (See the
RSTS / E System Installation and Update Guide.)
You can disable and reenable data caching with the DCL command SET CACHE.
Moreover, DCL commands let you specify caching for a particular file because it
lets you mark the file's UFD entry.
Under the BASIC-PLUS or MACRO programming languages, you can use MODE
values in the OPEN statement or monitor directives to open a file for cached data
access. However, the use of caching MODE values and monitor directives requireE
TUNE privilege. As system manager, you can use the DCL command SET FILE
(see Chapter 12) to designate a file for caching by marking its UFD entry. Once
you mark a file's UFD entry for caching, it is cached on OPEN regardless of the
user's privilege, as long as caching is enabled on the system. See the RSTS / E
Programming Manual for information on the BASIC-PLUS OPEN statement and
the RSTS / E System Directives Manual for information about MACRO directives.
Device Management and Caching Control

11-7

When you enable data caching, the cache receives all data transfer requests that
are otherwise directed to the disk driver. Read operations on data that is in the
cache occur without placing a load on the disk driver. The monitor constantly
updates the cache so that it contains the most recently requested data for cached
files.

11.3.1

Size of the Cache
The RSTSIE monitor allocates space for the cache from the Extended Buffer Pool
(XBUF). You specify the amount of this allocation with DCL command qualifiers
that set the size of a cache cluster (/CLUSTERSIZE=n) and the nwnber of
clusters in the cache (IMAXIMUM=n, IDIRECTORY_MAXIMUM=n, and IDAT~
MAXIMUM=n). The size of a cache cluster (1, 2, 4, or 8 blocks) determines the
amount of data that is treated as a unit in a read request and, in many cases,
the nwnber of read requests that can be 'resolved in the cache before access to the
disk driver is required.
For example, when the cache cluster size is eight blocks, any read operation that
installs data in the cache causes eight physically contiguous blocks (including the
requested blocks) to be installed. To ensure that only requested data is read in to
the cache, make the cache cluster size equal to or less than the pack cluster size
(set during disk initialization) of the disk with the most files to be cached.

11.3.2

Sequential and Random Caching Modes
You can cache the data in a file in two ways: sequential or random. You can
specify either of these modes for a particular file with the SET FILE command
qualifiers (see Chapter 12).
If a job executes a read operation on a cached file and the data is not in the
cache, that data is installed in the cache if there is free space or if a current cache
cluster is eligible for replacement.
A cluster is eligible for replacement if either of the following conditions is in
effect:

11.3.3

•

The last block of the cluster has been read in sequential mode.

•

The cluster has been in the cache without being read for more than the
minimwn residency time, which is set using the SET CACHE command.

Random Mode Caching
DIGITAL recommends caching a file in random mode if the file is RMS index
structured or a heavily overlaid program. With an RMS indexed file structure,
the monitor does not access the data in the file sequentially but must check the
file index each time it gets more data. Thus, it reads the index, the data, and the
index again.
As an example of random cache operation, consider a read operation executed
on a file whose UFD entry is marked for random caching. When a read on the
cached file occurs, the monitor examines the contents of the cache to determine if
the requested data is present. If the data is in the cache, the data is copied from
the cache cluster that contains it. The data is made available to the program and
the cache cluster time of last access is updated.

11-8

Device Management and Caching Control

If the requested data is not in the cache, the monitor first attempts to allocate
more ofXBUF to install the new data (cache limits permitting). IfXBUF cannot
accommodate more data or the request exceeds the caching limits set with SET
CACHE, then the monitor examines the list of cache clusters to find one that is
eligible for replacement. If a cluster is not available, the monitor cannot install
the requested data in the cache; a normal disk read is automatically performed.
If the cache cluster is eligible for replacement, the monitor installs the requested
data in that cache cluster and makes the data available to the program.

11.3.4

Sequential Mode Caching
DIGITAL recommends caching a file in sequential mode if you access the contenb
of the file sequentially. As an example of sequential cache operation, consider
a read operation executed on a file whose UFD entry is marked for sequential
caching. When a read on the cached file occurs, the monitor examines the
contents of the cache to detennine if the requested data is present. If the data
is in the cache, the data is copied from the cache cluster that contains it. The
data is made available to the program. If, in the process, the last block of a cache
cluster was read, that cluster is made available for replacement in the cache.
Furthermore, when you request a read operation on any block of a cache cluster
(except the last block), the monitor installs a full cluster of data in the cache.
That is, if you use the SET CACHE command qualifier ICLUSTERSIZE=n to set
a cache cluster size of 8 blocks and then execute a read on the first block of a file,
the requested data plus the next 7 blocks from the disk are read into the cache.
Thus, the contents of the next 7 reads can be satisfied from the cache. When the
last block of the cache cluster :is read, the cluster is immediately made available
for the installation of new data as requested by other read operations.
When a read operation from a sequentially cached file causes more than one
cache cluster to be read, all of the requested data blocks are made available to
the program. However, no data is installed in the cache for any cluster whose last
block was read. (That is, only the last cluster can be installed and only if it was
incompletely read.) If the cache cluster size is defined as one block and sequentia1
mode is specified, no data blocks are installed in the cache. Thus, every data
block is the last block in a cache cluster.

11.4 Cclching Guidelines
The relationship between cache cluster size, which you set with a SET
CACHE/CLUSTERSIZE=n cODlmand, and the pack cluster size of the file
can have an effect on the efficiency of caching. The cache cluster size determines
the number of blocks on a disk copied into the cache when a cluster is installed.
If the cache and pack cluster sizes are equal, there is direct correspondence
between the data that will probably be read and the data copied from the disk
to the cache. If the cache cluster size is larger than the pack cluster size, the
data you request plus some additional (and possibly unrelated) data is copied to
the cache, wasting cache clusters. If the cache cluster size is less than the pack
cluster size, a read request may require multiple disk accesses to move all of the
data into several different cache clusters.

Device Management and Caching Control

11-9

Figure 11-1 shows the flow of UFD entry and MODE value checks perlormed by
the monitor. The result of these checks determines the type of caching to be used
on the file.
Figure 11-1:

Caching Mode Checks

Yes

Caching
Performed

No
Caching
Performed

Yes

H-MK-00075-00

The size relationship between the cache cluster and RMS indexed file bucket is
similar to the size relationship between the cache cluster and the pack cluster.
Efficient caching is obtained when the cache cluster and bucket sizes are equal
because RMS always reads an entire bucket at once.
Consider the following caching guidelines:

11-10

•

When setting defaults, the amount ofXBUF you allocate to data and directory
caching should be as large as possible while maintaining sufficient memory for
user jobs, resident libraries, and run-time systems. However, an exceptionally
large allocation can increase swapping activity, especially on systems with
less than 256K words of memory.

•

On systems that use RMS heavily, consider the use of the RMS resident
library and the library's effect on XBUF allocation. Typically, on systems with
less than 256K words of memory, the resident library should receive priority
in memory allocation, to the extent of limiting XBUF allocation. In such a
case, data caching may become ineffective and you probably should disable it.

Device Management and Caching Control

•

1b optimize directory caching, set all UFD cluster sizes to 16 blocks if possible.

•

Files that are frequently accessed sequentially (including RMS sequential and relative files and heavily overlaid programs) should have their
UFD entries marked for sequential caching with the DCL command SET
FILE/CACHE=SEQUENTIAL.

•

Files that are frequently accessed randomly (including RMS relative and
indexed files) should have their UFD entries marked for random caching
with the DCL command SET FILE/CACHE=RANDOM. Because of the way
RSTSIE maintains RMS indexed files, they should always be flagged as
random even if they are accessed sequentially. Where possible, RMS indexed
file bucket sizes should be a power of two to correspond with caching cluster
sizes.

•

Use data caching judiciously on RMS indexed files. You gain the most benefit
frolm caching an indexed file when it is subject to many read requests and is
open for long periods of time. In addition, you should allocate more XBUF to
compensate for the lower caching "hit" rate (the probability that the data you
want to read is in the cache) on indexed file data. The hit rate is generally
lower on indexed files than it is on sequential files or directory blocks.

•

Where possible, all files should be contiguous to reduce window turning.
Where contiguous files are not possible, you should calculate the file cluster
size as follows:
1.

Divide the file size by seven

Round the result to the next higher power of two (not greater than 256
nor less than the pack cluster size)
NOTE

If you preextend the file and specify a cluster size of zero, RSTS/E
does this calculation for you.
•

The minimum residency tiule is meant to keep some useful data in the cache,
even when the cache is very small. Minimum residency time is also meant to
make sure that if you want to access previously accessed data the system will
find it in the cache, rather than having to read information off disk each time
you request more. For systems with large caches or rapidly changing disk
access patterns, it may be advantageous to reduce the minimum residency
time from its default value of 60 seconds. Use the IKEEP=n qualifier to
modify minimum residency time.

Caching is not the most efficient data access method for all files under all circumstances. But, in general, a high-access file is a logical candidate for caching. A
high-access file meets one or more of the following conditions:
•

Accessed by more than one user at a time

•

Opened for long periods of time

•

Frequently accessed

•

Accessed sequentially (indexed files)

•

Primarily read

•

Is an overlaid program (such as a user program, TKB, or DTR)

However, because of the large number of different applications, you must experiment to arrive at the best use of caching on your system.
Device Management and Caching Control

11-11

11.5 Virtual Disks
You can create a virtual disk that acts like a physical disk, only faster. Use the
DCL commands CREATENIRTUAL_DISK and DELETENIRTUAL_DISK to
create and delete the virtual disk. You must have the HWCFG and INSTAL
privileges to use these commands. A syst.em can have no more than one virtual
disk.
The virtual disk speeds up 110, but it uses up a great deal of CPU time. Do not
create a virtual disk on a system that is short of CPU time. Use the virtual disk
when 110 is your principal bottleneck.
You can create a virtual disk at any time, but it takes up a large amount of
memory. Therefore, DIGITAL strongly recommends that you create the virtual
disk immediately after booting, before any libraries or auxiliary run-time systems
get loaded. This minimizes the amount of time spent swapping things in and out
of memory at the time of creation.

11.6 DCl Command Descriptions
The following sections describe the DCL commands and qualifiers for Device
Management and Caching Control.

11-12

Device Management and Caching Control

1"1.6.1

ALLOCATE Command
The ALLOCATE command allocates a device and optionally assigns it a logical
name. ALLOCATE does not require any privileges to allocate a device to your
own job.

Format
ALLOCATE

dev: [/ogicaLname[:]]

Command Qualifiers
IJOB=n
ISEIZE

Defaults
none
none

Prompts
Device:

Command Parameters
dey:
Specifies the device to be allocated.
loglcaLname[:]
Specifies the logical name to be assigned to the device.

Comrnand Qualifiers
IJOB=n
Allocates the device to the specified job. You need HWCTL privilege to use this
qualifier.
ISEIZE

Reallocates a device currently allocated to another job. The /SEIZE switch
returns an error message if a file is open on the device. You need HWCTL
privilege to use this qualifier.

Device Management and Caching Control

11-13

11.6.2

CREATE/VIRTUAL_DISK Command

The CREATENIRTUAL_DISK command reserves a section of memory for use
as the virtual disk. Once the virtual disk has been created, you can initialize
and mount it using the INITIALIZE and :MOUNT commands. Use the SHOW
MEMORY command to verify the location of the virtual disk after it has been
created. Note that the virtual disk appears as locked out in the memory list
until it has been initialized and mounted. You need the INSTAL and HWCFG
privileges to use the CREATENIRTUAL_DISK command. Example:
$ CREATE/VIRTUAL DISK 38
Virtual Disk created with size of 40 blocks
$ INITIALIZE/EXERCISE=O DVO: V
$ MOUNT DVO: V
$ SHOW MEMORY

Memory allocation table:
End Length Permanent
Start
Temporary
99K ( lOOK) MONITOR
OK l02K (
3K) MONITOR
lOOK l03K 426K ( 324K)
(User)
436K ( 10K) Virtual Disk
427K (User)
437K 446K ( 10K)
DAPRES LIB
(User)
447K 450K (
4K)
RMSRES LIB
451K 51lK ( 6lK) ** XBUF **
5l2K *** END ***
$ SHOW DISK
Disk Structure:
Dsk Open
Size
DVO
0
40
DUl
18
83204

Free
Clu
28 70%
1
47718 57%
2

Err Name

o V

o RD32

Level
1.2
1.2

Comments
Pri, DLW
Pub, DLW

Format
CREATE/VIRTUAL_DISK
Command Qualifiers
f ADDRESS[ =address]
f[NO] lOG

Defaults
highest available memory
flOG

Prompts
Virtual disk size:

Command Parameters
disk_size
Specifies the size, in blocks, of the virtual disk. If the disk size is not a multiple
of 4, the system rounds it up to the next multiple of 4 blocks. A confirmation
message tells you the actual size of the virtual disk.

11-14

Device Management and Caching Control

Command Qualifiers
/ ADDFIESS[=add ress]

Specifies the address, in K-words, at which to locate the virtual disk. The address
must be between 1 and 2043. Use the SHOW MEMORY command to select an
available address to locate the virtual disk. The system returns an error message
if the address is above the memory on the system, or if any part of the memory is
in use.
If you do not specify an address, the system puts the the virtual disk at the
highest section of memory in which it will fit. The system returns an error
message if there is not enough available memory to place the disk.
/[NO]LOG

Specifies whether or not you get a confirmation message saying the virtual disk
was created. The confirmation message shows the actual size of the virtual disk.
The default is /LOG.

Device Management and Caching Control

11-15

11~6.3

DEALLOCATE Command

The DEALLOCATE command deallocates the specified device. DEALLOCATE
does not require any privileges.

Format
DEALLOCATE

dev[:]

Command Qualifiers

Default~.

tALL

none

Prompts
Device:

Command Parameters
dey:
Specifies the device to be deallocated.

Command Qualifiers
tALL

Deallocates all devices that were allocated to the current job.

11-16

Device Management and Caching Control

11.6.4

DEL.ETE/VIRTUAL_DISK Command

Frees up the section of memory reserved for the virtual disk. 1b delete the
virtual disk, you must dismount it. It must not have files opened on it or an
error will be returned. You need the INSTAL and HWCFG privileges to use the
DELETENIRTUAL_DISK command. Example:
$ DELETE/VIRTUAL_DISK
DELETE/VIRTUAL_DISK

?Virtual disk is mounted or has open files
$ SHOW DISK
Disk Structure:
Dsk Open
Size
DVO
0
40
DUl
18
83204

Free
Clu
28 70%
1
47718 57%
2

Err Name
0 V
0 RD32

Level
1. 2
1.2

Comments
Pri, DLW
Pub, DLW

$ DISMOUNT DVO: V
$ DELETE/VIRTUAL_DISK
Virtual disk deleted

Format
DELETE/VIRTUAL_DISK
Command Qualifiers
/[N 0] lOG

Defaults
flOG

Prompts
none

Command Qualifiers
/[NO]lOG
Specifies whether or not the system displays a confirmation message telling you
the requested action was taken. The default is /LOG.

Device Management and Caching Control

11-17

11.6.5

SET CACHE Command
The SET CACHE command lets you set caching parameters for the system as a
whole. The SET CACHE command lets you enable data and directory caching on
the system. Note that commands in the system startup command file normally
enable caching automatically. You need TUNE privilege to use the SET CACHE
command.
When you use SET CACHE with no qualifiers, the last settings specified are
applied. The initial system defaults are:
/CLUSTERSIZE

4 blocks

!MAXIMUM

No limit on total cache allocation

!DIRECTORY_MAXIMUM

No limit on directory caching

!DATA_MAXIMUM

No limit on data caching

!KEEP

60 second8

If you use SET CACHE/ALL, all read requests are cached. The type of caching
(sequential or random) for a particular file is detennined by the file's UFD entry
as specified in the SET FILE command or by an OPEN MODE specification.
DIGITAL recommends that you avoid using SET CACHE/ALL. It generally
reduces performance except for a small number of specific applications.
If you use SET CACHEIFILE, a file is cached on the basis of its UFD entry and/or
the specified OPEN MODE. That is, a file is cached if either the UFD entry or
OPEN MODE specifies caching; it is cached sequentially if either the UFD entry
or OPEN MODE specifies sequential.
If you use SET CACHEINOFILE, RSTSIE does not cache files. This qualifier is

used to enable directory caching but disable data caching on the system.
Data and directories on a virtual disk (DV:) are never cached, since they are
already in memory.

11-18

Device Management and Caching Control

Figw'e 11-2 shows the mechanism by which the monitor decides the amount
of caching to be performed. The figure shows the flow of DCL commands and
caching specifications examined by the monitor.
Figure 11-2:

Monitor Caching Checks

Yes

Cache
Sequential

Yes

[

Cache
Random
- - - - '

H-MK-00076-00

Use the SET CACHE/CLUSTERSIZE=n command to specify the cache cluster
size as 1, 2, 4, or 8 blocks for each cluster. The cache cluster size controls the
nwnber of contiguous data blocks that are copied from disk to the cache whenevel
file data is cached. The cache cluster size should be small enough to contain only
the target blocks but large enough to reduce the nwnber of disk accesses. This
means you must anticipate data requests and ensure that the cache cluster size
is equal to the file cluster size of the files most frequently accessed.
The SET CACHEIMAXIMUM=n command (where n is in the range 0 to
UNLIMITED) specifies the total number of cache clusters used by both directory and data caching. The default limit is UNLIMITED, which means that cachE
cluster allocation is limited only by the amount of XBUF available.
The SET CACHEIDIRECTORY_MAXIMUM=n and SET CACHEIDAT~
MAXIMUM=n commands (where n is in the range 0 to UNLIMITED) specify
the cache cluster allocation for directory and data caching, respectively. The
specified allocation is an upper limit. Thus, if a 40K-word XBUF is defined at
system

Device Management and Caching Control

11-19

installation and !DIRECTORY_MAXIMUM and !D~rA-MAXIMUM both specify
a number of cache clusters equivalent to .25K, data can use space in the cache up
to a maximum of 25K, which leaves a 15K minimum for directory caching. The
reverse is also true. In this manner, data and directory caching are guaranteed
a minimum allocation and the amount of overlap is controlled, which permits the
cache to dynamically adjust to system and program requirements.
The SET CACHEIKEEP=n command specifies the cache replacement time in
seconds. This is the amount of time that RSTSIE keeps an unused cache cluster
in memory before the monitor makes it elligible for replacement by another cache
cluster.
The SET CACHE!DISABLE command disables all data and directory caching
on the system. If you reenable caching following a disable and do not specify
caching parameters (/CLUSTERSIZE, !MAXIMUM, !DIRECTORY.-MAXIMUM,
!DATA_MAXIMUM, and IKEEP), the default parameters are those that were in
effect before the disable.

Format
SET CACHE
Command Qualifiers

Defaults

IAll
IClUSTERSIZE=n
IOATA_MAXIMUM=n
IOIRECTORY_MAXIMUM=n
IOISABlE
IENABlE
I[NO]FllE
IKEEP=n
IMAXIMUM=n

none
IClUSTERSIZE=4
none
none
IENABLE
IENABLE
IFllE
none
none

Prompts
none

Command Qualifiers
IAll

Caches all data transfers.
IClUSTERSIZE=n

Specifies a cache cluster size of n with values equal to 1, 2, 4, or 8 blocks. The
default is ICLUSTERSIZE=4.
IDATA_MAXIMUM=n

Specifies the maximum number of clusters (based on available memory) used for
data caching in the range 0 to 65535 or the keyword UNLIMITED.
IDIRECTORY_MAXIMUM=n

Specifies the maximum number of cache clusters used for directory caching in the
range 0 to 65535 or the keyword UNLINHTED.
IDISABlE

Disables data and directory caching.
11-20

Device Management and Caching Control

IENABLE
Enables caching on the system using the previous settings as the defaults. The
default is !ENABLE.
I[NO]FILE
Indicates if file data should be cached. The /FILE qualifier indicates that file
data is cached depending on their file attributes or OPEN MODE. The /NOFILE
qualifier indicates that no file data is cached. The default is /FILE.
IKEEP="
Specifies the cache replacement time in seconds that an unused cache cluster is
kept in memory before the space can be used for another cache cluster. The value
of n is in the range 0 to 65535.
IMAXIMUM="
Specifies the maximwn nwnber of cache clusters used for directory and data
caching in the range 0 to 65535 or the keyword UNLIMITED.

Device Management and Caching Control

11-21

11.6.6

SET DEVICE Command
The SET DEVICE command sets the characteristics for the specified device. You
need HWCTL privilege to use tENABLE, HWCFG to use !RESTRICT, HWCFG
and HWCTL to use !DISABLE, and HWCFG and DEVICE to use INORESTRICT.

Format
SET DEVICE

de v:

Command Qualifiers

Defaults

IDISABLE
IENABLE
I[NO]RESTRICT

none
none
none

Prompts
Device:

Command Parameters
dev:
Specifies the device whose characteristics you want to change.

Command Qualifiers
IDISABLE

Disables the specified device. No other qualifiers can be specified with !DISABLE.
IENABLE

Enables a device that has been disabled by the SET DEVICE!DISABLE command.
I[NO]RESTRICT

Tells the system which users can open oX' allocate the specified device. The
!RESTRICT qualifier lets only users with DEVICE privilege open or allocate the
device. The INORESTRICT qualifier lets any user open or allocate the device.

11-22

Device Management and Caching Control

1'1.6.7

SET PRINTER Command
The SET PRINTER command sets the characteristics of the specified line printel
You need HWCFG privilege to use the SET PRINTER command.

Format
SET PRINTER

dev[.]

Command Qualifiers
IBACKSPACE[=mode]
I[NO]CONTROLS
I[NO]CR_PREFIX
IDEVICE_TYPE=type
I[NO]EIGHT_BIT
I[NO]EOT
I[NO]FILL
ILOV"ERCASE
I[NO]OMIT_CR
IPAGE_LENGTH=n
I[NO]SPECIAL_CHARACTER[ ==char]
I[NO]TAB
IUPPERCASE
!WIDTH

Defaults
IBACKSPACE=REAL
none
none
none
none
none
none
none
none
none
none
none
none
none

Prompts
Printer:

Command Parameters
dev[:]
Specifies the line printer whose characteristics you want to set.

Command Qualifiers
IBACKSPACE[=mode]
Tells the system how to handle backspace characters for thls printer. The possible
modes are:

•

CONTROL-Treats a backspace like other nonprinting characters.

•

REAL-Uses the actual backspace feature on the specified printer.

•

SIMULATE-Simulates the action of backspace on the specified printer by
issuing a carriage return (CR) followed by the appropriate number of spaces
to position the print head one colwnn to the left of its original position.

The default is IBACKSPACE=HEAL.
I[NO]CONTROLS
Tells the system how to handle nonprinting characters. /CONTROLS sends
nonprinting characters to the line printer. /NO CONTROLS discards nonprinting
characters or uses up arrow mode.
Device Management and Caching Control

11-23

I[NO]CR_PREFIX

Tells the system how to handle line feed (LF), vertical tab (VT), and form feed
(FF) characters. The options are:
•

ICR_PREFIX inserts a CR before a LF, a VT, or a FF.

•

fNOCR_PREFIX tells the printer to do an implied CR before a LF, a VT, or a
FF.

IOEVICE_TYPE=type

Sets the characteristics of the line printer to the specified type:
•

LA180-Sets the characteristics of the line printer to those of an LA180.

•

LNOI-Sets the characteristics of the Hne printer to those of an LNOl.

•

LPll-Sets the characteristics of the line printer to those of an LPll.

When you use !DEVICE_TYPE, you automatically set all the other printer
characteristics as well.
I[NO]EIGHT_BIT

Tells the system how to handle 8-bit characters. The lEIGHT_BIT qualifier lets
you send 8-bit characters to the line printer. The fNOEIGHT_BIT qualifier does
not let you send 8-bit characters to the line printer.
I[NO]EOT
Tells the system how to handle an end-of-transmission character. The /EOT
qualifier sends an end-of-transmission (CtrllD) to the line printer. The INOEOT
qualifier treats an end-of-transmission (CtrllD) like other nonprinting characters.
I[NO]FILL

Tells the system how to handle fill characters. The IFILL qualifier inserts fill
characters after form feeds. The fNOFILL qualifier does not insert fill characters
after form feeds.
ILOWERCASE
Sends lowercase characters to the line printer.
I[NO]OMIT_CR

Instructs the system on how you want the CR character handled. The IOMIT_
eR qualifier tells the system to omit the eR if the next character is a LF. The
fNOOMIT_CR qualifier tells the system to always send CR to the printer.
IPAGE_LENGTH=n

Sets the form length of the line printer to n. The form length can be in the range
1 to 255.
I[NO]SPECIAL_CHARACTER[=char]

Instructs the system on how you want the printer to handle its special character.
The /SPECIAL_CHARACTER=char qualifier tells the printer to print its special
character whenever it receives the specified char. Char can be given by its
numeric value in the range 0 to 255 or by enclosing the character in quotation
marks (It) if it is printable. The fNOSPECIAL_CHARACTER qualifier tells the
system to disable the special character feature for the specified line printer.
I[NO]TAB

Instructs the system on how you want the TAB character handled. The /TAB
qualifier indicates that the printer has hardware TAB capability. The fNOTAB
qualifier indicates that the printer does not have hardware TAB capability and
the system has to translate TAB characters to the appropriate number of spaces.
11-24

Device Management and Caching Control

IUPPERCASE

Tells the system to convert lowercase characters to uppercase before sending
them to the line printer.
IWIDTH=n

Sets the width of the line printer to n. The width can be in the range 0 to 254.

Device Management and Caching Control

11-2f

11.6.8

SHOW CACHE Command
The SHOW CACHE command displays the current caching parameters for your
system. SHOW CACHE does not require any privileges.
The listing includes information on cluster sizes, and data caching. For example:
$ SHOW CACHE
Caching configuration

Caching enabled for Directories and Data
Cache Keep timer = 60 seconds
Cluster size = 4
Data caching = File selectable
XBUF usage (in K-words)
Total cache:
Directory cache:
Data cache:
ASSIGN/SYSTEM tables:
LOAD/OVERLAY tables:
LOAD/INDEX tables:
DeL Buffers:
Quota tables:
Window turn blocks:

229.00
200.00
29.00
0.94
3.16
9.22
1.13
0.38
0.00

(maximum clusters allowed: Unlimited)
(maximum clusters allowed: Unlimited)
(maximum clusters allowed: Unlimited)

(small buffers in use: 0)
(maximum blocks used: 5)

In this example, SHOW CACHE reports that:

•
•
•
•
•
•

Directory and data caching are currently enabled
The cache keep timer is set to 60 seconds
The cache cluster size is 4 blocks
Only selected files are cached
No limit is set on the use ofXBUF for directory and data caching
229 K words ofXBUF are currently being used for directory and data caching

• 200 K words of XBUF are currently being used for directory caching

•
•

0.94 K words of XBUF are currently being used to store system logicals
(ASSIGN/SYSTEM)

•

3.16 K words of XBUF are currently being used to store monitor overlays
(LOAD/OVERLAY)

•

9.22 K words of XBUF are currently being used to store disk allocation
indexes (LOADIINDEX)

•
•

1.13 K words ofXBUF are currently being used for DCL buffers

•
•
•

11-26

29 K words of XBUF are currently being used for data caching

0.38 K words ofXBUF are currently being used to store logged-in disk quota
tables
No small buffers are being used to store logged-in disk quotas
No XBUF is currently being used for storing file window turn blocks
A maximum of 5 window turn blocks have been used since the system was
started

Device Management and Caching Control

Format
SHOW CACHE
Command Qualifiers

Defaults

none

Prompts

none

Device Management and Caching Control

11-27

11.6.9

SHOW DEVICE Command

The SHOW DEVICE command displays information about the specified device.
If you do not specify a device, RSTS/E displays information about all devices.
SHOW DEVICE does not require any privileges.

Format
SHOW DEVICE

[dev[:]]

Command Qualifiers
fAll
fAllOCATED

Defaults
fAll

Prompts
none

Command Parameters
dev[:l
Specifies the device whose characteristics you want to display.

Command Qualifiers
fAll
Displays information about all devices configured on the system. A device
specification overrides the fALL qualifier.
fAllOCATED
Displays information about all devices that are currently open or allocated. The
display includes devices that are open as non-file-structured devices.

11-28

Device Management and Caching Control

11,,6.10

SHOW DISKS Command

The SHOW DISKS command displays the status of all mounted disks. SHOW
DISKS does not require any privileges.

Format
SHOW DISKS
Command Qualifiers
none

Defaults
none

Prompts
none

Device Management and Caching Control

11-29

11.6.11

SHOW PRINTER Command
The SHOW PRINTER command displays the characteristics of the specified
printer. SHOW PRINTER does not require any privileges.

Format
SHOW PRINTER

dev[:]

Command Qualifiers
fAll

Defaults
fAll

Prompts
none

Command Parameters
dev:
Specifies the printer whose characteristics you want to display.

Command Qualifiers
fAll
Displays information about all the printers configured on the system. A printer
specification overrides the fALL qualifier.

11-30

Device Management and Caching Control

Chapter 12

Managing System Files and Logical Names
This chapter describes the DCL commands that you use to manage RSTSIE
system files and logical names. Table 12-1 lists and briefly describes the
DCL commands available to you. The rest of the chapter lists each command
alphabetically and explains it in more detail.
Table 12-1:

DCl Commands for Managing System Flies and logical Names

Command

Meaning

----------------------------------------Run Time System Control

INSTALURUNTIME_SYSTEM

Adds the named run-time system to the list of installed
run-time systems.

REMOVEIRUNTIME_SYSTEM

Removes the name of a run-time system module as an
entry in the run-time system taMe.

SHOW RUNTIME_SYSTEM

Displays information about all installed run-time
system.

UNLOADIRUNTIME_SYSTEM

Unloads the named run-time system from memory.

Resident Library· Control
INSTALULIBRARY

Adds the named library to the list of installed resident
libraries.

REMOVEILIBRARY

Removes the named file as a resident library.

SHOW LIBRARY

Displays information about all installed libraries.

UNLOADILIBRARY

Removes the named resident library from memory.
System Logical Names Control

ASSIGN/SYSTEM

Assigns a system-wide logical name to the specified device type and unit. If you include an account number,
the ASSIGN/SYSTEM command also associates that
PPN with the logical name.

(continued on next page)

Managing System Files and Logical Names

12-1

Table 12-1 (Cont.):

DCl Commands for Managing System Flies and logical
Names
System Logical Names Control

DEASSIGN/SYSTEM

Removes the entire entry for the logical name from
the system logical table. It can also remove logical
names for devices on your system. Device mnemonic
names for system devices have permanent entries in
the logical table. If you use the DEASSIGN/SYSTEM
command to remove a logical name associated with a
device, only the logical name goes away, not the device
mnemonic. You can then assign a new name to that
device.

SHOW LOGICAIlSYSTEM

Displays information about a system-wide logical
name.

System wide Command Language Control
DEFINE/COMMAND/SYSTEM

Creates or 'redefines a system-wide command.

DELETE/COMMAND/SYSTEM

Deletes the specified command from the list of defined
system-wide commands.

SHOW COMMAND/SYSTEM

Displays information about the specified system-wide
command.

System FilE~ Control
DUMP/SYSTEM

Records in the CRASH.SYS file the current state of the
system so you or field service has a way to investigate
reasons for system malfunctions.

INSTALUERROR_FILE

Installs the specified file as the system error message
file.

INSTALUOVERLAY_FILE

Installs the specified file as the system overlay file.

I NSTALUSWAP_FILE=n

Adds the specified file to the list of installed swap files.
Valid values of swap-index are 0, 1, or 3.

REMOVEIERROR_FILE

Removes the system error message file.

REMOVE/OVERLAY_FILE

Removes the system overlay file.

REMOVE/SWAP_FILE=n

Removes the specified swap file from the list of
installed swap files. Valid values of swap-index are 0,
1, or 3.

SHOW FILE

Displays information about the specified file(s).

User File Control

12.1

SET FILE

Sets the file characteristics.

SHOW LOGICALS

Displays information about a user-defined logical
name.

Run-Time System Control
A run-time system on RSTS/E is the CODlmon, shareable part of a user job
and, in some cases (for example, BASIC-PLUS), is the interface between the
user's executable code and the monitor. Run-time systems control execution
of jobs when the monitor allows the jobs to run. The monitor sets up jobs and

12-2

Managing System Files and Logical Names

establishes an environment in which a run-time system can function. The
BASIC-PLUS run-time system interprets requests made by a user job, translates
those requests into a format the monitor understands, and transmits the requestf:
of the monitor. The monitor processes the requests and passes either data or an
error to the job. The run-time system decides whether to interpret the data or
log an error. Jobs running under the BASIC-PLUS run-time system cannot make
direct requests to the monitor.
Other run-time systems may not be language interfaces in the same sense that
the BASIC-PLUS run-time system is. Jobs under their control can make direct
requests of the monitor. In such cases, the major role of the run-time system is to
load compiled programs to handle system errors, and return control, on program
termination, to the job's keyboard monitor. Some of these run-time systems can
also emulate the action of certain system directives from other operating systems,
such as RT11.
A run··time system interprets user requests when it is a language interface for a
user job. A computer language provides a standard means by which many users
can define data, process it, and obtain results. The language relieves each user of
the need to learn the workings of the monitor. The run-time system under which
the language operates is more efficient because multiple users can share the same
code. This ability to share code is more efficient for the system because each user
need not have a copy of the common code.
All RSTSIE installations have DCL as the default keyboard monitor and can
optionally have auxiliary run-time systems. DCL:
•

ABows users access to system resources

•

Interprets system command requests

•

Performs housekeeping chores

An auxiliary run-time system:
•

Provides additional shared computer processing functions

•

Enables processing in other languages

Such an auxiliary run-time system may compile and/or execute BASIC-PLUS-2,
COBOL, DIBOL, FORTRAN-Iv, FORTRAN-77, or other language operations.
The system account [0,1] stores run-time systems as contiguous files with .RTS
file types. You must store DCL on the system disk because, at the start of
timesharing, the system disk is the only disk mounted on the system. You can
store auxiliary run-time systems on any disk, either public or private. (Auxiliary
run-time system files can also be stored in accounts other than [0,1].) All runtime system files are contiguous because they must be loaded into memory in the
fastest time possible.
Auxiliary run-time systems are created either during system installation or
afterwards. The monitor structures that describe these run-time systems do not
exist automatically at the start of timesharing. Therefore, you must provide the
commands to add monitor structures for the auxiliary run-time systems that you
want to make available for any given time-sharing session. The structure for an
auxiliary run-time system is dynamic and can be removed during a time-sharing
session.
The DeL run-time system commands let you control operations of auxiliary
run-time systems. The DCL commands perform additional processing for
certain commands. 1b assist the system programmer, the command descriptions
distinguish between the monitor call action and the DCL command action.
Because run-time systems under development need certain special controls,
Managing System Files and Logical Names

12-3

some features of run-time system control are intended for Digital personnel only.
Nevertheless, all features are described to give you a broader perspective.
Digital recommends that you make auxiliary run-time systems available by
commands in the START.COM file. By reading the descriptions in this section,
you can 'understand what actions should occur for each run-time system
command.

12.1.1

Adding and Removing Auxiliary Run-Time Systems
Because the structures that define auxiliary run-time systems are temporary,
you must create them for each time-sharing session. The INSTALURUNTIME_
SYSTEM command creates the necessary structure to enable an auxiliary runtime system to function. INSTALLIRUNTIME_SYSTEM command qualifiers can
alter predefined characteristics of the run-time system.
The INSTALLIRUNTIME_SYSTEM comInand creates a run-time system description block with a general small buffer. A file with the name of the run-time
system and a file type of .RTS can be in any account on the disk specified in the
command. A run-time system with the same name cannot already exist on the
system. The system makes sure that the file found is contiguous and has the
proper format and parameters.
You can specify qualifiers to override predefined characteristics. The
INSTALLIRUNTIME_SYSTEM command extracts information from the run-time
system file to establish the proper entries in the run-time system description block. To establish the entries, the lnonitor opens the file, reads it, and
sets up the predefined characteristics to be placed in the description block.
You can change the predefined characteristics by the qualifiers: fPREFIX=n,
I[NO]UNLOAD, I[NO]ERROR_LOGGING, ![NO]SHAREABLE, ![NO]READ_
ONLY, and I[NO]KEYBOARD_MONITOR.
You can also change the predefined characteristics by negating these qualifiers.
For example, if you were testing a run-time system with no special command
decoding capabilities, you would use the qualifiers INOSHAREABLE, INOREAD_
ONLY, INOERROR_LOGGING, and !UNLOAD to add the run-time system. The
INOSHAREABLE makes sure that no other job tries to share the untested code;
INOREAD_ONLY is necessary for setting breakpoints with ODT; INOERROR_
LOGGING prevents false errors generated by the run-time system from cluttering
the system error log; /UNLOAD makes sure that the run-time system's image is
not left in memory but is reloaded from disk immediately preceding every entry
to it.
The qualifiers and negations of the qualifiers do not alter the predefined characteristics in the run-time system file; only the characteristics defined in the
description block are altered. The characteristics remain in effect until you
remove the run-time system.
The ![NO]ADDRESS[=n] qualifier lets you take advantage of high speed memory
on the system or avoid fragmentation of memory. Without the IADDRESS=n
qualifier, the monitor decides where to load the run-time system each time
residency is required. With the qualifier? the monitor loads the run-time system
at the specific lK-word section of memOly. You must include the IADDRESS=n
qualifier whenever you use the INOREAJD_ONLY qualifier.
The IADDRESS qualifier without an argument indicates that the monitor should
compute a starting address by finding the highest memory segment large enough
to hold the run-time system.

12-4

Managing System Files and Logical Names

The /NOADDRESS qualifier lets the run-time system "float" in memory; that is,
RSTSIE computes a new address every time the run-time system is loaded into
memory. The default is INOADDRESS.
Because 1K-word section numbering begins at 0 and ends at n-1 (where n is
the total size of memory), the 1K section number in the IADDRESS=n qualifier
is one less than the physical section number. For example, to load the RT11
run-time system (4K words in size) into the 61st through 64th 1K sections of
memory, specify IADDRESS=60. The run-time system is loaded from low memory
to high memory at its defined initialized size. Th be loaded without error, enough
contiguous user space must be available starting at that location.
The location specified in the IADDRESS=n qualifier becomes the default location
at which the run-time system is loaded during the current time-sharing session.
You need to change the location only if you exchange the allocation of the section
of memory with either the DEFAULT or the START initialization options.
One precaution is necessary when specifying the address at which the run-time
system is loaded. The section of memory you choose must not fragment the user
job space to prevent the run-time system from executing a job. For example,
assume a system has 24K words of user space available between the 36K and
60K sections of memory. Assume also that a job requires 18K words of user space
to run and that the run-time system requires 4K words when resident. If the
loading address is 36K, the space between 40K and 60K remains available for an
18K job to run. If the loading address is 42K, the user space is fragmented into
two sections - one from 36K to 42K and one from 46K to 60K. An 18K-word job
area is not available to execute a job using this auxiliary run-time system.
The system verifies that the memory section given in the IADDRESS=n qualifier
is reasonable. If the entire range of memory starting at the load address is not
available, the system prints:
?Memory not available

You should consult a memory status report (for example, the report you get with
the SHOW MEMORY command) to select an available range of memory. If the
range of memory results in fra~,..menting that may cause a swapping violation, the
system prints the error message:
?Memo:ry would be fragmented

A swapping violation occurs if the memory to be occupied by the run-time system
does not allow enough contiguous space for a maximum sized job to run.
Appendix D contains a summary of the errors that can occur when you are
installing a run-time system.
The IPOSITION=n qualifier indicates the position in the linked list of blocks in
which the run-time system block is to be placed. The default keyboard monitor
block is always first in the list. If the run-time system you are adding is to be
accessed frequently, you can reduce system overhead by placing it in the next
position in the list. Without the qualifier, the system adds the block to the end of
the list.
If n in the IPOSITION=n qualifier is one, the block is placed immediately after
that of the default keyboard monitor. If n is zero or a value greater than the
number of currently defined run-time systems, the block is added to the end of
the list. The default is /POSITION=O.
The position of a run-time system description block in the linked list affects how
the system treats a RUN request for a file without a file type. On receiving such
a RUN request, the system checks the indicated directory for all files with the
specified name and an executable file type.
Managing System Files and Logical Names

12-5

For example, if the directory contained three files of the same name with file
types .BAC, .TSK, and .SAV, the system ehecks for the run-time system nearest
the default keyboard monitor in the linked list of description blocks. If RTII was
nearer than BASIC, the system runs the .SAV version of the file.
The ITYPE=filetype qualifier on the INSrrALURUNTIME_SYSTEM command
changes the default file type used for an executable file. If you do not specify this
qualifier, the monitor extracts the file type defined in the run-time system file.
The system applies this default file type when a user types a RUN command for
a program and does not specify a file type.
The fMINIMUM=s and fMAXIMUM=s qualifiers alter the minimum and maximum job sizes (in K words) defined in the file for the run-time system. Generally,
the minimum value allowed is one and the maximum is 32, the size of RTS,
but a given run-time system may not allow this full range of job sizes. For
BASIC-PLUS, the minimum and maximum sizes are 2 and 16, respectively.
The /LOCK qualifier makes the run-time system permanently resident. Usually,
an auxiliary run-time system is temporarily resident. It occupies memory as
long as the currently active job is running under its control. The system frees
the memory it occupies when that memory is needed to load another run-time
system or to load a job running under another run-time system. (Thus, the
run-time system may be nonresident if there are active jobs using it.) The system
automatically loads the run-time system when a user needs it. If you specify the
fLOCK qualifier, the auxiliary run-time Elystem becomes permanently resident, as
the default keyboard monitor is. The only way to make such a run-time system
nonresident is to use the DCL command UNLOADIRUNTIME_SYSYTEM.
The REMOVEIRUNTIME_SYSTEM command reverses the steps the
INSTALLIRUNTIME_SYSTEM command performs. Before you can shut down
the system, you must remove all auxiliary run-time systems. The SHUTUP program performs this removal operation automatically. The system checks to make
sure that no jobs are currently running under control of the run-time system to
be removed. If the usage count is zero, the description block is removed from the
list and the run-time system file is closed.
The UNLOADIRUNTIME_SYSTEM command frees the memory occupied by the
run-time system. If a job is currently resident or being loaded and intends to use
the run-time system, the system prints:
?Run-time system in use

12.1.2

Associating a File with a Run-Time System
The SET FILEIRUNTIME_SYSTEM=name command changes the name of the
run-time system associated with a file. Every disk file on RSTSIE has in its
directory the name of the run-time system under which it was created, except for
files that are larger than 65,535 blocks. On a RUN request for the file, RSTSIE
checks the name to find out what run-time system to use for the job. RSTS/E
automatically passes control to that run-time system which, in turn, executes the
file.
To change the name, RSTSIE attempts to open the file to establish write access.
If write access is denied, it prints:
?Protection violation

12-6

Managing System Files and Logical Names

The name of the run-time system is written in the directory of the file specified. RSTSIE performs this naming operation during the system library build
procedures when a file designed to run under an auxiliary run-time system is
transferred from the distribution medium.

12~.2

Resident Library Control
A resident library is a collection of shareable routines or data that the task
builder links together into a task image file on disk. The MAKSIL program
formats this disk file into Save Image Library (SIL) format. You can store
resident libraries on any disk, either public or private. (Resident library files can
be stored in accounts other than [0,1].) All resident library files are contiguous
because they must be loaded into memory in the fastest time possible. See
the RSTS / E Programmer's Utilities Manual for more information on resident
libraries.
You then use the INSTALLILIBRARY command to assign the task image portion
of the SIL file to a contiguous region of physical memory. Note that you can
also use monitor SYS call -18 to assign the task image portion of the SIL file to
memory. Once the body of shareable routines or data is linked, formatted, and
assigned to memory, it becomes a resident library that is accessible to user tasks
as part of their virtual address space.

12.2.1

Resident library DCl Commands
You can use the DCL commands. to install, remove, load, or unload a resident
library. For example, the DCL command to add a resident library is:
$ INS'l'ALL/LIBRARY library-file-spec

You can specify any valid RSTSIE file specification. You can also specify a
protection code in addition to the library file-spec, using the !PROTECTION
qualifier. If you do not specify a device, then the system defaults to the system
disk. The defalt PPN is account [0,1], and the default protection code is <42>
(read-only). The file type must be .LIB; otherwise an error occurs.
The /[NO]ADDRESS[=n] qualifier locates the resident library at a specific point
in memory. The n variable can be any number from the lowest to the highest
available memory locations in increments of 1K-word sections. The lAD DRESS
qualifier without an argument indicates that the monitor should compute a
starting address by finding the highest memory segment large enough to hold thE
library.
The INOADDRESS qualifier lets the library "float" in memory; that is, RSTSIE
computes a new address every time it loads the library into memory. However,
this method of loading has the following restrictions:
•

A program can only map one such library at any time

•

The program must be running under the RSX emulation run-time system

•

The maximum size of the library is 28K words

•

The highest address for mapping the library is 32K

The /LOCK qualifier lets you make a resident library permanently resident.
Usually, a resident library remains in memory only temporarily; that is, it
occupies memory only as long as currently active jobs run under its control.
However, when you need to access the library again, the system automatically

Managing System Files and Logical Names 12-7

reloads it. To force the library to remain permanently in memory, you must
specify the ILOCK qualifier in the INSTALULIBRARY command.
If you want to remove a resident library from memory, delete the monitor structure that defines the library, and close the library file, use the DCL command:
$ REMOVE/LIBRARY name

Note that the REMOVEILIBRARY command makes the resident library inaccessible for sharing. However, the !UNLOAD qualifier, which can be used only with
the INSTALLILIBRARY command, only temporarily removes the library from
memory while it is not in use.
The UNLOADILIBRARY command removes a resident library from memory and
thus frees memory that can then be used to load another library or be available
for other system purposes. If you attempt to unload a resident library that is in
the process of being loaded or is in use by the currently running job, the system
prints the message:
?Library in use

To prevent inadvertent loss of data, the ITlonitor does not automatically unload a
library you have loaded with the /NOREAD_ONLY qualifier. Unloading a library
removes it from memory but does not write it back to disk. This means the
system discards the library and retrieves a new copy from the disk when it is
needed again. Thus, anything you write to a library that you mark read/write is
lost after you unload it with the UNLOADILIBRARY command. You must use
the UNLOADILIBRARY command to unload a resident library that is designated
read/write.
Whether a library remains in memory depends on the qualifiers you attach to
the INSTALULIBRARY command. The !LOCK and /NOREAD_ONLY qualifiers
prevent the monitor from removing it automatically. In addition, the monitor
does not remove a library to which a user job is attached; instead, the system
prints:
?Library in use

12.3 System Logical Names
RSTSIE lets users access devices by logical names as well as by physical names.
Logical names that apply to all users are called system logical names. On all
systems, users can refer to a disk by its pack identification or by a name that
replaces the pack identification. Thus, each disk unit configured on RSTSIE
systems has the capability of being accessed by a system logical name. Logical
names that apply to a single job are called user logical names.
You can define system logical names for nondisk devices and additional names for
disk devices. The nwnber of names is lim.ited by the size of XBUF. Such a system
logical name can have an account number associated with it. Consequently, use
of the name refers not only to the related device but also to the account on the
device. The system treats device names in the following manner:

12-8

Checks the list of user logical names first. Each job can have up to three
assigned logical names. (If the logical name has a $ character in it, this step
is omitted.)

Scans the table of disk logical names for a matching pack identification or
equivalent.

Managing System Files and Logical Names

Checks the additional names.

Checks the list of valid physical device names.

If individual users have not defined user logical names that duplicate currently
defined system logical names, all users have access to devices by the logical
names the system manager adds.
A system logical name must be unique. A pack identification or its logical
name cannot conflict with any other pack identification or system logical name.
However, you can define multiple system logical names for the same device.
If the pack identification (or a logical name to replace the pack identification) of
a disk you are to mount is the same as an existing system logical name, then
RSTSIE mounts the disk with no system logical name. Unless you specify a
different, unique name, the disk, at mount time, has no logical name associated
with the device and unit being used.

System logical names apply fo:r a single time-sharing session; however, you can
remove or change them during the session. DIGITAL recommends that you place
the proper commands in START.COM to define names automatically at system
startup time. See the RSTS / E Programming Manual for more details on system
logical names.

12:.3.1

Assigning Logical Names
The ASSIGN/SYSTEM dev:[PPN] logical-name command defines a system
logica~ name for nondisk devices or an additional name for a disk device. You
need INSTAL privilege to use ASSIGN/SYSTEM. If the name duplicates one
currently defined and you specify the IREPLACE=QUERY qualifier or you omit
the ![NO]REPLACE qualifier, DCL prints a warning message and prompts to see
if you want to replace the logical. If you specify INOREPLACE and the logical
name is already defined, DCL prints the error message:
?Logical already installed

To change a currently defined name, use ASSIGN/SYSTEMIREPLACE command.

The logical name must contain from 1 to 9 alphabetic or numeric characters.
You can include the dollar sign ($) character, however, it must not be the first
character. If the name contains illegal characters, DCL prints one of the the
following error messages:
?Invalid logical name
?Unexpected character

pointing to the first nonalphanumeric character. If there is no more room in
XBUF, the system prints:
?No buffer space available

The device with which the name is associated must exist on the system. If it does
not, the system prints:
?Not a valid device

Managing System Files and Logical Names

12-9

12.3.2

Removing Logical Names
The DEASSIGN/SYSTEM command deletes the association defined for the logical
name specified. You need INSTAL privilege to use DEASSIGN/SYSTEM. Because
a system logical name must be unique, you need to specify only the name in the
command. If the name you include is not currently defined, the system prints:
?Logical not found

If you use an illegally formed name, the system prints one of the following:
?Invalid logical name
?Unexpected character

pointing to the first nonalphanumeric character. 'Th delete all logical names for
a device, you must use the DEASSIGN/SYSTEM command for each currently
defined name.

12.3.3

Listing System Logical Names
The SHOW LOGICAL/SYSTEM command displays, for the specified logical name,
the device designation and unit nwnber, any account nwnber associated with the
logical name, and the logical name itself. If you do not specify a logical name,
RSTS/E lists all the system logical names. If you specify the first few characters
of a system logical name, RSTS/E lists all the names beginning with those letters.
RSTS/E displays the entries in the order in which they are found in the monitor
tables.

12.4 Defining Concise Command Language Commands
Concise Command Language (CCL) comtnands on RSTS/E let users type systemlevel commands that load and run programs from disk. 'Th take advantage of
this feature, you must code your programs to recognize a CCL entry and to
extract any command string passed to them. The RSTS / E Programming Manual
describes the operation and interpretation of CCL commands.
Certain RSTS/E programs can interpret a standard set of CCL commands. A
RSTS/E installation can have any number of unique CCL commands. You make
all CCL commands available to users by the DEFINE/COMMAND/SYSTEM
command which creates the monitor structure required. You need INSTAL
privilege to use the DEFINE/COMMAND/SYSTEM command.
Because the monitor structure that defines a CCL command is temporary, CCL
commands must be defined at the start of each time-sharing session. It is
suggested that you place the proper command definitions in START.COM.
Each CCL command definition occupies one small buffer on the system. All
definitions are in a linked list of small buffers. The system executes a SYS call
to add or remove CCL definitions and provides a way to list all currently defined
CCL commands.

12-10

Managing System Files and Logical Names

1 ~~.4.1

Adding a CCl Definition
'lb add SY-STAT as a CCL command, execute the DEFINE/COMMAND/SYSTEM

command as follows:
$ DEFINE/COMMAND/SYSTEM SY-STAT

[1,2]SYSTAT.*

This command adds the definition for the SYSTAT system program. The hyphen
(-) designates the abbreviation point so typing SY runs the SYSTAT program. If
you do not include a hyphen in your CCL definition, the system places a hyphen
at the end of the command and allows no abbreviation. RSTS/E assumes the
public disk structure (SY:) when you do not specify a device. Programs you run
with a CCL command can reside on any disk device, but if it is not in the public
structure, you must include the device in the command definition. The account
designation [1,2] in the example means SYSTAT must reside in the system
library. If you do not specify an account with the CCL command, the system
assumes the program is located in the account in which the job (that is, the job 01
the user later invoking the CeL) is running.
The asterisk (*) in SYSTAT. * indicates a wildcard file type. In the case of multiple
versions of SYSTAT, the version of SYSTAT that is run depends on the order of
the run-time systems in your run-time system list.
You can specify the line number at which execution starts by appending the
ILINE:_NUMBER=n qualifier to the command. The value of n can be any integer
from 0 to 32767. There are standard entry points that you can specify by the use
of keywords:
•

CCL---3 0000

•

DCL---30500

•

CHAIN-31000

The system performs error checking before adding the CCL command. If the
command is invalid, RSTSIE prints one of the following error messages:
?Invalid command
?Unexpected character

pointing to the first nonalphanumeric character. A valid command can consist of:
•

A string of 1 to 9 alphanumeric characters

•

The first character must be a letter, or one of the following single characters:
@ or #

Lowercase letters are equivalent to uppercase in the definition and use of CCL
commands. The $ CCL is reserved; it always points to DCL. The command can
have a maximum of nine characters. If more than nine characters are present,
the system prints:
?Command too long

When you add two or more commands that begin with the same character or
set of characters, you must define the command with the largest number of
characters (to the left of the hyphen) first. You then define the command with
the next fewest characters to the left of the hyphen, until you have made all the
assignments. For example, you must define MACR-O before adding MAC-, or
RSTSIE prints:
?Command not unique

Managing System Files and Logical Names

12-11

This message also appears if you attempt. to define a command that was already
defined.
If you need to add a longer CCL definition after you have already added a similar
but shorter one, you must first remove the shorter entry. You then add the longer
CCL command which lets you reinstall the CCL with the shorter command name.
For example, if you had installed MAC- and at some later date needed to add
MACR-O, you would have to remove MAC-, install MACR-O, and then reinstall
MAC-. The section "Removing a CCL Definition" describes how to remove CeL
entries.

12.4.2

listing Currently Defined CCl Commands
The SHOW COMMAND/SYSTEM command prints a listing of all currently
defined CCL commands. You can also list a single eCL command with the
SHOW/COMMAND/SYSTEM <name> command. You need INSTAL privilege to
see the program name and characteristics in the SHOW COMMAND/SYSTEM
command. The list of CCL commands on your system may contain some of the
following:
$ SHOW COMMAND/SYSTEM
SY: [ 1,2
ATT-ACH
SY: [ 1,2
BYESY: [ 1,2
DIR-ECTORY
DIS-MOUNT
SY: [ 1,2
EDTSY: [ 1,2
HELL-O
SY: [ 1,2
SY: [ 1,2
HE-LP
LBRSY: [ 1,2
LIBRSY: [ 1,2
LIN-K
SY: [ 1,2
MACR-O
SY: [ 1,2
MACSY: [ 1,2
MOU-NT
SY: [ 1,2
PIPSY: [ 1,2
PL-EASE
SY: [ 1,2
QU-EUE
SY: [ 1,2
SUB-MIT
SY: [ 1,2
SY-STAT
SY: [ 1,2
TE-CO
SY: [ 1,2
TKBSY: [ 1,2

/LINE=CCL /PRIVILEGE
] LOGIN *
/PRIVILEGE
] LOGOUT. *
/LINE=O
]DIRECT.*
/LINE=CCL /PRIVILEGE
] UMOUNT. *
/LINE=CCL /PRIVILEGE
.TSK /LINE=O
]EDT
/LINE=O
/PRIVILEGE
] LOGIN *
] HELP
/LINE=CCL /PRIVILEGE
*
]LBR
.TSK /LINE=O
]LIBR .SAV /LINE=8208
]LINK .SAV /LINE=8208
] MACRO .SAV /LINE=8216
] MAC
.TSK /LINE=O
/LINE=CCL /PRIVILEGE
]UMOUNT.*
]PIP
.SAV /LINE=8208
]PLEASE.*
/LINE=CCL /PRIVILEGE
/LINE=CC:L /PRIVILEGE
]QUE
*
]QUE
/LINE=CCL /PRIVILEGE
*
/LINE=CCL /PRIVILEGE
]SYSTAT.*
]TECO .TEe /LINE=3584
]TKB
.TSK /LINE=O

The system prints the commands in the order in which they were defined. If no
CCL commands are currently defined, the system prints the message:
%No Corrunands installed

NOTE

When you start the system, the monitor automatically installs the $
CCL. You cannot redefine the $ CCL.
The $ CCL lets you issue DCL comlnands from any command environment. For example, from BASIC-PLUS you can type:
Ready
$ MOUNT DLO:MINDY/NOSHARE <RET>
Ready

12-12

Managing System Files and Logical Names

1~~.4.3

Rernoving a CCl Definition
1b relnove a CCL definition, type the command DELETE/COMMAND/SYSTEM
followed by the definition. You need INSTAL privilege to execute the
DELETE/COMMAND/SYSTEM command. You can type either a full definition or
a valid abbreviation to indicate the command you plan to remove:
$ DELETE/COMMAND/SYSTEM MYPR-OG

If the definition does not currently exist, the system prints:
?Command not found

If the definition exists, the command removes the CCL definition immediately.
You can delete all the CCL cOlnmands by appending the /ALL qualifier to the
DELETE/COMMAND/SYSTEM command. For example:
$ DELETE/COMMAND/SYSTEM/ALL

1 ~~.5 Sys1tem File Control
You allocate and position the following system files on file-structured disks
by executing the INSTALUSWAP_FILE, INSTALL/OVERLAY_FILE, or
INSTALLIERROR_FILE commands:
•

SWAPO.SYS

•

SWAPl.SYS

•

SWAP3.SYS

•

OVR.SYS

•

ERR.SYS

These optional files optimize system perfonnance by taking advantage of higher
speed disks and the characteristics of user jobs running on the system. Because
the files are contiguous and contiguous space becomes scarce as user files are
created during timesharing, preallocation and prepositioning are necessary. The
RSTS / E Installation and Update Guide describes the planning ccnsiderations for
preallocating and positioning these system files.
RSTS/E does not automatically access these system files. You control RSTSIE
access to these files with DeL commands. Thus, the use of these files during
timesharing is dynamic. If a hardware problem occurs on a swap device, you can
remove the device without stopping timesharing. 1b maintain the same amount
of swap space, you can designate a private disk as the swap device.
This section discusses the DCL commands that control system file usage. The
commands in the startup control file allow access to system files at the start of
time-sharing operations.

12.5.1

Adding and Removing Swap Files
The swap files SWAPO.SYS, SWAPl.SYS and SWAP3.SYS in account [0,1]
are standardized names for the three optional swap files. While these names
are optional, Digital recommends you use them because they indicate the
characteristics of the files. You can, however, assign any file name to these three
swap files. The system restricts the file type to .SYS regardless of the file names
you include in the DCL commands.
Managing System Files and Logical Names

12-13

The INSTALL/SWAP_FILE command causes the system to access a specific file
(if a file-structured device is involved) or a. specific device (if a non-file-structured
disk is involved). You need INSTAL privilege to execute the INSTALL/SWAP_
FILE command. A typical command sequence is:
$
$

INSTALL/SWAP FILE=O DR3: SWAP 0
INSTALL/SWAP - FILE=l DR1: SWAP 1
INSTALL/SWAP :FILE=3 DR4: SWAP 3

NOTE

When you specify a non-file-structured device in the INSTALL/SWAP_
FILE command, the monitor destroys any existing file structure on that
device.
The /SIZE=n qualifier of the INSTALUSWAP_FILE command allows creation of
a swap file during timesharing. The qualifier is most useful when you want to
add a swap file to a private disk that contains sufficient contiguous space for the
file. The value n in the /SIZE=n qualifier represents the number of blocks in the
file. RSTS/E attempts to access the file wi.th a file type of .SYS in account [0,1] on
the associated device. If the file already exists and the file size differs from the
size specified in the /SIZE=n qualifier, RSTS/E prints a warning message in the
format:
%Swap file name already exists a sizl:l x blocks

The value x is the actual size of the swap file.
If the file does not exist, the system attempts to create a contiguous file of the
size specified. If there is not enough contiguous space available to create the file,

the system prints:
?Not enough contiguous space to create Swap file

The /POSITION=n qualifier lets you place the created swap file in a particular
location on the specified disk. The monitor ignores this qualifier if the already
exists.
Mter you type the INSTALL/SWAP_FILE command, the system tries to install
the particular file or device as the swap file. If the system finds any errors,
RSTS/E prints an error message. Appendix D contains a list of the possible
errors. If no errors occur, the system installs the swap file.
You can then activate the added swap file with the SET SYSTEMILOGINS
command. The SET SYSTEMILOGINS command forces the system to examine
all swap file space and to calculate the number of jobs the system is capable of
running.
During timesharing, the number of jobs that can run on the system depends on:
•

The job maximum (JOB MAX) set at system generation

•

The amount of swap space added

•

The number of logins allowed through the SET SYSTEMILOGINS command

At the start of timesharing, the number of logins allowed is one. One job is
necessary for the DCL START.COM file t.o run. You can increase this number
by using the SET SYSTEMILOGINS command, but the number is restricted
by the swap space available and by JOB MAX. When you add swap space with
INSTALL/SWAP_FILE commands, the c8lpacity to handle logins is increased.
The SET SYSTEMILOGINS command sets the number of logins to the maximum
allowed. This maximum cannot exceed JOB MAX.

12-14

Managing System Files and Logical Names

RSTSIE lets you dynamically remove as well as add swap files 0, 1, and 3. To
remove a swap file or device, first decrease the number of logins using the SET
SYSTEMINOLOGINS or SET SYSTEMILOGINS n command. The number of
logins to be decreased depends on the capacity of the swap file to be removed.
By decreasing the number of logins, you allow the system to safely remove swap
space.
The sequence of commands to remove a swap file is the reverse of the sequence
to add swap space. For example, assume that a system is running with a JOB
MAX of 63 and is using a swap file with a capacity of 16 jobs. In this case, you
must decrease the number of logins by 16 jobs before removing the swapfile. The
following sequence shows this procedure:
$
$

SET SYSTEM/LOGINS 47
REMOVE/SWAP _FILE=l

In practice, it may take some time for the system to adjust operations. If the
number of jobs currently running is greater than 47 or if the number of logins
allowed is too large, the system prints the error message:
?Too many logins allowed to remove Swap file

You must wait until users log out of the system. The SHOW BUFFERS command
or the DISPLY program prints the number of jobs currently on the system in the
free buffer status report. (The SHUTUP program automatically removes swap
files.)
It is possible to add more swap space than the system can use. In this case, you

GO not need to reduce logins to remove a swap file. However, enough room must
remain to swap all jobs currently logged in, or to swap all jobs that can log in,
whichever is greater.
The SHOW FILE/SYSTEM com.mand provides a way for you to verify that the
following files exist on your system:
•

Swap files (/SWAP_FILE)

•

Overlay file (/OVERLAY_FILE)

•

Error file (/ERROR_FILE)

•

DECnet/E Network Services Protocol (NSP) system file (INETWORK.-FILE)

This command is most useful when you want to verify the addition or removal of any of these files. Mter adding a file, you can use the SHOW
FILE/SYSTEM/SWAP_FILE command to verify, for example, that you gave
the file the correct file specification, or that the file was added properly.
If you add a swap file, such as DR3:[O,1]SWAPO.SYS, execute the SHOW
FILE/SYSTEM/SWAP_FILE cornmand as follows:
$

SHOW FILE/SYSTEM/SWAP_FILE

Swapfile
Swapfile
Swapfile
Swapfile

0:
1:
2:
3:

.SYS 10240 blks (40 jobs at 64K Swap max)
DR3: [0,1] SWAP
None
DR1: [0,1] SWAP
. SYS
256 blks (1 job at 641K Swap max)
DR4: [0,1] SWAP3 "SYS 1000 blks (3 jobs at 64K Swap max)

$
RSTSfI1~ prints None beside the files that do not exist.

Note that the system
cannot add or remove the DECnet/E file but can only verify that the file exists
on your system. There is no DCL command to add or remove this file. Refer to
the SET SYSTEM and CLEAR SYSTEM commands in the DECnet/E System
Manager's Guide for further information.

Managing System Files and Logical Names

12-15

12.5.2 Adding and Removing Overlay and Error Files
You can add and remove the overlay and error message files during timesharing.
OVR and ERR are standardized names to denote the separate file for the overlay
code and the alternate file for the default error message file. Although you can
use other file names, Digital recommend~J that you keep the standardized ones.
Regardless of the names you give the files, the system requires you to use a file
type of .SYS. You need INSTAL privilege to use the INSTALL command.
The INSTALL/OVERLAY_FILE command directs the system to use the specified
file for the system overlay code. Similarly, the INSTALLIERROR_FILE command
causes the system to use the specified file when it reads error messages. To add
these system files, specify the commands as in the following example:
$ INSTALL/OVERLAY FILE DU2:0VR
$ INSTALL/ERROR_FILE DBO:ERR

For the overlay file, the system copies the overlay code from the monitor SIL to
the file OVR.SYS on DU unit 2. For the error message file, the system copies into
ERR.SYS on DB unit 0 the contents of the error message file (ERR.ERR). You
must make sure that the disk on which an overlay or error file resides is mounted
when you issue this command. As with Bwap files, you can use the /SIZE and
!POSITION qualifiers for overlay and error files.
To install these files, the system makes sure that the files exist and are properly
formed. If the system finds any errors, it prints an error message.

Appendix D contains the text of all error messages. If no errors occur, the system
has successfully added the file and is accessing it. You do not need to enter any
further commands.
You can remove these files at any time with the REMOVE/OVERLAY_FILE and
REMOVEIERROR_FILE commands. Then, the system accesses the original data.
Note that the SHUTUP program automatically removes both files.

12.5.3

Using the DUMP/SYSTEM Command
The DUMP/SYSTEM command lets you take an online dump of the current
monitor image executing in memory. When you execute the DUMP/SYSTEM
command, RSTSIE copies the memory to the crash dump file CRASH.SYS in
account [0,1]. You need SYSIO privilege to use the DUMP/SYSTEM command.
Use the DUMP/SYSTEM command as follows:
$ DUMP/SYSTEM

Once the copy operation is complete, you can use the ANALYS program to analyze
the contents of the file (see Chapter 10).
The system returns the following error nlessage if crash dump was not enabled:
?Can't find file or account

This can happen only if sufficient disk space was not available at system startup
to allocate the CRASH.SYS file. To correct this condition, free up the necessary
amount of disk space, shut down and then restart the system.
If a user who does not have SYSIO priviilege attempts to dump memory with
DUMP/SYSTEM, the system prints:
?SYSIO privilege required

12-16

Managing System Files and Logical Names

Use of the DUMP/SYSTEM cOlnmand may also return various device dependent
errors. For example:
?Device hung or write locked

12.,6 DCl Command Descriptions
The following sections describe the DCL commands for managing system files and
logical names.

12.6.1

ASSIGN/SYSTEM Command
The ASSIGN/SYSTEM command assigns a system-wide logical name to
the specified device type and unit. If you include an account number, the
ASSIGN/SYSTEM command also associates that project-programmer number
with the logical name. ASSIGN/SYSTEM requires INSTAL privilege.

Format
ASSIGN/SYSTEM

dev: [p,pn] name[:]

Command Qualifiers

Defaults

I[NO]REPLACE[=OUERy]

IREPLACE=OUERY

IPrompts
Device:
Logical name:

Command Parameters
dey:
Specifies the device type and unit to which you want to assign a system-wide
logical name.
[p,pn]
~rhe optional PPN to which you want to assign a system-wide logical name.

DCL

does not allow any spaces between the device specification and the PPN.
name[:]
~rhe name that you want added to the system logical name table.

Command Qualifiers
I[NO]REPLACE[=QUERY]

Determines what action to take if the logical name already exists. !REPLACE
means replace the logical if it already exists. RSTS/E displays an informational
Inessage. INOREPLACE means display an error message and do not replace
the logical if it already exists. If neither is specified or if !REPLACE=QUERY is
fJpecified, RSTS/E displays a warning message and prompts to see if you want to
replace the logical.

Managing System Files and Logical Names

12-17

12.6.2

DEASSIGN/SYSTEM Command

The DEASSIGN/SYSTEM command removes a specified logical name from the
system logical table. It can also remove logical names for devices on your system.
Device mnemonic names for system devices have pennanent entries in the logical
table. So, if you use the DEASSIGN/SYSTEM command to remove a logical
name associated with a device, only the logical name goes away, not the device
mnemonic. You can then give a new name to that device. DEASSIGN/SYSTEM
requires INSTAL privilege.

Format
DEASSIGN/SYSTEM

name[:]

Command Qualifiers
none

Defaults
none

Prompts
Logical name:

Command Parameters
name[:]
Specifies the logical name to be removed from the system logical table.

12-18

Managing System Files and Logical Names

12,.6.3

DEF:INE/COMMAND/SYSTEM Command

The DEFINE/COMMAND/SYSTEM command creates or redefines a systemwide
command (CCL).
DEFINE/COMMAND/SYSTEM requires INSTAL privilege.

Format
DEFINE/COMMAND/SYSTEM
Command Qualifiers
IUNE_NUMBER=entry point
I[NO]PRIVILEGE

command file-spec

Defaults
none
none

Prompts
Command:
Program name:

Command Parameters
command
Specifies the command you want to define.
file-spec

Specifies the file you want to associate with the command.

Command Qualifiers
IUNE_NUMBER=entry point

Defines the entry point in the command program as follows:
•

n--Defines the entry point as the specified integer in the range 0 to 32,767

•

CCL-Defines the entry point as the CCL entry point (30000)

•

CHAIN-Defines the entry point as the CHAIN entry point (31000)

•

DCL-Defines the entry point as the DCL entry point (30500)

I[NO]PRIVILEGE

Specifies whether the program needs to retain its temporary privileges (see
Chapter 6).

Managing System Files and Logical Names

12-19

12.6.4

DELETE/COMMAND/SYSTEM Command

The DELETE/COMMAND/SYSTEM comnland deletes the specified command from the list of defined system-wide commands (CCLs). The command
DELETE/COMMAND/SYSTEM requires INSTAL privilege.

Format
DELETE/COMMAND/SYSTEM
Command Qualifiers
fAll

Defaults
none

Prompts
Command:

Command Parameters
command
Specifies the command to be deleted from the list of defined system-wide commands.

Command Qualifiers
fAll
Deletes all system-wide commands.

12-20

Managing System Files and Logical Names

12.6.5

DUl\np/SYSTEM Command

The DUMP/SYSTEM command records in the CRABH.SYS file the current state
of the system so you or field service has a way to investigate reasons for system
malfwlctions. DUMP/SYSTEM requires SYSIO privilege.

DUMP/SYSTEM
Command Qualifiers

Defaults

none

Prompts

none

Managing System Files and logical Names 12-21

12.6.6

INSTALL/ERROR_FILE Command

The INSTALLIERROR_FILE command installs the specified file as the system
error message file. INSTALLIERROR_FILE requires INSTAL privilege.

Format
INSTALL/ERROR_FILE
Command Qualifiers
/[N 0] lOG
/POSITION=dcn
/SIZE=n

[file-spec]
Defaults;
flOG
none
none

Prompts
Error file:

Command Parameters
file-spec

Specifies the file to be installed as the system error message file. If you do nC"'t
specify a file name, the system uses the default file ._SY:[O,l]ERR.SYS.

Command Qualifiers
I[NO]lOG

Indicates whether the system displays a confirmation Inessage after it installs the
error file. The default is ILOG.
IPOSITION=dcn

Places the error file at the specified position on the disk. You must also specify
the /SIZE=n qualifier. The value of position can be one of the following:
Value

Meaning

den

Locate the file at the first available position on the device past the
specified Device Cluster Number.

MIDDLE

Locate the file at the first available position past the middle of the
device.

INDEX

Locate the file at the first available position past the Master file
Directory (MFD) index.

ISIZE=n

Creates a system file with a size of n blocks. The /SIZE=n qualifier is required if
the file does not already exist.

12-22

Managing System Files and Logical Names

12.6.7

INSTALL/LIBRARY Command

The INSTALLILIBRARY comlmand adds the named library to the list of installe<
resident libraries. INSTALL/LIBRARY requires INSTAL privilege.

Format
INSTALL/LIBRARY

narne

Command Qualifiers
f[NO]ADDRESS[=n]
f[NO]ERROR_LOGGING
f[NO:ILOCK
fPROTECTION=n
f[NO]READ_ONLY
f[NO]SHAREABLE
f[NO]UNLOAD

Defaults
See discussion
fERROR_LOGGING
fNOLOCK
fPROTECTION=42
fREAD_ONLY
fSHAREABLE
fNOUNLOAD

Prompts
Library:

Command Parameters
name
Specifies the name of the library to add to the list of installed resident libraries.

Command Qualifiers
,'~NO]ADDRESS[=n]

Indicates that the resident lib:rary should be loaded into memory in lK-word
sections beginning at address n. Specifies the address in memory to load the
resident library. Values can be in the range lK to 2043K.
If n is omitted, RSTS!E computes the starting address by finding the hlghest
memory segment large enough to hold the named library. If this qualifier is
omitted, RSTSIE computes a new starting address each time it loads the named
library into memory.
I[NO]ERROR_LOGGING
Indicates whether errors that occur within the resident library should be recorded
in the system error log. The /ERROR_LOGGING qualifier logs any errors that
occur in the resident library to the system error log. The INOERROR_LOGGING
qualifier prevents errors that occur withln the resident library from being
recorded in the system error log. The default is !ERROR_LOGGING.
I[NO]L.OCK
Indicates whether the resident library should remain in memory. The /LOCK
qualifier locks the resident library in memory. The INOLOCK qualifier unlocks
the resident library from memory. The default is INOLOCK.
IPROTECTION=n
Sets the protection code of the resident library. The default code is 42.

Managing System Files and Logical Names

12-23

/[NO]READ_ONLY

Indicates the allowable type of access to the resident library. The fREAD_ONLY
qualifier allows read-only access to the resident library. The INOREAD_ONLY
qualifier allows the resident library to be mapped read/write. The default is
fREAD_ONLY.
/[NO]SHAREABLE

Indicates how many users can access the resident library simultaneously. The
/SHAREABLE qualifier lets more than one user access this resident library at the
same time. The INOSHAREABLE qualifi.er lets only one user access the resident
library at any time. The default is /SHAREABLE.
/[NO]UNLOAD

Indicates whether the resident library should remain in memory when no jobs
are using it. The !UNLOAD qualifier rerrLOves the resident library from memory
whenever there are no jobs using it. The INOUNLOAD qualifier keeps the
resident library in memory even when there are no jobs using it. The default is
INOUNLOAD.

12-24

Managing System Files and Logical Names

12:.6.8

INSTALL/OVERLAY_FILE Command

The INSTALL/OVERLAY_FILE command installs the specified file as the system
overlay file. INSTALL/OVERLAY_FILE requires INSTAL privilege.

Fc)rmat
INSTALL/OVERLAY_FILE

[file-spec]

Command Qualifiers
f[N0] l.OG
fPOSITION=dcn
fSIZE==n

Defaults
flOG
none
none

Prompts
Overlay file:

Command Parameters
file-spec

Specifies the file to be installed as the system overlay file. If you do not specify a
file name, the system uses the default file, _SY:[O,l]OVR.SYS.

Command Qualifiers
I[NO]LOG

Indicates whether the system displays a confirmation message after it installs th~
overlay file. The default is !LOG.
IPOSITION=dcn

Places the overlay file at the specified position on the disk. The /SIZE=n qualifiel
is required. The value of position can be one of the following:
Value

Meaning

---------------------------------------------------------dcn
Locate the file at the first available position on the device past the
specified Device Cluster Number.
MIDDLE

Locate the file at the first available position past the middle of the
device.

INDEX

Locate the file at the first available position past the Master file
Directory (MFD) index.

ISIZE::n

Creates a system file with a size of n blocks.

Managing System Files and Logical Names

12-2!

12.6.9

~NSTALL/RUNTIME_SYSTEM Command

The INSTALLIRUNTIME_SYSTEM command adds the named run-time system
to the list of installed run-time systems. INSTALLIRUNTIME_SYSTEM requires
INSTAL privilege.

Format
INSTALL/RUNTIME_SYSTEM

narne

Command Qualifiers

Defaults

/[NO]ADDRESS[=n]
/[NO]ERROR_LOGGING
/[NO]KEYBOARD_MON ITOR
/[NO]LOCK
/MAXIMUM_SIZE=s
/MINIMUM_SIZE=s
/POSITION=n
/PREFIX[=x]
/[NO]READ _ ONLY
/[NO]SHAREABLE
/TYPE=filetype
/[NO]UNLOAD

See discussion
/ERROn_LOGGING
INOKEYBOARD_MON ITOR
INOLOCK
none
none
at end of list
IPREFD(=255
IREAD_ONLY
ISHAREABLE
See discussion
INOUNLOAD

Prompts
Run-time system:

Command Parameters
name
Specifies the name of the run-time systeYD to be added to the list of installed
run-time systems.

Command Qualifiers
I[NO]ADDRESS[=n]

Specifies the address in memory to load the run-time system. The value of n
can be in the range lK to 2043K. If n is omitted, RSTS/E computes the starting
address by finding the highest memory segment large enough to hold the runtime system. If the qualifier is omitted, RSTSIE computes a new starting address
each time it loads the run-time system into memory.
I[NO]ERROR_LOGGING

Indicates whether the run-time system slhould log errors occurring under its
control to the system error file. The /ERROR_LOGGING qualifier indicates that
this run-time system should log errors. The INOERROR_LOGGING qualifier
indicates that it should not log errors. The default is /ERROR_LOGGING.

12-26

Managing System Files and Logical Names

I[NO]KEYBOARD_MONITOR

Indicates whether the run-time system has a keyboard monitor. The
/KEYBOARD_MONITOR qualifier indicates that this run-time system
is a keyboard monitor. The INOKEYBOARD_MONITOR qualifier indicates that this run-time system is not a keyboard monitor. The default is
INOKEYBOARD_MONITOR.
I[NO]LOCK

Indicates whether the run-time system should reside permanently in memory.
The /LOCK qualifier indicates that the run-time system should reside permanently in memory. The INOLOCK qualifier indicates that the run-time system
should not reside permanently in memory. The default is INOLOCK.
IMAXIMUM_SIZE=s

Uses s in K words as the maximum job size.
IMINIMUM_SIZE=s

Uses s in K words as the minimum size job allowed.
IPOSITION=n

Places the run-time system block in the nth position after the default keyboard
monitor in the linked list. By default, RSTSIE positions the run-time system at
the end of the list. N can be any number in the range from 0 to 255.
IPREFIX[=x]

Causes the run-time system to use the RSTSIE special prefix EMT feature. The
value x can be any number from 0 to 255. RSTSIE selects the code 255 by defaul
if you do not specify a value for x.
I[NO]READ_ONLY

Indicates whether the run-time system should allow read-only memory access. The fREAD_ONLY qualifier indicates read-only memory access. The
INOREAD_ONLY qualifier indicates read/write memory access. The default is
fREAD_ONLY.
I[NO]SHAREABLE

Indicates whether the run-tinle system supports multiple users. The
/SHAREABLE qualifier indicates that the run-time system supports more than
one user at the same time. The INOSHAREABLE qualifier indicates that the
run-time system supports only one job at a time. The default is /SHAREABLE.
ITYPE=fi letype

Specifies the default file type for files executed under the run-time system. By
default, DCL uses the file type defined in the run-time system file.
I[NO]UNLOAD

Specifies whether RSTSIE unloads the run-time system from memory when
no one is using it. The !UNLOAD qualifier removes the run-time system from
memory when all users are using another run-time system or when no jobs
are using the run-time system. The INOUNLOAD qualifier keeps the run-time
system in memory even when no jobs are using it. The default is INOUNLOAD.

Managing System Files and Logical Names

12-2'

12.6.10

INSTALL/SWAP_FILE Command

The INSTALL/SWAP_FILE command adds the specified file to the list of installed
swap files. Valid values of swap-index are 0, 1, or 3. INSTALL/SWAP_FILE
requires INSTAL privilege.

Format
INSTALL/SWAP_FILE

[=}swap-index [file-spec}

Command Qualifiers
/OEVICE=disk-name[:]
I[NO] lOG
IPOSITION=dcn
IS IZE=n

Defaults
none
IlOG
none
none

Prompts
Swapfile number:
Swapfile name:

Command Parameters
swap-Index
Specifies the index of the specified file. Valid values are 0, 1, and 3. If you do not
enter a file spec, the system uses the default file, _SY:[0,1]SWAPn.SYS, where n
is the index nwnber.
file-spec
Specifies the file to add to the list of installed swap files. You must specify the
disk.

Command Qualifiers
IDEVICE=dlsk-name[:]
Specifies a disk to be used as a swapping device. RSTSIE uses this disk as a
non-file-structured device; any data on the disk is lost.

If you include the !DEVICE qualifier, you cannot include a file-spec parameter or
include the /SIZE or !POSITION qualifierB.

I[NO]lOG
Indicates whether the system displays a confinnation message after it installs the
swap file. The default is !LOG.

12-28

Managing System Files and Logical Names

IPOSITION=dcn
Places the created system file at the specified position on the disk. You must also
specify the /SIZE=n qualifier. The value of position can be one of the following:
Value

Meaning

dcn

Locate the file at the first available position on the device past the
specified Device Cluster Number.

MIDDLE

Locate the file at the first available position past the middle of the
device.

INDEX

Locate the file at the first available position past the Master file
Directory (MFD) index.

ISIZE::n
Creates a system file with a size of n blocks. The /SIZE=n qualifier is required if
the file does not already exist.

Managing System Files and Logical Names

12-2~

12.6.11

REMOVE/ERROR_FILE Command

The REMOVEIERROR_FILE command re'moves the system error message file.
REMOVEIERROR_FILE requires INSTAL privilege.

Format
REMOVE/ERROR_FILE
Command Qualifiers

Defaults

none

Prompts

none

12-30

Managing System Files and Logical Names

12.6.12 REMOVE/LIBRARY Command
The REMOVEILIBRARY command removes the named file as a resident library.
REMOVEILIBRARY requires INSTAL privilege.

Fc)rmat
REMOVE/LIBRARY

name

Command Qualifiers

Defaults

none

Prompts
Library:

Command Parameters
name
Specifies the file to remove from the list of resident libraries.

Managing System Files and Logical Names

12-31

12.6.13

REMOVE/OVERLAY_FILE Command

The REMOVE/OVERLAY_FILE command removes the system overlay file.
REMOVE/OVERLAY_FILE requires INS11\L privilege.

Format
REMOVE/OVERLAY_FILE
Command Qualifiers

Defaults

none

Prompts

none

12-32

Managing System Files and Logical Names

1 ~~.6.14

RE~MOVE/RUNTIME_SYSTEM Command

The REMOVEIRUNTIME_SYSTEM command removes the named run-time
system as an entry in the run-time system table. REMOVEIRUNTIME_SYSTEM
requires INSTAL privilege.

Fc)rmat
REMOVE/RUNTIME_SYSTEM
Command Qualifiers

name

Defaults

none

Prompts
Run-time system:

Command Parameters
name
Specifies the name of the run-time system to remove from the list of installed
run-time systems.

Managing System Files and logical Names

12-33

12.6.15

REMOVE/SWAP_FILE Command

The REMOVE/SWAP_FILE command relmoves the specified swap file from
the list of installed swap files. Valid values of swap-index are 0, 1, or 3.
REMOVE/SWAP_FILE require,s INSTAL privilege.

Format
REMOVE/SWAP_FILE

[=]swap-index

Command Qualifiers

Defaults

none

Prompts
Swapfile number:

Prompts
Command Parameters
swap-index
Specifies the index of the swap file to renlOve. Valid values are 0, 1, or 3.

12-34

Managing System Files and Logical Names

1~~.6.16

SET FILE Command
The SET FILE command changes the characteristics of the specified file(s). SET
FILE requires read and write access to the file(s).

Format
SET FILE

file-spec

Command Qualifiers

Defaults

I[NO]CACHE
INOCONTIGUOUS
I[NO]DELETABLE
I[NO]LOG
I[NO]PLACED
IPROTECTION=n
IRUNTIME_SYSTEM=name

none
none
none
INOLOG
none
none
none

PromlPts
Files:

Command Parameters
file-spec
Specifies the file whose characteristics you want to change. You can use wildcard
characters in the file specification.

Command Qualifiers
I[NO]CACHE[=SEQUENTIAL I RANDOM]

The /CACHE qualifier specifies that the file is automatically cached when open.
The type of caching depends on the file's UFD entry and the specified OPEN
MODE. If you use /CACHE alone on a file with no UFD setting or MODE
specification, the default is random caching. TUNE privilege is required to use
this qualifier. The optional argmnents are:
•

SEQUENTIAL-Causes RSTSIE to mark the file's UFD entry such that, if
the file is cached, it is cached sequentially.

•

RANDOM-Causes RSTS/E to mark the file's UFD entry so, if the file is
cached, it is cached randomly. Note that you can override the random cache
UFD setting with a sequential caching MODE specification.

The /NO CACHE qualifier specifies that the file is not automatically cached when
open. To cache such a file, you must specify SET CACHE/ALL or specify a MODI
value (see the RSTS / E Programming Manual).
INOCONTIGUOUS

Lets you extend a contiguous file by changing it to noncontiguous.
I[NO]DELETABLE

Indicates whether a file can be deleted or renamed during timesharing. In a
directory listing, RSTSIE shows NODELETABLE files with the flag P (protected)
after their size. You need SYSIO privilege to use this qualifier.
Managing System Files and Logical Names

12-3!

I[NO]LOG

Indicates whether the file specification of ,each modified file is displayed at the
terminal. If the file specification contains any wildcards, the default is /LOG.
Otherwise, the default is INOLOG.
I[NO]PLACED

Indicates whether, after modification, a file is placed at its present position on the
disk.
IPROTECTION=n

Sets the protection code of a file. The protection code must be in the range 0 to
255. You can also use the SET PROTECTION command to perform this function.
You need SYSIO privilege to set the privil,ege bit of the protection code.
IRUNTIME_SYSTEM=name

Associates a run-time system with the file. The name can be any valid run-time
system name from the SHOW RUNTIME_SYSTEM display. When the system
tries to execute this file, it uses the named run-time system to load the executable
file. You need SYSIO privilege to use this qualifier for file(s) whose protection
code is 192 or higher.

12-36

Managing System Files and Logical Names

12:.6.17

SHOW COMMAND/SYSTEM Command

The SHOW COMMAND/SYSTEM command displays information about
the specified systemwide command. SHOW COMMAND/SYSTEM requires
INSTAL privilege for a full listing. Without the INSTAL privilege, SHOW
COMMAND/SYSTEM only displays the command you name.

Format
SHOW COMMAND/SYSTEM
Command Qualifiers
IAll

command-name

Defaults
/All

Prompts
none

Command Parameters
command-name
Specifies the systemwide command whose information you want displayed. If
you do not include this parameter, RSTSIE displays all system commands. If you
specify the first few letters of a command, RSTSIE displays all the commands
beginning with those leters.

Command Qualifiers
fAll
Displays information about all the systemwide commands.

Managing System Files and Logical Names 12-3'

12.6.18

SHOW FILE/SYSTEM Command

The SHOW FILE/SYSTEM command displays infonnation about the specified
system :fileCs). The default is /ALL. SHOWIFILE/SYSTEM does not require any
privilege.

Format
SHOW FILE/SYSTEM
Command Qualifiers
fAll
fERROR_FllE
fNETWORK_FllE
fOVERLAY_FilE
/SWAP_FllE[=swap-index]

Defaults
fAll
none
none
none
all swap files

Prompts
none

Command Qualifiers
fAll

Displays infonnation about all installed £lles.
fERROR_FllE

Displays infonnation about the system error message file.
INETWORK_FllE

Displays infonnation about the network database file.
IOVERlAY_FilE

Displays infonnation about the system overlay file.
ISWAP_FllE[=swap-lndex]

Displays infonnation about the specified swap file, where swap-index is 0, 1, or 3.
If no argument is given, RSTSIE lists all swap files.

12-38

Managing System Files and Logical Names

12.6.19

SHOW LIBRARY Command

The SHOW LIBRARY command displays information about all installed libraries.
SHOW LIBRARY does not require any privilege.
The following is an example of the SHOW LIBRARY display:
$ SHOW LIBRARY

Resident Libraries:
Name Prot
Acct
[
HISTOG < 42>
1,2
CSPLIB < 42> DR1: [ 0,1
RMSRES < 42> DR1: [ 0,10
RMSLBB < 42> DR1: [ 0,10
RMSLBA < 42> DR1: [ 0,10
RMSLBD < 42> DR1: [ 0,10
RMSLBC < 42> DR1: [ 0,10
RMSLBE < 42> DR1: [ 0,10
RMSLBF < 42> DR1: [ 0,10
DAPRES < 42> DR1: [ 0,10
EDT
< 42> DR1: [ 0,11
B2MRES < 42> DR1: [ 0,1

Size
36K
8K
4K
3K
4K
2K
2K
3K
4K
10K
38K
19K

Users

8
1
1

°
°
°
°
°
°1
°

Comments)
Perm, Addr:1628
Temp, Addr:117
Temp, Addr:1624
Temp, Addr:1621
Temp, Addr:1617
Temp, Addr:1615
Non-Res, Addr:1613
Temp, Addr:1610
Temp, Addr:1606
Temp, Addr:876
Temp, Addr:1568
Temp, Addr:830

F()rmat
SHOW LIBRARY
Command Qualifiers

Defaults

none

Prompts

none

Managing System Files and Logical Names

12-3!

12.6.20

SHOW LOGICALS Command
The SHOW LOGICALS command displays information about a user-defined
logical name. SHOW LOGICALS does not require any privilege.

Format
SHOW LOGICALS

[name]

Command Qualifiers
fAll

DefaulU~

fAll

Prompts
none

Command Parameters
[name]
Specifies the user-defined logical name whose infonna tion you want displayed. If
you specify the first few letters of the logical name, RSTS/E displays information
on all the logical names beginning with those letters.

Command Qualifiers
fAll
Displays information about all user-defined logical names.

12-40

Managing System Files and logical Names

12~.6.21

SHOW LOGICAL/SYSTEM Command
The SHOW LOGICAL/SYSTEM command displays information about a
systemwide logical name. You need no special privileges to use SHOW
LOGICAL/SYSTEM.

F()rmat
SHOW LOGICAL/SYSTEM

name

Command Qualifiers
fAll

Defaults
fAll

Prompts
none

Command Parameters
name
Specifies the systemwide logical name whose information you want displayed.

Command Qualifiers
fAll
Displays information about all systemwide logical names.

Managing System Files and Logical Names

12-41

12.6.22

SHOW RUNTIME_SYSTEM Command

The SHOW RUNTIME_SYSTEM command displays information about all installed run-time systems. SHOW RUNTIME_SYSTEM does not require any
privilege. The following is an example of the display:
$ SHOW RUNTIME_SYSTEM
Run-Time Systems:
Name
Ext Dev
... RSX TSK
DCL
COM DR1:
RT11
SAV DR1:
TECO
TEC DR1:
BAS4AL BAC DR1:
BAC DR1:
BAS4F
BASIC
BAC DR1:
FORTH
4TH DR1:
GMBODT COM DRO:
MOMODT COM DRO:

Size
Users Comments
0(64)K 14
Monitor, KBM
28 (4) K
8
Temp, Addr:77, DF KBM
4(28)K
0
Temp, Addr: 109, KBM, CSZ, EMT:255
10(20)K
0
Temp, Addr:383, KBM
17(12)K
0
Non··Res, KBM, CSZ
16(16)K
0
Temp, Addr:322, KBM, CSZ
15(16)K
1
Temp, Addr: 133, KBM, CSZ
6(24)K
0
Temp, Addr: 125, KBM, NER, CSZ
0
27(4)K
Non--Res, Addr: 1541, KBM, 1US, R/W
27(4)K
Temp, Addr: 849, KBM, 1US, R/W, NER
1

Format
SHOW RUNTIME_SYSTEM
Command Qualifiers

Defaults

none

Prompts

none

12-42

Managing System Files and Logical Names

1~~.6.23

UNLOAD/LIBRARY Command

The UNLOADILIBRARY command removes the resident library you specify from
memory. UNLOADILIBRARY requires the INSTAL privilege.

F4lrmat
UNLOAD/LIBRARY

name

Command Qualifiers
none

Defaults

Prompts
Library:

Command Parameters
name
Specifies the name of the resident library to remove from memory.

Managing System Files and Logical Names

12-43

12.6.24

UNLOAD/RUNTIME_SYSTEM Command

The UNLOADfRUNTIME_SYSTEM command unloads the named run-time
system from memory. UNLQADfRUNTIME_SYSTEM requires the INSTAL
privilege.

Format
UNLOAD/RUNTIME_SYSTEM
Command Qualifiers

name

Defaults

none

Prompts
Run-time systems:

Command Parameters
name
Specifies the name of the run-time systmn to re'move from memory.

12-44

Managing System Files and Logical Names

Chapter 13

Job Management
This chapter describes the DCL commands that you use to manage RSTS/E jobs.
Table 13-1 summarizes these commands. The rest of this chapter explains the
commands in detail and gives examples of their use.
Table "13-1:

13.1

DCl Commands ffor Job Management

CommlllDd

][)escription

BROADCAST

Displays a message at the specified destination(s).

DETACH

Detaches a job from its terminal.

FORCE:

Forces a "string" to the specified destination(s).

HANGUP

Disconnects the specified remote line.

Creates a job at the specified terminal, logs your current
job in under a different account, or resets your current
job to its initial logged-in state.

REMOVE/JOB

Immediately terminates the specified user job and logs
it off.

SET JOB
SET SYSTEM

Defines the characteristics of a job.

SHOW JOBIPRIVILEGE

Displays the set of current privileges for the current job.

Sets the default system characteristics.

Controlling the Number of Logged-In Jobs
You can monitor and control system operation while logged in to a RSTS/E
system. With the SYSTAT or the DISPLY system program, you can observe how
the system is performing. If performance declines, you can remedy the problem,
in some cases, by preventing more users from logging in to the system. You
adjust the number of logged-in jobs allowed with the SET SYSTEMILOGINS, and
SET SYSTEMINOLOGINS com,mands.
SET SYSTEMINOLOGINS sets to one the number of jobs that are allowed to
log in to the system. Mter RSTS/E executes the SET SYSTEMINOLOGINS
command, the system allows no more users to log in to the system. If a user
attempts to log in, the monitor prints the following error message:
?Logins are disabled - please try again later

Jobs already logged in to the system can continue running. However, jobs such
as BATCH that create logged in jobs cannot successfully process further requests.
(The SHUTUP system program, also disables further logins in preparing to shut
down timesharing.)

Job Management 13-1

If you try to connect to RSTSIE over a LAT line and get the message "?Logins
are disabled-please try again later", RS~rSIE does not disconnect the session.
Instead, the terminal waits for logins to become enabled; then you can start the
login sequence by pressing Return. If you want to get back to the LAT prompt,
use the local key sequence, usually the Break key.
1b keep you from being locked out of the system, RSTSIE allows jobs to be logged
in at the system console terminal (KBO:), regardless of the number of logins
currently allowed. (This feature is a special characteristic of the terminal you
designate as the system console; the default is KBO:. By installing a feature
patch, you can change the terminal which has this characteristic. However, only
one terminal on a system can have this capability.) This capability remains in
effect as long as a job slot is available to handle a new job. You need SWCTL
privilege to use the SET SYSTEMINOLOGINS command.
The SET SYSTEMILOGINS command places a ceiling on the number of users
that can log in to the system. This comm.and is useful for limiting or extending
the load allowed on the system. For example:
$
$

SET SYSTEM/LOGINS=35

This command limits the number of logins to 35; if there are 35 or more users
logged in to the system, the next user who tries to log in gets the error message
"?Logins are disabled-please try again later," and must wait until enough users
log off before being allowed to log in.
The system does not let you set the nwnber of logins to zero or to a number
greater than the value of JOB MAX set at system installation time. If you specify
zero in the command, RSTSIE sets the number to one. The default number
is JOB MAX. You need SWCTL privilege to use the SET SYSTEMILOGINS
command.
The nwnber of logins allowed can never exceed the capacity of the swap space
installed on the system or the value specified during system generation. If you
enter a nwnber greater than the maximum allowed, RSTSIE sets logins to the
maximum possible. The SHOW BUFFEHS command or the DISPLY program
includes the maximwn nwnber of logins currently allowed.
You normally use the SET SYSTEMILOGINS command with the INSTALL/SWAP_FILE
command to adjust the nwnber of logins allowed. After you add swap space to
the system, the SET SYSTEMILOGINS command performs these steps:
1.

Recalculates the swap capacity

2. Factors in the currently allowed maximwn swap size (SWAP MAX)
3. Increases the nwnber of allowed logins
If you are preparing to remove swap space, use the SET SYSTEMILOGINS
command to decrease the number of allowed logins.

13-2

Job Management

1 ~1.2 Broadcasting Messages to Terminals
The BROADCAST command lets you communicate with one user or with all users
on the system. You need SEND privilege to execute the BROADCAST command.
This command places a specified text string in the output buffer of a terminal, or
of all terminals, and displays the text on the terminal. For example, if a user at
terminal KB32: assigns a peripheral device for an unreasonably long time, you
can send a message asking the user to deassign the device:
$ BRO.ADCAST KB32: "Bob, can you deassign MMO: for me; I need it ASAP."
$

By specifying /LOGGED_IN instead of a terminal designator, you can broadcast
the message to each online terminal in the RSTSIE system (unless the terminal
is set for NOBROADCAST). For example:
$ BROADCAST/LOGGED_IN "Read $NOTICE.TXT; the system is coming down!!"
$

The system displays your message as follows:
*** From [1,214] "Account Name" on KB25: at 02:17 PM 27-May-85
** Read $NOTICE.TXT; the system is coming down!!

This message appears on all terminals that are on line and set to BROADCAST,
except your own.
You can direct your message to specific users by substituting a PPN for the
destination specification. Wildcard characters are also permissible. For example,
to broadcast a message to all users logged in under the same project number:
$ BROADCAST [20,*] "Please attend a meeting at 3:00 PM today"

You can send a multiline message with BROADCAST as follows:
1.

Press Return after the destination specification. RSTSIE displays the prompt
Message: on your terminal.

Enter your message, terminate each line by pressing Return.

Enter CtrllZ to send your message.

In a multiline message, you do not need to enclose the message in quotes. For
example:
$ BROADCAST/ALL<RET>
Message: The power will be shut off in the building this weekend.<RET>
Message: Therefore, this system will be taken down Friday night.<RET>
Message: No logins will be allowed after 5PM on Friday.<RET>
Message: <Ctrl/Z>
$

13.3 Controlling Jobs
The FORCE command lets you send a text string to another terminal (or all
terminals). You need SYSIO privilege to use the FORCE command. The system
places the text string in the input buffer of the specified terminal as if it were
typed by the user. The command has the format:
FORCl~ destination "command"

Job Management

13-3

If you want to send a CtrllC to a terminal, KB32 in this case, to stop the
execution of whatever is running at the terminal and then to log the user off the
system, type:
$ FORCE KB32:
$

" . . . LOGOUT"

The circumflex (") character before the command forces a Ctrl/C to the terminal,
and the LOGOUT command logs the user off the system.
You can also send only a CtrllC ("C) to a terminal by placing a circumflex
character in the first position after the keyboard number. For example:
$ FORCE KB32: " . . . "
$

Do not place any text after the circumflex if you want to send only a CtrllC to the
terminal to stop the job that is running. RSTS/E sends a CtrllC to the terminal
and returns control to the keyboard monitor prompt. For example:
... C
$

To force a control character combination (Ctrl/Z for example) to a terminal, enter

the circumflex (") character as the first character of the text, followed by the
proper control letter (such as Z). For example:
$ FORCE KB32: " . . . Z"
$

Typing "Z and pressing Return executes a CtrllZ at KB32: No other text
should follow the control character combination. If the circumflex (") is the first
character of more than two characters of text, however, RSTS/E forces a CtrllC to
the terminal before sending the text that immediately follows the first character.
For example:
$ FORCE KB32: " . . . ! Harry, sorry but your job is out of control!"
$

The "c characters appear on Harry's terminal to end the job he was running and
the system prints the text of the message on the following line. (The text after
the first character can contain up to 80 characters.)

13.4 Removing Jobs
The REMOVE/JOBINOQUERY command lets you terminate a user's job
and remove it from the system. You need JOBCTL privilege to use the
REMOVE/JOBINOQUERY command. Before using REMOVE/JOBINOQUERY,
use the SHOW USER command to learn the user's job number, then type:
$ REMOVE/JOB/NOQUERY 10
$

The monitor clears job 10 and frees it for other system use. The user is no
longer logged in to the system and must log in again before attempting any other
system activity. Use the SHOW USER command immediately after issuing the
REMOVE/JOBINOQUERY command to verify that the job is no longer on the
system.

13-4 Job Management

13.5 Placing a Job on HOLD
There are times when you do not want to terminate a job, but rather want to
place it in a hold state, a state from which it can resume normal processing. A
job may be using up too many system resources, or you may be running a batch
job that you need to correct. The SET JOBIHOLD command automatically sets
the priority of a job to ~128. This causes the system to stop giving the job any
time to run. Use the SET JOBIHOLD command as follows:
$ SET JOB/HOLD 10
$

Job 10 stops processing but does not terminate. SET JOBIHOLD always sets
the priority to ~128, which means you cannot include a priority number with the
command. You must reset the priority of the job to ~120 or higher with the SET
JOBIRELEASE command to reactivate the job. Use the SET JOBIRELEASE
command as follows when you want to restart the job:
$ SET JOB/RELEASE 10
$

Now job 10 has the default priority of ~8, which under normal system load allows
the job to successfully run to completion.
SET JOBIHOLD and SET JOBIRELEASE are the logical commands to use when
you want to stop a job temporarily and then start it again. You need TUNE
privilege to use the SET JOB/HOLD or the SET JOBIRELEASE commands.

13;.6 Setting Job Priority, Run Burst, and Maximum Size
The system lets you set the priority, run burst, and maximum size of an existing
job. You can change any of the current values to:
•

Increase or decrease the chance of gaining run time in relation to other
running jobs

•

Determine how much CPU time the job can have when it is compute bound

•

Increase the area a job can occupy

The system runs jobs on the basis of priority. The higher a job's priority, the
better its chances of obtaining run time in relation to other jobs that are running.
Figure 13-1 shows how an 8-bit priority byte determines priority.
Figure 13-1:

Priority Byte Format

x
Sign
Bit

U ser- Deti nable
Portion

System- Definable
Portion

Job Management 13-5

Using the SET JOBIPRIORITY command, you or another user can set the userdefinable portion of the priority byte for any job on the system. Because the three
system-definable bits are normally zero, standard priorities are multiples of 8
between -120 (lowest priority) and +120 (highest priority). Zero is a legal priority.
When the system changes a priority, it truncates any value not a multiple of eight
to the next lowest legal priority. For exantple, if you specify a priority of +10,
RSTIE sets the value to +8.
You need TUNE privilege to use the SET JOBIPRIORITY command.
All of the system-definable bits are normally off (zero). The least significant bit is
set when a keyboard delimiter is typed and the job is waiting for keyboard input
or when any 1/0 completes. Keyboard delimiters are Carriage Return (CR), Line
Feed (LF), Form Feed (FF), Control Z (CtrllZ), and Escape Sequence Character
(ESC). The next significant bit is set whenever a CtrllC is entered and can be set
at any time. Finally, a system function call sets the most significant bit of the
system-definable portion. RSTSIE clears all system-definable bits when it chains
to another program or when the job is at the system command level.
The system-definable portion of the priority byte is always less significant than
the user-definable portion. Therefore, the system-definable bits affect priority
only within the user-definable priority range. For example, if two jobs are running under priority -8, the user who enters a CtrllC has a higher priority (that is,
priority -6 in this case) than the user who does not. However, a third user with
priority 0 supersedes two users whose priorities are -8 and -6.
When you log in, LOGIN runs with priority 0 and automatically sets your job to
priority -8. This is the default priority with which most or all jobs are run. Do
not assign priorities other than -8 except in unusual circumstances (for example,
compute-bound or detached jobs). On occasion, you may want to run a program
that requires a great deal of computation. If time is not a factor in obtaining
results, and you have TUNE privilege, you can decrease the job priority of the
compute-bound job to improve efficiency for the other users on the system. On
the other hand, infrequently used detached programs often have higher priorities
(typically priority 0) because they must run quickly when needed but do not run
compute bound for an extended period and do not run often.
Run burst is the maximum time a job can run compute bound before another job
obtains access to the CPU. On systems using the KW11L line frequency clock,
each unit of run-burst time is equal to 1/60th or 1/50th of a second, depending
on the system's power line frequency. SYBtems running with the KW11P clock at
crystal speeds, rather than at line frequency, have a run-burst unit of 1I50th of
a second. If the system is operating from a 60 Hz power line, one run-burst unit
equals 1I60th of a second. In that case, six units equal 1I10th of a second, which
is the run-burst default value.
If you assign a run burst of six to a spedfic job that does not require much
compute-bound time, the system automatically transfers control to the next user
before the six units have been used. One tenth of a second is generally considered
the best run-burst time period to ensure efficient overall system operation. If a
job is guaranteed to become 1/0 bound (that is, 1/0 stalled) after a certain amount
of computation, use SET JOBIRUN_BURST to specify a run burst larger than
six. In many cases, a run burst greater than six has a significant effect on long
computational programs. Their percentage of the overall CPU time would be
increased.
You need TUNE privilege to use the SE'l1 JOBIRUN_BURST command.

13-6

Job Management

The maximum size assigned to a job refers to the allowable memory space.
You can restrict certain jobs to run only programs smaller than 32K words (or
SWAPMAX) by assigning a job a size smaller than the current SWAPMAX. This
assigned limit does not affect privileged, compiled programs. Thus, a user with
a small amount of space can still run system programs that would normally
exceed the memory limit. 1b alter the maximum job size of job 14, use the SET
JOB/SIZE command as in the following example:
$ SET JOB/SIZE=16 14

You need the TUNE privilege to use the SET JOB/SIZE command.

1 ~L 7 Controlling Keyboards and Remote Lines
If you determine that a dataset line is in use but no keyboard activity is taking
place (by using SYSTAT or the DISPLY program job status report), you can
disconnect the dataset. The HANGUP command disconnects the remote line
specified by the KBn: keyboard. The hangup capability prevents a user from
monopolizing the line without being charged for connect time and frees the
line for other remote users. You need HWCTL privilege to use the HANGUP
command.
The DETACH command frees a terminal from its job. If the job does not have thE
terminal open on a nonzero channel, DETACH frees the terminal for other use.
The command normally forces the system to close all nonzero channels on which
the te:nninal is open. The fNOCLOSE switch suppresses this action. When the
job specified is currently detached, the monitor prints the error message:
?Job number x is already detached.

You need JOBCTL privilege to use the DETACH command (except for your
current job).

1 ~L8 DCL Command Descriptions
The following sections describe the DCL commands that you can use for job
management.

Job Management

13-7

13.8.1

ATTACH Command
The ATTACH command attaches a terminal to a job. This command does not
require any privileges.

Format
ATTACH

Command Qualifiers

Defaults

none

Prompts
Job number:

Command Parameters
n
The number of the job you want to attach to.

13-8 Job Management

1~J.8.2

BROADCAST Command
The BROADCAST command sends a message to:
•

A specified terminal (KBn:)

•

All users logged in under a specified PPN (the PPN can contain wildcard
characters)
NOTE

This command displays a message only on terminals that are not set to
NOB ROADCAST.
You need SEND privilege to use BROADCAST.

FClrmat
BROADCAST

destination ["message']

Command Qualifiers
IAll
I[NO]BEll
I[NO]HEADER
IlOGGED_IN

Defaults
none
INOBElL
IHEADER
Inone

Prompts
To:
Message:

Command Parameters
destination
The device(s) or account(s) where the message is to be displayed. For multiple
destinations, enter the individual keyboards, separated by commas, for example:
$ BROADCAST [4,*],KB23:,KB5 "Please log off."

"message"
The message to be displayed on the destination device(s). If you type Return
after entering the destination specification, DCL prompts you for each line of
a message. Enter CtrllZ to send the message. Note that when you enter a
multi-line message in this way, do not enclose the message in quotes.

Command Qualifiers
IAll
Broadcasts the message to all terminals.
I[NO]BELL

Broadcasts a BELL character before the message. The default is INOBELL.

Job Management

13-9

f[NO]HEADER

Broadcasts a header before the message. The header is in the format:
From ARK:: [1,194]

"Account Name" on KB52: at 11:52 AM 30-May-85

The default is !HEADER.
flOGGED_IN

Broadcasts the message to all logged-in terminals.

13-10

Job Management

1:1.8.3

DETACH Command
The DETACH command detaches a job from its terminal. You need JOBCTL
privilege to detach any job other than your current job.

Format
DETACH

Command Qualifiers
I[NO]CLOSE

Defaults
ICLOSE

Prompts
Job number:

Command Parameters
n
The number of the job you want to detach from its tenninal.

Command Qualifiers
I[NO]CLOSE

Indicates whether the system closes all nonzero channels on which the job's
terminal is open before detaching. The default is ICLOSE.

Job Management

13-1

13.8.4 FORCE Command
The FORCE command forces a string to:
•

A specified terminal

•

All users logged in under a specified PPN. The PPN can contain wildcards.

You need SYSIO privilege to use FORCE.

Format
FORCE

destination "string"

Command Qualifiers
fAll
flOGGED_IN

Defauitsl
none
none

Prompts
To:
Command:

Command Parameters
destination
The device(s) that you want the "string" forced to. For multiple destinations,
enter the individual devices, separated by commas, for example:
$ FORCE [4,*]KB23:,KBS "'" $LOGOUT"

"string"
The string that you want forced to the destination device(s).

Command Qualifiers
fAll
Forces the string to all terminals.
flOGGED_IN
Forces the string to all logged-in terminals.

13-12

Job Management

1~J.8.5

t-IANGUP Command
The HANGUP command disconnects the specified remote line. You need HWCTI
privilege to use HANGUP.

Format
HANGUP

KBn:

Command Qualifiers
none

Defaults
none

PromlPts
Keyboard:

Command Parameters
KBn:
Specifies the remote line you want to disconnect.

Job Management

13-13

13.8.6

Create a new job at the specified terrninal. You can use this feature in your
system startup procedure to log in various users automatically. You need
WACNT or GACNT privilege to create a new job. If the terminal is restricted,
you also need DEVICE privilege.

•

Log your current job in under a different account. This feature lets you switch
between accounts without having to first issue the LOGOUT command (see
RSTS/E System User's Guide). You need WACNT or GACNT privilege to log
in to a different account without being prompted for that account's password.

•

Reset your current job to its initial logged-in state (the state your job was
in after logging in). This feature is useful for resetting logical assignments,
DCL symbols, prompts, and so on to their initial state. This feature does not
require any privilege.

Format
LOGIN

[p,pn]

Command Qualifiers

Defaults

I[NO]OVERRIDE[=NOLOG INS]
ffERM INAL=terminal(:]

INOOVERRIDE
Your terminal

Prompts
Password:

Command Parameters
[p,pn]
The PPN for the account you want to log in to. The default PPN is your PPN.

Command Qualifiers
IOVERRIDE[=NOLOGINS]
INOOVERRIDE

Indicates whether to create the new job if logins are currently disabled. Specify
/OVERRIDE or /OVERRIDE=NOLOGINS to create the new job even if logins are
disabled. If you attempt to create a new job with the INOOVERRIDE qualifier
and logins are disabled, the monitor displays an error message. The default is
INOOVERRIDE. The /OVERRIDE qualifier requires that you also specify the
!TERMINAL qualifier.
NOTE

Even with /OVERRIDE in effect, you cannot create a new job unless
there are job slots available on the system.

13-14 Job Management

ITERMINAL=termlnal[:]

Indicates that a new job should be created at the specified terminal. The terminal
argument can be any valid device name that refers to a keyboard device. If the
terminal you specify is restricted, you must have DEVICE privilege. You must
have WACNT or GACNT to log: in to another account on the specified terminal.
The following examples show the different features of LOGIN. To reset your job
to its initial logged-in state, enter:
$

To log in to account [10,10], enter:
$ LOGIN [10,10]
Password:

You must then supply the correct password. If you have WACNT or GACNT
privilege and you are switching from any [10,*] account, the monitor lets you log
in without prompting you for the password.
To log in user [100,3] at terminal KBll: while you are working at another

termingal, enter the following:
$ LOGIN/TERMINAL=_KB11:

[100,3]

You must have WACNT privilege, or have GACNT privilege and be logged in to
a [100,*] account to execute the previous command. You must have the DEVICE
privilege, or you must have _KBll: set to NORESTRICT.

Job Management

13-1 ~

13.8.7

REMOVE/JOB Command
The REMOVE/JOB command immediately tenninates the specified user job and
logs it out. You need JOBCTL privilege to use this command.

Format
REMOVE/JOB

Command Qualifiers

Defaults

/[NO]CONFIRM
/[NO]QUERY

/CONFIRM
/QUERY

Prompts
Job number:

Command Parameters
n
Specifies the job to remove.

Command Qualifiers
/[NO]CONFIRM

Causes a SHOW JOB display for the job you want to remove, then asks if you
really want to remove the job just displayed.
/[NO]QUERV

Synonym for the /[NO]CONFIRM qualifier.

13-16

Job Management

1~1.8.8 SET JOB Command
The SET JOB command defines the characteristics of the specified job. The
/KEYBOARD_MONITOR and !PRIVILEGE qualifiers can only be used for the
current job. If you specify those qualifiers with a job number other than your
own, RSTSIE displays the following error message:
?Only current job allowed.

You need TUNE privilege to use any of the qualifiers except !PRIVILEGE and
/KEYBOARD_MONITOR.

Fc)rmat
SET JOB

[n]

Command Qualifiers
IHOLD
IKEYBOARD_MONITOR[=name]
IPRIORITY[=p]
IPRIVILEGE=(priv,priv, ... )
IRELEASE
IRUN_BURST[=q]
ISIZE=q
Prompts
Job number:

Defaults
No change
none
none
none
none
none
none

Displayed if you use the IHOLD qualifier

Command Parameters
[n]

Specifies the job whose characteristics you want to define. If you leave off the job
number, the default is your own job number.

Command Qualifiers
IHOLD

Sets the priority of job n to -128. This qualifier cannot be used for the current job.
IKEYBOARD_MONITOR[=name]

Sets the current job's keyboard monitor to name. The allowable values of name
are any installed run-time system that has a keyboard monitor. Use the SHOW
RUNTIME_SYSTEM command to list the run-time systems on your system. If
you omit the optional argument, DeL becomes the current job's keyboard monitor.
IPRIORITY[=p]

Sets the priority of the specified job. Priority can be any number from -128 to
+120. The system rounds the priority down to a multiple of eight. (For example,
if you select a priority of -19, the system sets the priority to -24.) If you do not
specify a priority value, the default is -8.
IPRIVILEGE=(priv,prlv, ... )

Modifies the set of current privileges for the current job only.

Job Management

13-17

IRELEASE
Resumes job n at priority of -8.
IRUN_BURST[=q]
Changes the run burst of the specified job to q. The value for q can be any
number from 1 to 127. If you do not specify a run burst value, the default is 6.
ISIZE=q
Changes the maximum size of the specified job to q. You can set q to a maximum
of 32K, or 64K for I&D space machines. The system automatically lowers any
higher value to the maximum allowed.

13-18

Job Management

1~1.8.9

SET SYSTEM Command
The SET SYSTEM command sets the default system characteristics. You need
SWCTL privilege to use the SET SYSTEM command.

F()rmat
SET SYSTEM
Command Qualifiers

Defaults

/[NO]LOGINS[=n]

See Discussion

Prompts
none

Command Qualifiers
I[NO]LOGINS[=n]

Defines the maximum number of users who can log in to the system at one time.
Set ILOGINS=n to the number of user jobs that can log in to the system at one
time. The default is JOB MAX. Specify INOLOGINS if you want to prevent users
from logging in to the system by setting the number of logins to one.
See Chapter 14 for additional SET SYSTEM command qualifiers.

Job Management

13-19

13.8.10

SHOW JOB/PRIVILEGE Command

The SHOW JOBIPRIVILEGE command displays the set of current privileges for
the specified job. You need JOBCTL privilege to display the privileges of another
job.

Format
SHOW JOB/PRIVILEGE

[n]

Command Qualifiers

Defaults

none

Prompts

none

Prompts
Command Parameters
[n]
Specifies the job whose current privileges you want to display. The default is the
current job.

13-20

Job Management

Chapter 14

Miscellaneous System Management Functions
This chapter describes miscella.neous DCL commands that you use to manage
your RSTSIE system. Table 14-1 summarizes the DCL commands available to
you. The rest of this chapter explains each command in more detail.
Table "14-1:

14.1

DCl Commands for Miscellaneous System Functions

Comm,and

Description

LOADIINDEX

Loads SAT into memory.

LOAD/OVERLAY
SET DATE

Loads one or more monitor overlays in memory.
Sets the RSTSIE system date to the value you specify.
Optionally, you can specify a new value for the system time.

SET SYSTEM

Sets the default system characteristics.

SET TIME

Sets the RSTSIE 24-hour clock to the value you specify.
Optionally, you can set the system date to the value you
specify.

SHOW SYSTEM

Displays the system-default characteristics.

UNLOADIINDEX

Unloads SAT from memory.

UNLOAD/OVERLAY

Unloads one or more monitor overlays from memory.

Loading and Unloading Resident SAT
The LOADIINDEX and UNLOADIINDEX commands let you load and unload the
disk Storage Allocation Table (SAT) into memory during timesharing. Each disk
on the system has a SAT, which is called SATT.SYS.
The SAT is a 1 to 16 block permanent structure on all file-structured RSTSIE
disks. Its purpose is to provide the file processor with a table of allocated or
unallocated pack clusters, so that the processor can obtain new disk space for
file structures. By default, whenever new disk space is allocated on the disk, the
system updates the SAT and rewrites it to the disk, showing the true state of
in-use disk space.
The LOADIINDEX and UNLOADIINDEX commands let you optionally create
a SAT that is resident in memory during timesharing. This feature improves
system performance because the RSTS monitor does not need to perform disk
1/0 to the SAT on a per-disk unit basis. By allocating the appropriate amount of
XBUF, you can make the SAT for each disk unit completely memory resident, so
that a write to the disk occurs only when the disk unit is logically dismounted.

Miscellaneous System Management Functions

14-1

NO'TE

Because the resident SAT feature inlproves system performance at the
expense of memory, be sure to calculate the amount of XBUF needed
according to the number and size of any resident SATs desired before
starting timesharing.

14.2 Loading and Unloading Monitor Overlays
The LOAD/OVERLAY and UNLOAD/OV1~RLAY commands let you load and
unload monitor overlays during timesharing. You can use these commands in
conjunction with the DISPLY program and the SHOW CACHE command to
monitor the effects on your system's performance of making certain overlays
resident or nonresident.
The monitor overlays that you can specif:y with these commands are:
•

ATTRIBUTE-Performs file and account attribute read/write operations

•

DCL-Performs file-related operations for DCL

•

DELETE_RENAME-Performs file deletion and renaming

•

DIRECTORY-Performs disk file lookup operations

•

SYSTEM_CALLS-Performs specific monitor directives

•

TERMINAL-Performs terminal characteristic SET and SHOW operations

14.3 Suspending System Operations
The SET SYSTEMIHOLD and SET SYS'rEMIRELEASE commands let you
temporarily suspend all disk I/O and other system operations. When you issue
the SET SYSTEMIHOLD command, all jobs on the system are suspended, except
for the job issuing the command. Operat.ions reswne when you issue the SET
SYSTEMIRELEASE command.
The SET SYSTEMIHOLD and SET SYS'rEMIRELEASE commands are used
primarily on systems where the system disk is an RC25. The main reason for
using these commands is to suspend operations when you need to stop the fixed
disk drive so you can mount a different disk cartridge (both disks share the
same spindle). With SET SYSTEMIHOLD and SET SYSTEMIRELEASE, a total
system shutdown is not necessary. You need HWCTL privilege to use the SET
SYSTEMIHOLD and SET SYSTEMIRELEASE commands. To use the commands,
type:
$ SET SYSTEM/HOLD

System now on hold
$

The system displays the message Systenl now on hold to show that the suspend
is in effect. When the disk stops, you can change the removable cartridge. When
you are ready to resume system operations, type:
$ SET SYSTEM/RELEASE
System released

14-2

Miscellaneous System Management Functions

14.4 Changing System Date and Time
The SET DATE and SET TIME commands let you change the system date and
the 24-hour clock. These commands are important on systems where strict
accounting data is kept and incorrect values may have been entered at system
startup time. You can enter the date in either of two formats:
SET DATE dd-mmm-yy
SET DATE yy.mm.dd

The following example shows how to enter the date using the first format:
$ SET DATE 30-MAY-85

You can enter the same date in the second format as well:
$ SET DATE 85.05.30

Be sure to place a space between the DATE command and the beginning of the
date field when entering a date with either format. You need DATES privilege to
use the SET DATE or SET TIME command.
You can enter the time in either of two formats:
SET TIME 01:30PM
SET TIME 13:30

You change the date and/or the time with either command as follows:
For SET DATE;
$ SET DATE 30-MAY-85

$ SET DATE 30-MAY-85:01:30PM

For SET TIME;
$ SET TIME 01:30PM
$ SET TIME 30-MAY-85:01:30PM

Note that when you specify both parameters, you must first enter the date and
separate the two parameters with a colon.
You can also specify a relative date and time with the SET DATE command. For
example:
SET DATE -lDAY
SET TIME +lHOUR

The first command, when executed, sets the system date back a day. The second
command, when executed, adjust the system time ahead one hour. You can use
relative values to make minor adjustments in the system date or time. Relative
date and time syntax is as follows:
+nnDAYS
+nnHOURS
+nnMINUTES
You can string the keywords together in any order and you can abbreviate them
to one character. If you specify an absolute date and time, it must appear first in
the date-time specification. An example of mixed absolute and relative date and
time would be:
SET DATE 10-MAY-85:4:00PM+3DAYS+2HOURS-45MINUTES

This would be equivalent to:
SET DATE 13-MAY-85:5:15PM

Miscellaneous System Management Functions

14-~

14.5 Recording System Activities: EM-r Logging
An EMT is a PDP-II assembly language instruction through which a program
requests services from the monitor (for example, opening and closing a file,
performing 110, and logging in or out.) E:MT logging is an optional feature that
provides a "window" on the process by which timesharing jobs request and
receive services from the RSTS/E monitor. Thus, EMT logging lets you gather
information about the activity on your syl3tem.
For example, you might want to know the following information about your
system:
•

The number of logins on a particular terminal

•

How many files are accessed on a certain drive

•

Which nonresident FIP overlays get the heaviest use

Such information can help you improve system performance, identify bottlenecks,
establish fee setting algorithms, and watch for potential security problems.
This section describes the EMT logging feature, how it works, and how to inel ude
it in your system. However, because of the wide variety of RSTS/E systems, you
must decide which data is most useful in your environment, and how to collect
and use the data the EMT logger provides.
EMT logging includes optional code in the monitor, uses XBUF to pass information, and requires an EMT logging program (which you must supply; a sample is
provided in UNSUPP$:), running as a tinlesharing job. The use of EMT logging
can affect performance. This effect depends on which EMTs you decide to log, for
which jobs you log them, and how much processing your logging program does for
each EMT.
EMT logging provides information on timesharing activity in terms of what the
monitor sees. The data returned to your llogging program is in terms of FIRQB
and XRB contents, regardless of the programming environment of the job that
issues the directives. See the RSTS / E System Directives Manual for information
on the FIRQB and XRB, as well as descriptions of the MACRO form of system
directives.

14.5.1

Programming for the EMT Logger
To use EMT logging, you must:

•

Include optional code in your monitor at system installation time.

•

Supply a program to process the data extracted by the monitor code. This
program retrieves extracted data by send/receive calls.

During the system installation dialogue, RSTS/E asks you: EMT Logging?
Answer YES to create a monitor that can extract EMT information and pass it to
your program.
Mter creating the proper monitor, you must enable logging for the particular
EMT information you want. Use the SET SYSTEMIEMT_LOGGING command to
do this. You can use the same command to disable logging for EMT information
you no longer want. For example:
$ SET SYSTEM/EMT_LOGGING=(NOCRE,NODLN,ASS,DEA)

14-4

Miscellaneous System Management Functions

This command disables logging for creating files (CRE) and deleting files (DLN),
at the same time enabling logging for assigning devices (ASS) and deassigning
devices (DEA). You need the SWCTL privilege to use the IEMT_LOGGING
qualifier on the SET SYSTEM command.
The exact logging you choose affects performance. If you choose very common
actions, the monitor must take out more time to log them. To keep track of the
current EMT logging, use the SHOW SYSTEMlEMT_LOGGING command. This
command displays the current logging status of all EMTs. For example:
$

SHOW SYSTEM/EMT_LOGGING

Code Mnemonic
CAJ~FIP

0
4

8
16
~~ 0
24

Description

functions:
CLS
Close a channel
CRE
Creat.e a file
REN
Rename a file
RST
Reset. channels
ASS
ASSIGN a device
DAL
DEASSIGN all
CRB
Create cmpiled

UUO functions:
-29
TB3
Tables III
-27
DMP
Snapshot dump

Status

Code Mnemonic

Description

Status

Disabled
Disabled
Disabled
Disabled
Enabled
Disabled
Disabled

2
6
10
18
22
26
30

OPN
DLN
DIR
LOK
DEA
CRT
RUN

Open a channel
Delete a file
Directory info
File lookup
DEASSIGN dvice
Create .TMP
Run a program

Disabled
Disabled
Disabled
Disabled
Enabled
Disabled
Disabled

Disabled
Disabled

-28

SPL
FIL

SPOOL request
File utility

Disabled
Disabled

-26

You need no special privilege to use the SHOW SYSTEMlEMT_LOGGING
command.
Your logging program is a nonnal time-sharing job. You can design the program
to do such activities as:
•

Select pertinent data

•

Maintain a log file

•

Signal events on one or more terminals

•

Control your system while it is running

RSTSIE never logs EMTs for certain jobs, including the Error Logger (ERRCPY)
and your EMT logging program itself; in addition the monitor does not allow you
to log certain EMTs.

14.5.2

How EMT Logging Works
If EMT logging is available and active, the monitor inspects each EMT as it is
received. ("Available" means installed in the system; "active" means that your
EMT logging program has properly declared itself as a receiver.) If the newlyreceived EMT is one that RSTSIE logs, the monitor builds a packet of information
about the directive and stores the packet in XBUF. Note that EMT logging does
not use small buffers for message transfer.
The packet contains:
•

Context data, such as date and time, user job number, and keyboard number

•

Directive data, such as information from the job's FIRQB and/or XRB

Miscellaneous System Management Functions

14-5

When the directive is complete, additional information (for example, the RSTSIE
error code) is added to the packet. The packet is then made available to your
program. Your program can inspect the parameters passed from the user's job
to the monitor, determine the relevance of the directive being logged from the
packet, and take the appropriate action.
Your program retrieves EMT logging packets by issuing message receive calls.
Parameters in the declare receiver call of your program specify:
•

How many packets constitute a message

•

How many packets can be outstanding at any time

•

How much XBUF the EMT logger can use

Each message received by your logging program consists of the packets (one per
selected EMT) plus control information. This control information includes a count
of EMTs that may have been lost because one or more of the limits you set up at
declare time has been exceeded.
See the RSTS / E Programming Manual for a description of the use of message
send/receive calls as well as specific information about the send/receive calls used
for EMT logging.

14.5.3

Data Returned by EMT Logging
The meaning of the data returned by the logging program you write depends on
the internal functioning of the monitor. ]~or this reason, both the format and
meaning of data returned are subject to ehange in future releases of RSTSIE.
DIGITAL reserves the right to change int.ernal mechanisms and, therefore, makes
no commitment to continue providing any specific part of the data described for
this release of RSTSIE.

14.5.3.1

Loggable EMTs Enabled or Disabled by SET SYSTEM

For RSTSIE Version 9.7, the following EMTs are available for logging:
CALFIP functions:

14-6

Code

Mnemonic

Description

Code

Mnemonic

Description

0
4
8
16
20
24
28

CLS
CRE
REN
RST
ASS
DAL
CRB

Close a channell
Create a file
Rename a file
Reset channel~'
ASSIGN a device
DEASSIGN all
Create cmpilecl

2
6
10
18
22
26
30

OPN
DLN
DIR
LOK
DEA
CRT
RUN

Open a channel
Delete a file
Directory info
File lookup
DEASSIGN dvice
Create .TMP
Run a program

Miscellaneous System Management Functions

UUO functions:

1 ~~.5.4

Code

Mnemonic

Description

Code

Mnemonic

Description

-29
-27
-25
-18
-15
-13
-11
-8
-3
-1
1
4
6
8
11
13
15
17
20
24
26
29
34

TB3
DMP
ATR
RTS
ACT
PRI
BCK
FCB
TB1
YLG
DLU
LIN
ATT
CHU
DEA
ZER
DIR
LOK
CNV
JOB
SYS
STL
CFG

TablE~s III
Snapshot dump
File attributes
RTS/J:-eslib ctl
Acctng data
Priority, etc.
File stats ctl
FCB/DDB info
Tables I
Logins <= max
Delete PPN
Login
Attach
Password/quota
DEASSIGN
Zero a device
Directory data
Wildcard lookup
Date => ascii
Spawn a job
Job status
Stall system
Configure sys

-28
-26
-19
-17
-14
-12
-9
-6
-2
0
3
5
7
10
12
14
16
19
23
25
27
33

SPL
FIL
LOG
NAM
OAT
TB2
HNG
POK
NLG
PAS
MNT
BYE
DET
ASS
DAL
RAD
TRM
CHE
SWP
PPN
KMC
ONX

SPOOL request
File utility
Set Logins
Set RTS name
Date/time ctl
Tables II
Hangup dataset
Poke memory
Logins <= 1
Create PPN
Mount/dismount
Logout
Detach
ASSIGN
DEASSIGN all
Read acctg data
Terminal char
Cache ctl
Swap+ files ctl
PPN lookup
Connect KMC
Open next file

EMT Logging and System Security
The information that EMT logging provides may be helpful in checking and
maintaining system security. However, the use of EMT logging as a system
security tool depends on many site-specific factors:
•

The types of events that can be watched for

•

The degree of security required

•

The way that EMT logging is set up

•

The experience and judgment of the system manager who sets up EMT
logging and interprets its data

•

The general security practices of the site

In summary, although EMT logging does not guarantee improved system securit~
it can provide a very useful tracking mechanism.

14.6 Monitoring System Status: SYSTAT
During normal timesharing, there are many occasions when you need to monitor
the status of your system. RSTS/E provides two system programs for this
purpose:
DISPLY

Automatically displays system status information on your terminal.
By default, the program displays the system status every 15 seconds;
however, you can specify a different interval. See the following section
for a description of the DISPLY program.

SYSTAT

Displays the status of different parts of the system when you use
various program switches. See the RSTS / E Utilities Reference Manual
for a complete description of SYSTAT and the switches associated with
it.

Miscellaneous System Management Functions

14-'

The description of SYSTAT in this section gives a set of guidelines for using the
SYSTAT program for more effective system managelnent. Use SYSTAT to:
•

Prepare for system shut-down to learn what jobs are active and which disk
devices and assignable devices are in use.

•

Determine when the nwnber of gene'ral small buffers becomes too
low. You can then decide whether to use the DCL commands SET
SYSTEMfNOLOGINS and SET SYSTEMILOGINS to restrict the number of users that can log in to the system. Note that you can also use the
SHOW BUFFERS command to display the nwnber of available general small
buffers.

•

Check on the amount of free disk space. The disk status report in SYSTAT
reflects the apparent nwnber of free blocks on each disk on the system. For
practical purposes, however, such as for allocating a file on the device, all free
blocks that SYSTAT reports may not be usable. You may get the message ?No
room for user on device when SYSTKr reports there are free blocks available.
Note that you can also use the SHOW DISKS command to display the amount
of free disk space.
The file cluster size or the nwnber of clusters you need can prevent a file
from fitting on a device. For example, a file whose cluster size is 16 and
whose length is 10 blocks may not fit on a device that SYSTAT reports to
have 50 free blocks of file space remaining. The cluster size of 16 demands
that 16 contiguous blocks of free space exist on the device before the file can
be allocated to the device. In some cases, 16 contiguous blocks do not exist on
a device and RSTSIE does not allow a file to extend to another physical device
even though SYSTAT can give you so:me indication of the availability of free
disk space.

•

Follow the progress of user jobs. You can determine if a job is stalled, waiting
for resources on your system. If you notice that a run-time value of a job is
not increasing (the value is printed in a job status report), it indicates that
the job is stalled and waiting for an I/O device. One user job can allocate
a device or keep an assignable device locked by having one file open. You
can determine who the user is by examining the device status report that
associates the busy device with the job number of the user controlling that
device. You can then ask the user to free the device or, if that is not possible,
you can use DCL commands to force the job off the system (FORCE) or
seize the device (ALLOCATE/SEIZE). Note that you can also use the SHOW
USERS and SHOW JOBS commands to display user job status.

•

Follow the progress of detached jobs. When SYSTAT reports a detached job in
the HB (hibernate) state, you know the detached job is attempting to communicate with its terminal. In this case, log in to the system at a free terminal,
and then use the ATTACH command to attach the job to the terminal. Once
you attach to the detached job, RSTSIE can display the messages from the
job. Note that you can also use the SHOW JOBSIDETACHED command to
display the status of detached jobs.

•

Check on the number of jobs on the system. The free buffer report lists the
number of jobs currently logged in to the system and the maximum nwnber
of jobs allowed. These nwnbers are useful when you are adding and removing
swap files. After adding the swap files and raising the logins allowed, you can
check the free buffer report to confirnl the success of the procedure. Before
removing a swap file, check this repo:rt to make sure that the nwnber of
logins allowed is low enough to enable you to remove the swap file at some
later time. Note that you can also use the SHOW SYSTEM command to
display the nwnber of jobs on the system.

14-8 Miscellaneous System Management Functions

14.7 Dynamic Display of System Status: DISPLY
The DISPLY system program displays the system status on VT52, VT100, and
VT200 series terminals. Mter you start the program and decide how frequently
you want it to update the system status on the screen, the program runs until
you interrupt its execution. The information DISPLY prints is similar to that of
the SYSTAT program. Commands you type during the execution of the program
can alter the information displayed on the screen. These commands let you
choose only those portions of the system status you need to see.
The program resides in the AUXLIB$: account with the protection code of <232>
This protection code lets all users run the program. If you want to restrict
running the DISPLY program to users that have the WREAD privilege, change
the protection code to <124>.

14.7.1

Running the DISPLV Program
The program prints an identification line and a question after you type the
command line RUN AUXLIB$:DISPLY. They appear as follows:
$ RUN AUXLIB$:DISPLY
DISPLY V9.7 RSTS V9.7 EDERE
Interval?

If you want updates to occur every 15 seconds, press Return to accept the default
Otherwise, type the number of seconds you want to elapse between screen
updates. You can include any combination of the switches in Table 14-2 in your
response. If you select a program switch but do not specify an interval, the
program again assumes the default and updates the screen every 15 seconds.
Table 14-2 lists the DISPLY program switches.
Table 14-2:

DISPLY Program Switches

Switch

Meaning

I[NO]DCA

Indicates if the tenninal can use direct cursor addressing to position
the cursor on the screen. By default, DISPLY uses IDCA.

IDET

Detaches the job from this tenninal or from the tenninal you
specify.

IKBn:

Prints the output at keyboard unit n if it is available. If you specify
the IDET switch, the program runs detached.

Specifies the number of lines to display. The value of n can range
from 12 to 24. By default, DISPLY displays 24 lines.

!PRIORITY

Runs the program at a special priority instead of at the nonnal -8
priority. Requires TUNE privilege.

I[NO]TAB

Indicates if the tenninal can correctly process TAB characters. By
default, DISPLY uses II'AB.

When you include the IDET switch in response to the INTERVAL question
to run the program detached, you interrupt execution and you can use the
terminal for other work. As soon as you enter CtrllC to interrupt the display,
the program prints a message telling you the terminal is available. When you
release the terminal by logging out, the program automatically displays the
status information on the screen as if it had not been interrupted.

Miscellaneous System Management Functions

14-9

Use CtrllC to stop the execution of the DISPLY program. The procedure you
follow after stopping the program depends on whether you were running the
program attached or detached. If you are running the program attached, entering
CtrllC to end the program returns control to the keyboard monitor. When running
the program detached, entering CtrllC causes the program to print the following
message before terminating:
The terminal is all yours now

You must then attach the job running the program to the terminal as follows:
HELLO 1,2 <RET>
PASSWORD:
Jobs detached under this account:
Job What Size State
Run-time
10 DISPLY 16
SL
5.6
Job number to attach to? <RET>

RTS
... RSX

$ Continue <Yes>? N<RET>

When the program attaches to the terminal as a result of the HELLO command,
it prints the CONTINUE question. Finally, type NO or any other string not
beginning with Y and press Return to tenninate the program.
The program displays on your terminal an identification line at the top line of
the screen, skips a line, and fills the left portion of the screen with job status
information and the right side with infornlation about:

•
•

Busy devices

•
•
•
•

Run-time systems

Disks

Message receivers
Free buffers
Resident libraries

Mter the screen is full, the program moves the cursor to the first character on the
second line of the screen. The program is then idle, waiting to cycle through the
display again.
At the specified interval, the program checks the system tables and updates
the status information on the screen with any changed data. While executing
routines to extract update information, the program prints the message:
WORKING ...

It leaves the cursor to the right of the message. Mter the update is complete, the
cursor returns to its idle position.
While the cursor is in the idle position, you can type commands to modify the
contents and arrangement of items on the screen. You can end any command by
pressing the Escape key. Although any line terminator works, Escape leaves the
cursor positioned on the blank line. The program prints no message or takes no
action on invalid statements.
Some commands that add items to the screen take the minus (-) sign as a prefix.
A command with a minus sign preceeding it causes DISPLY to delete rather than
add a display of the information that would normally appear without the minus
sign. That is, the minus sign negates the effect of the command.

14-10

Miscellaneous System Management Functions

Table 14-3 lists the commands you use (usually while the cursor is in its idle
position) to delete or add system status information.
Table '14-3:

DISPLY Program Commands

Command Type

Format and Description

General

C
Clears the screen and displays new status.
Sn
Displays memory status in place of job status. Starts with the
8K-word section less than or equal to n. If n is not given, starts at
the beginning of memory.
J
Displays job status in the standard manner.

In
Displays job status starting with active job n+1. Overcomes
physical limitation of the screen.
Xn
Changes the interval to n seconds.

XO
Updates the display with an interval of 0 seconds (that is, runs
continuously) but lowers the priority so that other jobs are not
stalled.
Job Status

T
Displays total CPU time each job has used. The time is displayed
as number of hours, minutes, seconds, and tenths of seconds under
the RUN-TIME column.

+
Displays the increment of CPU time each job has used since the
display program last updated the screen. The user can return to
total CPU time by typing T.
%

Displays the amount of CPU time each job has used as a percent
of the total CPU time expended. The user can return to total or
increment of CPU time by typing, respectively, T or +.

J-D
Does not display detached jobs.
J+D
Includes detached jobs in display.
JD
Displays only detached jobs.
J-S
Does not display sleeping jobs.
J+S
Includes sleeping jobs in display.
N
Indicates the program name in the WHAT column.

-N
Removes the program name from the WHAT column and replaces
it with name of the RTS under which the job is running.

(continued on next page)

Miscellaneous System Management Functions

14-11

Table 14-3 (Cont.):

DISPLV Program Commands

Command Type

Format and Description

Job Status (con't)

P
Indicates, under the Pr column, the exact priority (modulo 8) of
the jobs.

-p
Indicates, under the Pr column, the relative priority of the jobs.
See the section "Job Status Statistics."
W
Indicates, under the STATE column, the last WAIT state rather
than actual state.

-w
Removes the last WAIT state and indicates the actual state of
each job.
K
Displays, under the SIZE column, the amount of memory occupied
by each job.

-K
Displays, under the SIZE column, the amount of memory remaining for each job.
Disk Structure

D, Dn,-D
Displays disk structure statistics. If n is 1, places the item first
on the screen. A pre(:eding minus sign removes the disk structure
statistics from the screen.

L, -L
Displays, under the COMMENTS column, the logical name of
each device. The preceding minus sign replaces logical names
with standard PUB, PRI, NFS, or LCK notations.
Busy Devices

B, Bn, -B
Displays busy device statistics. If n is 1, the program places the
statistics first on the screen. A preceding minus sign removes
busy device statistics from the screen.

Free Buffer

F, Fn, -F
Displays free buffer statistics. If n is 1, the program places the
statistics first on the screen. A preceding minus sign removes free
buffer statistics from the screen.

Message

M, Mn,-M
Displays message receiver statistics. If n is 1, the program places
the statistics first on the screen. A preceding minus sign removes
message receiver statistics from screen.

Run-Time System

R, Rn, -R
Displays run-time system data. If n is 1, the program places the
data first on the screen. A preceding minus sign removes run-time
data from the screen.

Resident Libraries

14-12

H, Hn,-H
Displays resident lib:rary data. If n is 1, the program places the
data first on the screen. A preceeding minus sign removes the
resident library data from the screen.

Miscellaneous System Management Functions

14,,7.2 Screen Layout
The program divides the screen into four major sections:
Header line

Contains the RSTSIE version number, the name of the
system, the current date and time of day, and the number of
hours, minutes, and seconds since the start of time-sharing
operations (termed "up-time"). The header line appears on
the first line of the display.

Total statistics line

Tells the percentage of time that is expended by users (User),
input and output processing (110), and the monitor (Exec) as
well as the amount of idle time (Idle) and lost time (Lost).
The statistics print on the line below the header line but
are replaced at the interval you specify by the WORKING...
message. (The statistics are only printed if you use the SET
SYSTEM/STATISTICS command.)

Left-hand portion

Contains either the job status statistics or, if you use the S
command, the memory usage status.

Right-hand portion

Contains statistics for busy devices, the disk structure, runtime systems, message receivers, system buffers, and resident
libraries.

The DISPLY program prints information similar to the information SYSTAT
includes in its status reports. The description of the major sections of the
dynam.ic status report includes information about how the DISPLY program
reports differ from those of the SYSTAT program. If you need a more complete
description of SYSTAT, see the RSTS / E Utilities Reference Manual.
The following example displays all four divisions. While the example does not
represent an actual log of the DISPLY program, it does give you a picture of how
information is displayed. When you run the program on your system, not all the
illustrated information will fit on your terminal screen. You will often need to
use commands described in Table 14-3 to delete reports so unseen portions of the
display become visible. The example lets you see an entire display, unhindered by
screen limitations (usually 24 rows).
RS'I'S V9. 7-()4 NOODLE: :
Status on 06-Feb-89 02:10 PM Up: 118:16:34
58.6%User,
15.2%I/O,
25.6%Exec,
.2%Idle,
.4%Lost
Job Who Where What Size State Run-time Pr
Busy devices
1 197,1
Det
ERRCPY 5 SR
Dev Job Why Dev Job Why
2
1,2
Det
OPSRUN 23 SL
MM1 26 Open MSO 26 Open
3
1,2
Det
EVTLOG 18 SL
4
Det
PBS ... 20 SL
1,2
Disk Structure
5
1,2
Det
QUMRUN 24 SL
+ Dsk Open
Free Clu Err Comments
6
1,2
Det
SPLIDL 23 SL
DVO
0
23 1
0 Pri,DLW
7
1,2
Det
BATIDL 20 SL COl
- DRO
3
14888 8
0 Pri,DLW
8
1,2
Det
BATIDL 20 SL
- DUO 134 177696 16
0 Pub,DLW
11
1,2
Det
MAILSN 32 SR
DU1
0 194544 16
0 Pri,R-O
12 194,1
KB47 DCL
1 "'C
13
1,2
Det
TWATCH 5 SL
Run-time systems
14 202,1
KB46 DeL
4 "'C
... RSX
OK 16 Mon,KBM
15 248,1
KB44
... EDT 8 KB
.3
DCL
24K 13 Trnp,KBM
16 248,1
KB45 DCL
4 "'C A08
RT11
4K
1 Tmp,KBM,CSZ
20 206,1
KB49 DCL
4 "'C
BASIC
16K
1 Tmp,KBM,CSZ
22 225,1
KB52 ... EDT 8 KB
BP2
1K
0 Non-res,KBM
24 223,1
KB51 DISPLY 17 RN
.8
FORTH
6K
0 Tmp,KBM,NER
29 208,6
KB59 DCL
4 "'C
30 206,1
KB56J4BACKUP 18 RN
1.2 K- Gen FIP Jobs TTY Err
31 208,6
KB60J4LINK
28 RN
.3
187 214 30/45
0 3995
35 196,1
KB63 DCL
4 "'C

Miscellaneous System Management Functions

14-13

Message receivers
ERRLOG(Prv)
1 0
OPSER (Loc)
2 0
EVTLOG(Prv)
3 0
EVTLSN(Prv,Nt) 3 1
QM$CMD(Prv)
4 1
QM$SRV(Prv)
4 2
QM$URP(Loc)
4 3

0/40
0/30
0/32
0/16
0/20
0/30
0/10

Resident Libraries
HISTOG 36K
o Prm,R/W,NER
8K
CSPLIB
4 Tmp
EDT
39K
1 Tmp
RMSRES
4K
2 Tmp
DAPRES 10K
o Non-res
o Tmp
BP2RES 20K
B25SHR
4K
o Tmp

14.7.3 Job Status Statistics
The job status report displays infonnation in eight separate columns on the
left-hand portion of your tenninal. The report is similar to the infonnation the
SYSTAT program prints in its display of job status. Unlike the SYSTAT program,
the job status report does not include the name of the run-time system under
which the job is running but does include abbreviations indicating the priority of
the running job. The titles that the DISPLY progra:m prints at the top of a job
status report have the following meanings:
Job

Job number that the system assigns when the job starts timesharing
activities.

Who

Account number under which each job runs. If the job is not logged in to
the system, **, ** appears in this column.

Where

Keyboard number of the job. DET appears in place of the keyboard
number for jobs that run detached from the keyboard. The abbreviation
KBxJy can appear for a job running on a pseudo keyboard. The value KBx
identifies pseudo keyboard unit x; and the value Jy denotes job number y,
under which the controlling job is running.

What

Program name that the job is executing.

Size

Current size in K words of the job.

State

Current state of the job indicated by the set of abbreviations in Table 14-4.

Run-time

Hours, minutes, seconds, and tenths of seconds of central processor (CPU)
time the job has consumed.

The job status report includes a PR column, which is not displayed by SYSTAT,
that identifies the priority of the running job. The PR column can display the
following abbreviations:
If -P is in effect:
+

Higher than normal priority
Lower than normal priority

14-14

Special run priority

CtrllC temporary priority

Keyboard delimiter temporary priority

Miscellaneous System Management Functions

If P is in effect:

Positive priority n * 8

Zero priority

-n

Negative priority n * 8

Table 14-4 contains a brief description of the abbreviations that can appear in
the STATE colwnn of the job status report. The following status descriptions can
appear after one or more of the job state abbreviations in Table 14-4:
Lck

Job is locked in memory for the current operation.

Nsw

Job has requested that it not be swapped from memory and cannot
be swapped unless it requests additional memory.

Swi

Job is currently being swapped into memory.

Swo

Job is currently being swapped out of memory.

Xnn

Job is swapped out and occupies slot nn in swap file X; file is
denoted A, B, C, D to represent files 0 through 3 of the swap
structure.

Table 14-4:

STATE Column Attributes

Abbreviation

Meaning

Job state cannot be determined.

Job is waiting for buffers (no space is available for 110 buffers).

Job is in CTRUC state, awaiting keyboard monitor input.

Job is waiting for card reader input.

DB,DK,DL,DM,
DR,DU,DV

Job is waiting to perform disk 110.

DX
FP

Job is waiting for floppy diskette 110.

Job is detached and waiting to perform 110 to or from a terminal.

Job is waiting for file processing action by the system (opening or
closing a file, file search).

Job is waiting for input from a terminal.

Job is waiting to perform line printer output.

MT,MM,orMS

Job is waiting for magnetic tape 1/0.

Job is waiting for RJ2780 1/0.

Job is running or waiting to run.

Job is waiting for residency.

Job is sleeping.

Job is sleeping and is a message receiver.

Job is waiting to perform output to a terminal.

Miscellaneous System Management Functions

14-15

14.7.4

Busy Devices
The busy device report lists the devices that are assigned or opened by a specific
user. Items reported are the device specification, the job owning that device, and
the condition of the device. The disk status information reports assigned disk
units. The busy device report prints on the right-hand side of the screen and
looks as follows:
Dev Job
PRO 11

Busy Devices
Why Dev Job
Open PR1
8

Why
Open

The DISPLY program reports the same information as SYSTAT reports in its busy
device report, except that DISPLY prints two side-by-side reports to accommodate
the limited screen space. Table 14-5 contains the abbreviations that can appear
in the WHY column of the busy device report.
Table 14-5:

14.7.5

Busy Devices Status Abbreviations: WHY Column

Abbreviation

Meaning

Device is explicitly assigned to a job.

OPEN

Device is open on a channel.

DOS

Magnetic tape is assigned with DOS labeling format.

ANSI

Magnetic tape is assigned with ANSI standard labeling format.

Disks
The disks report describes each disk in Ulse on the system. The report contains
the same information that SYSTATID displays. The DISPLY program prints the
following information; however, because of screen limitations, it does not print
column headings or as much detail as SYSTAT does:
•

Disk device specification

•

Number of open files

•

Number of free 512-byte blocks

•

Pack cluster size

•

Disk hardware error count

•

Comments on the status of the disk

•

Pack identification name or system logical names (if any) assigned for the
devices in use on the system

The disks report appears on the right-hand side of your terminal and looks as
follows:
Disk Structure
DRO 0 17252 4
0 Pri,DLW
3044 4
2 Pub,DLW
DR1 42
DR2 0 12496 4
0 Pri,R-O
DR3 5 32144 8
3 Pri,DLW
3272 4
DR4 1
1 Pri,DLW
6 Pri,R-O
DR5 o 532032 16

14-16

Miscellaneous System Management Functions

1b display the pack identification names or system logical names, you must use
the L command described in Table 14-3. This command displays the disk label
information in place of the comments that the program normally displays. For
example:
Disk Structure
0 ROGER
ORO 0 17252 4
DR1 42
3044 4
2 SYS
DR2 0 12496 4
o ZEBRA
3 ABLE
DR3 5 32144 8
DR4 1
3272 4
1 MIKE
DR5 0 532032 16
6 YOKE

If you want to display the comments again, use the -L command. The limited siz~
of the screen requires the use of the Land -L commands.
Table 14-6 lists the abbreviati.ons in the COMMENT column of the disks report.
Table 14-6:

14.7.6

Disk Status Abbreviations: COMMENT Column

Abbreviation

Meaning

Pub
Pri

Cartridge or pack is public.
Cartridge or pack is private.

NFS

Disk is open as a non-:file·structured device.

R-O

Disk unit is read-only (write-locked)

DLW

Date of last write (modify), rather than date of last access, is stored
in :file accounting entries.

Lck

Disk i$ in a restricted state.

NFF

New files on this disk are put at the beginning of the directory.

Message Receivers
The message receivers report:
•

Provides the job name of the receiving job

•

Provides the job number of the receiving job

•

Includes the number of messages queued for the job

•

Gi.ves the declared maximum number of messages the job can queue

•

Tells whether local and network senders are allowed

•

Indicates whether local senders must have SEND privilege

The message receivers report appears on the right-hand portion of the screen and
often does not appear until you delete the disk and run-time system reports with
the -D and -R commands, respectively. A sample of the message receivers report
is:
Message receivers
ERRLOG(Prv)
1
OPSER (Loc)
2
3
QUEMAN(Loc)
LPOSPL(Prv)
4
5
BAOSPL (Prv)
BA1SPL(Prv)
6
EVTLSN(Prv,Nt) 7
7
EVTLOG (Prv)

0
0
0
0
0
0
0
1

0/40
0/30
0/60
0/5
0/5
0/5
0/16
0/32

Miscellaneous System Management Functions

14-17

Due to limited space on the screen, some of the abbreviations are shorter than
those printed by the 1M switch of SYSTAT. The receiver report does not include
the:
•

Receiver Identification Block (RIB)

•

Object type

•

Number of links used and the maximum links allowed.

Table 14-7 describes the abbreviations found in the message receiver status
report.
Table 14-7:

14.7.7

Message Receivers Abbreviations

Abbreviation

Meaning

Loc

Local senders are allowed for this receiver ID.

Prv

Local senders must have SEND privilege to send to this receiver ID.

Network senders are allowed for this receiver ID.

Receiver can handle one and only one link.

Both of the above. Network senders are allowed for this receiver ID.
Receiver can handle one and only one link.

Free Buffers
The free buffers report provides the following infonnation:
•

Number of general small (16-word) buffers not currently in use

•

Number of FIP buffers not currently in use

•

Number of jobs currently running

•

Maximum number of jobs allowed to run

•

Number of hung terminal errors

•

'Ibtal number of errors logged on the system

The report prints the same information t.he IF switch of the SYSTAT program
produces. An example of a free buffers report is:
Gen FIP
183

14.7.8

Jobs

TTY

Err

1 14/50

Run-Time Systems
DISPLY shows the run-time systems report on the right-hand portion of the screen.
The report includes the same information as SYSTATIR displays, except for:

•

The maximum size in K words that a job running under the run-time system
can take

•

Some information in the comments column because of limited screen space

The run-time systems report prints:

14-18

•

Name of each run-time system

•

Size of the run-time system in K words

Miscellaneous System Management Functions

•

Number of user jobs currently executing under the control of the run-time
system

•

Comments regarding the status of the run-time system

A sample run-time systems report is:
Rs'rs V9. 7 ERgRE::
Status on 31-Jun-89 04:34 PM Up: 120:08:06
5.8%User,
2.7%I/0
.5%Exec,
91.0%Idle,
.0%Lost
Busy devices
Job Who Whc:lre What Size State Run-time Pr
1
1,2
Dat
ERRCPY 5 SR
+
*** None ***
2
1,2
DC:lt
OPSRUN 23 SL
.1
3
1,2
Dat
EVTLOG 18 SL
Disk Structure
PBS ... 21 SL
4
1,2
DC:lt
Dsk Open
Free Clu Err Corrun
+ DVO
0
21 1
5
1,2
Dat
QUMRUN 24 SL
o Pri,
6
1,2
Dat
SPLIDL 23 SL
DRO
3
46640 8
o Pri,
7
1,2
Dat
BATIDL 20 SL
- DUO 116 175472 16
o Pub,
8
1,2
DC:lt
BATIDL 20 SL
- DU9
0 390752 16
o Pri,
9
1,2
Dat
BATIDL 20 SL
- DU1
0 230320 16
o Pri,
- DU1
o 194544 16
10
1,2
Dat
MAILQ 32 SR
o Pri,
11
1,2
DC:lt
MAILSN 32 SR
12 223,1
KB43 DCL
4 "C
Run-time systems
- ... RSX
OK 13 Mon,KBM
13
1,2
DC:lt
T~iATCH
5 SL
14 217,1
KB44 DeL
4 "C
DCL
24K 10 Tmp,KBM
4K
0 Tmp,KBM,CSZ
RT11
15 236,1
KB53 DCL
1 "C
TECO
10K
0 Tmp,KBM
16 202,1
KJ346 NONAME 2 "C
1.3
BAS4AL 16K
2 Non-res,KBM
18 194,1
KB47 DISPLY 17 RN

Table 14-8 contains a list and description of each abbreviation that DISPLY
prints in the COMMENTS column.
Table 14-8:

Run-Time Systems and Resident libraries Report Abbreviations

Abbreviation

Meaning

N on-res

Run-time system or library is nonresident.

Run-time system or library is being loaded int.o memory.

Temp

Run-time system or library is removed from memory when not.
being used.

Perm

Run-time system or library stays in memory when not being used.

Addr:xxx

The value of xxx denotes the starting address of the run-time
system or library.

[DF] KBM

Run-time system or library can serve as a keyboard monitor. The
optional prefix DF indicates default keyboard monitor.

1US

Run-time system or library can serve only one user.

RIW

Run-time system or library allows read/write access.

NER

Errors occurring within the run-time system or library are not sent
to the system error log.

Rem

Run-time system or library is removed from memory as soon as all
of its jobs switch to another run-time system or library.

CSZ

Proper job image size (in K words) to run a program can be computed as K-size=(file size+3)/4.

EMT:yyy

Denotes the EMT code for special EMT prefix.

Miscellaneous System Management Functions

14-H

14.7.9

Resident Libraries
The resident libraries report includes:
•

Name of the resident library

•

Protection code of the resident library

•

Size of the resident library

•

Number of user jobs currently executing under its control

•

Comments regarding the status of the resident library

DISPLY shows the resident libraries report on the right-hand portion of the
screen. The report shows the same information as SYSTATIL displays. Some
information in the COMMENTS column may be omitted due to limited space
on the screen. See Table 14-8 for a description of the abbreviations used in the
COMMENTS column.

14.7.10

Memory Status
The S command causes the program to print a table that shows the use of each
1K-word portion of memory. The memory status report replaces the job status
report on the lefthand half of the screen. Use the J command to display the job
information again.
Type the S command to have DISPLY print the memory status report as in this
example:
Memory usage (Starting at OK)
o MON MON MON MON MON MON
8 MON MON MON MON MON MON
16 MON MON MON MON MON MON
24 MON MON MON MON MON MON
32 MON MON MON MON MON MON
40 MON MON MON MON MON MON
48 MON MON MON MON MON MON
56 *1* *1* *1* *1* *1* *1*
64 *1* *1* *1* *1* *1* *1*
72 XBF XBF XBF XBF XBF XBF
80 XBF XBF XBF XBF XBF XBF
88 XBF XBF XBF XBF XBF XBF
96 XBF XBF XBF XBF XBF XBF
104 XBF XBF XBF XBF XBF XBF
112 XBF XBF XBF XBF XBF XBF
120 XBF XBF XBF XBF XBF XBF
128 XBF XBF XBF XBF XBF XBF
136 XBF XBF XBF XBF XBF XBF
144 XBF XBF XBF XBF XBF XBF
7
7
152 7
7
7
7
160 7
7
7
7
7
7

MON
MON
MON
MON
MON
MON
MON
*1*
*1*
XBF
XBF
XBF
XBF
XBF
XBF
XBF
XBF
XBF
"1

7
7

Due to limited screen space, the program may not be able to display all memory
status information simultaneously. The Sn command lets the user determine the
starting 8K section. The display progranl prints the starting section number for
the row and eight abbreviations per row. Each of the abbreviations relates to the
status of a 1K section of memory. The number of rows the program prints and
the extent of memory covered is limited by the terminal. The program indicates
the starting 1K section by printing a header line in the following format:
MEMORY USAGE (STARTING AT nK)

14-20

Miscellaneous System Management Functions

Table 14-9 contains a description of the abbreviations the memory status report
uses.
Table "14-9:

Memory Status Report Abbreviations

Abbreviation

Meaning

MON

Occupied by the RSTSIE monitor.

*n*

Occupied by run-time system n, where n is the position in the runtime system list. (A question mark is printed if the run-time system
name cannot be determined.)

-n-

Occupied by resident library n, where n is the position in the
resident library list.

Occupied by job number n.

nLK

Job number n is locked in this lK portion.

Job number n is being swapped out of memory.

nSI

Job number n is being swapped into memory.

NXM

Memory space is nonexistent.

END

End of physical memory for user jobs.

LCK

Memory is locked.

XBF

Memory is reserved for the extended buffer pool (XBUF).

14.8 Electronic Messaging Systems
RSTS/E can tie in to electronic messaging systems. Electronic messaging
systems, such as Telex and TWX, resemble electronic mail systems. However, the
messages are not sent between accounts on computers, but between machines
(computers or simply terminals). The senders and most of the receivers are
subscribers to the electronic messaging service, which sends the messages over
the public telephone system.
Every subscribing machine in a messaging system has an identification.
When the machine sends or receives messages, it sends this identification
to the messaging service to verify its identity. Use the DCL command SET
SYSTEM/ANSWERBACK, to store your system's identification. For example:
$ SET SYSTEM/ANSWERBACK="221BBAKERSTREET"

The identification text must be 24 alphanumeric characters or less. Most
messaging services do not allow more than 15 characters. RSTSIE allows spaces
and punctuation marks, but most messaging services do not.
You can set any terminal, or any number of terminals, to receive these messages.
Use the DCL command SET TERMINAUANSWERBACKIPERMANENT to
designate a terminal. For example:
$ SET TERMINAL KB12:/ANSWERBACK/PERMANENT

When messages arrive, RSTSfE stores each of them in a file with a unique
name based on the time, for example, 118432.001. The first four digits count
the number of minutes since midnight (1184); the next two count the number
of seconds since the beginning of the current minute (32). The number in the
filetype gives the order in which the message was received. Thus the second
message to be received in the same second goes into 118432.002.

Miscellaneous System Management Functions

14-21

RSTSIE keeps all these files in the EMS$: account. You must create an account
and assign the logical name EMS$: to it by including this DCL command in your
[O,l]START.COM file:
$ ASSIGN/SYSTEM dev:[p,pn] EMS$:

You can print these files for manual disttibution, or an application program can
scan them to determine which user on the system should receive which file.
The Unsupported Software package contains a sample procedure under the name
EMS.COM.
You can write application programs for other systems that deliver messages to
the RSTSIE system in the same manner as an electronic messaging service. 1b
use such a program, you must:
•

Have a phone line to a modem connected to the RSTSIE system

•

Have the program send CtrllE to the RSTSIE system

•

Know what identification to expect in response to the CtrllE

•

Dump an ASCII file containing the message to the RSTSIE system

14.9 DCl Command Descriptions
The following sections describe the DCL Icommands for the miscellaneous system
functions.

14.9.1

LOAD/INDEX Command

The LOADIINDEX command lets you load SATs into memory during timesharing.
By making SATs resident in memory, you can improve the system's performance
at the cost of increased memory (XBUF) usage. LOADIINDEX requires SWCTL
privilege.

Format
LOAD/INDEX [device-name]
Command Qualifiers

Default!»

none

Prompts

none

Prompts
Command Parameters
device-name
Specifies the disk whose SAT you want to load into memory. If you do not specify
a device, the default is the system disk.

14-22

Miscellaneous System Management Functions

14.9.2

LOAD/OVERLAY Command

The LOAD/OVERLAY command lets you load monitor overlays in memory during
timesharing. By making certain monitor functions resident in memory, you can
improve the system's performance at the cost of increased memory (XBUF) usage.
LOAD/OVERLAY requires SWCFG privilege.

F()rmat
LOAD/OVERLAY

overlay-name[, ... ]

Command Qualifiers
fAll

Defaults
none

Prompts
Overlay name:

Prompts
Command Parameters
overlay-name
Specifies the monitor overlay that you want to load into memory. The allowable
values are:
•

ATTRIBUTE-Indicates the monitor overlay that performs file and account
attribute read/write operations.

•

DCL-Indicates the monitor overlay that performs file-related operations for
DCL.

•

DELETE_RENAME-Indicates the monitor overlay that performs file deletior
and renaming.

•

DIRECTORY-Indicates the monitor overlay that performs disk file lookup
operations.

•

SYSTEM_CALLS-Indicates the monitor overlay that performs all of the
following monitor directives:
Get monitor tables (parts 1, 2, and 3)
Return open file information
Date and time conversion
Return job status

•

TERMINAL-Indicates the monitor overlay code that performs terminal
characteristics SET and SHOW operations.

Command Qualifiers
fAll
Indicates that all possible overlays should be loaded into memory. If you specify
this qualifier, RSTSIE does not allow any overlay names.

Miscellaneous System Management Functions

14-2:

14.9.3 SET DATE Command
The SET DATE command sets the RSTS/E system date to the value you specify.
Optionally, you can specify a new value for the system time. You can specify relative values for the date or time. See the previous section Changing System Date
and Time for an explanation of relative dates and times. SET DATE requires the
DATES privilege.

Format
SET DATE

date[:timeJ

Command Qualifiers

Defaults

none

Prompts
Date:

Command Parameters
date
Specifies the system date. Insert a space between the date field and the DeL
command. You can specify a relative date with the keyword DAYS.
[:tlme]
Specifies the system time. You can specify a relative time with the keywords;
HOURS and MINUTES. Insert a colon (:) between the date and time fields.
NOlrE

When you specify the AMlPM suffix, there is no space between the
minutes and the suffix. For example:
$ SET DATE 5-JAN-90:1:30PM

14-24

Miscellaneous System Management Functions

14.9.4 SET SYSTEM Command
The SET SYSTEM command sets the default system characteristics. You need
one or more privileges to use this command (see the discussion for each command
qualifier).

Fc)rmat
SET SYSTEM
Command Qualifiers
IANSWERBACK="identification
IDATE_FORMAT=format
IDENSITY=default
IDYNAMIC_REGION_LlMIT=n
IEMT_LOGGING=(emt1 ,emt2, ... }
I[NO]FMS
IHANGUP={DELAYED I IMMEDIATE}
IHOLD
lLABEL=default
I[NO]LAT
I[NO]LOG
I[NO]LOGINS[=n]
IMON ITOR_NAME=file.name
INAME="instaliation name"
I[NO]PASSWORD_PROMPT[=(type[, ... ])]
IPOWERFAIL_DELAY=n
IPSELJDO_KEYBOARDS=n
IRElEASE
I[NO]STATISTICS[=RESET]
ISWAP_MAXIMUM=n
!TIME_FORMAT=format
lU

Defaults
none
See Discussion
See Discussion
none
none
none
IHANGUP=DELAYED
none
See Discussion
none
IlOG
IlOGINS=job max
none
none
See Discussion
none
IPSEUDO_KEYBOARDS=4
none
INOSTATISTICS
none
See Discussion

Prompts
None

Command Qualifiers
I ANSWERBACK="ldentlfication"
Sets the system identification for electronic messaging systems such as Telex or
TVVX. The exact format of the identification depends on the messaging system
you subscribe to. RSTSIE requires that it have no more than 24 characters, all
alphanumeric. Most messaging services do not allow identifications longer than
15 characters, or with spaces or punctuation marks. You need SWCTL to use this
qualifier.

Miscellaneous System Management Functions

14-25

IDATE_FORMAT=format

Sets the default date format:
•

ALPHABETIC-Sets the default date format to alphabetic (DD-MMM-YY)

•

NUMERIC-Sets the default date fonnat to nwneric (YY.MM.DD)

You need SWCFG privilege to specify this qualifier.
IDENSITY=default

Specifies the magnetic tape default density; n can be any nwnber between 2 and
32766, or the keywords MINIMUM or MAXIMUM may be used. Generally, the
available tape drives support 800, 1600, or 6250 bpi densities. If the system
default density is less than the drive's lowest possible density, then RSTSIE uses
the drive's lowest possible density.
If you do not specify a density for the INITIALIZE, or MOUNT commands, the
system chooses the nearest legal density for that tape! drive that is not greater
than the system default density. If you specify MINIMUM or MAXIMUM for
the default density, the system chooses the tape drive's minimum or maximwn
density for the INITIALIZE, or MOUNT commands.
You need HWCFG privilege to specify this qualifier.
IDYNAMIC_REGION_LlMIT=n

Sets the amount of memory available to use as dynamic regions for those users
and applications without the INSTAL privilege. Enter a nwnber n for nK words.
The available memory must be shared arnong all the users an4 applications on
the system, so there may be less than nK available .at any given time. You need
the SWCFG privilege to use this qualifier.
IEMT_LOGGING=(emt1,emt2, ... )

Enables and disables logging for the particular EMT information you want. For
example:
$

SET SYSTEM!EMT LOGGING= (NOCRE, NODJ:'N, ASS, DEA)

This command disables logging for creating files (CRl~) and deleting files (DLN),
at the same time enabling logging for aSl3igning devices (ASS) and deassigning
devices (DEA). See the section Loggable EMTs Enabled or Disabled by SET
SYSTEM earlier in this chapter. You need the SWCTL privilege to use the IEMT_
LOGGING qualifier on the SET SYSTElVI command.
I[NO]FMS

Enables FMS the next time the system restarts. The /NOFMS qualifier disables
FMS. You need the SWCFG privilege to llse this qualifier.
IHANGUP={DELAYED I IMMEDIATE}

Sets whether or not the system delays hanging up dial-up lines when carrier is
dropped. The default is DELAYED, which means the system waits approximately
five seconds before hanging up a dial-up line. Setting IMMEDIATE means the
system hangs up a dial-up line as soon as the carrier is dropped. IMMEDIATE is
mainly used for European dial-up lines. You need SVVCFG privilege to use this
qualifier.
IHOLD

Stops execution of all jobs except the one issuing the command. You need HWCTL
privilege to use this qualifier.

14-26

Miscellaneous System Management Functions

ILABEL=default

Sets the default magnetic tape label:
•

ANSI-Sets the default magnetic tape label to ANSI

•

DOS-Sets the default magnetic tape label to DOS

You need SWCFG privilege to use this qualifier.
I[NO]LAT

Enables LAT support the next time the system restarts. The INOLAT qualifier
disables LAT support, saving 2K words of memory. You need the SWCFG
privilege to use this qualifier.
I[NO]LOG

Use with the IEMT_LOGGING qualifier. Controls whether or not a confirmation
message displays when you enable or disable EMT logging. The default is /LOG
(confirmation displays).
I[NO]LOGINS[=n]

Specifies the maximum number of users who can log in to the system at one time.
SET SYSTEM/LOGINS=n sets the number of jobs that can be logged in to the
system at one time to n. The default for n is JOB MAX. Specify INOLOGINS if
you want to prevent users from logging in to the system by setting the number of
logins to one.
IMONITOR_NAME=file.name

Sets the monitor name to the name you specify. The monitor name is a file name
and has the usual file name format. The new monitor must reside in _SYO:[0,1]
and have a file type of .SIL. The system uses the new monitor at the next
automatic restart, or as the default name the next time you start timesharing.
You need SYSIO privilege to use this qualifier.
INAME="lnstaliatlon name"

Sets the system name to the name you specify. You can ~pecify up to 15 characters in the name. You need SWCFG privilege to use this qualifier.
I[NO]PASSWORD_PROMPT[=(type[, ... ])]

Specifies the types of users that must enter the system password. If you do
not specify any types, all users must enter the password. If you want to limit
prompting to remote users, use the types NETWORK and DIALUP. You need
WACNT privilege to use this qualifier.
IPOWERFAIL_DELAY =n

Sets the power fail delay to n. When a power failure occurs, the system waits n
seconds before rebooting the system. The value of n can range from 1 to 300. You
need SWCFG privilege to use this qualifier.
IPSEUDO_KEYBOARDS=n

Specifies the maximum number, n, of static pseudo keyboards available on the
~ystem the next time the system restarts. n can range from 1 to 127. You need
the SWCFG privilege to use this qualifier.
IRELEASE

Resumes execution of all jobs on the system. You need HWCTL privilege to use
this qualifier.

Miscellaneous System Management Functions

14-27

I[NO]STATISTICS[=RESET]

Turns monitor statistics on and off. The system starts with statistics turned
off. Once statistics have been turned on, the SET SYSTEMINOSTATISTICS
command freezes statistics. The SET SYSTEM/STATISTICS command unfreezes
them. SET SYSTEM/STATISTICS=RESET resets the statistics counters to zero.
You need the SWCTL privilege to use the qualifier.
ISWAP_MAXIMUM=n

Sets the maximum size of the swap file to the value you specify. Enter a number
n between 32 and 64 for a size of 32K to 64K words. ~rhe system uses the new
maximum size at the next restart. You need TUNE privilege to use this qualifier.
ITIME_FORMAT=format

Sets the default time format:
•

AM_PM-Sets the default time format to AMlPM time (1:30PM)

•

24_HOUR-Sets the default time forrnat to 24-hour time (13:30)

You need SWCFG privilege to use this qualifier.

14-28 Miscellaneous System Management Functions

14.9.5 SET TIME Command
The SET TIME command sets the RSTSIE 24-hour clock to the value you specify.
Optionally, you can set the system date to the value you specify. You can specify
relative values for the date or time. See the previous section Changing System
Date and Time for an explanation of relative dates and times. SET TIME requires
the DATES privilege.

Fc)rmat
SET TIME

[date.1time

Command Qualifiers
none

Defaults
none

Prompts
Time:

De!scription
The SET TIME command has no qualifiers.

Command Parameters
[date:]
Specifies the system date. If you specify this optional field, insert a space between
the DeL command and the date field and insert a colon between the date and
time fields. You can specify a relative date with the keyword DAYS.
time
Specifies the system time. You can specify a relative time with the keywords;
HOURS and MINUTES.
NOTE

When you specify the AMlPM suffix, there is no space between the
minutes and the suffix. For example:
SET TIME 1: 30PM

Miscellaneous System Management Functions

14-29

14.9.6

SHOW SYSTEM Command

The SHOW SYSTEM command displays the system default characteristics as
follows:
$ show system
System name:
Currently installed Monitor:
Swap Maximum:
Date format:
Time format:
Magtape label default:
Magtape density default:
Power fail restart delay:
Dynamic Region Limit:
Hangup:
Monitor Statistics:
Job limit:
Current jobs:
Password Prompting:
Last System Password change:
LAT is installed
FMS is not currently loaded
Current Pseudo keyboards:
Answerback message:

RSTS V9.7-04 Managers Only
BOSSMN
Upon reboot: V97BL3
64K
Upon reboot: 32K
Alphabetic
AM PM
DOS
1600 BPI
300 seconds
OK words
Delayed
Disabled
13
4

All users
15-Nov-88 at 10:35 AM
FMS will be enabled on reboot
4
Upon reboot: 6
None

The SHOW SYSTEM command does not require any privileges.

Format
SHOW SYSTEM
Command Qualifiers
IEMT_LOGGING

Defaults
none

Prompts
none

Command Qualifiers
IEMT_LOGGING
Displays the current logging status of all EMTs. You need no special privilege to
use this qualifier.

14-30

Miscellaneous System Management Functions

14.9.7

UNl.OAD/INDEX Command

The UNLOADIINDEX command lets you Wlload SATs that were loaded into
memory using the LOADIINDEX command. UNLOADIINDEX requires SWCTL
privilege.
Fc~rmat

UNLOAD/INDEX

[device-name]

Command Qualifiers
none

Defaults
none

Prompts
none

Command Parameters
device-name
Specifies the disk whose SAT you want to Wlload from memory. If you do not
specify a device, the default is the system disk.

Miscellaneous System Management Functions

14-31

14.9.8

UNLOAD/OVERLAY Command
The UNLOAD/OVERLAY command lets you unload lnonitor overlays that were
loaded into memory with the LOAD/OVERLAY comnland. UNLOAD/OVERLAY
requires SWCFG privilege.

Format
UNLOAD/OVERLAY

overlay-name[, ...]

Command Qualifiers
fAll

Defaults
none

Prompts
Overlay name:

Command Parameters
overlay-name
Specifies the monitor overlay that you want to unload from memory. The
allowable values are:
•

ATTRIBUTE-Indicates the monitor overlay that performs file and account
attribute read/write operations.

•

DCL-Indicates the monitor overlay that perfornls file-related operations for
DCL.

•

DELETE_RENAME-Indicates the rnonitor overlay that performs file deletion
and renaming.

•

DIRECTORY-Indicates the monitor overlay that performs disk file lookup
operations.

•

TERMINAL-Indicates the monitor overlay code that performs terminal
characteristics SET and SHOW operations.

Command Qualifiers
fAll
Indicates that all possible overlays should be unloaded from memory. If you
specify this qualifier, RSTSIE does not allow any overlay names.

14-32

Miscellaneous System Management Functions

Chapter 15

Managing Disks and Tapes
This chapter describes the DCL commands that you use to manage RSTSIE
disks and tapes. First, Table 15-1 lists and briefly describes the DCL commands
available to you. Then, the chapter provides background information about
disk and tape handling on RS'rSIE. Finally, the chapter lists each command
alphabetically and explains it in more detail.
Some functions of INITIALIZE, MOUNT, and DISMOUNT require privilege.
This manual describes all the functions. The RSTS / E System User's Guide also
describes those functions that do not require any privileges.
Table 15-1 :

1S.1

DCl Commands for Disk and Tape Handling

Command

Meaning

INITIALIZE

Initializes a disk or magnetic tape. Use this command to prepare
the medium for writing new files.

MOUNT

Logically mounts a disk or magnetic tape. Use this command after
you physically mount the disk or tape on the device.

DISMOUNT

Logically dismounts a disk or magnetic tape. Use this command
before you physically dismount the disk or tape from the device.

Working with Disks and Tapes
The system manager or opera tor is often responsible for mounting and dismounting disk packs and magnetic tapes. Indeed, only a user with MOUNT
privilege can load disks on the public structure. Thus, you need to be familiar
with commands that prepare tapes and disks for use: INITIALIZE, MOUNT, and
DISMOUNT.

151.1.1

Initializing Disks and Tapes
Initializing a disk or tape makes it "like new." You should consider initializing if
you have:

•

An old disk or tape to recycle; you initialize it to eliminate any existing files
that it contains.

•

A new disk or tape; you initialize it to perform all the necessary actions that
make it usable on your system. (The INITIALIZE command descriptions
explain these necessary actions.)

Managing Disks and Tapes

15-1

Certain disks require fonnatting before they are initialized. Fonnatting writes
timing and sense marks onto the disk and destroys any infonnation that the
disk contains. You fonnat disks using th€l DSKINT option of INIT.SYS; for a
description see the RSTS / E Installation and Update Guide.
You must shut the system down to fonnat the following disks before using them
on a RSTSIE system: RK05, RK05F, RP02, RP03, and RX33.
NOTE

These disks need to be fonnatted only once. All other disks have been
formatted at the factory and need not be formatted again.
The DSKINT option also creates a RSTSfE file structure on the disk. You can
later use the DCL INITIALIZE command to clear the disk of files and recreate
the RSTSIE file structure.
The INITIALIZE command:
•

Has different qualifiers for disk initialization from. those for tape initialization.

•

Requires WRTNFS privilege when used for disks.

The section "INITIALIZE Command for Tapes" later in this chapter describes
tape initialization. See the RSTS / E System User's Guide for nonprivileged uses
of the INITIALIZE command.

15.1.2

Mounting and Dismounting Tapes and Disks
Whenever you want to use a different tape or disk from that which is mounted,
you must logically dismount the current disk or tape (using the DISMOUNT
command), and physically remove the disk or tape from the drive. Then you
physically mount the new disk or tape on the drive and place the drive on line.
Mter physically mounting and activating the medium:. you must logically mount
the new disk or tape (using the MOUNT command). This allocates the drive for
appropriate use. For example, a tape drive is allocated for use only through your
account. A disk drive with a pack initialized and mounted as public becomes part
of the public structure. It can be used by anybody on the system.
The MOUNT command prepares an initialized disk or tape for processing by
system commands or user programs. The DISMOUNT command releases a disk
or tape that was previously accessed with a MOUNT command. Both MOUNT
and DISMOUNT have different qualifiers when used with disks or tapes.

15-2

Managing Disks and Tapes

15;.2 Del Command Descriptions
The following sections describe the DCL commands for managing RSTSIE disks
and tapes. See Chapter 11 for descriptions of the DCL commands for managing a
virtual disk.

15,,2.1

DISMOUNT Command for Disks

The DISMOUNT command for disks releases a disk previously accessed with a
MOUNT command. You issue this command before you take the drive off line
or before you physically dismount the disk. You can dismount any disk with
MOUN"T privilege. Without MOUNT privilege, you can only dismount a disk
mounted INOSHARE for your job.
The DISMOUNT command deallocates the disk if it was allocated to you. You
cannot DISMOUNT a device if there are open files on it. If you try, the system
displays the error message:
?Account or device in use

Format
DISMOUNT

disk-device-name[:] [pack-id]

Command Qualifiers
/PUBUC

Defaults
none

Prompts
Device: device-name[:]

Command Parameters
d isk-device-name[:]
The physical or logical name of the disk unit containing the pack to be dismounted. If you do not specify a unit number, the following error message
appears:
?Unit number needed

[pack-id]
The pack-id specified when the disk was initialized and mounted. The pack-id is
deassigned as a system-wide logical. The pack-id is optional.

Command Qualifiers
IPUBUC
You need MOUNT privilege to dismount a public disk. (The disk must have been
mounted and initialized as public.)

Managing Disks and Tapes

15-3

If you do not specify /PUBLIC, the systenl assumes that the disk was mounted as
private. If you try to dismount a public disk without lthe /PUBLIC qualifier, you
get the error message:
?Disk is mounted public

If you try to dismount a disk by using the /PUBLIC qualifier and the disk was

mounted as private, you get the error message:
?Disk is mounted private

If you do not have MOUNT privilege you can only dismount private disks if they
were mounted INOSHARE. If you try to dismount a public disk, you get the error
message:
?You must have MOUNT privilege to dismount a public disk

If the disk was mounted with the INOSHARE qualifier, and any unprivileged
job except the one to which the disk is assigned attelnpts to dismount it, the
following error message appears:
?Account or device in use

15-4

Managing Disks and Tapes

15;.2.2

DISMOUNT Command for Tapes
The DISMOUNT command for tapes releases a tape previously accessed with a
MOUNT command. You issue this command before you take the drive off line or
before you physically dismount the tape. The DISMOUNT command deallocates
the tape drive if it was allocated to you. You cannot DISMOUNT a tape if there
are open files on it. Any attempt to do so returns the error message:
?Account or device in use

When used for tapes, the DISMOUNT command does not require any privileges.

F()rmat
DISMOUNT

tape-device-name[:] [label]

Command Qualifiers
/[NO]UNLOAD

Defaults
/UNLOAD

Prompts
Device: device-name[:]

Command Parameters
tape-devlce-name[:]

Specifies the physical or logical name of the drive on which the tape is to be
dismounted. If you do not specify a unit number with a physical device name, thE
default is unit O.
label
Specifies the label on an ANSI tape. Note that it is not necessary to specify a
label with the DISMOUNT command. However, if you are using a hard-copy
terminal, you can specify a label to keep a log of the tape being dismounted. The
label you specify is ignored; it is not checked against the label on the tape.

Command Qualifiers
I[NO]lJNLOAD
Specifies whether to unload the tape from the drive. Once a tape is unloaded,
it must be manually loaded before it can be mounted again. The default is

!UNLOAD.

Managing Disks and Tapes

15-5

15.2.3

INITIALIZE Command for Disks

The INITIALIZE command creates a RSTSIE file structure on a disk; that is,
it prepares a disk so files can be written to it and read from it. (Be aware that
INITIALIZE overwrites any data previously written to the disk unless you specify
the INOEXERCISEINOERASE qualifiers.)
In creating a RSTSIE file structure for the disk, INIT][ALIZE writes several
structures to the disk. For example, INITIALIZE creates a User File Directory
(UFD) for account [0,1] that contains entries for the files SATT.SYS (the storage
allocation table) and BADB.SYS (the bad block file).
NOTE

When you issue the INITIALIZE conlmand, the disk should only be
physically mounted in the drive; it should not be logically mounted.
You must create DVO: first.
When used for disks, the INITIALIZE cODlmand requires WRTNFS, RDNFS, and
WREAD privileges.

Format
INITIALIZE

disk-device-name[:] pack-id

Command Qualifiers
ICLUSTER_SIZE=n
IDATE=ACCESSED
IDATE=MODIFIED
I[NO]ERASE
I[NO]EXERCISE=n
IINDEX=position
IMFD_ CLUSTER_SIZE=n
INEW_FILES=[FIRST I LAST]
IPRIVATE
IPUBLIC
I[NO]QUERY
I[NO]RETAIN
I[NO]WRITE

Defaults
n=devico cluster size
IDATE=MODIFIED
IDATE=MODIFIED
IERASE
IEXERCISE=FULL
IINDEX:=MIDDLE
n=16
IN EW_FILES=LAST
IPRIVATE
IPRIVATE
IQUERY
IRETAII'I
/WRITE

Prompts
Device:
Pack-id:

Command Parameters
d Isk-devlce-name[:]
Specifies the name of the drive on which the disk is physically mounted.

15-6

Managing Disks and Tapes

packEld

One- to six-alphanumeric characters to be used when logically mounting the
disk. By default, RSTS/E uses this pack-id as a systemwide logical name for the
mounted disk.

Command Qualifiers
ICLUSTER_SIZE=n

Declares the pack cluster size, which is the minimum allocation unit, in 512-byte
blocks, for the disk. All files written to this disk use multiples of n blocks, where
n is the pack cluster size.
In general, the cluster size specified (n) must be a power of 2 (1, 2, 4, 8, 16), equal
to or greater than the device cluster size, but not greater than 64. If you do not
specify a cluster size, RSTS/E assumes the device cluster size.
The /CLUSTER_SIZE=n option affects performance. In general, a large pack
cluster size speeds disk I/O operations, but wastes disk space. For example,
if you declare /CLUSTER_SIZE=16 for a disk, a file consisting of 8193 bytes
(16 512-byte blocks plus one byte) requires 2 pack clusters, or 32 blocks. A
/CLUSTER_SIZE=8 for the same file would require three clusters, or 24 blocks, 01
disk storage.
IOATE=ACCESSEO
IOATE=MODIFIEO

Lets you specify whether to record a file's date of last access (=ACCESSED) or of
last modification (=MODIFIED). The default is IDATE=MODIFIED.
I[NO]ERASE

Indicates whether the data on the specified disk is erased. The INOERASE
qualifier applies only if you specify INOEXERCISE. The default is /ERASE.
NOTE

Using INOERASE may cause sensitive data to remain on the disk. If
the disk contained files marked as "erase on delete" (protection code of
128 or higher), you should not use INOERASE unless you are sure that
no security problems will result.

Managing Disks and Tapes

15-7

Table 15-2 lists the device cluster sizes for the disks that RSTSIE supports.
Table 15-2:

Disk Size and Cluster Size

Disk
Type

Minimum
Device Cluster
Size

Acceptable Pack
Cluster Size
(/CLUSTE:R_SIZE)

Total Device
Size 1 (Blocks)

RX50
RX33

1
1

1,2,4,8,16,32,64
1,2,4,8,16,32,64

800
2400

RK05
RK05F

1,2,4,8,16,32,64

4800

1,2,4,8,16,32,64

4800 2

RL01
RL02
RD51

1
1

1,2,4,8,16,32,64

10220
20460

RD52

1,2,4,8,16,32,64
1,2,4,8,16,32,64

21600

1,2,4,8,1.6,32,64

60479

RC25
RK06
RK07

1
1
1

1,2,4,8,1.6,32,64
1,2,4,8,16,32,64
1,2,4,8,16,32,64

RD31

1,2,4,8,16,32,64

50902 2
27104
53768
41559

RP02
RP03
RD32

2
2
2

2,4,8,16,32,64
2,4,8,16,32,64
2,4,8,16,32,64

40000
80000
83204

RD53

4,8,16,32,64

138668

RM02
RM03
RP04
RP05

4
4
4

4,8,16,32,64

131648

4,8,16,32,64
4,8,16,32,64

131648
171796

RA80

4
4

4,8,16,32,64
4,8,16,32,64

171796
237208

RM80
RD54

4
8

4,8,16,32,64
8, 16, 32, 64

242575
311200

RP06
RA60

8,16,32,64
8,16,32,64

340664

RM05

8
8
8

RA81

8,16,32,64
16,32,64

RA82

32,64

400175
500352
888012
1216640

lSize when disk is initialized at lowest acceptable cluster size
2For each unit; 2 units per drive

IEXERCISE=n
INOEXERCISE

Controls whether to check for bad blocks by "exercising" the disk. This means
that the nwnber of patterns you select (n) are written to the disk, and then
each block is read to make sure that the patterns were written correctly. If the
patterns read do not match the patterns written, RSTSIE asswnes the block is
bad. RSTSIE adds any bad blocks found to the "bad block" file (BADB .SYS in
account [0,1] on the disk). RSTSIE does not store data in any blocks allocated to
the BADB.SYS file.

15-8

Managing Disks and Tapes

You can have the INITIALIZE command run from zero to three patterns by
specifying 0, 1, 2, or 3 for n. Each pattern consists of three octal words; the
pattern used for each of these numbers is:
Pattern

Three Octal Words

155555 133333 066666

133333 066666 155555

066666 155555 133333

For example, IEXERCISE=2 causes pattern 1 to be written to and read from the
disk, followed by pattern 2 being written to and read from the disk.
You can also select the default, IEXERCISE=FULL, which uses all three patterns:
one at a time. That is, each block is written to and read from the disk three
times, each time with a different pattern. Using the FULL argwnent increases
the probability that all bad blocks are found, and decreases the probability
that you will lose information later by writing into a bad block. (Note that
IEXERCISE=FULL is equivalent to IEXERCISE=3.)
The INOEXERCISE qualifier indicates that you do not want to check for bad
blocks. For example, if you previously initialized the disk and feel that the current bad block file is accurate, you can save time by specifying the INOEXERCISE
qualifier. This still writes one pattern all over the disk unless you specify
INOERASE, but it does not read it back in. (Note that /NOEXERCISE is
equivalent to IEXERCISE=O.
IINDEX=position

Positions the SATr.SYS file on the disk. SATT.SYS is the storage allocation table
created in account [0,1] during initialization.
The position argument can be::
BEGINNING

Puts SATT.SYS at the beginning of the disk to avoid fragmenting
the disk. That way, more contiguous space is available to store
other files on the disk. This is the default position for SATT.SYS fo]
disks of 4500 blocks or less if you do not specify the !INDEX option.

MIDDLE

Puts SATT.SYS in the middle of the disk. For moving-head disks, i1
helps to locate SATT.SYS near the middle of the disk, which reduce
average seek times for the disk heads. MIDDLE is the default for
disks of more than 4500 blocks if you do not specify the !INDEX
option.

Puts SATI'.SYS at device cluster number n, where n can range fron
1 to the total device size divided by the device cluster size. See
Table 15-2 for these values.

IMFD_CLUSTER_SIZE=n

Declares the minimwn allocation unit, in 512-byte blocks, for the Master File
Directory (MFD). The MFD is a special "catalog" structure that RSTSIE creates
during disk initialization. The INITIALIZE command puts the MFD at the
beginning for disks of 4500 blocks or less, or at the middle for disks of more thar
4500 blocks.
Along with other key structures, RSTSIE updates the MFD each time users add
or delete accounts and files on the disk. Because RSTSIE must access the MFD
frequently, you increase performance when the MFD cluster size is large. On thE
other hand, a small MFD cluster size can save some disk space.

Managing Disks and Tapes

15-

The MFD cluster size (n) can be 4, 8, or 16; it must be greater than or equal to
the value entered for the CLUSTER_SIZI~ qualifier, unless the CLUSTER_SIZE
is 32 or 64; in that case, the MFD cluster size remainl3 16. The default is IMFD_
CLUSTER_SIZE=16.
INEW_FILES

Indicates whether files stored on the disk should be JPlaced at beginning of the account (INEW_FILES=FIRST) or at the end of the account (INEW_FILES=LAST).
The default is INEW_FILES=LAST.
IPRIVATE

Allows only users who have accounts on a disk to access it. If you do not specify
/PUBLIC or /PRIVATE, then RSTSIE assumes /PRIV.NTE.
You can create a system disk by initializing the disk as /PRIVATE and then using
the CREATE/ACCOUNT command to create and position accounts [1,1] and [1,2].
In this case, note that you must transfer files to the dlisk and use the HOOK
utility to make it a RSTSIE system disk.
/PUBLIC

Allows anyone with an account on the system to access the disk.
You can use /PUBLIC to identify any disk on the system as public, except the
system disk. When the disk is mounted, it is considered part of the public
structure. You need MOUNT privilege to mount a public disk. Mter you mount
it, the disk is a logical extension of the system disk; thus, any user can create
files on it.
In general, DIGITAL recommends against using /PUBLIC, because it can degrade
performance over time. Consider the following points if you decide to select
/PUBLIC for disks on your system:
•

If your system disk is nearly full, and you want to dedicate one or more
additional disks to general file storage, you might add public disks to your
system. Users are not generally aware of whether their files are on the
system disk or a public disk.

•

A public disk is an extension of the system disk; therefore, it should always
remain mounted during timesharing. In addition, you must remount it every
time you restart the system. If you dismount the disk during timesharing,
some users' files will mysteriously (to them) disappear. Dismounting a public
disk causes disruption unless you transfer all of the files on the disk to the
system disk or to some other public disk.

•

Before adding a public disk to your system, considler the alternative of
adding a private disk for use by specific users. DIGITAL recommends
this alternative because it yields significantly better system performance.
However, this means that you must create account.s explicitly on the private
disk. In addition, users who have accounts on the disk must refer explicitly to
that disk in file specifications.

Because of these considerations, the system provides several safety checks on
initializing and mounting a public disk. First, you must have MOUNT privilege
to do so. Second, you must specify /PUBLIC two times: when you initialize the
disk, and when you mount the disk. Third, you must specify when you want
a file placed on DVO:, the virtual disk; the system never puts files on DVO:
automatically.

15-10

Managing Disks and Tapes

IQUERY
INOQUERY

When you use the IQUERY qualifier, INITIALIZE displays the characteristics of
a disk you have selected that already has a RSTS!E file structure, and asks if y01
still want to initialize the disk.
Thus t IQUERY lets you verify that you are initializing the disk you intended,
thereby keeping you from accidentally wiping out the wrong disk. If you specify
INOQUERY, INITIALIZE simply displays the characteristics of the disk before
overwriting its contents. The default is IQUERY.
The display of disk characteristics is similar to:
This disk pack appears to be a RSTS/E formatted
disk with the following characteristics:
Pack ID :
Pack Cluster Size
Pack is currently

THELMA
4

Private,
Level 1.2

Proceed (Y or N)?

You must respond to the prompt with YES or NO, or some abbreviation of these.
NO cancels the initialization; YES requests that it proceeds. There is no default
response to the prompt.
IRETAIN
INORETAIN

Controls whether to use a previously initialized disk's bad block file. A disk that
has not been previously initialized has no bad block file. In this case, !RETAIN
and INORETAIN are meaningless, and are ignored if specified. !RETAIN is the
default on a previously initialized disk.
If you use both the !RETAIN and !EXERCISE qualifiers, the INITIALIZE

command keeps the old bad block file and adds any additional bad blocks it findl
while it is checking the disk for bad blocks.
If you specify INOEXERCISE, then you should use !RETAIN. Because
INOEXERCISE does not exercise the disk for bad blocks, you should use !RETAIJ
to keep the existing bad block information.
The INORETAIN qualifier is useful if problems with a disk drive have been
causing apparent bad blocks on a disk. Mter you learn that the problem is with
the drive, you can recover space on the disk by using INORETAIN.
IWRITE
INOWRITE

Detennines whether the disk should default to read-only (INOWRITE) or
read/write (!WRITE) access when it is mounted. This is useful when you initialh
a disk that is usually mounted read-only. The default is !WRITE.
When mounting a disk, you can override the read-only access by specifying the
!WRITE qualifier with the MOUNT command.

Managing Disks and Tapes 15-1

15.2.4 INITIALIZE Command for Tapes
The INITIALIZE command for tapes prepares a new tape or recycles a used tape
that contains no useful files. The INITIALIZE command for tapes makes any
existing data on the tape unavailable to file-structured access. For ANSI tapes, it
also writes a label that RSTSIE uses to identify the tape in a MOUNT command
(ANSI tapes require a label). INITIALIZE allocates the tape drive if it is not
already allocated.
When used for tapes, the INITIALIZE command does not require any special
privilege unless the device is restricted. If so, DEVICE privilege is required.

Format
INITIALIZE

tape-device-name[:] [label]

Command Qualifiers
IDENSITY=nnn
IFORMAT=ANSI
IFORMAT=DOS

Defaults
See Discussion
See Discussion
See Discussion

Prompts
Device: magtape[:]
Label: label (if ANSI)
Proceed (Y or N)?

Command Parameters
tape-devlce-name[ :]

Specifies the name of the drive on which the tape is physically mounted.
[label]

Specifies the identification label to be encoded on the tape. The label can consist
of a maximum of six alphanumeric characters.
An ANSI-format tape requires a label; RSTSIE prompts you for a label if you do
not specify one. RSTSIE checks this label against the label you specify when you
later use the MOUNT command (see the section "MOUNT Command for Tapes").

DOS tapes do not allow labels: RSTSIE ignores any label you specify for them
and displays a warning message.

Command Qualifiers
IDENSITY=nnn

Specifies the density in bits per inch (bpi). The argwnent nnn can be any number
between 2 and 32766, or the keywords MINIMUM or MAXIMUM may be used.
Generally, available tape drives support 800, 1600, or H250 bpi.
If you specify MINIMUM or MAXIMUM, the system checks the tape drive and
sets the density to the minimum or maxilnum the drive supports. If you specify a
density that the tape drive does not support, RSTSIE displays an error message
and aborts the command. If you do not specify a density, the default is the system
default density.
15-12

Managing Disks and Tapes

IFORMAT=ANSI
IFORMAT=OOS

Specifies the tape fonnat. If you do not specify a fonnat, the system uses the
current system default (SHOW SYSTEM command displays the current system
defaults). You can change the default format using the DCL command, SET
SYSTJtjMILABEL= (see Chapter 14). As mentioned earlier, ANSI fonnat requires
a label; DOS fonnat ignores any label specified.

Managing Disks and Tapes

15-1:

15.2.5

MOUNT Command for Disks
The MOUNT command prepares an initialized disk for processing under timesharing. 1b mount a disk, you must:
1.

Physically mount the disk on a device drive

Put the drive on line

3. Use the MOUNT command to logically mount the disk
You need the MOUNT privilege to mount a disk that has been initialized as
public. For users with MOUNT privilege, the MOUNT command rebuilds a
"dirty" disk - one that has been physically dismounted (removed from the drive)
or taken off line without being logically dismounted fi'rst with the DISMOUNT
command. This is the only time that a disk's "dirty" bit is set, although you may
choose to rebuild the disk at other times to ensure integrity of its structure.
If MOUNT discovers the disk needs rebuilding, it displays the message:
Disk is being rebuilt - wait ...

The rebuild operation proceeds and the disk is logically mounted when the dollar
prompt ($) for the next DeL command appears. However, if the rebuild operation
discovers blocks that have been allocated to more than one file (doubly allocated
blocks), MOUNT displays a message asking which filE~ to allocate each doubly
allocated block to. You can choose a file at that point or stop and mount the disk
without rebuilding, which lets you examine the situation further.

Format
MOUNT

disk-device-name[.1 pack-id [Iogical-name[:]]

Command Qualifiers
IOVERRIDE[=identification]
IPRIVATE
IPUBLIC
I[NO]QUOTA
I[NO]REBUILD
I[NO]RESTRICT
I[NO]SHARE[=n]
I[NO]WRITE

Defaults
Inone
IPRIVATE
IPRIVATE
See Discussion
See Discussion
INORESTRICT
See Discussion
See Discussion

Prompts
Device: device-name[:]
Pack-id: pack-id

Command Parameters
d Isk-devlce-name[:]
Specifies the physical or logical name of the drive containing the disk you want to
logically mount.

15-14 Managing Disks and Tapes

packald

Specifies the one to six character alphanwneric pack identification label written
to the disk during initialization. (See the INITIALIZE command description in
the section "INITIALIZE Command for Disks.") MOUNT then. verifies that the
pack-id written on the disk is the same as the label you specified in the MOUN1
This process helps make certain that you physically mounted the right disk.
You must include a pack-id to mount a disk, unless you have the MOUNT
privilege and use the /OVERRIDE qualifier. If you use this qualifier, do not ente
the pack-id parameter, but only the device-name parameter and the logical-name
parameter (if any).
[Ioglcal-name[:]]

Lets' a user with INSTAL privilege assign a system-wide logical name to the disl
drive specified by a device name. If you do not specify a logical name and you
have INSTAL privilege, the system uses the pack-id label as a system-wide logic~
name for the drive. You may want to use this logical name to keep the pack-id
from unauthorized users. Unauthorized users can see logical names assigned
to devices in SYSTAT. RSTSIE ignores the specification of a logical name by an
unauthorized user (without INSTAL privilege), but the mount succeeds.

COlTlmand Qualifiers
/OVERRIDE[=ldentlficatlon]

Eliminates the need to specify the pack-id when mounting the disk. The identifi·
cation argwnent is optional and does not change the meaning of the qualifier. Yo
must have MOUNT privilege to include this qualifier. If you include /OVERRIDI
then MOUNT assumes that the second parameter, if specified, is the logical nam
for the device instead of the pack ID.
By default, you must include the pack ID of the disk.
/PRIVATE
/PUBLIC

Declares that the disk you are mounting is accessible to all users on the system (
public disk), or only accessible to users who have accounts on the disk (a private
disk). The default is /PRIVATE.
The action taken for /PRIVATE and /PUBLIC depends on two things:
•

Whether you have MOUNT privilege when issuing the MOUNT command

•

Whether you initialized the disk as public or private

In general, the disk is safeguarded, even against a user with MOUNT privilege,
from being inadvertently made available for public use.
For Users with MOUNT prlvilege
If you have MOUNT privilege, you can MOUNT a disk initialized as private. Y01
can explicitly declare the usage with the /PRIVATE or /SHARE qualifier, but thh
is not required. You can also :mount the disk with the INOSHARE=n qualifier, t(
make the disk accessible only to job n (see the discussion for INOSHARE[ =n]).

Managing Disks and Tapes

15-1

Table 15-3 summarizes the possibilities of the /PUBLJ[C, /PRIVATE, and
I[NO]SHARE qualifiers. (If none of these qualifiers is EJpecified, RSTS/E assumes

!PRIVATE.)
Table 15-3:

The IPUBLlC, IPRIVATE, and I[NO]SHARE Qualifiers for the MOUNT Command
INITIALIZE for disk was:

Privileges

Public

Private

User does not have
MOUNT privilege

Any MOUNT returns an error.

Only a MOUNTINOSHARE succeeds.

User has MOUNT
privilege

MOUNT with /PUBLIC succeeds: the
disk is accessible to all users~
MOUNT with /PRIVATE or ISHARE
succeeds: the disk is accessible to only
those users with an account on the
disk.
MOUNT with INOSHARE[=n] succeeds; the disk is accessible only to job
n (the user's job, by default).
MOUNT with no qualifier succeeds as
private disk: the disk is accessible to
only those users with an account on
the disk: RSTSIE displays a warning
message.

MOUNTIPUBLIC returns an error.
MOUNT with no qualifier or with
/PRIVATE, ISHARE, or INOSHARE
succeeds.

IQUOTA
INOQUOTA

Controls whether quota checking is performed when the disk is mounted. This
qualifier applies only to Level 1.2 disks. 1he default iEl IQUOTA. DVO: is always
INOQUOTA.
IREBUILD
INOREBUILD

Controls the rebuilding of a disk, regardless of whether the disk was initialized
as read/write or read-only. A dirty disk is one that ha.s been physically dismounted (removed from the drive) without being logically dismounted first by a
DISMOUNT command.
A DISMOUNT checks to see that no files are currently open on a disk. If files
are open, the DISMOUNT does not succeed. If someone makes a mistake and
physically removes a disk pack from a disk drive while programs are still using
files on the disk, the usual "clean-up" operations tha.t the RSTSIE monitor
performs may not have occurred. For example, the monitor keeps the file
SATT.SYS (the storage allocation table for the disk) in memory and updates it
when files are added to or deleted from the disk. If S01neone physically dismounts
a disk before the monitor has written SA1'T.SYS back to disk, the data in the old
SATT.SYS will probably be incorrect.
Rebuilding a disk requires MOUNT privilege. This operation performs necessary
clean-up before the disk can be used again. Specifically, a rebuild operation:

15-16

•

Locates blocks that have been allocated to more than one file. The system
displays a message listing the files to which such blocks have been allocated,
and asks you to delete all but one of the files.

•

Deletes invalid directories.

•

Deletes all files that have the type .TM:P, are marked for deletion, or have no
accounting entry.

Managing Disks and Tapes

•

Builds a new storage allocation table (SATr.SYS), to show current file
allocations (after deletions accomplished by rebuilding).

•

Zeroes all blocks that were in the old storage allocation table (SATT.SYS), btl
not in the new one. This is done as a security precaution: in case these block
belonged to files with a protection code of <128> or higher (which are always
zeroed when they are deleted).

A MOUNT operation for disk automatically checks to see if the disk is dirty. If
so, and you have MOUNT privilege, a rebuild occurs automatically, regardless oi
whether you specify /REBUILD. The following message appears to indicate that
the disk is being rebuilt:
Disk is being rebuilt - wait ...

However, if a user without MOUNT privilege tries to mount a dirty disk, RSTS/1
displays the error message:
?Disk needs rebuilding but you do not have MOUNT privilege

The /REBUILD qualifier forces a rebuild on a disk. It lets the user with MOUN~
privilege rebuild a disk that the system has not identified as "dirty." For exampl4
some programs leave temporary files in the user's account, rather than deleting
them before the program exits. Such a disk is not flagged as dirty. If you want
to get rid of the temporary files on a disk, you can mount the disk using the
/REBUILD qualifier.
If you specify /REBUILD and the drive is write-protected, RSTSIE displays the
following error message:
?Can't rebuild disk because device is write-protected

If a user without MOUNT privilege specifies /REBUILD, RSTSIE displays the
following error message:
?You must have MOUNT privilege to rebuild the disk

The INOREBUILD qualifier overrides an automatic rebuild of a dirty disk. You
may not want to take the time to rebuild the dirty disk. Or, the disk may have
doubly allocated blocks, and you do not want to risk deleting them. In this case,
you <:an mount the disk with the INOREBUILD qualifier (thus granting only
read··only access) and try to correct the situation. For example, you could use
the COPY command to copy the files to new locations (where they do not contair
doubly-allocated blocks) and try to sort out the situation from there.
If you have MOUNT privilege, and specify INOREBUILD to mount a dirty disk,
the disk is mounted restricted and read-only, and two warning messages appear:
%Disk needs rebuilding
%Disk is restricted and mounted no-share, read-only

Thus, the disk is accessible only to the job that mounted it.
When a user without MOUNT privilege specifies INOREBUILD to mount a
"clean" (rebuilt) disk, the mount succeeds and the disk is uneffected.
I[NO]RESTRICT

Indicates whether the disk is restricted. If you specify the /RESTRICT qualifier
in the MOUNT command line, only users with DEVICE privilege can access the
specified disk. The default is INORESTRICT.

Managing Disks and Tapes

15-1

ISHARE
INOSHARE[ =n]
Controls whether to limit disk access to:
•

Job n (INOSHARE=n)

•

The job that mounted the disk (INOSHARE)

•

Any user with an account on the disk (lSHARE)

A disk mounted as shared is the same as a disk mounted as private. A user with
MOUNT privilege can mount a private or public disk as shared or nonshared.
If you do not specify /PRIVATE, /PUBLIC, or I[NO]SHARE, then the default
action is to mount the disk as private (shared).
The INOSHARE qualifier conflicts with either /PUBLIC or /PRIVATE. You can use
ISHARE with either /PUBLIC or /PRIVATE; its presence in the same command
line as the other two qualifiers has no effect.

IWRITE
INOWRITE
Controls whether data can be written to the mounted disk.
If you specify INOWRITE, the disk is write-protected, which means that no users

can write to the disk. The INOWRITE qualifier protects files on the disk; they
can only be read from, not written to.
If you specify !WRITE, users who access the disk can write to it.
When you do not specify !WRITE or INOWRITE, the default depends on whether
the disk was initialized as read/write or read-only, and whether the disk is clean
or dirty. For disks initialized as read-only, the default is to give read-only access
(INOWRITE) to the disk. For disks initialized as read/write, there are a few more
conditions to consider. If the disk is clean or can be rebuilt (INOREBUILD was
not specified), and the drive is not write-locked, !WRITE is the default. If the disk
is clean and the drive is write-locked, INOWRITE is the default, and the following
warning message appears:
%Device write protected

If the disk is dirty and INOREBUILD was specified, the disk is mounted readonly, noshare.

15-18

Managing Disks and Tapes

15.2.6

IViOUNT Command for Tapes

The MOUNT command for tapes verifies that the tape is properly loaded onto tb
unit and checks an ANSI tape for the label specified in the MOUNT command
(ANSI tapes must have labels). You issue a MOUNT command after you have
physically mounted a tape on a tape unit and placed it on line.
If the device is restricted, you need DEVICE privilege to use the MOUNT command.

MOUNT

tape-de vice-nam e[:][/abe/]

Command Qualifiers
IDENSITY=nnn
IFORMAT=argument
IOVERRIDE[=identification]
I[NO]WRITE

Defaults
See Discussion
See Discussion
none
See Discussion

Prompts
Device: device-name[:]
Label: label

Prompts
Command Parameters
tapendevlce-name[:]
Specifies the physical or logical name of the drive on which the tape is physicall:
mounted. If you do not specify a unit nwnber with a physical device name,
RSTS/E asswnes unit O.
[label]
This parameter is necessary only for ANSI tapes. If you try to mount an ANSI
tape without specifying a label, RSTS/E prompts you for a label.

A label is not required (and is ignored) when you mount a tape in DOS or foreig
fonnat. In this case, RSTS/E displays a warning message:
%Label ignored

The label specifies the one- to six-character alphanumeric label written to the
tape when it was initialized (see the INITIALIZE command description in the
section "INITIALIZE Command for Tapes"). MOUNT then verifies that the labe
on the tape is the same as the one you specified in the MOUNT command. This
process helps make certain that you have mounted the right tape.

Command Qualifiers
IDENSITY=nnn
Specifies the density in bits per inch (bpi). The argwnent nnn can be any nwnb.
between 2 and 32766, or the keywords MINIMUM or MAXIMUM may be used.
Generally, available tape drives support 800, 1600, or 6250 bpi.

Managing Disks and Tapes

15-1

If you specify MINIMUM or MAXIMUM, the system checks the tape drive and
sets the density to the minimum or maximum the drive supports. If you specify a
density that the tape drive does not suppo:rt, RSTS/E displays an error message
and aborts the command. If you do not specify a density, the default is the system
default density.
IFORMAT=ANSI
IFORMAT=OOS
IFORMAT=FORE~GN

Indicates whether the tape is in a standard format used by the RSTS/E operating
system. The default format is the current system default (see SHOW SYSTEM,
Chapter 14).
The RSTS / E System User's Guide describes ANSI and DOS formats. A tape is
FOREIGN if it is not in ANSI or DOS fonnat.
NOTE

If you mount a tape with IFORMAT=FOREIGN, the program you use
to read the tape must know what types of labels there are on the tape
or be able to process any labels on the tape.
10VERRIOE[=ldentlficatlon]

Lets you mount ANSI magnetic tapes without specifying the label name. Since
DOS format tapes have no label, the /OVERRIDE qualifier is ignored if the tape
format is DOS. The identification argument is optional and does not change the
meaning of the qualifier. You need no privJlleges to use the /OVERRIDE qualifier.
IWRITE
INOWRITE

RSTS/E checks to see whether the tape is read/write (the write ring is present) or
read-only (the write ring is not present).
If the write ring is not present and you specify !WRITE, RSTS/E displays the
following error message:
?Device is write protected

If the write ring is not present and you do not specify 'either qualifier, RSTS/E

displays the following warning:
%Device is write protected

15-20

Managing Disks and Tapes

Chapter 16

Backing Up and Restoring Files
Backing up disk files and accounts means copying the contents of the disk to
another volume or set of volumes. Backing up disks is a precautionary measure
to let you recover from the loss or destruction of valuable information.
Most sites establish a policy and a schedule for regularly backing up disk files.
This chapter provides the operating procedures for backing up both selected files
and entire disks.
It is just as desirable to back up information on private disks as it is to back up
public disks. However, responsibility for backing up the files on private disks
could be left to the individual owners of those files and disks.

There are two kinds of backups of disk files:
•

Incremental, or partial backups

•

Full, or all-inclusive, backups

In either case, the backup medium can be disk or magnetic tape. Incremental
backups save only those files that have been modified recently. Periodic full
backups are necessary to provide the basis for reconstruction of an entire disk.
As a rule, incremental backups are undertaken more frequently than full
backups. After consulting with users of the system, you decide how often to back
up files and volumes and how long to retain backup media.
Generally, you are responsible for setting up a schedule for backing up files and
accounts, and for maintaining this schedule. The following schedule for backing
up public disk volumes on magnetic tape affords adequate protection of data for
many installations:
•

Daily-An incremental backup retained for seven days. This schedule
requires seven daily tapes (or sets of tapes) that are rotated once a week.

•

Weekly-An incremental backup retained for four weeks. This schedule
requires four weekly sets of tapes that are rotated once every four weeks.

•

Monthly-An all-inclusive backup retained for a year. This schedule requires
twelve monthly sets of tapes that are rotated once a year.

Despite all precautions, there is always the risk of losing a file. Frequent backupi
and longer retention periods reduce this risk. You may also want to consider an
off-site storage facility for the monthly save sets.

Backing Up and Restoring Files

16-'

You can perform full backups to magnetie tape or to another disk. Each has
its advantages and disadvantages. The advantage of using magnetic tape for
backups is the much lower media and storage cost, which may permit you to
retain backups longer than keeping full backups on disk.
However, there are several advantages to keeping copies on disk, which in
some cases outweigh the higher cost. Disks exhibit better data reliability than
magnetic tapes. Furthermore, disks tend to degrade 118SS in storage.

16.1

Recovering Your System Disk (RECOVR)
You can use the BACKUP utility and the command file RECOVR.COM to
recreate your system disk in case it becomes corrupted. RECOVR.COM is located
in account [0,1].
BACKUP lets you backup your system disk once it is ready for normal timesharing; RECOVR.COM creates a bootable disk or tape with the minimum files
needed to start your system. Use them in the following procedure to create a
recovery medium and restore your systeln disk:
Creating the recovery medium:
1. Mter the system is installed and running, use the DCL BACKUP command
to create a full backup of all accounts and files on the system.
2.

Once the backup is finished, use the command file RECOVR.COM to create
the bootable disk or tape from which the RESTORE command can recreate
the system disk.

Perform system backups on a regular basis.

Recreating the system disk:
1.

Boot the recovery disk or tape you created with the RECOVR.COM file. See
Chapter 2 for a description of this procedure.

Begin installation procedures. Note that this will reinitialize the system disk.

3. Mter you answer "yes" to the "Start timesharing?" prompt, the system
displays a message stating it is ready for you to begin using the RESTORE
command.
4. At the DCL prompt, physically mount your latest full backup set. Enter the
RESTORE/ACCOUNTING command.
5. Next, mount the latest incremental backup (if you. do incremental backups)
and enter the RESTORE/ACCOUNTINGIREPLACE command.
6.

When the restore completes, run the SHUTUP program and restart the
system using your normal monitor (restored from the backup set).

The system disk is now back to the point of the last backup.

16-2

Backing Up and Restoring Files

16.2 DCl Command Descriptions
You can use two DCL commands: BACKUP and RESTORE. Files created by
BACKUP are called Backup sets. A Backup set is a RSTSIE file that can be
copied just like any other RSTSfE file. However, only BACKUP and RESTORE
can interpret the data in a Backup set. The rest of this chapter describes
BACKUP and RESTORE in detail.
NOTE

Because the BACKUP and RESTORE operations can use significant
amounts of system resources, you need the TUNE privilege to use
these commands. Additional privileges may be required to use certain
qualifiers (for example, IACCOUNT_DATA requires GACNT or WACNT
privilege). If you want to delegate the responsibility for backing up
certain files (for example, files on private disks) to the individual
owners, they will need GREAD or WREAD privileges.

116.2.1

BACKUP Command

The BACKUP command creates a Backup set on a specified disk or magnetic tap
from one or more RSTSfE disk files. Once the BACKUP utility determines the
syntactical validity of the command specification and issues any warning messag
relative to either the command or the environment, BACKUP prompts with:
Please mount volume 1 of Backup set backup-set-name
Where can this volume be located <default>?

At the prompt, enter the device name to write the Backup set to. The default is
the output device you specified in the BACKUP command line. You can use the
/INITIALIZE and INOINITIALIZE qualifiers to control initialization of the outpu
rlevice.

Format
BACKUP

input-file-spec output-backup-set-spec

Command Qualifiers
I[NO]ACCOUNT_DATA
IBLOCK_SIZE=value
IBRIEF
IBUFFER_SIZE=n
ICREATED=(dateltime or range)
IMODIFIED=(date or range)
I[NO]CRC
IDENSITY=n
IDIRECTORY
IEND=[NO]REWIND
[,[NO]DISMOUNT]
IEXCLUDE[=(filespec list)]
/[NO]FILE_DATA
IFULL
/[NO]GROUP_SIZE[=n]
IINCLUDE=(filespec list)

Defaults
IACCOUNT_DATA
IBLOCK_SIZE=2048
none
See Discussion
ALL if neither ICREATED
or IMODIFIED is present
INOCRC
Highest available
none
IEND=DISMOUNT
None excluded
IFILE_DATA
none
IGROUP_SIZE=1 0
None included
Backing Up and Restoring Files

16-~

IIN_LABEL=name
I[NO]IN ITIAUZE
I[NO]UST_FI LE[=filespec]
lOUT _LABEL=name
I[NO]OUTPUT[=filespec]
I[NO]PROMPT
I[NO]QUERY
I[NO]REWIND
ISELECT=(filespec list)
I[NO]VERIFY

Volume ID is not checked
See Discussion
lUST_FILE=KB:
See Discussion
IOUTPUT=KB:
IPROMPT
INOQUERY
INOREWIND
See Discussion
INOVERIFY

Prompts
See Discussion

Restrictions
Backup sets produced by the BACKUP utility are not bootable. Reboot using a
recovery medium (for example, a disk or tape produced. by RECOVR.COM or the
latest installation kit) then use the RESTORE command to restore the backup
set.

Command Parameters
input-file-spec
Identifies the files to be backed up. The file specification should be a standard
RSTSIE file specification or a list of specifications separated by the plus (+) sign
or a comma (,):
dev:[p,pn]filename.typ,. ..
You must specify the device. The PPN is optional and defaults to the user's
current account. The file specification is optional and defaults to *. *. You can use
standard wildcards. If you specify a filename but not t.he type, BACKUP assumes
a * type. If the filename ends with a period, BACKUP assumes a null type. If
you specify a file type, you must also specify a filenam,e.
output-backup-set-spec
Identifies the device where the Backup set is to be created. The Backup set
specification should be a standard RSTSIE~ device specification of the format:
dev:[PPN]filename.typ
You must specify the device. PPN is optional and defaults to the user's current
account. The file specification is optional and defaults to BACKUP.BCK. If the
backup operation needs more than one volume in the Backup set, BACKUP
prompts for subsequent volumes as follows:
Please dismount volume 1 of Backup s'at backup-set-name
P lease mount volume 2 of Backup set .backup-set -'name
Where can this volume be located <default>?

16-4

Backing Up and Restoring Files

Command Qualifiers
I[NO]ACCOUNT_DATA

If you specify /ACCOVNT_DATA and you have the necessary privileges (GACNT
or WACNT), BACKUP preserves all pertinent information about the account
(quotas, password, privileges, and so on). RSTS/E uses this data to recreate or
update accounts during a RESTORE.
If you do not have accounting privileges (GACNT or WACNT), BACKUP only
saves the attributes of accounts that you have access to. In particular, you can
not access the password attribute or restricted user attributes.
If you specify fNOACCOUNT_DATA, RESTORE cannot create accounts that do
not already exist on the output volwne. The default is /ACCOVNT_DATA.
IBLOCK_SIZE=value

The [BLOCK_SIZE qualifier lets you select the block size used for the output
Backup set. You must specify the value argwnent. Value can be in the range
2048 to 4080 bytes for tape, and 2048 to 7680 bytes for disk. Further, the value
must be a multiple of 16 for tape and 512 for disk.
If you do not specify /BLOCK_SIZE, BACKUP uses 2048 bytes.
NOTE

Use of this qualifier can affect the performance and reliability of
JBACKUP. If you use large values, BACKUP operates faster and can
store more data into a volwne. Smaller values cause less data to fit
into a given volwne; however, they can increase the reliability (see the
description of the /GROUP_SIZE qualifier).
IBRIEF

Use this qualifier with /DIRECTORY to specify one file per line.
IBUFFER_SIZE=value

The /BUFFER_SIZE=value qualifier controls the size of the dynamic region used
by BACKUP. The valid range is 3 to 127. The value represents the nwnber of
K-words of memory to be used by BACKUP as buffer space. In general, the large'
you specify n, the faster BACKUP performs. However, large nwnbers increase
the system loading and resource usage, and can seriously impact other users.
If /BUFFER_SIZE=value is omitted, BACKUP uses the minimwn buffer space
consistent with the other qualifiers, or a value of 55 for MU: and MS: devices. If
the alnount of memory requested is not available, BACKUP uses as much memo~
as possible and displays an informational message. /BUFFER_SIZE=MAXIMUM
uses as much available memory as possible without displaying any message.
ICREATED=(BEFORE=date-time, AFTER=date-tlme)
ICREATED=date-tlme
IMODIFIED=(BEFORE=date, AFTER=date)
IMODIFIED=date

The /CREATED and /MODIFIED qualifiers let you select the files to be backed
up, as chosen by the file specification criteria (see the input-file-spec description),
based on the creation date/time and/or revision date. If no file specification
criteria exist, these qualifiers apply to all files on the input volwne for this
operation. You can use /CREATED and /MODIFIED in conjunction with each
other; they are not mutually exclusive. To be selected, a file has to satisfy all the
criteria you specify.

Backing Up and Restoring Files

16-E

You can specify an exact date or range of dates for either qualifier. In the range
format, each qualifier must take an argument of one or both of the subqualifiers
BEFORE or AFTER. The subqualifier order in the argument is immaterial.
The date-time argument is of the standard form (for example, dd-mmm-yy or
yy.mm.dd and hh:mm AlPM or hh:mm). You must specJlfy the date. If you do not
specify the time, it defaults to 00:00. In addition, if you specify the time, you
must be separate it from the date by a colon (:).
The following generic terms can be used in the subqualifier:
•

TODAY-meaning 00:00 on the current (system) date

•

YESTERDAY-meaning 00:00 on the date previous to the current date

When used with ICREATED, the generic terms also acc:ept a time argument.
If you do not specify either ICREATED or !MODIFIED, BACKUP selects all
specified files for transfer to the Backup set.
NOTE

Files transferred by BACKUP retain their original creation and
revision dates (that is, the dates on the source volume). You need the
DATES privilege to keep these original dates; otherwise, RSTSIE lists
the current date as the creation date of the backed-up file.
I[NO]CRC

Controls the generation of 32-bit cyclic redundancy check (CRC) codes to better
detect data errors. Every block header includes a 16··bit CRC, but this only
detects errors in the header. The 32-bit CRC produced by the ICRC qualifier detects errors anywhere in the block. Generating CRC code blocks uses significant
amounts of CPU time, and may degrade streaming performance on some systems.
When restoring a backup set written with ICRC, RESTORE automatically checks
the CRC. If RESTORE finds a CRC error, it attempts to recover the data. The
default is !NOCRC.
IDENSITY=n

Specifies the recording density on the output tape. The argument n can be
any number between 2 and 32766, or the keywords MINIMUM or MAXIMUM.
Generally, available tape drives support 800, 1600, or H250 bpi.
If you specify MINIMUM or MAXIMUM, BACKUP checks the tape drive and
sets the density to the minimum and maximum the drive supports. If you select
a density that is not valid for the drive addressed, BACKUP displays a warning
message and selects the highest density (below the specified value) that the drive
supports. If you do not specify a density, the default value is the highest density
available on the drive addressed.
You can only specify this qualifier for the first VOlumE! of a Backup set.
Subsequent volumes are written at the same density as the first volume unless the addressed drive is different from the first drive and the original density
is not available on the addressed drive. In this case, the density is the highest
density available for the addressed drive.
IDIRECTORY

Tells BACKUP to list the files within a BACKUP saveset. When you use the
/DIRECTORY qualifier, you can also use the qualifiers that go with the separate
DIRECTORY command, such as ISELEC'l~ IEXCLUD)~, !FULL, and !BRIEF.
!BRIEF is the default.

16-6

Backing Up and Restoring Files

IEND={REWIND I NOREWIND I DISMOUNT I NODISMOUNT}
!REWIND rewinds the tape before searching for the logical end of tape to create
the backup file.

BACKUP must locate the logical end of tape so that it can create the backup file
after any files already on the tape. While searching for the end of tape, BACKUI
will check the name of each file it encounters and report an error if it finds a file
with the same name as the backup file. This prevents having two files with the
same name on the tape.
If you specify !REWIND, BACKUP rewinds the tape and scans the entire tape fm
a duplicate file. If you specify INOREWIND, then BACKUP scans from the tape'!:
current position, and will not detect duplicate files located earlier on the tape.
The default is INOREWIND.
IEXCLUDE=(filespec list)

Lets you specify files that you want excluded from the BACKUP operation (for
example, [0,1]SWAP?SYS). You can include a list of file specifications in the
command line up to the maximum length of a DCL command line.
I[NO]FILE_DATA

Controls the transfer of file data during a backup operation. If you specify
!FILE_DATA, the operation proceeds with actual file data being transferred to thE
Backup set from the RSTS/E volume. If you specify INOFILE_DATA, no files are
written to the Backup set. Use INOFILE_DATA with IACCOUNT_DATA to back
up only accounting data.
The default is !FILE_DATA.
IFULL

Use this qualifier with /DIRECTORY to specify an expanded directory of each file
in the saveset. Several lines of information displays for each file.
I[NO]GROUP_SIZE[=n]

Specifies the group size for XOR redundancy. The optional argument can be a
decimal number from 0 to 100. If you specify zero (equivalent to INOGROUP_
SIZE)~ BACKUP creates the Backup set without XOR redundancy. The default is
IGROUP_SIZE=10.
The IGROUP_SIZE qualifier causes additional error recovery data to be stored
in the Backup set. RSTS/E uses this information to recover from media errors
during a RESTORE operation.
The smaller the number, the greater the chance of recovering from data errors on
RESTORE. Note that a smaller number also makes the Backup set larger.
RESTORE can recover from one error in each group. So, for example, if you
specify IGROUP_SIZE=20, RESTORE can recover from one error in every 20
blocks. That is, with IGROUP_SIZE=20, RESTORE cannot recover from 2 or
more errors every 20 blocks. Note that the /BLOCK_SIZE qualifier sets the size
of each Backup set block. So, if RESTORE cannot recover 2 blocks of data, you
lose 2 blocks of that size. In general, if something cannot be recovered, you lose
less data with a smaller /BLOCILSIZE.
IINCLLJDE=(filespec list)

Lets you specify files or accounts that you want to be unconditionally processed. You can include a list of file specifications in the command line up to the
maximum length of a DCL command line.
Files in this set override those excluded with the /EXCLUDE qualifier.

Backing Up and Restoring Files

16-7

IIN_LABEL=name

Specifies the name to be compared to the Bource volume ID. If you use this qualifier, the source volume ID must equal the specified naTne before the BACKUP can
be perform the requested operation.
I[NO]INITIALIZE
Specifies the disposition of the output media prior to the data transfer operation
and sets the default device specification in the "Where can this volume be located
<default>?" prompt.

For disk volumes, /INITIALIZE specifies that any file structure on the disk is to
be disregarded and the volume is to be reinitialized with a RSTSIE file structure
as if it were done with the INITIALIZE command.
BACKUP also creates a nonuser account on the disk to hold the Backup set. That
account is the account explicitly specified in the output specification, or implicitly
specified as the current account of the user issuing the command. The pack
cluster size is 16, the UFD cluster size is 16, and the HIe cluster size for the save
set is set to the amount of free space on the disk divided by 100, rounded up to
the next highest power of two. BACKUP does not define any accounting data for
the created account.
For magnetic tape volumes, /INITIALIZE specifies that BACKUP initializes the
tape as a standard ANSI tape.
You can only specify INOINITIALIZE when doing a BACKUP to tape to create
a new Backup set to be appended to the end of a tape containing other files
(possibly other Backup sets).
I[NO]LlST_FILE[=filespec]

Controls the production of a directory-like listing file that catalogs the sequence of
events in the BACKUP operation. The file specification can be any valid RSTSIE
file specification (wildcards are not allowed). The default is your terminal.
You can use this qualifier to produce a listing of files in the Backup set.
lOUT_LABEL=name
Specifies the desired name for the output volume ID. The resultant name can
vary depending on the interaction between lOUT_LABEL and I[NO]lNITIALIZE.

Table 16-1 summarizes the possible outcomes of the interaction of I[NO]lNITIALIZE
and lOUT_LABEL during BACKUP.
Table 16-1:

IINITIALIZE and lOUT_LABEL Interaction During BACKUP

Qualifiers
Specified

Output Volume ID

lOUT_LABEL=name
/INITIALIZE

The specified name.

lOUT_LABEL=name
INOINITIALIZE

The specified name; however, the operation is only
successful if the output volume label has the same name.

/INITIALIZE

BACKUP supplies the name.

(only)

INOINITIALIZE
(only)

16-8

Backing Up and Restoring Files

The name is unchanged.

I[NO]OUTPUT[=filespec]

The same as the !LIST_FILE qualifier.
I[NO]PROMPT

Normally, the Backup utility issues a MOUNT prompt, for each BACKUP
operation even if the Backup set's device is already mounted. You can eliminate
the prompt for the first volume by including the INOPROMPT qualifier in your
command line. The default is !PROMPT.
I[NO]QUERY

Causes BACKUP to prompt the user for each file that meets the selection criterio
as to whether or not the file should be transferred. A y[ES] response effects the
transfer. A response of N[O] or pressing Return bypasses the file. Any other
response causes the prompt to be reissued. The default is INOQUERY.
I[NO]REWIND

For tape backup files, indicates whether to rewind the tape before searching for
the logical end of tape to create the backup file.
BACKUP must locate the logical end of tape so that it can create the backup file
after any files already on the tape. While searching for the end of tape, BACKUl
will check the name of each file it encounters and report an error if it finds a file
with the same name as the backup file. This prevents having two files with the
same name on the tape.
If you specify /REWIND, BACKUP rewinds the tape and scans-the entire tape fo:
a duplicate file.
If you specify INOREWIND, then BACKUP scans from the tape's current position
and will not detect duplicate files located earlier on the tape. The default is
INOREWIND.
ISELECT[=(filespec list)]

Specifies the files or accounts to restore from the backup file. Use this qualifier t4
restore a single file or a set of files from the backup file.
If you don't include this qualfier, then all files and accounts in the backup file arE
restored, unless you specify other selection qualifiers.
I[NO]VERIFY

The NERIFY qualifier specifies that a comparison operation is to be performed
on all data transferred. This post-process verification reports the presence of
any data differences. For any differences found, BACKUP displays the PPN, file
name~ and file type at the user's terminal. The default is INOVERIFY.

Backing Up and Restoring Files

16-9

Examples

Table 16-2 shows command formats for BACKUP operations and some of the
qualifiers associated with a backup operation.
Table 16-2:

16-10

BACKUP Quick Reference T,lble

Command Action

Command Example

Save a file to a save set on tape

BACKUP DRl:PAYROL.DAT MTO:PAYROL.BCK

Save a disk to a save set on tape

BACKUP/ACCOUNT_DATA DUl:[*,*]*.*MMO:1409MA.BCK

Save a list of files to a save set on
tape

BACKUP DR2:PAYROL.DAT,DR2:BENEFI.DAT,DR2:VACATI.DAT MrrO:DAT17M.BCK

Full backup of a disk

BAC KUPIBUFFER_SIZE= 75IINITIALIZEILIST_FILE=SYDISK [ic. *]*. * MUO:IEND=REWINDIEXCLUDE=[O, I]??????. SYS/NOPROMPTIVERIFY

Incremental backup of a disk
(following earlier full backup)

BACKUPIBUFFER_SIZE=75IINITIALIZE/CREATE=AFTER=4-F:EB-90:9:28- .
ILIST_FILE=SYDISK [*.*]*.* MUO:IEND=REWINDIEXCLUDE=[O,l]??????SYS/NOPROMPTIVERIFY

Backing Up and Restoring Files

16,,2.2

RESTORE Command

The RESTORE command restores a RSTS/E disk or subset thereof from a tape or
disk Backup set. Once RSTS/E determines the syntactic validity of the command
specifieation and issues any warning message relative to either the command or
the environment, RESTORE prompts with:
Please mount volume 1 of Backup set backup-set-name Where can this
volume be located <default>?

At the prompt, enter the device to use or press Return or Line Feed to accept the
default value. The default value is the device specified as the input parameter. If
you specify INOPROMPT, RESTORE omits the prompt message.
If you are doing a partial restore, and you know what volume to start with,
start with that volume (although you may start with any volume). RESTORE
issues a confirmation message and starts with the volume you specify. Note that
RESTORE may issue a warning message in this case.
NOTE

If you have sufficient privilege and you specified IACCOUNT_DATA in
the RESTORE command, and if the Backup set contains accounting
information, BACKUP creates accounts if they did not previously exist.
The account characteristics are the same as they were for the Backup
set accounts. If the account already exists, BACKUP replaces the
existing account characteristics.
Use RECOVR.COM or the latest installation kit to reboot the system
before you restore system disks.

Format
RESTORE

input-backup-set-spec output-file-spec

Command Qualifiers
I[NO]ACCOUNT_DATA
IBRIEF
IBUFFER_SIZE=n
ICREATED=(date/time or range)
IMODIFIED=(date or range)
IDIRECTORY
IEND=[NO]REWIND
[,[NO]DISMOUNT]
IEXCLUDE=(filespec list)
I[NO]FILE_DATA
IFULL
IINCLlJDE=(filespec list)
IIN_LABEL=name
I[NO]INITIALIZE
I[NO]LlST_FILE[=filespec]
IOCCURRENCE=number
I[NO]OPTIMIZE[=(list)]
lOUT _LABEL=name
I[NO]OUTPUT[=filespec]
IPLACED_POSITION[=keyword]

Defaults
IACCOUNT_DATA
none
See Discussion
ALL if neither ICREATED
or IMODIFIED is present
none
IEND=DISMOUNT
None excluded
IFILE_DATA
none
None included
Volume ID is not checked
See Discussion
INOLIST_FILE
IOCCURRENCE=1
See Discussion
See Discussion
INOOUTPUT
IPLACED_POSITION=INDEX
Backing Up and Restoring Files

16-11

I[NO]PROMPT
I[NO]QUERY
I[NO]REPLACE[=QU ERY]
I[NO]REWIND
ISELECT=(filespec list)
I[NO]VERIFY

IPROMPT
INOQUERY
IREPU\CE=QUERY
INOREWIND
See Discussion
INOVERIFY

Prompts
See Discussion

Command Parameters
Input-backup-set-spec
Identifies the device where the first volume of the Backup set is mounted. The
Backup set specification should be a standard RSTSIE device specification of the
format:
dev: [PPN]file-spec
You must specify the device. PPN is optional and defaults to the user's current
account. File-spec is optional and defaults to BACKUP.BCK. If the Backup set
has more than one volwne, RESTORE prompts you for subsequent volumes as
follows:
Please dismount volume 1 of Backup 13et backup-set-name
Please mount volume 2 of Backup set backup-set-name
Where can this volume be located <default>?

output-tile-spec
Identifies where to put the files or accounts to be restored. The file specification
should be a standard RSTSIE file specification of the format:
dev: [PPN]file-spec
You must specify the device. PPN is optional and defaults to the account you
logged in to. File-spec is optional and defaults to the file specification(s) recorded
on the Backup set.
If you specify a PPN, RESTORE transfers all selected files to that account, even
if the Backup set has multiple accounts, asswning you have sufficient privilege
to create files in that account (GWRITE or WWRITE). If file name conflicts
occur as a result of the RESTORE (for example, TEMP.BAS exists in two or
more accounts on the Backup set), the second and subsequent transfers require
operator intervention unless you specify the !REPLACE qualifier.
If you specify a wildcard PPN, RESTORE transfers an selected files to:
•

[*, *]-the same accounts as on the Backup set

•

[n, *]-forces the project number to n, uses the programmer nwnber as on the
Backup set

•

[* ,n]-forces the programmer nwnber to n, uses the project nwnber as on the
Backup set

If the accounts do not exist and you specify IACCOUNT_DATA, RESTORE creates
them with the same characteristics as they have on the Backup set, providing you
have the necessary privilege to create the accounts (GACNT or WACNT).

16-12

Backing Up and Restoring Files

Comlnand Qualifiers
I[NO]ACCOUNT_DATA

If you have sufficient privilege (GACNT or WACNT), this qualifier determines
what level of information from a UFD is restored from a Backup set.
Use /ACCOUNT_DATA to restore all accounting data that is recorded in the
Backup set both to existing and to newly created accounts. Use INOACCOUNT_
DATA to inhibit the creation of accounts and restoration of UFD data. The
default is /ACCOUNT_DATA.
NOTE

If an account already exists, RESTORE supersedes the accounting data
when restoring to the original account.
IBRIEF

Use this qualifier with /DIRECTORY to specify one file per line.
IBUFFIER_SIZE=value

The /BUFFER_SIZE=value qualifier controls the size of the dynamic region that
RESTORE uses. The valid range is 3 to 127. The value represents the number
of K-words of memory that RESTORE can use as buffer space. If you omit
/BUFFER_SIZE=value, RESTORE uses the minimum buffer space consistent
with the other qualifiers.
The larger value you specify, the faster RESTORE performs. However, note that
large values cause a greater impact to other users.
If the amount of memory you request is not available, RESTORE uses as much

memory as possible and displays an informational message. If you specify
/BUFFER_SIZE=MAXlMUM, RESTORE uses as much memory as possible;
however, it does not display any message.
ICREATED=(BEFORE=date-time, AFTER=date-tlme)
ICREATED=date-time
IMODIFIED=(BEFORE=date, AFTER=date)
IMODIFIED=date

The /CREATED and /MODIFIED qualifiers let you select the files to be restored,
as chosen by the file selection criteria (see the output-file-spec description), based
on the creation dates/times and/or revision dates. If no file selection criteria exist,
these qualifiers apply to all files on the input volume to which you have access for
this operation. You can use /CREATED and /MODIFIED in conjunction with each
other; they are not mutually exclusive. To be selected, a file has to satisfy all the
cri teria you specify.
You can specify an exact date or a range of dates for either qualifier. In the
range of dates format, each qualifier must take an argument of one or both of the
subqualifiers, BEFORE or AFTER. The order of the sub qualifiers in the argument
is immaterial. If you specify both BEFORE and AFTER dates for either one of
the qualifiers, the AFTER date must be earlier than the BEFORE date.
If you do not use either /CREATED or /MODIFIED, all specified files are selected
for transfer from the Backup set.
NOTE

BACKUP retains both the creation and revision dates as they were on
the source volume for the files transferred if you have DATES privilege.

Backing Up and Restoring Files

16-13

IDIRECTORY

Tells BACKUP to list the files within a BACKUP saV'eset. When you use the
/DIRECTORY qualifier, you can also use the qualifiers that go with the separate
DIRECTORY command, such as ISELECT, /EXCLUDE, /FULL, and !BRIEF.
!BRIEF is the default.
IEND={REWIND I NOREWIND I DISMOUNT I NODISMOUNT}

(For magnetic tapes only) REWIND causes the magnetic tape to be rewound to
the load point and dismounted following successful completion of the RESTORE
operation. NO REWIND causes the magnetic tape to remain positioned after the
most recent Backup set. The default is NOREWIND.
If you want the Backup set's device to remain mounted at the end of the

BACKUP operation, specify NODISMOUNT in the command line. The default is
DISMOUNT.
IEXCLUDE=(filespec list)

Lets you specify files that you want to be excluded from the RESTORE operation. You can include a list of file specifications in the command line up to the
maximum length of a DCL command line.
I[NO]FILE_DATA

Controls whether file data is transferred during a RI~STORE operation. If you
specify !FILE_DATA, RESTORE transfers the file data from the Backup set to
the RSTS/E volume. If you specify INOFILE_DATA, RESTORE proceeds without
restoring any of the specified files. The default is !FILE_DATA.
If you have accounting (GACNT, WACN1') privileges, you can use INOFILE_DATA

with the IACCOUNT_DATA qualifier to restore only account information from a
Backup set.
IFULL

Use this qualifier with /DIRECTORY to specify an expanded directory of each file
in the saveset. Several lines of information displays for each file.
IINCLUDE=(filespec list)

Lets you specify files that you want to be unconditionally processed. You can
include a list of file specifications in the command line up to the maximum length
of a DCL command line.
Files in this set override those specified with the /EXCLUDE qualifier.
IIN_LABEL=name

Specifies the name to be compared to the source volume ID. If you use this qualifier, the source volume ID must equal the specified name before the RESTORE
operation can be performed.
I[NO]INITIALIZE

Specifies the disposition of the output media prior to the data transfer operation.
The /INITIALIZE qualifier specifies that RESTORE should disregard any file
structure on the disk and reinitialize the volume with a RSTS/E file structure as
if it were done with the INITIALIZE command.
RESTORE takes the characteristics froIn the source device summary. (If the
device cluster size of the original volume is too smaH, the next higher valid
cluster size is used for the output volurn.e.)
You can only specify INOINITIALIZE when the RESTORE operation is from a
Backup set to an existing RSTS/E disk. In this case, RESTORE places the files
on the existing file structures instead of replacing them.
16-14

Backing Up and Restoring Files

I[N 0] LIST_ FI LE[=file-spec]

Generates a directory-like listing file, cataloging the sequence of events in the
RESTORE operation. If you specify !LIST_FILE, you must include the file
specification. It can be any legal RSTSIE file specification (wildcards are not
allowed). The default is !LIST_FILE=KB:.
10CCURRENCE=number

Indicates the particular occurrence of a named Backup set to use when more tha
one Backup set of the same name exists on a magnetic tape. You must specify th
argument which refers to the position rehltive to either the current tape positiol1
or the beginning of the tape depending on the use of I[NO]REWIND.
RSTSIE Backup sets are interchangeable with VAX Backup sets. However, VAX
Backup set names can be greater than six characters, while RSTSIE looks at
only the first six characters. Therefore, RSTSIE cannot distinguish between two
VAX names that have the same first six characters (for example, ABCDEF and
ABC]DEFG). To transfer ABCDEFG, you must use ABCDEF/OCCURRENCE=2.
The default is IOCCURRENCE=l; the first occurrence of the named Backup set.
I[NO]OPTIMIZE[=(lIst)]

The I[NO]OPrIMIZE=(list) qualifier controls RESTORE's optimization of the
directory structure. Two optimizations are available:
•

Directories can be extended to the size they were on the original volume

•

File cluster sizes can be increased to the maximum necessary to minimize
calls to the file processor

The INOOPTIMIZE qualifier inhibits these opt.imizations. The IOPTIMIZE
qualifier selects all optimizations. The default is IOPTIMIZE.
The /OPTIMIZE=(list) qualifier selects optimizations from the following:
•

CL USTERSIZE-Optimize cluster size

•

DIRECTORY-Optimize directory size

In general, you would want to perform these optimizations to increase overall
system performance. However, if there is not much contiguous disk space
available, you would not want to make all executable files contiguous. Similarly,
if there is little free disk space, you would not want to optimize cluster sizes
because this takes more disk space.

Backing Up and Restoring Files

16-1

lOUT LABEL=name

Specifies the desired name for the output volume IDo 'rhe resultant name can
vary depending on the interaction between lOUT_LABEL and I[NO]INITIALIZE.
Table 16-3 summarizes the possible outcomes from the interaction of
I[NO]INITIALIZE and lOUT_LABEL dllring RESTORE.
Table 16-3:

IINITIALIZE and lOUT_LABEL Interaction During RESTORE

Qualifiers Specified

Output Volume ID

lOUT_LABEL=name
/INITIALIZE

The specified name.

IOUT_LABEL=name
INOINITIALIZE

The specified name; however, the operation is only
successful if the output volume label has the same name.

/INITIALIZE (only)

The destination volume name is the same as the source
volume name.

INOINITIALIZE (only)

The name is unchanged.

I[NO]OUTPUT[=filespec]

The same as the !LIST_FILE qualifier.
IPLACED_POSITION [=keyword]

Controls the placement of files that were :marked as placed on the source disk.
The /PLACED_POSITION qualifier takes one of the following keyword arguments:
•

INDEX-Place close to the storage allocation table and master file directory

•

MIDDLE-Place close to the middle of the disk

•

ORIGINAL-Place close to the original position on the source disk.

The default is INDEX
The rules for contiguity apply first and foremost; that is, RESTORE sacrifices
placement in the interest of contiguity. The first sufficient contiguous space
closest to the selected position is the first choice for positioning the file on the
output volume.
If no contiguous space of sufficient size exists on the output volume, then the first

cluster(s) of the file are positioned as close to the original address as possible and
the remaining clusters are positioned as required.
I[NO]PROMPT

Normally, the BACKUP utility issues a MOUNT prompt for each RESTORE
operation even if the Backup set's device is already mounted. You can eliminate
the prompt for the first volume by including the /NOPROMPT qualifier in your
command line. The default is !PROMPT.
I[NO]QUERV

Causes RESTORE to prompt you as to whether each file that meets the selection criteria should be transferred. A Y[]~S] response effects the transfer. A
response of N[O] or pressing Return bypasses the file" Any other response causes
RESTORE to reissue the prompt. The default is /NO(~UERY.

16-16

Backing Up and Restoring Files

I[NO]REPLACE[=QUERY]
The /REPLACE qualifier specifies that if a file being transferred in a restore
operation already exists on the destination disk, the file from the Backup set
replaces the one already on the disk. The /REPLACE=QUERY qualifier tells
RESTORE to query the user before replacing a file.

The /NO REPLACE qualifier specifies that if a name conflict occurs, the file
transfer is not to take place under any circumstances.
The default is /REPLACE=QUERY.
I[NO]REWIND
Specifies the disposition of a magnetic tape Backup set volume before processing.
The default is /NOREWIND.

If you specify /REWIND, the search for the Backup set begins at the beginning
of the tape. The /NOREWIND qualifier means the search starts at the present
position.
If the logical end-of-tape on the volume is found before the desired Backup set,
the tape is rewound and an error message appears at the terminal.
ISELECT[=(filespec list)]
Lets you select specific files from the input Backup set to be restored. You can use
the ISELECT qualifier in conjunction with the /EXCLUDE qualifier and/or the
/INCLUDE qualifier. If you use ISELECT in combination with other qualifiers,
the file transfer candidates are chosen according to the following steps:

1. All files on the volume for which the user has read access privilege are
initially considered
2.

If you use the ICREATED or !MODIFIED qualifiers, only those files from step
1 which meet the time bracket criteria are considered

If you use the ISELECT qualifier, only those files from step 2 that match one
of the ISELECT file specifications are considered

If you use the /EXCLUDE qualifier, any files from step 3 that match one of
the /EXCLUDE file specifications are removed from consideration

If you use the /INCLUDE qualifier, RESTORE also considers any files
meeting the inclusion criteria, provided that you have access to them. Note
that RSTS/E does not consider the ICREATED, !MODIFIED, ISELECT, or
/EXCLUDE qualifiers when processing the /INCLUDE list.

I[NO]VERIFY
The NERIFY qualifier specifies that a comparison operation is to be performed on
all data transferred. This post-process verification (for multi-volume Backup sets,
all volumes are transferred before any volume is verified) reports the presence of
any data differences. For any differences found, RESTORE displays the PPN, file
name, and file type at the user's terminal. The default is /NOVERIFY.

Backing Up and Restoring Files

16-17

Examples

Table 16-4 shows command fonnats for RESTORE operations and some of
the qualifiers associated with restore operations.
Table 16-4:

16-18

RESTORE Quick Reference Table

Command Action

Command. Example

Restore a file from a save set on
tape

MTO:PAYROL.BCK DRl:*.*

RESTORE/SELECT=[*~*]PAYROL.DAT -

Restore a disk from a save set on
tape

RESTORE MMO: 1409lVLA.BCK DUl:[* ,*]*. *

Restore a list of files from a save
set on tape

RESTORE MTO:DAT17M.BCKlSELECT=(PAYROL.DAT,BENEFI.DAT,VACATI.DAT)
DR2:[*.*]*.·'"

Restore a disk from a save set, if
file exists, replace it.

RESTOREILIST_FILE=928FEB.LOGIACCOUNT_DATA MMO:928FEB.BCK DRl:[*. *]*. *

Backing Up and Restoring Files

Appendix A

Managing the OPSER Spooling Package
This appendix describes the OPSER-based spooling package. This package, like
the new PrintlBatch Services (PBS) package (see Chapter 9), provides print and
batch services. Most of the programs require SWCFG privilege to run. General
aspects of related user programs, such as QUE and BATCH, are mentioned only
briefly in this appendix. See the RSTS / E Utilities Reference Manual for a full
description.
The information presented here is for backwards compatibility only. New
applications should use the PBS package. Move existing applications to PBS over
time. PBS is:

A.1

•

More efficient

•

More flexible

•

Easier to manage

•

Handles DCL command files

•

Permits 8-bit and control characters on printers

•

Has job limits

Overview of Operator Services
Operator services on RSTS/E involve the OPSER program and controlled
programs in OPSER tables. Controlled (online) programs are:
•

QUEMAN

•

SPOOL

•

BATCH

•

RESTOR

SPOOL and BATCH are spooling programs executing queued requests for either
a line printer device or batch processor. QUEMAN is the queue manager program
that passes queued requests to spooling programs and adds, updates, and deletes
requests in the various queues. RESTOR is the optional system data restoring
package that OPSER can control.

Managing the OPSER Spooling Package

A-1

A.1.1

OPSER Program Overview
The OPSER program establishes intetjob communication for all of its controlled
programs. OPSER declares itself a message receiver with a system-wide identification that the controlled programs can recognize. When a controlled program
starts, it declares itself a message receiver and supplies OPSER with certain
data by means of the system message send/receive SYS call. OPSER places the
program in its table of online jobs. See the RSTS / E Programming Manual for
information about SYS calls.
The identification that OPSER uses to communicate with online programs is
called the message receiver identification, or simply, the receiver identification.
Eachidentification exists in the system tnessage receiver table and must be
unique. Because of this uniqueness, only one copy of OPSER can be running
on your system. Additionally, because all control and intetjob communication of
spooling jobs depends on OPSER, the operator services program must be running
before any controlled jobs can be run.
Mter it establishes initial intetjob communication, OPSER makes it possible
for an operator to interact with the controlled jobs. OPSER becomes the main
interface between the operator and a system controlled program. Figure A-I
shows the interaction among system controlled programs and the operator.
OPSER broadcasts information to the operator throul~h a terminal designated as
the Operator Services Console (OSC). For flexibility, the operator can make any
keyboard line on the system the the OSC. 1b isolate the keyboard control from
unwarranted tampering, OPSER itself does not need to be connected to the OSC,
but may broadcast data on the physical keyboard line. Given that intermittent
output is generated on the OSC, any other user (including the operator) could be
logged into the system on that terminal.
OPSER itself establishes a data base by which it controls online jobs. The data
base resides in the OPSER package account OPSER$: and consists of three work
files:
•

OPSERO.WRK-Contains the current activities and message control directory
tables

•

OPSERl.WRK-Has the table of jobs on line to OPSER, the valid operator
table, and tables of messages and leg-al commands

•

OPSER.LOG (which can be optionally renamed)--Provides a history of
operating activity

1b provide a nonvolatile data base during a time-sha.ring session, as well as
continuity of operations between time-sharing sessions, OPSER stores the
work file information on disk. Thus, if the system crashes or if OPSER alone
unexpectedly terminates, the work files retain the most current processing
information. Similarly, these files maintain data continuity from one time-sharing
session to another.
Whenever OPSER starts, it examines its files to determine whether entries in
the online job and valid operator tables are still valid. Generally, OPSER retains
active and valid entries and purges any entries that do not satisfy validation
requirements. Upon restarting, therefore, OPSER attempts to preserve data from
its previous operational state.

A-2

Managing the OPSER Spooling Package

The lnost convenient way to interact with OPSER is through the PLEASE
program. PLEASE transmits commands to the OPSER program. OPSER checks
that the sender is a valid operator and, if so, takes the requested action. Becaus4
of PLEASE, the operator need never directly run or attach to OPSER to perform
operator functions on controlled jobs. See the section "Operator Communication
Program: PLEASE" for a description of the PLEASE program.

Figure A-1:

System Controlled Programs and Operator Interaction

OPSER$:OUEUE,WRK OPSER$:TEMPnn.OUE
File
File

----u

OPSER$:BATDCD.DCD
File

L___ .-, ___ J
\ I

-=---_=_-= =-::...-=--=-

L ~--:...--:...-_~-_

G LJ
il

I I

__ ...J I

_ __ J

OPSER$:OUEUE.SYS
File

OPSER$:CHARS.OUE
File

(detached)

u--~

OPSER

BACKUP.PRM
File

Communication method:

------z..-----

broadcast
message send I receive
direct access

OPSER$:OPSER.LOG
OPSER$:OPSERO'wRK
File
File
OPSER$:OPSER1'wRK
File
I-MK-00077-01

OPSER provides processing checks for online jobs. OPSER makes sure that all
online jobs are still active, are still valid receivers, and are not hibernating. If
any job is found to be hibernating, OPSER notifies the operator, who can attach
the job to a terminal, relnedy the cause of the hibernation, and/or restart the job.

Managing the OPSER Spooling Package

A-3

OPSER interaction between a controlled job and the operator is recorded in one
of three ways:

A.1.2

•

As a message-A message is specifically formatted data to which OPSER
assigns a sequence number for operator reference.

•

As an action request-An action request is a special message that requires
operator action and a response before a controlled job can resume processing.
A request usually involves both performing an action (such as mounting
a volume) and typing an answer (such as a device designator or program
interrupt command) in direct response to the request.

•

As an information line-An information line is a free form report of some
internal operation performed. Information lines are typically not important to
the operator function but merely provide a history of online events.

QUEMAN Program Overview
The QUEMAN program manages the system queue file OPSER$:QUEUE.SYS.
The file retains all pending user requests and can store requests between timesharing sessions. Users make requests of spooling programs through the QUE
system program. To form a request for a user, QUE builds a message and sends
it to QUEMAN, which updates the QUEUE.SYS file.
QUEMAN maintains a table of online spooling programs in a work file,
QUEUE.WRK, which is stored in the OPSER package library (OPSER$:).
The program passes a queued request to a destination spooling job not currently
busy with a request. As a spooling job completes a request, it notifies QUEMAN,
which updates the queue and work files and checks for another request to pass to
the spooling job.
A spooling program typically handles one request at a time from QUEMAN. If
no request is pending for a spooling program, the job enters an indefinite sleep
state. The job is not awakened until the system queues a message for it. The
message is the result of QUEMAN or OPSER transmitting data to the spooling
job by means of the system message send/receive mechanism.
To retain as much data as possible between time-sharing sessions, QUEMAN,

like OPSER, examines its files upon restarting. After restarting, QUEMAN
should resume processing with no loss of data from its previous run. QUEMAN
examines the queue file structure and contents and reestablishes synchronization
with any spooling jobs already on line and active.

A.1.3

SPOOL Program Overview
The SPOOL program handles requests made for line printer output. SPOOL
maintains dual communication paths: one with QUEMAN and one with
OPSER. QUEMAN transmits queued requests to SPOOL. SPOOL interacts
with QUEMAN only to notify it that a transmitted request is completed. Under
normal circumstances, communication between SPOOL and OPSER is necessary
only when an operator requires some spedal action (for example, aligning forms)
of the spooling job.
To identify output of a user's queued request, SPOOL prints heading burst pages.
The program accesses the character generation file OPSER$:CHARS.QUE to form
the large, easily readable letters necessary to distinguish discrete requests and
files within a request.

A-4

Managing the OPSER Spooling Package

A.1.4

BATCH Program Overview
The BATCH program executes user requests by running a job for a user on a
static pseudo keyboard defined with the SET SYSTEM command. By using a
pseudo keyboard, BATCH eliminates the requirement for a physical terminal and
relieves the user from typing by processing commands from a disk file.
BATCH, like SPOOL, maintains dual communication paths: one with QUEMAN
and one with OPSER. QUEMAN sends a message to BATCH to process a job
request. BATCH interacts with QUEMAN only to notify it that a transmitted
request is completed. BATCH typically depends on OPSER to form requests for
operator action. To mount a user volume, for example, BATCH notifies OPSER.
In turn, OPSER generates the request for the operator and processes the response
from the operator.

To process commands from a user-queued disk file, BATCH requires the command
decoding file OPSER$:BATCH.DCD. The BATCH.DCD file contains all the codes
and parameters BATCH needs to decode user-specified control statements,
perform syntax checking, and form a special intermediate file (BAnJmm.COM) of
encoded BATCH commands.
Both the SPOOL and BATCH programs charge processing (execution) time to the
account under which the user queues the request. Overhead time, used to set up
a request, is charged to the account under which the program is running.

A.11.S

RJ2780 Program Overview
The RJ2780 program is an optional layered product, available through DECUS.
Unlike BATCH and SPOOL, &J2780 maintains only one communication path.
QUEMAN sends a message to RJ2780 to process the next request (send a file).
The RJ2780 program can also send a message to QUEMAN requesting that
SPOOL print a file that RJ2780 has received. RJ2780 has no communication
with OPSER.

A.11.6

Controlling RESTOR with OPSER
An operator can control the RESTOR program through OPSER. If RESTOR runs
detached, control through OPSER is the only means by which the operator can
communicate with the RESTOR job. RESTOR, when detached, transmits all
operator requests through OPSER and receives responses only from OPSER.

A.11.7

Overview of OPSER Shutdown
The entire operator services package is designed to terminate in a controlled
manner using a "shutdown level." The defined shutdown level for each job
determines the order in which OPSER terminates controlled jobs during system
shutdown. The SHUTUP system program can be run so OPSER terminates
controlled jobs with the lowest shutdown level first. OPSER informs each job at
a certain level to complete processing at a logical end point. Only after OPSER
shuts down all jobs at a given level does it proceed to shut down all jobs at the
next highest level. SHUTUP itself does not proceed with final system shutdown
until OPSER terminates all of its controlled jobs and kills itself. By using the
controlled shutdown of OPSER, you can ensure continuity of processing user
requests from one time-sharing session to another.

Managing the OPSER Spooling Package

A-5

Since RJ2780 does not communicate with OPSER, OPSER does not automatically
shutdown RJ2780 as it does SPOOL and BATCH.

A.2 Operator Services Program: OPSER
The operator services program OPSER consists of two modules:
•

OPSER-Establishes the initial conditions on startup, checks initial conditions upon restart, and chains to the second module, OPRUN

•

OPRUN-Executes commands and periodically checks status.

Each module is stored in the OPSER Package Library (OPSER$:), has a
protection code of <232>, and requires SWCFG privHege to run. For simplicity,
this appendix refers only to one program, OPSER, which functionally includes the
two modules.
You must run the OPSER program before the queue manager and spooling
programs. If your applications need to UBe the OPSER package, you should start
OPSER by including the RUN command in your system startup command file. To
start OPSER, type the following command:
$ RUN OPSER$:OPSER
OPSER V9 RSTS V9 TIMESHARING
#

At this point, the OPSER program perfonns a number of checks before it prints
the number sign (#) prompt. The remaining portion of this section describes these
initial program checks. Once you are familiar with this procedure, you can learn
about OPSER operator commands, descri.bed in the following sections.
If the you have SWCFG privilege, the program prints an identification line at the
terminal to indicate it is running. If you do not have SWCFG privilege, OPSER
displays an error message to indicate that you do not have access to OPSER.

OPSER tests for the presence of its two work files (OPSERO.WRK and
OPSER1.WRK) in OPSER$:. If the work files are not present, OPSER prints the
warning message:
%OPSER files not found - will initialize ...

OPSER then initializes all its tables. At this point, the valid operator table has
a single entry that allows the current user (on anytE~rminal) to communicate
through OPSER. If ERRCPY is active, the online job table has ERRCPY as its
only entry.
When OPSER finds the work files, it makes sure that it has write access to
the files. Should another job have write access to the files, OPSER prints the
following warning message and terminates:
%OPSER does not have write privileges to its files

To allow OPSER to gain write access, the operator rn.ust terminate the job
currently having write access and run OPSER again.
The program performs further special processing for eurrently existing work files.
The processing makes sure of data integrity in case OPSER is being restarted
after a system crash. A check is made of the entries Jin the online job and valid
operator tables.

A-6

Managing the OPSER Spooling Package

For each entry in the online job table, OPSER checks the following conditions:
•

The job number indicated in the entry must be active on the system.

•

The receiver identification in the entry must exist in the system message
receiver table for that job n.umber.

•

The account in the entry must be a valid account number.

An entry for an inactive or otherwise illegitimate job is removed from the online
job table. For each active and legitimate job, OPSER requests retransmission of
the job's last message when processing begins. When processing begins, OPSER
generates a command to display the online job table.

OPSER checks the entries in the valid operator table. Additional actions take
place depending on whether jobs found in the online job table are active. If
the online job table has no active jobs and a meaningless entry is in the valid
operator table, OPSER clears the table and sets initial conditions. The resulting
operator table has one entry that allows, as valid operators, the current user
from any keyboard. If the online job table has an active job, the program merely
removes those entries that have no meaning from the valid operator table. To
signal a change in the cw-rent table, OPSER creates an internal command to
print the valid operator list when processing begins.
Mter processing of the work files is complete, the program attempts to declare
OPSER as a receiving job. If the declaration fails, the program prints the error
message:
?OPSER cannot declare itself a receiver

This message indicates that another copy of OPSER has been run and must be
properly terminated to allow the current copy to run successfully.
OPSER establishes the current keyboard as the Operator Services Console (OSC)
and a file named OPSER.LOG as the default log file. OPSER.LOG is optional,
but OPSER uses it until an explicit command closes it or assigns another log
file. When OPSER begins normal operation, it executes any internally generated
commands to list online jobs and valid operators. Mter executing the commands,
OPSER is ready to accept data sent to it by any program. OPSER prints the
number sign (#) character to show that it is ready to accept commands.
A.~~.1

OPSER Operator Commands
You can abbreviate a valid OPSER command to three or more characters and
separate the command from any operands by a space character. No embedded
spaces are allowed within a command. Delimiters within the text of the operands
are the semicolon (;) and colon (:) characters and must be present to delimit
elements of text. Throughout the following discussion, note that square brackets
enclose optional items.

Managing the OPSER Spooling Package

A-7

Table A-I summarizes the commands and formats. Note that the circumflex (A)
character marks the location of a required space.
Table A-1:

OPSER Commands

Command Name

Syntax and Meaning

ANSWER

ANS[WER]"msgnumber:text
Conveys the text following the colon as a response to an
action request denoted by mensage number. Also deletes the
action request.

CHANGE CONSOLE

CHA[NGECONSOLE]"KBn:
Changes the operator services console to the keyboard unit
designated by n. The unit should be on line but need not be
free. It must alBo be set for IBROADCAST (see Chapter 8).

DELETE

DEL[ETE]"msgnumber
Deletes from an OPSER internal table an unanswered action
request denoted by its message number. DEL[ETE]"#m[:n]
Deletes the n oldest (from 1 to 32) action requests for a given
job. The command requires the number sign (#), the number
(m) of the sending job, a cololl1 (:), and the number (n) of
requests to delete.

DETACH

DET[ACH]
Detaches the OPSER program from its keyboard, after which
the operator can run PLEASl~ to communicate with OPSER.

EXIT

EXI[T]
Closes the log file and work files, removes OPSER from
the system message receiver table, and terminates the
program. If thEl job is attached, the program returns you
to the keyboard monitor prompt. If detached, the program
broadcasts a message to the 13ystem console terminal (KBO:)
and kills itself.

INTERRUPT

INT[ERRUPT]l\rcvrid:text
INT[ERRUPT]J\#n:text
Sends the unsolicited text to an on-line job specified either
by its message receiver identification or by #n (where n
is the number of the job under which the on-line program
is running). The LIST JOBS command listing gives job
numbers and receiver identifications.

LIST

LIS[T]"JO[BS]
LIS[T]"OP[ERATORS]
Prints a listing of all jobs on line to OPSER or a listing of
valid operator accounts and keyboards.

(continued on next page)

A-8

Managing the OPSER Spooling Package

Table A-1 (Cant.):

OPSER Commands

Command Name

Syntax and Meaning

LOGFILE

LOG[FILE]/\[file];[msglevel]
Establishes, as the OPSER log file, the specified file or devic~
and sets the message level to the value specified. If a file or
device is not specified, a semicolon followed by a level simply
changes the message level of the open log file. If neither a
file nor message level is specified, OPSER closes the current
log file and records nothing more.
LOG[FILE]/\[file];AL[L]
ALL records all messages and action requests.
LOG[FILE]/\ [file];RE [QUESTS]
REQUESTS records only action requests.
LOG[FILE]/\[file];NO[NE]
NONE records no messages, action requests, or OPSER
information lines.

MESSAGE

ME S[SAGE]/\ [level]
Sets the level of messages for the OSO. The level controls what types of messages are printed at the OSO. If
no level is included in the command, an error is generated. MES[SAGE]/\AL[L] ALL includes all messages, action
requests, and OPSER information lines.
MES[SAGE]/\ RE[QUESTS]
REQUESTS includes only action requests and information
lines and excludes messages.
MES[SAGE]/\ NO[NE]
NONE means that nothing is printed on the ose. (See the
section "Message Types" for a discussion of message level.)

OPERATOR

OPE [RATOR]/\ KBn:[p,pn]
Updates the valid operator list with the designated keyboard
unit and account combination. The asterisk (*) character
can replace the keyboard unit, project number(p), and
programmer number (pn) to include all of that element.
OPE[RATOR]/\-KBn:[p,pn]
If a minus (-) sign precedes the specification, that specific
element is deleted from the list.

RETYPE

RET[YPE]/\msgnumber
Reprints, at the requesting terminal, an unanswered action
request denoted by its associated message number.
RET[YPE]/\[#n[:n]]
Several unanswered action requests can be printed by giving
the number of the sending job. If:n is given with the job
number, the n oldest (from 1 to 32) action requests are
printed for that job.

NOTE

Entering an account that does not have SWCFG privilege in the
OPSER operator table with the OPERATOR command allows anyone
logged in to that account access to all accounts. You must use caution
when deciding which accounts you place in the OPSER operator table.

Managing the OPSER Spooling Package

A-9

A.2.2

Message Types
OPSER displays two types of information on the Operator Services Console
(OSC): messages and action requests. Messages are simply informative. For
example, a message might tell you that a control file has been queued for batch
processing. Action requests require an operator to perform some action before the
job continues (for example, mounting a tape).
The MESSAGE command lets you define the level of information displayed at the
OSC. For example:
•

MESSAGE ALL-You want both messages and action requests displayed at
the OSC.

•

MESSAGE REQUESTS-You want only action requests displayed at the
OSC. Messages are held in a table until you specify MESSAGE ALL.

•

MESSAGE NONE-You do not want any messages displayed at the OSC.
Messages and requests are held in a table until you specify MESSAGE ALL
or MESSAGE REQUESTS and then the RETYPE command.

When OPSER starts, the default is MESSAGE ALL. H you suspend the display
of messages, or messages and requests, by typing ME~SSAGE REQUESTS or
MESSAGE NONE, the suspended items are held in a table. The table can hold a
maximum of 32 messages or requests. (The actual number held depends on the
length of the messages and requests.) If OPSER receives a message or request
after the table is full, the new message or request is added to the table, and the
oldest message or request is deleted.
Thus, you should not suspend message printing for long, or you may lose
messages and requests. When you want to see messages and requests again, type
MESSAGE ALL. Then use RETYPE to print the messages and requests currently
held in the table.
If you see "Message Table Full" displayed at the OSC, it means that a message
was received, and the table was full. Some unknown number of old messages and
requests had been lost while MESSAGE NONE or MESSAGE REQUESTS were
in effect.
To help you distinguish between messages and action requests on the OSC,

output appears in distinct formats. A single TAB character denotes message text;
two TAB characters denote action request text.
Message requests have the format:
MESSAGE

nnnnn : date time JOB:NN KB:MM jobname [p,pn]
text

If the first character of the message text is a question mark (?) character, the
words FATAL MESSAGE replace MESSAGE in the identification line.
Action requests have the format:
REQUEST

A-10

nnnnn : date time JOB:NN jobname [p,pn] rcvrid
text

Managing the OPSER Spooling Package

In an action request, the first character of text is a Ctrl/G (BEL) that sounds a
bell in the tenninal to alert the operator. Table A-2 summarizes the contents of
messages and action requests.
Table A-2:

OPSER Message and Action Request Contents

Item

Meaning

nnnnn

Sequence number of message starting at 1 and incremented by 1
during processing; use with the ANSWER command to respond to a
specific action request. Af\;er restarting with currently existing work
files, OPSER rounds the last assigned sequence number up to the next
multiple of 10. This number is the next assigned sequence number.

date

Current system date.

time

Current system time of day.

JOB:nn

Job number under which sending program is running; use it, preceded
by the number (#) sign, to refer to the job in an INTERRUPT, DELETE,
or RETYPE command.

K.B:mm

The keyboard number of the job that sent the message. The KB:mm
field is displayed as DET if the job was detached, and as KB:?? if
OPSER is unable to determine the keyboard number. Typically, the
KB:?? message appears when the job has terminated.

jobname

Name of the program that an online sending job is running; this name
is printed for information purposes. OPSER prints six question marks,
??????, in place of the jobname if it is unable to determine the jobname.
Typically, the ?????? message appears when the job has terminated.

[p,pn]

Project-programmer number (PPN) under which the sending online job
is running.

rcvrid

For action requests only. The message receiver identification given to
the job when it is started. The identification can be standard (LPnSPL
for the SPOOL program or BAnSPL for the BATCH program) or can
be optional (as specified in the NAME startup switch for SPOOL
or BATCH). Use the receiver identification with the INTERRUPT
command to refer to the spooling program.

An action request is the result of a spooling job's requesting interaction with the
operator: the spooling job sends a message to OPSER. The spooling job perfonns
no further processing until operator action is taken or until the condition that
generated the action request is satisfied. OPSER broadcasts action requests on
the OSC unless the message level is set to NONE.
'lb respond to an action request, the operator uses the ANSWER command.

Because OPSER usually runs detached, an operator can run PLEASE at valid
operator tenninals to send the response to OPSER. The ANSWER command
requires a message number to identify the action request to which the operator is
responding.
OPSER sends the text in the ANSWER command to the spooling program and
deletes the action request. If the response satisfies the request, the program
resumes processing. If the response does not satisfy the request, the spooling
program still expects an appropriate response. The operator must then send such
a response by means of the INTERRUPT command (see the section "Operator
INTERRUPT Command").

Managing the OPSER Spooling Package

A-11

A.2.3

Valid Operator and Online Job Lists
The LIST cO'mmand displays data cO'ncerning the valid O'peratO'r list O'r the O'nline
jO'b table maintained by OPSER:
•

LIST OPERATORS displays the keybO'ard number and PPN cO'mbinatiO'ns
currently defined as valid O'peratO'rs. The OPERATOR cO'mmand updates the
valid O'peratO'r list. Initially, O'nly users with SWCFG privilege can update the
valid O'peratO'r list.

•

LIST JOBS prints the header ONLINE JOB and then prints data cO'ncerning
O'nline jO'bs in the fO'IIO'wing fO'rmat:
#n [p,pn] 'rcvrid' SL=n

Table A-3 describes the items in the O'nline jO'b list.
Table A-3:

OPSER Online Job List

Item

Meaning

The job number under which the online program is running.

[p,pn]

The PPN under which the online job is running.

'rcvrid'

The message receiver logical identification used to identify the online
program in INTERRUPT commands.

SL=n

The OPSER shutdown level used wh~n SHUTUP notifies OPSER to
terminate spooling ope1'ations in an orderly fashion.

The jO'b data listed is impO'rtant to' the operatO'r. For example, certain OPSER
cO'mmands require either a receiver identificatiO'n or jO'b number to cO'mmunicate
with a sPO'O'ling prO'gram.

A.2.4

Operator INTERRUPT Command
The INTERRUPT cO'mmand sends unsO'licited messages and special cO'mmands
to' sPO'O'ling prO'grams. OPSER accepts any text in the INTERRUPT cO'mmand
and sends it to' the destinatiO'n jO'b. Only the destinatiO'n jO'b interprets the text.
If a reSPO'nse to' an INTERRUPT cO'mmand is generated, it is displayed O'n the
OSC. Tables A-6, A-12, and A-15 summarize special text interpreted by sPO'O'ling
prO'grams as cO'mmands.
The INTERRUPT cO'mmand allO'ws an operatO'r to' contrO'I and to' mO'nitO'r sPO'O'ling
prO'grams. Each sPO'O'ling prO'gram recO'gnizes cO'mmands that can be divided intO'
three sets:

A-12

•

CO'mmands that have nO' direct effeet O'n hO'W the prO'gram itself prO'cesses
a request queued by a user. Included in this set O'f cO'mmands are PAUSE,
CONTINUE, NOTICE, LAST, and STATUS. They allO'w the O'peratO'r to'
cO'ntrO'I the prO'gram in general and to' gain infO'rmatiO'n abO'ut the prO'gram.

•

CO'mmands that dO' have a direct effect O'n prO' gram O'peratiO'n but are unifO'rm
fO'r all sPO'O'ling prO'grams. END, ABORT, and OFFLINE represent this type O'f
command which directly influences the jO'b.

•

CO'mmands that are unique to' the individual spoO'ling prO'gram. A cO'mmand
fO'r fO'rms cO'ntrO'I O'n a line printer iEI an example.

Managing the OPSER Spooling Package

A,,2.5

OPSER Startup Procedure
Use commands in the system startup control file to start the OPSER program.
See "System Startup Command File: START.COM" in Chapter 3 for information
about creating a startup control file. This section describes the commands you
should include in that startup file.
The following is a typical sequence of startup file commands to start OPSER:
$ RUN OPSER$:OPSER
LOG OPSER$:OPSER.LOG;ALL

CRA KBO:
MESS ALL
OPER KB*: [1,2]

DETACH

The LOGFILE, CHANGECONSOLE, MESSAGE, OPERATOR, and DETACH
commands establish initial operating conditions for OPSER. The LOGFILE and
MESSAGE commands set conditions normally established as defaults and are
shown for clarity. An explanation of these commands follows:

A.2.6

•

The LOGFILE command retains the file OPSER$:OPSER.LOG with a
message level of ALL. As a result, all messages, action requests, OPSER
information lines, and operator responses are written to the file to provide a
complete historical reference.

•

The terminal on which OPSER starts is automatically defined as the OSC.
You can change this definition by placing the CHANGE CONSOLE command
in the startup procedure. Any valid operator may also alter the assignment 0'
the OSC. It is most convenient to keep keyboard unit 0 as the OSC because,
regardless of the number of logins currently allowed on the system, the
operator can always use the system console terminal to communicate with
OJPSER.

•

The MESSAGE command specifies that all messages, action requests, and
OJPSER information lines are printed at the OSC. Normally, the operator
is concerned only with action requests. If the operator is concerned with
any unanswered action request, the RETYPE command can be issued. In
addition, the log file can be closed with the LOG command and printed at an~
time to recover recorded information.

•

The OPERATOR command updates the valid operator list with user [1,2] on
any keyboard. Allowing the operator to communicate from any terminal on
the system gives flexibility to those who are performing operator functions.

•

The DETACH command causes OPSER to detach itself from the terminal on
which it is running. At the start of timesharing, this terminal is usually the
system console terminal (KBO:). While OPSER runs detached, it is immune
from tampering by unauthorized users. The operator can communicate with
OPSER through the PLEASE program.

OPSER Action Under Various Startup Conditions
OPSER startup actions depend on the way OPSER last terminated:
•

When OPSER runs after you have shut down operator services in an orderly
fashion through SHUTUP, a normal startup procedure occurs. When shutting
down in an orderly fashion, OPSER removes all entries from its online job
table and closes all files properly. On restarting, OPSER finds all data valid.
The online job table is empty, and entries in the valid operator table are legal
and therefore retained.
Managing the OPSER Spooling Package

A-1 ~

•

When OPSER starts after a system crash, it may find entries in its online job
table. Because OPSER starts before other online jobs, none of the jobs in the
table is active; therefore, the program clears the entries from the table. If any
entries in the valid operator table are illegal, OPSER clears that table and
sets one valid operator entry: the current user frOUI any keyboard.

•

When OPSER starts after an unexpected termination, it may find entries in
its online job table for jobs still active on the system. In this instance, the
program sends a command to each active job having an entry in the online
job table. The command requests the job to retransmit the last message it
sent to OPSER. OPSER does not wait for a response. If any job in the online
job table does not meet OPSER's intebTIity checks, OPSER removes its entry
from the table (takes the job off line) and generates an internal LIST JOBS
command. OPSER also scans the entries in the valid operator table. If any
entry is illegal, the program removes it. If OPSER removes any entry, it
generates an internal LIST OPERATORS command.

A.3 Queue Manager Program: QUEMAIN
The queue manager program QUEMAN rnaintains the file of queued requests
(QUEUE.SYS in the OPSER$: account an.d communicates with spooling programs
to execute queued requests. Queue management involves two modules stored in
the OPSER Package Library: QUEMAN and QUMRUN. Both modules have a
protection code of <232>, and require SWCFG privilege to run.
QUEMAN:
•

Sets initial conditions on starting

•

Checks initial conditions after restarting

•

Chains to QUMRUN

QUMRUN:
•

Executes commands

•

Manages the queues

For simplicity, the documentation refers only to one program, QUE MAN , which
functionally includes the two modules.
The QUEMAN program runs only if the job has SWCFG privilege and the OPSER
program is running (that is, only if the name OPSER is in the system table of
message receivers). Starting QUEMAN is a prerequisite to starting the spooling
programs. To start QUEMAN, type the following c01nmand while logged into the
system under an account SWCFG privilege:
$ RUN OPSER$:QUEMAN
QUEMAN V9 RSTS V9 TIMESHARING
STARTED AT: 01:52 PM ON 03-MAR-85

If you do not have SWCFG privilege, the system displays an error message to
indicate that you do not have access to QUEMAN.

If the job has enough privilege, the program prints its identification line and
startup lines and begins preliminary error checking.

A-14

Managing the OPSER Spooling Package

QUE1\1AN tests for the presence of OPSER which must be started before
QUEl\t1AN can run. If OPSER is not running, QUEMAN prints an informational
message and terminates:
QUEMAN cannot run without 'OPSER' active

QUEMAN opens the files QUEUE.WRK, OPSER1.WRK, and QUEUE.SYS in
the OPSER$: account. If QUEUE.WRK does not exist, the program prints the
following informational message and creates the file:
'QUEUE.WRK' not found - will initialize

If OPSER1.WRK does not exist, the program prints the following informational
message and terminates:
OPSERl.WRK not found --- can't run

To recover, the operator must start the OPSER program before starting
QUEMAN. If QUEUE.SYS does not exist, the program creates and initializes it.

QUEMAN signals this action by generating the informational message:
No queue file found -- will initialize

If QUEUE.SYS exists, QUEMAN makes sure that it has write access to the file.
If another program has write access to the QUEUE.8YS file, QUEMAN generate~
the following message:
Queue file opened by another program
Try again (YIN) <N>?

The operator must determine which job has the file QUEUE.SYS open and must
terminate that job. If you type Y in response to the question, QUE MAN retries
the open operation. Typing N terminates QUEMAN.
When the OPEN operation successfully gains write access to QUEUE.SYS,
QUEMAN declares itself a message receiver on the system. If QUEMAN is
already declared by a different job, the program prints the message:
QUEMAN cannot declare itself as a receiver ... can't run

The operator must terminate the other job so it is removed from the message
receiver table.
To make sure that the previous QUE MAN job closed the currently existing
QUEUE.SYS file properly, the program checks a flag value in the file. If the flag
is not properly set, the progratn prints the following informational message:
Queue file not closed properly --

now checking data for consistency

A later informational message signals that all integrity checks were successful:
Queue file data checked for consistency

If the flag is properly set, QUEMAN performs consistency checking but does
not print any messages unless it finds problems. If any problems are found,
QUEMAN generates the following informational messages:
Queue file data inconsistent - will initialize
Initialized

The INITIALIZED message indicates that QUEMAN has set all entries in
the QUEUE.SYS file to their initial conditions and has removed all queued
requests from the queues. See the section "QUEMAN Consistency Checking" for
a description of the queue file consistency checks QUE MAN performs.
Mter completing all initial checks, QUEMAN prints the number sign (#) prompt
indicating its readiness to accept a startup command or switch.

Managing the OPSER Spooling Package

A-15

A.3.1

QUEMAN Startup Commands and Switches
Table A-4 lists the startup commands that QUEMAN Jrecognizes.
Table A-4:

QUEMAN Startup Commandsl

Command

Syntax and Meaning

DETACH

DET[ACH]
Detaches QUEMAN from the terminal. You should run QUEMAN
detached.

INITIALIZE

INI[TIALIZE]
Sets all the entries in the qUEUE.SYS file to their initial conditions
and sets initial conditions for the spooling queues. Any currently
queued requests are lost.

Before QUEMAN executes the INITIALIZE command, it checks its online spooler
table. If the table has an entry for a spooling job, QUI~MAN prints the following
informational message:
Spoolers on line -- can't initialize

QUEMAN also recognizes startup switches. One or more switches may appear
following a command or may appear alone on a line in response to the prompt.
Tabl~ A-5 lists the startup switches that QUEMAN allows.

Table A-S:

QUEMAN Startup Switches

Option

Syntax and Meaning

PRIORITY

!PRI[ ORITY]:nnn
Sets the job priority to nnn, where nnn can be from -120 to +120.
Without the switch, the program automatically sets the priority to O.
You must have TUNE privilege to specify this qualifier.

RUNBURST

/RUN[BURST]:nnn
Sets the job nm burst to nnn, where nnn can be from 1 to 127. Without
the switch, the run burst value currently assigned is used. You must
have TUNE privilege to specify this qualifier.

If any data entered in response to the number sign prompt are not valid commands or options, QUEMAN prints the following informational message:
Invalid response -- text

The text is the data entered. The program prints the # prompt again.
'ro detach QUEMAN, the operator types the DET comlmand in response to the
prompt. For example:
#DET
Detaching ...

The program then prints the DETACHING message a.nd detaches itself from the
keyboard. Note that you should always run QUEMAN detached.
While detached, the job receives commands from the operator through OPSER
(see the following section). At no time should the operator enter CtrllC to the
QUEMAN program; this may corrupt the QUEUE.SYS file.

A-16

Managing the OPSER Spooling Package

A.3.2

QUEMAN ~nterrupt Commands
QUEMAN recognizes interrupt commands that allow an operator, through
the INTERRUPT command of OPSER, to get information on queue manager
processing and to control program operation.
Table A-6 summarizes these interrupt commands.
Table A-6:

QUEMAN Interrupt Commands

Comm.and

Option and Meaning

DIS[ABLE]

QUE[UING]<text>
SPO[OLING]<text>
ALL<text>
Disables QUEUING or SPOOLING or both operations with the
ALL switch. When you specify DISABLE QUEUING, the program
QUE stops sending messages to QUEMAN and stops listing the
queue. Any jobs already sent to QUEMAN are processed. If you
specify DISABLE SPOOLING, QUEMAN prevents any further
jobs from being sent to any spooler. Jobs sent to a spooler prior to
the DISABLE command are processed to completion. Specifying
DIS[ABLE] ALL tells QUEMAN to terminate both QUEUING and
SPOOLING. You can include a message to the user while specifying
the DISABLE command. Type the command, the switch, and then
a space followed by the message. After QUEMAN processes the
command, it prints the <text> message in one of the following
formats:
• Further QUEUING DISABLED <text>
• Further SPOOLING DISABLED <text>
• Further QUEUING AND SPOOLING DISABLED <text>
If you ignore the <text> field, QUEMAN prints the default text, "by
operator." The message appears when you try to queue a file.

ENA[BLE]

QUE[UING]<text>
SPO[OLING]<text>
ALL<text>
Enables QUEUING or SPOOLING or both operations with the
ALL option. When you specify ENABLE QUEUING, the program
QUEUE begins to send messages to QUEMAN and starts listing
the queue. If you specify ENABLE SPOOLING, QUEMAN begins to
send jobs to the spoolers. Specifying ENABLE ALL tells QUEMAN
to start both the QUEUING and SPOOLING functions. Use the
<text> field to send messages to the user. Type the command, the
switch, and then a space followed by the text of the message. After
QUEMAN processes the command, it prints the message in one of
the following formats:
• QUEUING ENABLED - <text>
• SPOOLING ENABLED - <text>
• QUEUING AND SPOOLING ENABLED - <text>
The default <text> message is "by operator." The message appears
when you try to queue a file.

(continued on next page)

Managing the OPSER Spooling Package

A-17

Table A-6 (Cont.):

QUEMAN Interrupt Commands

Command

Option and Meaning

END

Terminates QUE MAN in an orderly manner and disables queuing
and spooling. If any spooling programs are on line to QUEMAN,
END sends to OPSER the informational message "Spoolers still
online - will clear table" and clears the online spooler table. END
then sends a command to OPSER that takes the QUEMAN entry
out of OPSER tables. QUEMAN closes its work files and removes
its job from the system.

LAS[T]

Sends to OPSER the most recent message generated by QUEMAN.

NEXIT]

quenam=reqnam Places the pending user request identified by
reqnam at the head of the queue spedfied by quenam. A request
name can contain a job name, PPN, and sequence number. Request
names are found in the listing generated by the QUE program
command L dev:, where dev: is the appropriate queue name. For
example, LP:, BA:, or RJ: are valid queue names.

OFF[LINE]

Terminates QUEMAN us the END command does but does not
generate a notification message.

STA[TUS]

Prints a brief report of spooling jobs on-line to QUEMAN.

When a command is received from the operator throUlgh OPSER, QUEMAN
performs the action requested and either generates a message or an information
line. OPSER formats the message or information line and displays it on the ose.
An example of this interaction between an operator, '~UEMAN and the OPSER
program is shown using the STATUS cOlnmand. The following example assumes
the opera tor is running the PLEASE program and the ose is the keyboard on
which PLEASE is running:
:/I:/INT :/I:6:STATUS
COMMAND SENT TO 'OPSER'
:/I:
MESSAGE
29 : 10-MAR-85 10:48 AM
1 SPOOLER(S) ON LINE (25) LP1SPL LP1: FORMS=NORMAL;

JOB:6 DET QUMRUN[1,2]

PLEASE passes the full command line to OPSER and notifies the operator
before it reprints the prompt. OPSER recognizes the text INT as a request to
send unsolicited text to a spooling job. The characters "#6" in the command line
are recognized as the number of the destination job to which the text must be
sent. QUEMAN receives the text STATUS, generates a response, and sends the
response back to OPSER. OPSER formats a message and displays it on the ose
for the opera tor.
The STATUS command itself is the operator's way of getting information about
jobs that are on line to QUEMAN. QUEMAN'S statUE; printout gives the number
of spooling jobs currently on line and supplies data on each job (job number
within parentheses, receiver identification, physical lmit being spooled, and
program default conditions). If the particular spooling job is processing a request,
the data in the printout includes the following infonnation about the request: its
name, the PPN under which it is queued, and its sequence number. The STATUS
command also tells you if SPOOLING or QUEUING is disabled.
The NEXT command allows the operator to place a job request at the head of a
queue. Job requests are identified by the name, account, and sequence number
under which the request is queued. A request moved to the head of a queue is
the next one QUEMAN sends to a spooling program servicing that queue. See

A-18

Managing the OPSER Spooling Package

the RSTS / E Utilities Reference Manual for a description of the NEXT format in
the QUE program discussion.
If QUEMAN detects an error in the NEXT command, it generates an error
message in the format:
'NEXT' CMD: string - text

The string is the command you entered. Table A-7 lists the possible error texts.
Table A-7:

NEXT Command Error Text

Text

Meaning

ILLEGAL PARAMETERS

Only a queue name and request name,
separated by an equal (=) sign, are
allowed in the NEXT command.

A.3.3

ILLEGAL SYNTAX

A space or an = character is missing.

JOB IN PROCESS

The specified job request has already been sent to
the spooling program.

MULTIPLE JOB SPECIFIED

The specified request name matches two or more
jobs in the queue.

NO MATCH FOR JOB

The request name does not match any job in queue,

QUEMAN Startup Procedure
You should place commands in the system startup command file to start the
QUEMAN program. See "Creating and Using Command Files" in Chapter 3 for
inforrnation about creating a startup command file.
The following are typical commands to start QUEMAN:
$ RUN OPSER$:QUEMAN
/PRIORITY:O/RUNBURST=6
DETACH

The priority is set to zero because most jobs on RSTSIE run at priority minus
eight and QUEMAN, running at priority zero, can more readily process the queue
file. Include the DET command to make QUEMAN run detached on your RSTSIE
system.
NOTE

Without the PRIORITY switch, QUEMAN automatically sets its
priority to zero. You can explicitly specify the switch to provide a
record of what the program does automatically.
Because QUEMAN examines the QUEUE.SYS file when it starts, there is no
need explicitly to initialize the file. If the file contains bad data, QUEMAN resets
all entries to initial states. Therefore, you should specify the INI command only
in situations in which you want to explicitly initialize the queue file.

Managing the OPSER Spooling Package

A-19

A.3.4

QUEMAN Action Under Various Startup Conditions
QUEMAN startup actions depend on the way QUEMAN last terminated:

A.3.5

•

When QUEMAN runs after you shut down operator services in an orderly
fashion through SHUTUP, a normal startup procedure occurs. QUEMAN
finds that the QUEUE.SYS file was closed properly. The program performs
consistency checking but does not print any messages unless it finds problems. If the data is consistent, QUEMAN retains all job requests in the
queue file and examines each request for completeness and status. Because
QUEMAN normally starts before any spooling programs, there are no entries
in the online spooler table. QUEMAN checks to see that this table is clear of
entries.

•

When QUEMAN starts either after the system crashes or after QUEMAN
itself terminates unexpectedly, it finds the queue file improperly closed and
generates a message to that effect. The program performs consistency checking on the queue file structure and on the data in the QUEUE.SYS file itself,
generates messages telling the results, and reestablishes communication with
all spooling jobs.

QUEMAN Consistency Checking
The QUEUE.SYS file has room for 250 queued requests. QUEMAN creates an
entry for a queued request from an entry in a free list. A request for a line
printer or batch queue requires one entry from the free list. A request queued
with an AFTER date and time, however, requires two entries from the free
list-one for the proper queue and one for the AFTER queue.
In performing consistency checking on the QUEUE.SYS file, QUEMAN initializes
all entries if it finds one of the following eonditions:
•

An entry in the free list is also in a queue or AFTER list

•

The root of the free list is outside the legal range

•

A request queued with an AFTER date and time that has not expired does
not have an entry in the AFTER queue

For other inconsistencies, QUEMAN either may remove a single entry or perform
some related action. If an entry in the AFTER queue has no corresponding entry
in a spooling queue, QUEMAN removes the AFTER entry. If an entry in either
an AFTER or spooling queue is not completely set up. the program removes the
entry. (QUEMAN may have terminated while it was processing the request.) To
signal this event, QUEMAN generates a :message in the format:
%LPn: 'reqnarn'

[p,pn];QUEUED JOB INCOMPLETE - REMOVED FROM QUEUE

If an entry to be killed still remains, QUEMAN removes it and generates a

message in the format:
%LPn: 'reqnarn'

[p,pn];QUEUED JOB IN 'KILL' STATUS - REMOVED FROM QUEUE

If QUEMAN finds that an entry has been processed (sent to a spooling program)
but not yet completed, it places the request in a hold status and generates' text in
the format:
%LPn: 'reqnarn' [p,pn];JOB PREVIOUSLY SENT TO SPOOLER;
WILL BE PUT INTO HOLD STATUS

The requester can remove the job request from hold status.

A-20

Managing the OPSER Spooling Package

QUEMAN checks the online spooler table in the QUEUE.WRK file for consistency.
If the program finds the table empty, it generates the informational message:
# SPOOLERS ON LINE = 0; WILL CLEAR TABLE

If the program finds that the count of online spooling jobs is larger than the limit
(16), it clears the table and generates the informational message:
ONLINE SPOOLER TABLE CORRUPT - WILL CLEAR TABLE

If QUEMAN clears the online spooler table but spooling jobs are still running,
the operator must terminate those jobs through OPSER and restart each spooling
program. This action allows QUEMAN to put each spooling job on line again
proper~y.

For each spooling program found in the online job table, QUEMAN removes the
entry and generates the informational message:
'rcvrid'

(nn) FOUND ON LINE TAKEN OFF LINE

QUEMAN sends the job a message requesting that it declare itself on line to
QUEMAN again. The spooling job does not respond to the message while it is
processing a queued request. The answering message indicating that the job is
again on line to QUEMAN does not appear until the job completes the current
request. When QUEMAN receives the online declaration from a spooling job, it
generates the informational message:
'rcvrid'

(nn) PUT ONLINE

A.4 Line Printer Spooling Program: SPOOL
The liIlle printer spooling program SPOOL runs without operator intervention and
executes queued requests to transfer disk files to a line printer or terminal. The
program consists of three modules:
•

SPOOL-Establishes initial conditions on startup and checks initial conditions upon restart

•

SPLIDL-Executes when no job is being printed

•

SPLRUN-Prints any spooled file

Each module is stored in the OPSER package account OPSER$:, has a protection
code of <232>, and requires SWCFG privilege to run. For simplicity, this
appendix refers only to one program, SPOOL, which functionally includes the
three modules.
The SPOOL program runs only if the user has SWCFG privilege and the OPSER
and QUEMAN programs are running. To start SPOOL, type the following
command while logged in to the system under an account with SWCFG privilege:
$ RUN OPSER$: SPOOL
SPOOL V9 RSTS V9 TIMESHARING
#

Typically, this is done in the system startup file.
At this point, the SPOOL program performs a number of preliminary checks
before it prints the number sign (#) prompt. The remaining portion of this section
describes these initial program checks. Once you are familiar with this procedure,
you can then learn the options SPOOL accepts in response at its program prompt.
A description of these SPOOL program checks follows.

Managing the OPSER Spooling Package

A-21

If you do not have SWCFG privilege, the system dispJlays an error message to
indicate that you do not have access to SPOOL.
If the job has SWCFG privilege, SPOOL :runs and p:rints its identification line
and the # prompt. In response to the prompt, give a Elpecification in the following
format:
# logical device:/startup switch(es)

The logical device is the name of the queue and the unit nwnber within that
queue from which this copy of SPOOL takes job requests to execute. This
name ordinarily corresponds to the physical device on which SPOOL prints
the requests; the physical device, however, may be changed by the PHYSICAL
startup option (see the following section "SPOOL Startup Options").
The logical device name must have the form LPn: or LP:, where n is a unit
nwnber from 0 to 7. If the name is LPn:, the spooling job prints only requests
queued either to that explicit unit or to the general spooling queue LP:. If you
specify LP:, this copy of SPOOL prints requests in the printer queue regardless
of the unit to which they were originally queued. The logical device specified as
the queue name must be a line printer; the device name given must not have a
logical assignment to some other device.
NOTE

The general spooling queue LP: is useful only on systems having line
printers with similar characteristics.
You can use any combination of the startup switches described in the following
section to condition the operation of the SPOOL program. A switch is formed by
a slash (/) character and an option that may take an .operand or other switches.

A.4.1

SPOOL Startup Options
Startup options alter default processing conditions. You use these options to:

A-22

•

Spool output to a different physical device

•

Change form information

•

Control job environment

Managing the OPSER Spooling Package

Table A-8 summarizes these options. Note that the symbol L) is an underline
character and not a space.
Table A-8:

SPOOL Startup Options

Option

Syntax and Meaning

ASSIGN

IASS[IGN]
Reserves the physical device for this job. Without this option, SPOOL
retries continually whenever it cannot gain access to the unit and is
ready to print a job.

FORM

IFOR[M]:name switch
Defines the current form according to a name (NORMAL is the
default name) and switches as follows:
•

•

LPFORM

IALI[GN]
Causes SPOOL to execute a forms alignment procedure before
processing any further queued requests.
IDFL[ENGTH]:nnn
Declares the device form length as nnn lines, where nnn is a
number from 1 to 127. The length is that understood by the
hardware of the output device. For example, on a line printer
in the United States, the standard length is usually 66 lines per
page.
IHEA[DINGS]:n
Causes SPOOL to print n (from 0 to 3) heading burst pages
preceding each job request. Unless INH accompanies a file
specification in the QUE command, SPOOL also prints n burst
pages preceding each file in a job request. The default value is 1.
ILEN[GTH]:n
Defines the length of software form as n lines per page, where n
is a number from 1 to 127. If this switch is not given, SPOOL
assumes 66 lines per page.
IPAG[E_EJECT]:YES
Indicates that the device to which this copy of SPOOL directs
output has a hardware top of form capability and that the device
interprets the formfeed character (ASCII 12) as a top-of-form
command. This condition is the default case for line printer
devices.
IPAG[E_EJECT]:NO
Conditions the software to translate a formfeed character (ASCII
12) into the proper number of linefeed characters because the
hardware does not recognize the character as a top-of-form
command. This condition is the default case for SPOOL output
directed to terminal devices by the PHYSICAL option.
IWID[TH]:n
Defines the width of the heading burst page as n characters per
line. The default value is 132.

ILPF[ORM]:YES
Indicates that LPFORM characters are interpreted by the system
software for the output device. During normal transfer, the program
passes LPFORM characters unmodified to the device. This condition
is the default for line printer devices.

(continued on next page)

Managing the OPSER Spooling Package

A-23

Table A-a (Cont.):
Option

SPOOL Startup Options
Syntax and Meaning

ILPF[ORM]:NO
Indicates that the system software does not interpret LPFORM
characters. During normal transfer, the program converts each
LPFORM character to an appropriatE~ number of line feed characters.
This condition is the default for terminal devices. Note that this
option affects only jobs queued with the ILPFORM switch. The
SPOOL program does not automatically invoke ILPFORM processing.
NAME

!NAM[E]:rcvrid
Places the specified identification in t.he system message receiver
table and in OPSER's online job table. This identification rather
than the default identification is used. by the operator in OPSER
commands to access this spooling job" This identification must not be
in use by some other job.

PHYSICAL

!PHY[SICAL]:dev:
Uses this device as the physical devi<ce for this spooling program.
Requests in the queue specified by the logical device are printed on
this device. The default value is the logical device specified as the
queue name.

PRIORITY

!PRI[ORITY]:nnn
Sets the job priority to nnn, where nnn is a number from -120 to
120. Without this option, the priority remains as set by LOGIN.
You must have TUNE privilege to specify this qualifier. Do not use
!PRIORITY after a !FORM, !NAME, or /PHYSICAL option, or after
an LPn: designator.

RUNBURST

/RUN[BURST]:nnn
Sets the job run burst to nnn, where nnn is a number from 1 to 127.
Without this option, the run burst remains as set by LOGIN. You
must have TUNE privilege to specify this qualifier.

The PHYSICAL option alters the physieal device on which queued requests are
printed. For example, requests queued for line printer unit 1 could be processed
on another line printer unit or on a keyboard unit. In any event, the actual
device differs from the one for which the requests were queued.
The ASSIGN option reserves the actual spooled device to the job. Without this
option, SPOOL periodically tries to gain access to the spooled device whenever
SPOOL is ready to print and another job has ownership of the device. When
SPOOL terminates, it deassigns the device.
The NAME option allows the receiver identification to be different from that
normally assigned by SPOOL. The identification should be from one- to sixalphanumeric characters and must not exist in the system message receiver
table. The identification must not begin with a number because the operator uses
either this identification or a job number to identify a spooling job.
The FORM option allows the default fOIm information to be altered. For example,
individual characteristics of the NORMAL form can be altered, a new form
with a different name can be defined, or a forms alignment can be requested.
A forms alignment can be requested by itself or in conjunction with a form
alteration. When a form name other than NORMAL is in effect, user's queuing
requests must explicitly specify the fonn name to have files printed with its
characteristics.

A-24

Managing the OPSER Spooling Package

To alter the default characteristics of the NORMAL form, the operator can specify
the appropriate auxiliary switches without the !FORM: switch. For example, to
change the number and width of heading burst pages for line printer unit 0, type
LPO:/HEADINGS:2/WIDTH:80

When executed, this command causes SPOOL to output two heading burst pages
with a width of 80 characters per line print for each job request and each file
within a request. The width used applies only to the burst pages, not to the data
being printed,
Specifying a form with a name other than NORMAL usually means that the
operator must load special paper in the output device. The IALIGN switch with
the FORM option requests a forms alignment. The operator then must align the
form at system startup time. A SPOOL interrupt command with the IALIGN
switch can request a form alignment during timesharing. The section "Changing
and Aligning Forms" describes the forms alignment procedure.
The following message can occur when the line printer queues are empty, but jobe
with nondefault form names exist:
NO JOBS WAITING WITH FORMNAME 'XXXXXX' FOR SPOOLER 'LPn:'
OTHER JOB(S) WAITING.
***PLEASE INSPECT QUEUE AND TAKE APPROPRIATE ACTION***

The operator must then decide whether or not to use an interruption command to
change the form name so the other jobs can be queued.
The PAGE_EJECT option determines whether the SPOOL program counts lines
to effect forms control. (The DFLENGTH option can modify the effect of the
PAGE_EJECT option.) A value of NO with the PAGE_EJECT option indicates
that the output device (driver) does not understand a form feed character; that
is, a form feed character does not cause the device to position itself at top of
the (paper) form. With NO in effect, SPOOL counts lines to effect forms control.
A response of YES with the PAGE_EJECT option indicates that the device
(hardware) translates the form feed character to an appropriate number of line
skips to place the device at top of form. With YES in effect, SPOOL need not
count lines and may be able to pass all forms control data unaltered to the device
driver. (The DFLENGTH option can modify the effect of YES.)
NOTE

SPOOL issues three forms-control characters:
•

Form feed character (ASCII 12)

•

Carriage return character (ASCII 13)

•

Line feed character (ASCII 10)

No other forms-control characters are used. SPOOL has no provisions
for changing or extending the set of control characters for any special
device(s),
The terms line feed and line skip are not synonymous. Line feed refers
only to the line feed character, which either SPOOL or the device driver
sends to a device. Line skip refers to the movement of paper that the
device itself effects.
If NO is in effect for the PAGE._EJECT option, SPOOL ignores any value specified
in a DFLENGTH option. The operator should specify IPAGE_EJECT:YES for any
terminal device that has a top-of-form capability.

Managing the OPSER Spooling Package

A-25

SPOOL uses the value specified in the DFLENGTH option only if IPAGE_
EJECT:YES is in effect. The IPAGE_EJECT:YES switch can be in effect either
through default (on a line printer) or by explicit specification (for a terminal device having the fonn feed capability). The operator may specify the DFLENGTH
option either at startup time in a SPOOL startup switch or during processing in
a FORM interrupt command. For proper formatting, the DFLENGTH value must
be the maximum number of lines that the output device skips when it receives a
form feed command. The number of lines is called th,e device fonn length. This
number may differ from the length of the paper fonn being used.
Different devices have different fonn lengths. Three examples can show how the
device form lengths vary:
•

The LA180 device has a switch labeled "Length of Form," which has settings
for 4, 8.5, and 11 inches. These settings correspond to 24, 51, and 66 lines
per form. (The number of lines per form given for each setting assumes the
standard 6 lines per inch on the LA180.) If, when starting the spooling job,
the operator sets the switch on the LA180 to 8.5 inches, the operator should
also specify the DFLENGTH switch with a value of 51. If, at some later
time, the operator changes the setting on the LAl80 to 4 inches, the operator
should also specify a new DFLENGTH option with a value of 24 in a FORM
interrupt command to the spooling job.

•

Many line printers have only one form length. If the hardware length is
66, the value in the DFLENGTH switch should be 66. If the hardware form
length is 51, the operator should specify DFLENGTH=51 when starting the
spooling job. (The default value for DFLENGTH is 66.)

•

Some printing devices have an adjustable top-of-form. On these devices,
the operator can adjust the device to execute an arbitrary number of line
skips .when it receives a form feed character. Thil3 adjustment is made most
often on terminal-type devices with high quality print characteristics. When
resetting the top-of-form for this type of device, the operator should also
specify the DFLENGTH and LENGTH options with the new form length
information. The operator should also request a forms alignment for each
new form.

The effect of the DFLENGTH option on the SPOOL program differs for line
printers and terminal devices:
•

For a line printer, the DFLENGTH value affects only the procedure SPOOL
invokes when it receives an ABORT command while it is processing a job
request. If the DFLENGTH value equals the current length of paper fonn,
the program issues a form feed character to position the device at the top of
the next form. If the values are not equal, the program automatically invokes
the forms alignment procedure to reestablish the top-of-form position.
The automatic fonns alignment procedure requires operator action to align
the fonn and to respond to requests SPOOL makes. This is necessary only
when the program has no way to determine where the top-of-form lies after
an ABORT operation.

•

A-26

For a terminal, the DFLENGTH value controls th.e way SPOOL keeps its
place on a form and the way the operator handles recovery procedures. If
the value of the PAGE_EJECT option is YES, the equality of the device form
length and the paper fonn length is important. If the device form length
and paper fonn length values are not equal, the SPOOL program counts
lines to maintain a record of its position on a fonn and issues an appropriate
number of line feed characters to sinlulate form feed. If these values are
equal, SPOOL does not count lines and simply sends a form feed character to
position the paper at top-of-form.

Managing the OPSER Spooling Package

NOTE

A line printer is a device whose interface to the computer is through
a line printer controller and whose designation is in the form LPn:.
A terminal is a device whose interface to the computer is through
a keyboard line and whose designation is in the form KBn:. This
distinction is critical because some devices such as the LA180 are
called printers but may be connected to the computer through a
keyboard line.
If the value in effect for DFLENGTH does not correctly match the setting of the
hardware device, two undesirable conditions exist:
•

SPOOL must incur the overhead of counting lines.

•

Forms alignment may be lost during processing.

For these reasons, Digital highly recommends that the operator supply a new
value of DFLENGTH when changing the hardware setting on the device.

A.4.2

Line Printer Spooling
If the physical device specified at startup time is a line printer unit, SPOOL
applies the default values. Table A-9 lists these default values.
Table A-9:

SPOOL Line Printer Default Values

Switch

Result

IDFLENGTH:66

The device form length is 66 unless otherwise specified. If
you have patched the monitor to apply a different default
form length, you must always specify the correct value in a
DFLENGTH option.

IFORM:NORMAL
!HEADINGS: 1
ILENGTH:66
IWIDTH:132

The values shown are used as the form definition.

ILPFORM:YES

For any line printer, SPOOL assumes that the device driver
properly processes the special LPFORM characters (see the
RSTSIE Programming Manual).

IPAGE_EJECT:YES

For any line printer, SPOOL assumes that the device driver
properly processes top of form characters.

You can also set the characteri.stics of a line printer with the SET PRINTER
command (see Chapter 11).

Managing the OPSER Spooling Package

A-27

A.4.3

Keyboard Spooling
If the physical device specified at startup is a keyboard device, SPOOL applies

the default values. Table A-IO lists these default valllles.
Table A-10:

SPOOL Keyboard Default Values

Switch

Result

IDFLENGTH:66

DFLENGTH has no effect unless PAGE_EJECT is explicitly
specified as YES.

IFORM:NORMAL
!HEADINGS: 1
ILENGTH:66
IWIDTH:132

The values shown are used as the form definition.

ILPFORM:NO

The terminal driver does not process LPFORM characters.
For any file that contains LPFORM characters, SPOOL
simulates the LPFORM effect by issuing linefeed characters.

IPAGE __ EJECT:NO

Most terminals do not have a hardware top-of-form capability.
SPOOL simulates top-of-form by issuing the correct number
of linefeed characters.

For terminals with the hardware top-of-form capability:

A.4.4

•

Set the value of PAGE_EJECT to YES.

•

Set the value for DFLENGTH to the actual devic:e form length (see "SPOOL
Startup Options").

•

Set the proper characteristic of the device that aecepts and processes a fonn
feed. Use the SET TERMINAL command to perform this operation (see
Chapter 8).

Startup Error Processing
If you make an error when responding to the number sign (#) prompt, SPOOL

prints a message and the unparsed comlnand line in the format:
?ERROR MESSAGE
unparsed command line

A-28

Managing the OPSER Spooling Package

The unparsed command line is the right-hand part of the response beginning at
the element causing the error. Table A-11 summarizes the possible messages.
Table A-11:

SPOOL Syntax Error Messages
Message and Meaning

?Can't parse remaining string
Something illegal or undefined was found in the command line. An undefined switch or missing slash (/) character causes this error.

?Duplicate switch
Two occurrences of the same switch were found in the command line.

?megallogical device
The logical device specified was not in the form LP: or LPn: (where n is a
number from 0 to 7).

?megal operand
An illegal operand was found in an option or switch. For example, specifyinl
nonnumeric characters in a switch that requires a number (IWIDTH:n)
generates this error.

?megal physical device
The device specified in the PHYSICAL option is not a line printer or
keyboard device.

?Missing operand
A switch requiring an operand was specified without one. For example,
if a receiver identification is missing from the NAME option, this error is
generated.

If SPOOL encounters no syntax errors in the response, it begins setting initial
conditions. SPOOL reports errors in this phase by printing a message in the
format:
ERROR IN SOME OPERATION
error message

RESTARTING

The p:rogram reprints its identification line and the prompt.

An elTor message ?Duplicate receiver ID means that the receiver identification,
either the default one or the one specified in a NAME option, is already defined
for another job. The operator should type the command line again and specify a
unique receiver identification in the NAME option.
The error message ?No room in receiver table means that a general small buffer
is not available to let SPOOL declare itself a receiving job. A later retry with the
same command line should succeed.
The following message means that SPOOL attempted to send a message to that
program and failed:
*****OPSER HUNG***** or *****QUEMAN HUNG*****

Either the named program does not have an entry in the system message receivel
table or the program is not processing its messages and its message limit has
been I'eached. The operator must determine the cause of the problem and restart
the named program.

Managing the OPSER Spooling Package

A-29

A.4.5

SPOOL Interrupt Commands
The operator communicates with a SPOOL job by means of INTERRUPT
commands sent through the operator services progranl OPSER. Note that any
responses to an INTERRUPr command are displayed on the Operator Services
Console.
Table A-12 summarizes these SPOOL interrupt comrnands. Note that the
circumflex (") character marks the location of a required space.
Table A-12:

SPOOL Interrupt Commands

Command Syntax

Meaning

ABO [RT]

Immediately terminates the cur:rent process and removes the
request from the queue.

CON[TINUE]

Wakes up the spooling job to continue processing after a
PAUSE command.

END

Closes out processing after completing the current request.
The operator must then run the! spooling program again to
process further requests.

FOR[M]"name/switch

Changes the current output fonn to one identified by name
and defined by switches. These switches change current form
characteristics; those characteristics not changed by an option
remain at their current definitions. See the startup option
FORM description in Table A-8 for the allowable options.
Alignment is done only if the IALIGN switch is included.

FOR[M]

Displays characteristics of the c:urrent form if no name or
options are in the command.

FOR[M]/\/ALIGN

Requests a forms .alignment proc.edure for the current form.

LAS[T]

Prints the most recent message generated by the spooling job.

OFF[LINE]

Immediately terminates all processing by this spooling job
(same as ABORT followed by END).

PAU[SE]

Places the spooling job in a sleep state, during which it
responds to most commands and. resumes normal processing in
response to a CONTINUE command. l

REQ[UE]

Stops processing the current l'equest and replaces the request in the queue so that proc,essing later resumes at the
terminating point.

RES[TART]

Reprints the current copy of a file from the beginning, including heading burst pages.

RES[TART]/\JOB

Reprints the current iteration of the job, including heading
burst pages.

RES[TART]/\ :n

Reprints the current copy of thB file starting at page n. If n is
0, restarts from the beginning of the file but omits the heading
burst pages.

STA[TUS]

Prints a status report for the spooling job.

lNote that processing continues if any job modification command such as ABORT or REQUE is sent.

A-30

Managing the OPSER Spooling Package

A,,4.6

SPOOL Startup Examples
This section shows four examples of the startup procedure. They represent
typical cases.

A.4.6.1

Line Printer Startup with All Defaults

In response to the SPOOL program prompt, the operator types:
#LPO:

If SPOOL finds no errors, it prints the message:
DETACHING ...

SPOOL detaches itself and leaves the terminal logged off the system.
The following are the characteristics of this copy of SPOOL:
•

The program prints only job requests that were queued to line printer unit 0
with a form name of NORMAL or queued to the general line printer queue
with a form name of NORMAL.

•

The program prints on line printer unit O.

•

The program prints one heading burst page before each job and, unless a file
is queued with fNH, prints one heading burst page before each file. The widtl
of the burst page is 132 columns, and the form length is 66 lines.

•

The program assumes the device driver handles LPFORM characters.

•

The program assumes the device handles top-of-form.

•

The program assumes the device form length is 66 lines.

•

The run burst and priority of SPOOL are those in effect when the job starts.

•

When the program receives a job from QUEMAN, it attempts to assign the
printer. If the assignment fails, the program retries periodically until it
succeeds. When it finishes processing a job request, the program deassigns
the device.

The operator may later change all form characteristics (name, length, width, device form length, and headings count) through appropriate interrupt commands.
A.4.6.2

Line Printer Startup with Narrow Width

In response to the prompt, the operator types the command line:
#LP1:/FORM:NARROW/WIDTH:80/RUNBURST:12/ASSIGN

The following are characteristics of this copy of SPOOL:
•

The program prints only jobs queued with a form name of NARROW and
queued to either unit 1 or the general line printer queue.

•

The program prints on the line printer 1 device.

•

The program prints one heading page of width 80 columns; form length is 66.

•

The program assumes the device handles top-of-form.

•

The program assumes the device driver handles LPFORM characters.

Managing the OPSER Spooling Package

A-31

•

The program assumes the device form length is 66 .

•

The priority of the job remains unchanged. The program sets the run burst to
12.

•

The program does not begin spooling until it can assign line printer unit 1
and does not deassign the device until it terminates.

In this example, assume that line printer unit 1 is a relatively slow device and
that the SPOOL program can generate output data for this unit faster than the
device can print. Optimal use of the device results if the unit keeps running at
full speed for the entire job request it is printing.
'lb effect optimal use, it is sometimes necessary to raise either the priority or run

burst of the SPOOL program. The priority of a job generally affects how often
it runs, whereas the run burst affects the length of time the job runs once it
starts running. Because the SPOOL program is usually 1/0 bound and waiting
for the device to complete printing, there lis little reason for it to run more often.
Changing the priority is usually unnecessary and ineffective. Changing the run
burst, however, may improve processing.
When SPOOL starts executing, it remains in the run state until one of the
following events occurs:
•

The job's run burst expires. That is, the job actually executes for its full run
burst.

•

The job requests some type of 1/0 and the system Icannot immediately satisfy
the request.

For the spooling program, the second event is much more likely because the
system buffers only a fixed amount of data before it refuses to satisfy further
requests for output. When the refusal occurs, the system activates another job
and the spooler must wait until the syste·m sends at least some of the already
buffered data to the output device. (This wait condition shows as an LP or TT
STATE on a SYSTAT listing.)
As long as the buffers for the device are never completely emptied, the system
keeps the device running at or near full speed. 'lb a certain point, raising the
spooling job's run burst makes it more likely that the job will continue running
until it completely fills the buffers for its output devic:e. Mter a certain point,
raising the run burst has little benefit because the job can run only until the
buffers are full. Conversely, decreasing the run burst causes the system to
deactivate the program before the buffers are full. It is more likely that the
printer can empty the buffers before the system reactivates the job.
If a spooling job spends a significant time in the RN Btate while it is actually

printing, the run burst is too low. DIGITAL recommends that you raise the run
burst to a value at which the job spends most of its time in an output wait state.
A.4.6.3

Keyboard Startup on an LA36

For this example, assume that keyboard unit 5 is an LA36 terminal. At the
prompt, the operator types a command line as follows:
=11=

LP2: /PHYSICAL: KB5: /HEADINGS: O/ASSIGN

The following are the characteristics of this copy of SPOOL:

A-32

•

The program prints only jobs queued with a form name of NORMAL and
queued to the unit 2 or the general1ine printer queue.

•

The program prints on keyboard unit 5.

Managing the OPSER Spooling Package

•

The program prints no heading burst pages. The form length is 66.

•

The program assumes that. keyboard unit 5 does not handle top-of-form;
therefore, it counts lines as it prints them to simulate top-of-form.

•

The prograln assumes the device driver does not handle LPFORM characters;
therefore, it simulates LPFORM with line feeds.

•

The run burst and priority remain unchanged.

•

The program assigns keyboard unit 5 at startup and keeps it assigned until it
terminates.

Because the LA36 is a slow device, processing is not slowed any further with
burst pages. The IHEADINGS:O switch suppresses printing of burst pages.
1b retain any forms alignment on the LA36, the device is permanently assigned
to the spooling job. This assignment prevents any other job from using the device
and leaving the carriage at a position other than the top of form position. When
the spooler starts processing a request, it must assume the device is at its top of
form position because the program has no way to force the device to advance to
some known position. If the form is not aligned properly when SPOOL starts the
output, the entire output is misaligned.
A.4,.6.4

Keyboard Startup on an LA180

For this example, keyboard unit 2 is a serial LA180 DECprinter. This terminal
has a hardware top-of-form capability. In addition, the form length selector switch
on the terminal is set to 8.5 inches (51 lines per form) and the paper form is also
51 lines per page.
In response to the prompt, the operator types a command line as follows:
#LP3:/PHYSICAL:KB2:/PAGE EJECT:YES/OFLENGTH:51/LENGTH:51MORE>
/WIOTH:40/FORM:SPECIAL

The following are the characteIistics of the SPOOL copy:
•

The program prints only requests queued with a form name of SPECIAL and
queued to either the unit 3 or the general line printer queue.

•

The program prints on keyboard unit 2.

•

The program prints one heading page that is 40 columns wide and 51 lines in
length.

•

The program assumes the device handles top of form.

•

The program assumes that the device driver does not handle LPFORM
characters. The program therefore simulates" LPFORM with line feeds.

•

The run burst and priority remain unchanged.

•

The program assigns the keyboard before it prints each request and deassigns
it after printing the request.

Managing the OPSER Spooling Package

A-33

Because:
•

The LA180 is a much faster device than the LA3S, SPOOL can print a small
number of burst pages to simplify distinguishing jobs and files. The narrowest
width for the burst page is 40 columnfJ.

•

The paper form length and the device form length are equal, SPOOL can
position the device to top-of-form by issuing a forml feed character.

•

The program can position to top-of-fOIm, it does not need to keep the unit
assigned.

The ASSIGN option, therefore, does not appear in the example. If the paper fonn
length did not equal the hardware setting, however, SPOOL should keep the
device assigned to preserve the top-of-form position.

A.4.7

Recovery from Line Printer Errors
The interaction between the spooling job and the OPSER program controls error
handling in the SPOOL program. For example, if the paper runs out or jams,
SPOOL receives the error message:
?Device hung or write locked

SPOOL discontinues processing and sends OPSER a rnessage. OPSER generates
a message for the operator:
MESSAGE

3 : 28-MAR-85 04:33 PM JOB:21 DET SPLRUN[l,lOO]
LPO
: HUNG
JOB: [220,40]USERl
(PUT DEVICE ONLINE TO CONTINUE)

The message tells the operator that the spooling job for line printer unit 0 has a
problem and that the device needs to be Inade ready again before processing can
continue.
For a SPOOL program servicing a line printer, there are certain errors from
which you can easily recover. The device handler for the line printer tests the
ready status of an offline unit every 10 seconds. If the operator corrects the
error (for example, fixes a paper jam) and puts the line printer on line again,
the software detects the ready status and continues printing buffered characters.
The SPOOL program finds that the error condition is cleared and continues
processing.
If a special form is being processed and SPOOL receives an error, SPOOL may
request the operator to realign the forms before it continues processing on its
own.
If the operator cannot correct the error or wants to perform some other operation,
the spooled device may be left off line. This prevents the software from reprinting before the program processes an operator request. The operator uses the
INTERRUPT command with the proper text to access, the spooling job.
SPOOL allows the operator several ways to control the restart of a queued
request. The RESTART command can restart:
•

The current copy of a queued request (RESTART ,JOB)

•

The current copy of a file in a queued. request (RE:START)

•

The current copy of a file at a specific page (RES1~T:nnn)

The REQUE command can reenter the current job in the queue for later processing. On receiving a RESTART command? the program waits for any currently
buffered data to be printed before actually restarting.
A-34 Managing the OPSER Spooling Package

The REQUE command causes SPOOL to remember the point at which processing
was interrupted. Mter resuming the requeued request, SPOOL scans for the page
on which processing terminated and continues printing at the start of that page.
An en.·or may occur if a user modifies the requeued request before SPOOL can
resume the printing.
If the operator responds to an error condition with an ABORT command, which
terminates the job, SPOOL clears the output buffers on the line printer unit. If
the device is a terminal, SPOOL prints any buffered data. The program then
issues a message to show that it is terminating.
Before terminating a queued request, the SPOOL program makes sure that all
buffers have been emptied successfully.

A.4.8

Line Printer Output
SPOOL generates job header and file header burst pages to identify print requests
and files within a print request. Both types of header page contain identification
and general accounting information as follows:
•

The identification consists of large, easily readable block letters created from
the character generation file CHARS.QUE. The job identification contains the
account number of the user requesting the job and the name of the job. If no
job name appeared in the QUE command, SPOOL prints the name of the first
file in the request as the job name. The file identification shows the file name
and type.

•

General accounting information for the job header is on the burst page
and is offset from the identification by two rows of special characters. The
accounting information contains five lines of data:
The first line gives the job name, current date, current time of day, and
the requester's account number.
The second line gives the date and time of day of the request and the
device for which the request was queued.
The third line gives the system name.
The fourth line gives the QUE options used to process the request.
The fifth line gives the job copy number.
The accounting information for the file header burst page appears below
the identification and is framed (above and below) by two rows of special
characters. The first line of information gives the job name used when
SPOOL printed the file and shows the current date, time of day, and account
as the job header burst page does. The second line gives the copy number, the
QUE options used on the file, and the complete specification of the file. The
third line gives the record type and carriage control format used to print the
file.

If SPOOL does not print the file because of an error, file header identification and
accounting information is replaced by an error message framed (above and below)
by five rows of special characters. The error message is standard RSTSIE error
text.

Managing the OPSER Spooling Package

A-35

A.4.9

Error Messages During User Output
SPOOL reports errors it receives during printing in the line printer output.
SPOOL differentiates the error message from requested output by framing the
text (above and below) with five rows of special characters.
Table A-13 shows the possible SPOOL error messagel3 and related meanings.
Table A-13:

SPOOL User Requested Output Error Messages
Message and Meaning

?Can't find file or account

The related file was deleted between the time it was queued and the time
SPOOL tried to print it.
?File restart requested on page n

The operator requested a restart at the indicated page or placed the job
request back in the queue with. the REQUE command. If the operator
restarted the job request, SPOOL scans the file to find the correct page and
continues printing on a new page following the message. If the operator
requeued the job, this message appears when SPOOL restarts the job.
?Job aborted

Either the operator tenninated the job request with an ABORT command
or the K command of the QUE program was issued while the job was being
printed.
?Job aborted because of file error(s)

Errors in some file in the job pl'evented SPOOL from printing the remainder
of this job.
?Job requed at page n

The operator placed this job request back in the queue with the REQUE
command. SPOOL later continues the job request with a new header burst
page and the text FILE REST.ART REQUESTED ON PAGE n, where n is
the same page number shown in the REQUE message.
?<errortext>

SPOOL encountered the RSTSIE error shown, aborted processing the
request, and cleared the request from the qUl8ue.

A.4.10

Changing and Aligning Forms
The FORM command can:
•

Inform the spooler of a change in fonn on its device

•

Request an alignment operation on the device

•

Display information about the current form

Switches with the command specify the new form definition and request alignment. The command may request alignment when a new form is specified.
Th request a form change, the operator sends the FORM command with the
form name and the switches to define the form to the related spooling job. If an
alignment is wanted for the current fonr.l., the operator can include the IALIGN
switch with the FORM command. The SPOOL program does not recognize the
form change request until it becomes idle. If it is processing a queued request,
it completes the processing before it performs the forms change. (Th make the
A-36

Managing the OPSER Spooling Package

change immediately, request that SPOOL requeue the current request. The
program requeues the request and becomes idle.)
1b perform only an alignment, specify the FORM command without any form
name and with only the fALIGN switch. Again, the command takes effect only
when SPOOL becomes idle. The program follows these steps:
1. Prints characters to delineate left and right margins, and prints the text TOP
OF FORM centered on the same line.
2. Generates characters to position the paper on the last line of the form.
3. Prints characters to delineate the left and right margins, and prints the text
END OF FORM centered on the same line.
4. Positions the paper at top of the next form.
SPOOL generates an action request and awaits a response. The operator may
inspect the alignment and type one of two responses to the action request. If
the alignment is not correct, the operator can adjust the hardware and send
the response RETRY back to the spooler. In response to RETRY, the spooler
perfonns steps 1 through 4 again and generates another action request. When
the alignment is accurate, the operator types GO as a response to the action
request. The spooler returns to nonnal processing.
Whenever SPOOL processes any FORM command, it sends a message to the
operator indicating that the new fonn characteristics in effect. In any FORM
command, only those options specified in the command change the form characteristics. Thus, if the operator sends a FORM command alone, without a fonn
name or any options, the spooler displays the characteristics of the current fonn
on the Operator Services Console.

A.:5 Batch Processor Program: BATCH
The BATCH system program rims without user intervention on pseudokeyboards
and executes files of standardized commands queued for either a specific or the
general Batch Processor. The program consists of four modules:
•

BATCH sets initial conditions for batch processing.

•

BATIDL checks status when no requests are being handled.

•

BATDEC decodes and perfonns error checking on command files.

•

BATRUN executes the decoded batch commands.

For simplicity, this appendix refers only to one program, BATCH, which
functionally includes the four modules. Each module is stored in the OPSER
package account OPSER$:, has a protection code of <232>, and requires SWCFG
privilege to run.
1b run BATCH, type the following command while logged in to the system under
an account with SWCFG privilege:
$ RUN OPSER$ : BATCH
BATCH V9 RSTS V9 TIMESHARING
#

TypicaHy, this startup of BATCH is done in the system startup file.
If you do not have the SWCFG privilege, the system displays an error message
indicating that you do not have access to BATCH.

Managing the OPSER Spooling Package A-37

If you have the SWCFG privilege, BATCH runs and prints its identification
line and the number sign (#) prompt. In response to the # prompt, enter a
specification in the following format:
logical device:/startup switch(es)

The logical device is the name of the queue and the unit number within that
queue from which this copy of BATCH takes job requests. This name does not
specify the static pseudo keyboard on which BATCH executes requests. The
PHYSICAL startup option can specify thE! static pseudo keyboard unit on which
BATCH executes. The logical name must have the form BA: or BAn: where n
is a unit number from 0 to 7. If BAn: is f3pecified, thE! Batch Processor executes
requests queued to either that explicit unit or the general batch queue BA:. If
BA: is specified, the program executes job requests in the batch queue regardless
of the unit to which they are queued.
NOTE

The general batch spooling queue, BA:, is useful only on systems with
Batch Processors running with the same default conditions.
The logical device specified as the queue name must be a Batch Processor; the
device name given must not have a logical assignment to some other device.
If only a queue name is given in response to the pr01npt and BATCH finds
no etTors, the program establishes default processing conditions, prints the
DETACHING message, and detaches itself from the terminal.
BATCH forms the default receiver identification from the queue name. BATCH
also establishes other default conditions:
•

BATCH tolerates no etTors and terminates any queued request encountering
a fa tal or warning etTor.

•

User log files are queued to the system default line printer with deletion
specified.

•

The run burst remains unchanged.

•

The priority remains unchanged.

At startup, BATCH performs the same etTor checking and processing SPOOL
performs. See the section "Startup EtTor Processing" for the description of
SPOOL startup etTor processing.

A-3S Managing the OPSER Spooling Package

A"S.1

BATCH Startup Options
Include options with the BATCH processor designator to control its operation.
Table A-14 summarizes these options, which are in the form of switches.
Table A-14:

BATCH Startup Options

Option

Syntax and Meaning

ASSIGN

IASS[lGN]
Reserves a pseudo keyboard device to this job.

ERROR

IERR[OR]:severity
Sets the error severity default for this processor so that errors of
that severity are tolerated and queued requests having that or lesser
severity end successfully.
IERR[OR]:FAT[AL]
Tolerates all elTors.
IERR[OR]:WAR[NING]
Thlerates only warning errors.
IERR[O]R:NON[E]
Thlerates no errors (fatal and warning errors terminate queued requests). This is the default severity.

DELETE

!DEL[ETE]
Deletes user log files after printing. This is the default condition unless
the NODELETE option is specified.

NAME

INAM[E]:rcvrid
Places, in the system message receiver table, the specified receiver
identification for this job instead of the default identification. The
default for a Batch Processor is BAnSPL for specific processors and
BASPL for the general processor.

NODELETE

INODEL[ETE]
Does not delete user log files after printing. Without this option, all log
files are deleted. Requester may override this option with the INOQUE
switch in the $JOB command.

NOQUEUE

INOQUE[UE]
Does not queue user log files for printing. Without this option, all
log files are queued without a queue name and with deletion unless
changed by the IQUEUE and !DELETE switches. Requesters cannot
override this option.

PHYSICAL

!PHY[SICAL]:dev:
Uses the pseudo keyboard (for example, PK1:) as the device on which
this processor executes user batch jobs.

PRIORITY

!PRI[ORITY]:nnn
Sets the job priority of BATCH to nnn, where nnn is a number from
-120 to 120. Without this option, the priority remains unchanged. You
need TUNE privilege to specify this qualifier. Do not use !PRIORITY
after a INAME, or !PHYSICAL switch, or after a BAn: designator.

(continued on next page)

Managing the OPSER Spooling Package

A-39

Table A-14 (Cont.):

A.5.2

BATCH Startup Options

Option

Syntax and Meaning

QUEUE

IQUE[UE]:[quenam]
Queues all 'user log files to the specified spooling queue (LPO: through
LP7:) rather than to the general print queue. If no queue name is
specified, all jobs are queued to the gen(~ral print queue. Files are
deleted after printing. Requesters may override queuing with the
INOQUE switch in the $JOB command. Deletion of the files by the
SPOOL program may be overridden by including /NO DELETE as a
startup switch.

RUNBURST

IRUN[BURST]:nnn
Sets the job run burst of BATCH to nnn, where nnn is a number from
1 to 127. Without this option, the run burst remains unchanged. You
need TUNE privilege to specify this qualifier.

BATCH Interrupt Commands
The operator uses interrupt commands s·ent through the operator services program OPSER to communicate with a Batch Processor. Note that any responses to
an interrupt command are displayed on the Operator Services Console.
Table A-15 summarizes these BATCH interrupt comlnands.
Table A-15:

A-40

BATCH Interrupt Commands

Command Syntax

Meaning

ABO [RT]

Immediately terminates the current process and removes the
request from the queue.

CON[TINUE]

Wakes up the Batch Processor to continue normal processing after either a PAUSE command or after a user request for operator
action (for example, in the $MESSAGEIWAIT command).

END

Terminates processing after completing the current request. The
operator must then run the Batch Processor again to process
further requests.

LAS[T]

Prints the most recent message generated by this Batch
Processor.

NOT[ICE] text

Inserts the specified text in the user log file and precedes it with
the heading: NOTICE FROM OPERATOR.

OFF[LINE]

Immediately terminates all procl~ssing by this Batch Processor
(same as ABORT followed by END).

PAU[SE]

Places the Batch Processor in a sleep state, during which it
responds to most commands and resumes normal processing in
response to a CONTINUE command.

STA[TUS]

Prints a status report for this Butch Processor.

Managing the OPSER Spooling Package

A.~5.3

BATCH Startup Procedure
BATCH is most conveniently started by commands in the system startup command file. See the section "Creating and Using Command Files" in Chapter 3.

A.S.4

Operator Action Requests from BATCH
The BATCH program, in processing certain commands in a user command file,
generates requests for operator action. These requests come through OPSER
as action requests that require some operator action and a typed answer before
the Batch Processor continues processing. These action requests result from
commands requiring the operator either to mount and dismount volumes or to
respond to a special message with which a stall is involved. The operator answers
an action request with the ANSWER command through OPSER.
On a mount request, BATCH asks the operator to mount a device by an action
request with text in the following fonnat:
MOUNT xx: 'logical id'/'vid'/WRITE/NOWRITE/DEN:nn/PAR:nn
DEVICE?

If a specific volume is involved, its identification appears as text replacing 'vid.'
This text is the visual identification used to distinguish the volume.
Table A-16 lists the specific device type given by xx:. edt
Table A-16:
Designator

BATCH Device Type Designators
Meaning

CDII punched card reader

CRII high speed punched card reader or the CMII marked-sense
card reader

RP04, RP05 or RP06 disk pack drive

RK05 or RK05F disk cartridge drive

RK06IRK07 disk cartridge drive

RLOIIRL02 disk pack drive

RM02IRM03IRM05IRM80 disk pack drive

TU56 DECtape drive

All MSCP d.isks: RA60,lL\80, RA81, RA82, RD31, RD32, RD51,
RD52,RD53,RD54,RC25,RXa3,RXDO

RXOIIRX02 floppy disk drive

Any line pri.nter

TUI6, TEI6, TU45, or TU77 magnetic tape drive

TSII, TSV05, TK25, or TU80 magnetic tape drive

TMII, TUIO, TEIO, or TS03 magnetic tape drive

TK50 or TU81 magnetic tape drive

Any keyboard line

If the text of the action request indicates a specific volume, the operator should
mount the medium on a free unit of the device type specified by xx:. If a volume
is not involved, the operator should make sure a unit of the type specified by xx:
is ready and on line.
Managing the OPSER Spooling Package

A-41

When the unit is ready, the operator mu.st type the device specification (device
type and unit number followed by colon) as text in the ANSWER command for
that action request number. BATCH assigns that unit and continues processing
the user's queued request.
When BATCH is done with the device, it generates an action request to dismount
the volume or device. If the request involves a volunle, the operator must remove
the volume from the device specified and place the unit off line. To continue the
BATCH processing, the opera tor types the CONTINUE command as text in the
ANSWER command to OPSER for that action requel3t number. This response
tells BATCH to deassign the unit and to continue processing.

A.6 Operator Communication Program: PLEASE
The PLEASE program communicates directly with the operator services program
OPSER and, when OPSER is not running, sends text to the system console
terminal (KBO:). The program PLEASE is stored in ~the OPSER package account
OPSER$: with protection code of <232>.
Any user can run PLEASE, but only users with SWCFG privilege or users defined
as valid operators in the operator table can issue OPSER commands.
You, as the system manager or assigned operator, CSln run PLEASE to send
commands to OPSER. Users who do not have SWCFG privilege and are not valid
operators may run PLEASE to send text to the Operator Services Console.
NOTE

If you want to restrict PLEASE to only be run by users with WREAD

privilege, change its protection code to <124>.

A.6.1

Running and Terminating PLEASE
To run PLEASE, type the following command:
$ RUN OPSER$:PLEASE
PLEASE V9 RSTS V9 TIMESHARING

PLEASE prints its identification line and the number sign (#) character as a
prom pt to indicate it is ready to accept text.
PLEASE accepts one line of text terminated with the Return key. A line of text
exceeding the maximum length (255 characters) causes the program to print the
error message ?Line too long and reprint the prompt. If the first character of
text is not a slash (I) character, PLEASE tries to send the text as a message to
OPSER. If the first character of text is a slash, PLEASE tries to send the text as
a command to OPSER. To show the operator that it is sending the message or
command, PLEASE prints one of the following messages:
Message sent to 'OPSER'
Command sent to 'OPSER'

The program prints the # prompt again, after which the operator can type
another line of text. Entering CtrllZ in response to the prompt terminates the
program.

A-42

Managing the OPSER Spooling Package

When OPSER iF.: not active, PLEASE broadcasts messages on the system console
terminal; however, it does not send commands. In these situations, the program
notifies the user with one of two messages:
%OPSER not active - message broadcast to KBO:
%OPSER not active - command not sent

The message printed on keyboard unit 0 is preceded by the text:
%MSG for 'OPSER' but it is not active:

A.IS.2

OPSER Commands through PLEASE
Any user with SWCFG privilege or defined as a valid operator can communicate
with OPSER and with jobs on line to OPSER by typing a command through
PLEASE. The user denotes an OPSER command by typing the slash (/) character
as the first character in the line of text.
Table A-17 summarizes these PLEASE commands to OPSER. Note that the
circumflex (1\) character marks the location of a required space.
Table A-17:

PLEASE Commands to OPSER

Command Syntax

Meaning

ICHAI\KBn:

Changes operator services console to keyboard unit n.

IDELl\msgnumber

Deletes outstanding action request.

IDELI\#nn:n

Deletes n oldest action requests for job number #nn.

IDET

Detaches OPSER from the OSC.

lEX!

Terminates OPSER.

IINTl\rcvrid:text

Sends unsolicited text to online job.

ILISI\JO

Prints online job table.

ILISI\OP

Prints valid operator table.

1L0G"file;msglevel

Creates log file and sets message level.

1L0G";msglevel

Changes message level of current log file.

ILOG

Closes current log file and stops recording.

IMESl\msglevel

Sets message level for OSC.

lOPE 1\ KBn:[nnn,nnn]

Adds keyboard and account combination to valid
operator table.

10PEI\-KBn:[nnn,nnn]

Removes keyboard and account combination from valid
operator table.

IRETl\msgnumber

Prints associated message.

IRETl\m:n

Prints oldest m action requests for job number #n.

When OPSER receives a command from PLEASE, it makes sure that the user has
SWCFG privilege or that the user's keyboard and account numbers are present
in the valid operator table. If not, OPSER broadcasts the error message ?Invalid
oper on the terminal from which the command was made.
Commands sent by a valid operator can be executed by OPSER itself or by a
program on line to OPSER. If OPSER itself executes the command, any response
from the command is broadcast on the originating terminal. For example, because
OPSER itself processes the LIST JOBS command, the online job list generated
appears on the originating terminal.

Managing the OPSER Spooling Package

A-43

An INTERRUPT command, however, passes text to another program. Any
response from the other program is displayed on the ose and not on the
originating terminal. '1b see a message or response from an INTERRUPT
command, the operator should change the ose to the current keyboard (the line
on which PLEASE is running) before transmitting the! command. Action requests
and, messages (responses to commands or text sent to OPSER) are displayed only
if the current message level allows.

A.6.3

PLEASE as a CCl Command
If you install PLEASE as a eeL command, the operator can type PLEASE
followed by a text line to send commandEI to OPSER. PLE is the standard
abbreviation for the eeL command.

A.7 Terminating Operator Services and Spooling
Usually you terminate OPSER and programs on line to OPSER by running the
SHUTUP program. SHUTUP, however, terminates tilne-sharing operations. If
it is necessary to terminate operator services and spooling without shutting the
system down, the operator can type a sequence of cODlmands to stop spooling
operations in an orderly fashion. '1b understand the sequence, you must know
about OPSER shutdown levels.

A.7.1

OPSER Shutdown levels
When OPSER receives a command from SHUTUP to perform an orderly shutdown in logical end mode, OPSER sends an END com.mand to each online job in
a fixed, logical sequence. OPSER selects by shutdown level the order in which
jobs are ended. Online jobs at the lowest level terminate first. When all jobs
at one level have terminated and have been removed from OPSER'S internal
tables, OPSER begins sending END comrnands to job::! at the next highest level.
When all online jobs have properly terminated, OPSER kills itself and SHUTUP
proceeds with the system shutdown.

A.7.2

OPSER Manual Shutdown Procedure
You may need to shut down the SPOOL lprogram without shutting down the
entire system. The two possible procedures are as follows:
•

The operator can issue the END command to the spooler at any time. This
command causes the spooler to shut itself down the next time it is ready to
get a new job. Thus, if the spooler is printing a job, it completes the job's
output before shutting down. If SPOOL is waiting for a job, it shuts down
immediately. The operator can use the ABORT command to terminate the job
if the operator wants the job currently printing to end.
For example, to manually shut down the LPOSPL spooler in logical end mode,
use the command:
PLE/INT LPOSPL:END

A-44

Managing the OPSER Spooling Package

At the completion of the current job, the spooler kills itself. This shutdown
procedure merely mimics what OPSER does on command from SHUTUP. The
operator sends the END conlmand, in turn, to jobs at each shutdown level.
The output of the LIST JOBS command gives the shutdown levels for each
online job. When OPSER displays END messages for all jobs at one shutdown
level, the operator can send the END command to all jobs at the next highest
shutdown level. Mter the online job table is clear and OPSER has generated
END messages for all jobs, the operator can send the EXIT command to
OPSER.
•

A second way of shutting down the spooler is with an OFFLINE command.
The OFFLINE command immediately stops any job that is printing and
prints an ABORT message on the listing to indicate the action taken. The
OFFLINE command also causes SPOOL to shut down after the CUlTent job
finishes.
For example, to manually shut down the BAOSPL job immediately, use the
command:
PLE/INT BAOSPL:OFFLINE

BAOSPL aborts the current job, clears it from QUEUE.SYS, and kills itself.
If the operator wants to terminate a printing job for shutdown or any other
reason, the line printer must be on line. Under no circumstances does the
spooler consider a job completed until all the data that the spooler tries to print
(including termination messages) is actually printed. If the operator aborts the
job with the ABORT command, the spooler deletes (at the first opportunity)
whatever data is already buffered; the line printer must be on line to allow this.
If the operator wants no further output, use the REMOVE/JOB command to
remove the spooler job. In this case, no clean-up can be performed. This process
Jtnay be necessary when certain hardware fails (for example, if the line printer
interface does not raise the READY flag). It should only be used in a crisis.
The job that was being printed should be killed from the queue file by a QUEIK
command from the operator.

A.8 RESTOR as an OPSER Controlled Program
'rhe RESTOR program runs as a job on line to OPSER only after it has finally
detached from its terminal. When RESTOR is running detached, the operator can
send commands to it through OPSER. If, after having been detached, RESTOR is
later attached to a terminal again, the operator can communicate with RESTOR
either at its terminal or through the OPSER program.

Managing the OPSER Spooling Package

A-45

Table A-1S summarizes the commands that RESTOR recognizes. Note that the
circumflex (") character marks the location of a required space.
Table A-18:

RESTOR Commands through OPSER

Command Syntax

Meaning

ABO [RT]

Terminates the RESTOR run immediately and removes
RESTOR from OPSER internal tables.

CON[TlNUE]

Wakes up the RESTOR program to continue processing after
a PAUSE command or continuels processing after an operator
has performed some requested action.

DET[ACH]

Valid only under these conditions:
•
•
•

If RESTOR in attached
OPSER is running
The RESTOR listing file is not on the terminal to which
RESTOR is attached
•
No messages are currently pending for the specific
RESTOR job
If these conditions are met, RESTOR detaches and is given a
receiver identification of BACKnn where nn is its job number.
All further interaction with RESTOR is done through OPSER.

END

Finishes transferring the current file and terminates processing.

LAS[T]

At the OSC, prints the most recent message generated by this
RESTOR job.

LEG[AL]

At the OSC, prints a list of commands that can legally be given
for this phase of the RESTOR run.

NOT[ICE]"text

Inserts the specified text in thEl RESTOR listing file. Precedes
the text with the heading NOTICE FROM OPERATOR.

PAU[SE] STA[TUS]

Places RESTOR in a sleep state, during which it responds
to most commands, continues normal processing in response
to a CONTINUE command, Blnd terminates processing on an
ABORT command. Prints a status report for this RESTOR job.

While RESTOR is running detached, it generates all of its requests for operator
interaction through OPSER as messagel3. The operator uses the INTERRUPr
command to respond to such requests. See the RSTS / E Utilities Reference
Manual for the proper responses to RESTOR requests.

A-46

Managing the OPSER Spooling Package

Appendix B

Number Conversion
Many applications require a number based on bit values in a PDP-II word.
Table B-1 lists the octal and decimal values for each bit in the PDP-II word.
Table B-1:

Bit Values In the PDP-11 Word

Bit Number

Octal Value

Decimal Value

2
3
4
5

4
10
20

4
8
16
32

6
7
8
9

40
100
200

400
1000

128
256
512

2000

1024

4000

2048

12
13

10000
20000

40000

4096
8192
16384

100000

32768

Number Conversion

8-1

Appendix C

Disk Device Sizes
Table C-llists the device cluster size and device size (in 512-byte blocks) for ead
disk that RSTSIE supports. All values are in decimal.

Disk Device Sizes C-"

Table C-1:

Disk Device Sizes

Disk
Type

Minimum
Device
Cluster Size

RX50
RX33
RK05
RK05F
RLOI
RL02
RD51
RD52
RC25
RK06
RD31
RK07
RP02
RP03
RD32
RM02
RM03
RD53
RP04
RP05

1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
4
4
4
4
4

RA80
RM80
RD54
RP06
RA60
RM05
RA81
RA82

4
4
8
8
8
8
16
32

Total Available Size 1
(Blocks)

800
2400
4800
4800 2
10220
20457
21599
60479
50901 2
27104
41559
53761
39998
79998
83204
131648
131648
138668
171796
171796
237208
242572
311200
340664
400168
500352
891056
1216640

lSize when disk is initialized at lowest acceptable cluster size
2For each unit; 2 units per drive

C-2

Disk Device Sizes

Appendix D

General RSTS/E Messages
This appendix lists many of the general messages that you may encounter when
using RSTSIE. It is not a complete listing. Table D-l also indicates the probable
cause of the error in your use of a command or system program.
If you perform DECnetlE operations, you may receive messages that are
described in DECnetiE documentation but not in this appendix. In addition, the
utility programs that support DCL can produce messages not all of which are
listed here. Those messages are listed in the RSTS / E Utilities Reference Manual.

D.,1

Special Characters Used in Error Messages
The question mark (?), double question mark (??), and percent sign (%) characters
preceding a message indicate its severity. A single question mark precedes
standard error messages, which means that the command failed to do what you
asked but you can continue. For example:
?Too many parameters

In this case, the command failed because you included too many parameters.
A double question mark precedes severe error messages, which means that the
command failed to do what you asked and you cannot continue. For example:
??Fatal system I/O failure

By contrast, the percent sign identifies a warning message, which means that the
command may not have worked as you intended. For example:
%Logical name has not been assigned

In this case, the MOUNT comrnand worked but no logical name was assigned to
the device.
Informational messages do not have a preceding character; they provide extra
details about the effects of a command. For example:
Queue file being reorganized - please wait ...

This message informs you that, as a result of your command, the file is being
reorganized.
Angle brackets «
» surrounding text indicate a place holder for what the
system inserts when an error message occurs.

General RSTS/E Messages

0-1

Table D-llists the messages that you can get when using RSTSIE and the
probable causes.
Table 0-1 :

General RSTS/E Messages

Message

Meaning

? Account or device in use

The device cannot be mounted or dismounted because it is open or has
one or more open files. This elTor can also occur when a user attempts
to access a non-shared disk from a job other than the one specified in the
INOSHARE=n command.
? Additional argument required

You did not specify enough arguments in the DCL function.
? Additional qualifier required

You did not include a qualifier required in a command.
%All volumes must be of the same media. type

You attempted to write a Backup set on both disk and tape.
? Ambiguous keyword

You abbreviated a keyword to ahort; DCL cannot distinguish it from other
keywords.
%An unrecognized character was typed try again
u

There were extra characters in the command line that BACKUP did not
recognize as meaningful.
? Argument not allowed

You used an argument with a qualifier that does not accept one.
? Argument required

An argument was not supplied with a qualifier which required one. For
example, ISINCE=, without an argument, would generate this error.
The user did not supply an argument to a DCL function that required one.
??Bad directory for device

The directory of the device referenced is in an unreadable format. For
example, the tape format differs from the format you specified with the
MOUNT command.
%Block header CRC error detected for block

Although this block was read Imccessfully by the source device, BACKUP
discovered that there was an error in the block. BACKUP will attempt to
recover the data using recovery data if possible.
?Blocksize must be a multiple of <16 or 1)12>

An invalid blocksize was specified. Blocksize must be a multiple of 16 for
tape and 512 for disk.
?Can't access terminal<error text>

An error occurred while attempting to access the terminal to send the setup
file using the SET TERMINAL command.
?Can't find file or account

Either the account or file does not exist, or you typed a file specification
incorrectly.

(continued on next page)

0-2

General RSTS/E Messages

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Can't mount a private disk as public

You tried to mount as public (using the MOUNT command's /PUBLIC
qualifier) a disk that was initialized as private.
?Can't rebuild disk because device is write protected

A user with MOUNT privilege attempted to rebuild a disk (using the
MOUNT command's /REBUILD qualifier), and the drive is write protected
Therefore, the mount and rebuild operations fail.
?Command not installed

You typed a DCL command that has not been installed on your system.
The DCL keyboard monitor could not find the utility needed to carry out
your command. This message can also be displayed if you type a remote fi
specification in a command that allows remote file specifications, but your
system does not have DECnetJE.
?Command requires<privilege> privilege

The user does not have the appropriate privileges to complete an operatior
?Command too long

The length of the command, including continuation lines, exceeds the
maximum allowed for the command. Or, the length of the translated
command string exceeds the maximum allowed for the command. For
example, this message occurs with the LINK command in either of two
cases:
•
•

The text you typed in response to the $ prompt and the Files: or Root
files: prompt added up to more than 127 characters after translation.
With LINK/COBOL, the text was longer than 80 characters after
translation.

%Command will proceed as requested

The user's symbol tables could not be written out to the DCL work file, bu
the command will be processed. The user may be over quota. Corrective
action should be taken before the user continues processing.
?CoDlma required

A DCL function required two or more arguments and they were not separated by a comma,
?Conflicting arguments

Multiple arguments to a qualifier conflict. For example, both an AFl'ER
date and an exact date were specified to the /CREATED qualifier.
?Conflicting elements

You specified qualifiers or parameters in a command that do not agree in
type (for example, a print qualifier with a batch server).
?CPU limit exceeds queue's maximum

You specified a /CPU_LIMIT value in a SUBMIT command larger than thE
maximum allowed for the queue.
?Data error on device

One or more characters may have been transmitted incorrectly due to a
parity error, bad punch combination on a card, or similar error.

(continued on next pag«

General RSTS/E Messages

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Data error or incorrect density

This message occurs for one of the following reasons:
•
•
•

You specified the incorrect density.
The data on the tape is damaged.
The tape drive is faulty.

?Default macro for KBn: does not exist

The /INQUIRE command could not recognize the terminal and there was no
default macro specified for this keyboard in the default file, TERDFL.SYS.
You should supply a default macro for the keyboard. Until that is done, the
keyboard's characteristics have to be set manually.
Density of <density> BPI not available, using <density:> BPI instead

The requested density was not available on the drive. BACKUP reports the
density that is being used.
?Device does not exist

The device name you specified does not exist on your system.
%Device hung or write locked
%Dismount will proceed as requested

You attempted to dismount a disk that has been physically dismounted.
However, the logical dismount will succeed.
?Device hung or write locked

Check the hardware condition of the device requested. Possible causes of
this error include:
•
•
•

A line printer out of paper
A disk drive being off line
A DB or DR disk is dual-ported and is not write-protected

%Device in use

You specified INOSHAREABLE in an INITIALIZE/SERVER command;
however, the print server's device is not available or no pseudo-keyboards
are currently available for the batch server. PBS allocates the device as
soon as it becomes available.
?Device is not a terminal

A device other than a terminal was specified. Only specify keyboards.
?Device is restricted

This message occurs when a user without DEVICE privilege attempts to
open files on a disk restricted to users with DEVICE privilege.
?Device must be disk

You specified a file on a nondisk device, where a disk device is required.
?Device must be disk or tape

The device specified was not a disk or tape, and only disk or tape is valid.
?Device name may not be specified

A device name is not valid at this point, but one was specified.

(continued on next page)

0-4

General RSTS/E Messages

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Device not available
The specified device exists on the system, but an attempt to allocate or use
it is prohibited for one of the following reasons:
•
•
•
•

The device is currently reserved by another job.
The device requires privilege for ownership and you do have the
necessary privilege.
The device or its controller has been disabled by the manager.
The device is a keyboard line for pseudo keyboard use only.

?Device not file-structured
An attempt was made to access a device, other than a disk drive or magnet]
tape drive, as a file-structured device.
?Device not write protected
You specified the INOWRITE qualifier when you tried to mount a magnetic
tape, but the device is not write protected. Write protect the device by
removing the plastic ring from the magnetic tape hub.
?Device oMine
You tried to use a tape or disk, but the device is off line.
%Device write protected
The magnetic tape or disk is protected against write access. Therefore, you
can only read files on the device, but cannot write (perform output) to the
device. If you want to write to the device, first write enable the device, next
dismount the device, and then mount the device again.
?Device write protected
You requested write access to a device that is write protected. Write enable
the device and then retype the command.
%Directory attributes verification failure for directory <directory>
The directory attributes stored in the Backup set are different from the directory attributes as stored in the directory structure. This is a verification
mismatch. This can happen if a user logs in or out during a backup.
?Directory does not exist
A file specification i.ndicates a directory that does not exist on the particular
private disk. Or, a wildcard directory specification failed to produce a matcb
on the disk specified. This error can occur only with private disks.
??Disk error during swap
A hardware error occurs when your job is swapped into or out of memory.
The contents of the job area are lost, but the job remains logged in to the
system and returns to DeL. Report such occurrences to the system manager.
Disk is being rebuilt - wait.••
This informational message is displayed when a user attempts to rebuild
a disk. The disk is logically mounted when the rebuilding operation is
complete.
?Disk is mounted non-shared
A user attempted to dismount a disk that was mounted as nonshared (by a
different job) by specifying /PUBLIC in the DISMOUNT command.

(continued on next page)

General RSTS/E Messages

0-5

Table 0-1 (Cont.):
Message

General RSTSJE MfJSsages

Meaning

?Disk is mounted private or non-shared

A user attempted to dif3mount a disk that was mounted as non shared (by the
same job) or private by gpecifying /PUBLIC in the DISMOUNT command.
?Disk is mounted public

A user with MOUNT plivilege attempted to dismount a public disk that
was mounted as public, without I3pecifying /PUBLIC in the DISMOUNT
command, so the dismount fails.
%Disk is mounted read-only

This warning occurs if you do not specify eith.~r read or write access
(INOWRITE or !WRITE) when mounting a disk initialized as read-only.
Therefore, you are granted read access only.
?Disk is NOSHARE to another job; can't verify pack-id

A user with MOUNT privilege attempted to mount a disk that is mounted
INOSHARE to another job.
%Disk is restricted and mounted non-shared, read-only

A user specified INOREBUILD (to suppress rebuilding) and INOSHARE
in the MOUNT command and the disk was dirty. Therefore, the disk
is mounted nonshared and restricted, and read-only access is granted.
("Restricted" means that access "is allowed only to users with DEVICE
privilege.)
%Disk is restricted and mounted private, lread-only

A user specified INOREBUILD (to suppress rebuilding) and ISHARE in the
MOUNT command and the disk was dirty. Therefore, the disk is mounted
nonshared and restricted, and read-only access is granted. ("Restricted"
means that access is allowed only to users with DEVICE privilege.)
?Disk needs rebuilding but device is write protected

A user with MOUNT privilege attempted to mount a dirty disk (and did not
specify IREBUILD or INOREBUILD), but the device is write protected and
the automatic rebuilding cannot be performed. Therefore, the mount fails.
?Disk needs rebuilding but you do not have MOUNT privilege

This message occurs when a user without MOUNT privilege tries to mount
a private disk that was not logkally dismounted. The system manager can
correct the situation by rebuilding the disk.
?Disk needs rebuilding, so it can not be NORESTRICT

This message occurs when a user with MOUNT privilege tries to mount a
dirty disk with both INOREBUILD and INORESTRICT qualifiers.
?Disk pack is not mounted

The DISMOUNT command waf! attempted, hut the disk pack was not
mounted on the specified disk drive.
??Disk pack mount error

Fatal disk mounting error. The disk is corrupt and cannot be successfully
mounted with the MOUNT command.
?Dismount will proceed as requested

The disk was physically dismounted before the logical dismount was finished. RSTSIE will flag the device as dismounted anyway.

(continued on next page)

0-6

General RSTS/E Messages

Table D-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Division by zero

The expression specified by the user attempted to divide by zero.
?Do not specify file name or type

This message can occur with the first parameter of the ASSIGN command.
Remember to use a space between the first parameter (the string you assign)
and the second parameter (the name you assign it). For example, "ASSIGN
DR3: X", not "ASSIGN DR3:X".
%Dual-ported disk, may still be mounted on another system

You specified a DB or DR disk in a DISMOUNT command that is dualported. So, it may still be mounted on another system.
?Equal sign required

You used a qualifier that requires an argument, but you did not give an
argument.
?Error assigning Backup set device

The RSTSIE monitor reported an error in response to BACKUP's request
to assign the Backup set device. This message will be followed by another
message describing the problem.
?Error closing Backup set

The RSTSIE monitor reported an error in response to BACKUP's request to
close the Backup set. A message detailing the problem will follow.
?Error creating directory<directory>

The RSTSIE monitor reported an error in response to BACKUP's request to
create a directory. A message detailing the problem will follow.
?Error creating disk initialization command file

The RSTSIE monitor reported an error in response to BACKUP's request
to create the command file used to initialize disks. A message detailing the
problem will follow.
?Error creating dynamic region

The RSTSIE monitor reported an error in response to BACKUP's request to
create a dynamic region. A message detailing the problem will follow.
?Error creating job to initialize disk

The RSTSIE monitor reported an error in response to BACKUP's request to
create a job, used to initialize disks. A message detailing the problem will
follow.
?Error creating output directory

The RSTSIE monitor reported an error in response to BACKUP's request to
create a directory. A message detailing the problem will follow.
?Error deassigning Backup set device

The RSTSIE monitor reported an error in response to BACKUP's request to
deassign the Backup set device. A message detailing the problem will follow.
?Error dismounting disk to initialize it

The RSTSIE monitor reported an error in response to BACKUP's request
to dismount a disk so that it can be reinitialized. A message detailing the
problem will follow.

(continued on next page)

General RSTS/E Messages

0-7

Table 0-1 (Cont.):
Message

General RSTS/E Mes!sages

Meaning

?Error in forms definition<forms-name>
<error message>

An error, as the error message in the second line describes, was discovered
in the forms definition displayed.
?Error initializing Backup set
The RSTSIE monitor reported an error in response to BACKUP's request to
initialize the Backup set. This message will be followed by another message
detailing the problem.
?Error mounting output disk
This message indicates that thEl monitor has reported an error as a result
of BACKUP's mount request. It will be followed by an error detailing the
problem.
?Error mounting Backup set
The RSTSIE monitor reported an error in response to BACKUP's request to
mount the Backup set. A message detailing the problem will follow.
?Error number nn
An 110 error occurred while thEI system was attempting to retrieve an error
message. Possible causes could. be that the device containing the system
error file (ERR.SYS) is off line, or that the system error file contains a bad
block.
This is a serious error, and should be reported to the system manager.
?Error obtaining file flags for file <:ftIename>· for veriJication
The RSTSIE monitor reported :an error when BACKUP requested to find
out what the file flags for a file to verify were set to. This message will be
followed by another message describing the problem.
?Error opening disk non-:ftIe-structured to initialize it
<error message>
This message occurs when BACKUP is requested to initialize a disk and
the RSTSIE monitor is unable to open it in non-file-structured mode. This
generally happens because a disk unit is off line or no disk is mounted
in the drive. This can also occur for other reasons, such as no access to
non-file-structured disks or other device spedfic errors.
?Error opening :ftIe <filename> for input • skipping rest of directory
The RSTSIE monitor reported an error in response to BACKUP's request
to open a file in the directory (note that sinc!e the filename is unknown at
the time, this may contain wildcards). A mElssage detailing the problem
will follow. Since BACKUP scans the directory sequentially, the rest of the
directory will be skipped.
?Error opening file <filename> for output
The monitor reported an error to BACKUP when it was trying to create a
file. This message will be followed by a mesilmge describing the problem.
?Error opening file <filename> for verification
The RSTSIE monitor reported an error in rElSpOnSe to BACKUP's request
to open a file for verification. This message will be followed by another
message detailing the problem.

(continued on next page)

0-8

General RSTS/E Messages

Table [)-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?En-or opening PBS file <file-spec>
<en-or message>
An error, as the error message in the second line describes, occurred opening
the PBS file indicated.
?En-or opening Backup set
The RSTSIE monitor reported an error in response to BACKUP's request to
open the Backup set, Another message will follow, describing the problem.
?En-or opening Backup set non-file structured
The RSTSIE monitor reported an error in response to BACKUP's request to
open the Backup set device for control purposes. An error follows which will
detail the problem.
?En-or performing PPN lookup
The RSTSIE monitor reported an error in response to BACKUP's request to
look up PPNs to back up. A message detailing the problem will follow.
?En-or reading accounting data for directory <directory>
The RSTSIE monitor reported an error in response to BACKUP's request to
read the accounting data for a directory. A message detailing the problem
will follow.
?Error reading directory attributes for directory <directory>
The RSTSIE monitor reported an error in response to BACKUP's request to
read directory attributes. A message detailing the problem will follow.
?En-or reading file attributes for file <filename>
The RSTSIE monitor reported an error in response to BACKUP's request
to read file attributes for a file being backed up. A message detailing the
problem will follow.
?En-or reading file <filename> block <block>
The RSTSIE monitor reported an error in response to BACKUP's request to
read file data. A message detailing the problem will follow.
?En-or reading output volume label
The monitor reported an error to BACKUP in response to a request to
determine the volume label of the output device. This message will be
followed by a message detailing the problem.
?Error reading Backup set
This message indicates that an error was reported by the monitor to
BACKUP when attempting to read the Backup set. It will be followed by a
device specific error message describing the problem. In addition, it may be
followed by other messages if BACKUP is unable to recover from the lost
data.
?Error reading Backup set attributes
The RSTSIE monitor reported an error in response to BACKUP's request to
read the Backup set attributes. An error follows which will detail the

(continued on next page)

General RSTS/E Messages

0-9

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Error reading Backup set volume label
The RSTS/E monitor reported an error in res ponse to BACKUP's request to
read the Backup set label. A message detailing the problem follows.

?Error restoring file dates for file <filename>
The RSTSIE monitor reported an error in response to BACKUP's attempt
to restore file dates. Another message describing the problem follows this
message.

?Error rewinding Backup set
BACKUP encountered an error when asking the monitor to rewind the
Backup set. A message detailing the problem follows.

?Error setting end-of-volume Hag
The RSTSIE monitor reported an error in response to BACKUP's request to
set the end-of-volume flag. A message detailing the problem follows.

?Error setting file Hags for file <filename:~
The RSTSIE monitor reported an error in reslponse to BACKUP's attempt
to set file flags (caching, proteC1ted, etc.). Another message describing the
problem follows this message.

?Error setting RTS name for file <filenaDlLe>
The RSTSIE monitor reported an error in response to BACKUP's attempt
to set the runtime system name for a file. Another message describing the
problem follows this message.

<error text> opening setup file <filename>
An error occurred while trying to gain access to the setup file.

<error text> opening speed file
An error occurred while trying to gain access to the speed file. The speed is
not set and the rest of the SET TERMINAL c~ommand is executed.

<error text> opening/reading default file
An error occurred while procesBing the default file.

<error text> reading setup file <filename:>
An error occurred while processing the setup file.

<error text> reading speed file
An error occurred while processing the speed file. The speed is not set and
the rest of the SET TERMINAL command is executed.

?Error verifying directory attributes for directory <d,irectory>
The RSTSIE monitor reported an error in response to BACKUP's request to
read directory attributes for verification. Another message describing the
problem follows this message.

?Error verifying directory <directory>
The RSTSIE monitor reported lln error in response to BACKUP's request to
obtain directory information fo:r verification. Another message describing
the problem follows this message.

(continued on next page)

0-10

General RSTS/E Messages

Table D-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Error writing directory attributes for directory <directory>
The RSTSIE monitor reported an error in response to BACKUP's request to
write directory attributes. Another message describing the problem follows
this message.
?Error writing file attributes for file <filename>
The monitor reported an error to BACKUP when it was trying to restore file
attributes. Another message describing the problem follows this message.
?Error writing Backup set
The RSTSIE monitor reported an error in response to BACKUP's request to
write to the Backup set. Another message describing the problem follows
this message.
?Error writing Backup set attributes
The RSTSIE monitor reported an error in response to BACKUP's request to
write the Backup set attributes. Another message describing the problem
follows this message.
?Error writing to disk initialization command file
The RSTSIE monitor reported an error in response to BACKUP's request
to write to the command file used to initialize disks.Another message
describing the problem follows this message.
?Error writing to list file
The RSTSIE monitor reported an error in response to BACKUP's request to
write to the list file. Another message describing the problem follows this
message.
?Extra file data found in file <filename> • ignoring it
RESTORE found file data blocks for blocks beyond the written end of
file. If this happened with a SAVE set produced by RSTSIE and was not
accompanied by other messages, submit a SPR with a copy of the Backup
set.
??Fatal system 110 failure
An I/O error has occurred on the system level. The results of the last
command are unpredictable. This error is caused by a hardware condition.
Report such occurrences to your system manager.
?File attributes for file <filename> may be incorrect
If BACKUP has used recovery data to restore a file, there is a chance that
the file attributes may not be correct. If this is a possibility, BACKUP
informs the user with this message.

?File data block for file <filename> duplicated in Backup set
RESTORE found a file block duplicated in the Backup set. If this this
happened with a SAVE set produced by RSTSIE and was not accompanied
by other errors, submit a SPR containing a copy of the Backup set.
?File data block for file <filename> missing in Backup set
RESTORE found a file block missing in the Backup set. If this was not
accompanied by other errors, submit a SPR containing a copy of the Backup
set.

(continued on next page)

General RSTS/E Messages

0-11

Table 0-1 (Cant.):
Message

General RSTS/E Messages

Meaning

?File data in file may be incorrect

This message can occur if there are several missing blocks from a Backup
set. It should be accompanied by other messuges, describing the cause of
the problem. If not, submit a SPR with a copy of the Backup set.
?File does not exist

This message occurs for one of the following ]'easons:
•
•
•

An input file that must be present is not present.
A wildcard file specification does not match any files.
A file is not accessible because of its assigned protection code.

File <filename> not replaced

This is an informational message, which reports that a file was not restored
because there was a file of the game name already existing and you did not
use the !REPLACE qualifier.
File <filename> restored as non-contiguous

The RSTSIE monitor was unable to locate contiguous space to restore a
contiguous file as contiguous.
?File name attribute invalid - skipping file
If this Backup set was written by RSTSIE BACKUP, submit a SPR containing a copy of the Backup set. If this Backup set was written by any other
operating system, it indicates that the file name could not be mapped into
a RSTSIE name, and was skipped. Rename the file on the source system to
something which will fit into the RSTSIE file! naming scheme.

?File name attribute missing - skipping file

Submit a SPR with a copy of the Backup set.
?File name required

A file specification you typed does not includ·e a file name, but one is needed.
?File specification required

This error occurs if you specify node:: without a file specification (except with DIRECTORY). It also occurs if you have two commas with no
characters between them in a file specification list.
?Files cannot be on different nodes

This message occurs with network operations. Each input file specification
you include in a network command must be on the same node.
?Form <form-name> does not exist

The form name you specified was not defined by the system manager. See
your system manager to find out what form names are available.
?Forms De:flnition File does not exist

The forms definition file, PBS$:FORMS.SYS does not exist. See your system
manager.
?Forms not defined [for server <server-name>]

The forms you specified could not be found lin the forms definition file. See
your system manager.

(continued on next page)

0-12

General RSTS/E Messages

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

-------------------------------------------------------------------------%ID label ignored
You mounted a tape in DOS format and specified an ID label. Identification
labels are not encoded on DOS tapes; therefore, the label you specified is not
recognized by the MOUNT command.

%ID label should be specified when you mount an ANSI tape

You mounted a tape in ANSI format, but did not specify the tape identification label. DIGITAL recommends that you specify the identification label.
Thus, the MOUNT command can verify that the tape you selected has been
mounted.
?ID labels don't match

The identification label you specified does not match the identification
label encoded on the tape. To correct the problem, specify the correct
identification label. If you do not know what identification label is encoded
on the tape, you can omit the ID label from the MOUNT command.
?Illegal byte count for 110

The range of memory starting at the load address given is not available. Refer to the memory status report of a display program (SYSTAT or
DISPLY) to select an available range of memory. Note that you can also use
the SHOW MEMORY command.
?Illegal filename

The file name given in the command contains characters other than alphabetic or numeric characters.
?Illegal switch usage

A eeL command contains an error in an otherwise valid eeL switch
(qualifier). For example, you cannot use the /SI:n switch without a value for
n or a colon; or you cannot specify more than one of the same type of eeL
switch.
?Illegall value • n

The swap file number specified is not 0, 1, or 3. (The swap file 2 already
exists on the system disk and can not be added.)
?Impossible density for this device

You tried to mount or initialize a tape, but specified a density not available
on the tape drive you used. To correct the problem, use another drive or
specify a different density.
?Inconooect density

You specified a density different from the tape's density.
?Incorrect density or uninitialized tape

This message can occur for the following reasons:
You tried to mou.nta tape that had not been initialized.
You specified an incorrect density. If the tape has not been initialized,
use the INITIALIZE command. Otherwise, specify the correct density.
If the tape has not been initialized, use the INITIALIZE command.
Otherwise, specify the correct density.
•
•

(continued on next page)

General RSTS/E Messages

0-13

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Invalid account
The account specified was not valid either containing invalid characters (for
example, [1\1\1\]) or an invalid format (for example, [1,2345]).
?Invalid argument
This error can occur when you use a qualifier in the form /qualifier=argument. The qualifier you used is spelled properly, and the equal
sign (=) or colon (:) is present; however, the argument is either missing or
syntactically invalid.
RSTSIE displays this error message when there is no more specific message
to describe the syntax error.
?Invalid at interactive level
You specified a command that is invalid at the interactive level. (For
example, GOTO LABEL)
?Invalid attribute length - attribute skipped
If you get this message, submit an SPR with a copy of the Backup set that
generated the error.

?Invalid attribute - attribute skipped
This message indicates that the version number of a record was higher than
the highest version known to BACKUP. If this error occurs trying to restore
a Backup set written by a newer version of BACKUP on an older version of
BACKUP, the operation may not work prope!rly. If the version of BACKUP
writing the tape is older or the same as the one doing the RESTORE, submit
an SPR with a copy of the Backup set that generated the error.
?Invalid BATCH command
Although the command is valid in interactive mode, the command is not a
valid BATCH command because it implies o'r requires interaction with the
user. (For example, INQUIRE" SET NODA'I~)
?Invalid CCL command
You used the CCL prefix followed by a command that is not installed as a
CCL command on your system.
?Invalid channel number
The 110 channel number specified was invalid. Channel numbers can range
from 1 to 13 on OPEN, READ and CLOSE eommands, and from 0 to 13 on
WRITE commands.
?Invalid character
You typed an invalid punctuation character.
?Invalid command
The command name you gave is not a DCL command, and is not defined
on your system as a CCL command. Or, the line begins with a punctuation
character rather than with a keyword.
?Invalid date
A date either has improper syntax, represents a nonexistent date (for
example, 30-Feb), or represents a date before 1970 or after 1999.

(continued on next page)

D-14

General RSTS/E Messages

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Invalid density. n
You tried to mount or initialize a tape, but specified a density other than
800 'or 1600 bpi. The n is the density you specified.

?Invalid device
The device specified could not be found in the list of standard devices or in
the user-defined macros (TERMAC.SYS). Specify a valid device.

%lnvalid directory attribute record. skipping
This message indicates that a directory attribute code found in a Backup set
was unknown to BACKUP. If this error occurs trying to restore a Backup
set written by a newer version of BACKUP on an older version of BACKUP,
the operation may not work properly. If the version of BACKUP writing the
tape is older or the same as the one doing the RESTORE, submit an SPR
with a copy of the Backup set that generated the error.

?lnvalid entry name
The entry name you specified is invalid.

?Invalid expression
This covers a range of expression problems. The most common of these is an
expression ending in an operator without supplying a second operand (for
example, A = B *).

?Invalid file specification
A local file specification has improper syntax.

?Invalid fill factor
A fill factor was spedfied that was less than 0 or greater than 6.

?Invalid form definition
The definition of the specified form contains an invalid keyword or keyword
argument.

?Invalid form name
The form name you specified is either longer than six characters or consists
of one or more nonalphanumeric characters.

?lnvalid function
The function name specified after F$ is invalid. The minimum abbreviation
point is not met for the function (for example, F$LE is not valid because of
F$LEFT and F$LEN).

%lnvalid keyboard numbers n , n in default file
The numbers you specified in the default file were less than 0, greater than
127, or the first number of the range is greater than the second number of
the range.

%lnvalid keyboard numbers n , n in speed file
The numbers you specified in the speed file were less than 0, greater than
127, or the first number of the range is greater than the second number of
the range.

(continued on next page)

General RSTS/E Messages

0-15

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Invalid keyword

The keyword is not recognized. This error oc,curs with keywords that are
not qualifiers and are not command names. (For example, it can occur with
the options of SET and SHOW, and with qualifier values that are keywords.)
?Invalid label

The label specified in the command procedure contains invalid characters.
?Invalid node name

The node name field of the file specification contained invalid characters.
?Invalid operation on system disk

You specified an operation on the system disk which is invalid (for example,
DISMOUNT).
?Invalid operator

The operator you specified in the expression is not a defined logical, arithmetic, string, arithmetic comparison or string comparison operator.
?Invalid PPN

The project-programmer numbc~r (PPN) you specified does not have valid
syntax.
?Invalid print device

The device you specified is not a valid print device because it is not a line
printer or terminal.
?Invalid private delimiter

The delimiter you specified was not in one of the correct formats.
?Invalid qualifier

The qualifier keyword is not valid in the command you typed. (This message
may indicate an error in spelling or typing.)
?Invalid qualifier for disk

This message occurs when you specify an invalid qualifier for disks. For
example, IFORMAT=ANSI (applies only to magnetic tapes).
?Invalid qualifier for tape

This message occurs when you specify an invalid qualifier for tapes. For
example, !PRIVATE (applies only to disks).
?Invalid queue name

The queue name you specified is invalid.
?Invalid record length in Backup set· record skipped
If you get this message, submit an SPR with a copy of the Backup set that
generated the error.

%Invalid Backup set attributes· can't use it

The Backup set does not have the attributes correct for a BACKUP Backup
set.

(continued on next page)

0-16

General RSTS/E Messages

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Invalid server name
The server name you specified is invalid.
?Invalid speed
The speed specified was not a valid speed for the interface of the terminal.
Only specify speeds that the interface supports.
?Invalid symbol name
You specified a symbol name that contains invalid characters.
?Invalid time
A time either has improper syntax or represents a nonexistent time (like
25:00 or 13:00PM).
?Invalid width
You specified a width that was less than 1 or greater than 254.
?Invalid with network file specification
You gave a network file specification in one of the commands that accepts
them (RENAME, COPY, and so on), but you also gave a qualifier that can be
used only with local operations.
?IIO to detached keyboard
This message can result from one of two actions:
•

You tried to perform I/O with a terminal line that is used for dial-up
terminals, but nobody was dialed in.
• Your job became detached (perhaps because you were dialed in and your
line was later hung up) and then tried to perform 110 with the terminal.
The second situation either causes the job to hibernate or causes this error
condition, after which the job hibernates. You see this message when you
subsequently attach to the job.
?Keyword required
You typed nonalpha,numeric characters when a keyword is needed instead.
(If you type alphanumeric characters without a valid keyword, you receive
the error message, 'lInvalid keyword.)
?Label already defined
A label was defined more than once in an indirect command procedure.
?Label too long
The label you specified in the indirect command procedure was longer than
255 characters.
?Log fde already open
A log file was already currently open when you attempted to open another
log file using the OPENILOG_FILE command.
?Log file not currently open
You attempted to enable or disable logging using the SET LOG_FILE
command, and a log file did not exist.
?Log file print queue closed
The print queue required for a SUBMIT command's ILOG_QUEUE qualifier
is closed or marked for deletion. Use a different print queue for the log file.

(continued on next page;

General RSTS/E Messages

0-1 i

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Log file print queue does not exist
The print queue you specified with the /LOG_'QUEUE qualifier in a SUBMIT
command does not exist. Specify a different queue.
%Logical name has not been assigned
This warning can be displayed for one of two reasons:
•

•

You had INSTAL privilege and specified ;a logical name that was not
assigned as a system logical. (This means that neither the alternate
logical name nor the pack-id label were assigned.)
You did not have INSTAL privilege and Elpecified an alternate logical
name when attempting to mount a disk..

The mount succeeds in both cases, but the logical name is not assigned.
?Logins are disabled-please try again la,ter
This message can be displayed when you try t.o log in, for one of two reasons:
• The system is full, so it cannot accept additional users.
• The system manager has disabled logins.
(Possibly, logins are disabled because the system will be shut down shortly.)
?Magtape record length error
When performing input from magnetic tape, the record on tape was found to
be longer than the buffer designated to handle the record.
?Magtape select error
When access to a magnetic tapo drive was attempted, the selected unit was
found to be off line. This error can occur when you transfer data to or from
a tape.
?Map or executable file required
With LINK, you specified INOEXECUTABLE and did not specify a map file.
??Maximum memory exceeded
This is a nonrecoverable RSTS/E error caused by the following conditions:
•
•

While loading a program into memory, the job's private maximum
memory size was reached.
While executing a program, the system required more memory for string
or 1/0 buffer space, and the job's private maximum memory size or the
system maximum was reached.

?Missing closing apostrophe
The user did not specify matching apostrophes (' ... ') when requesting
apostrophe substitution.
?Missing closing bracket
This error can occur in a local or remote file specification. There is a left
bracket ([) or left angle bracket. «), but no right bracket (]) or right angle
bracket (».

(continued on next page)

0-18

General RSTS/E Messages

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Missiing closing quote

A quotation mark (") is not matched with another quotation mark. (This
error can occur in remote file specifications.)
?Missiing device or file name

The command must contain either a device specification or a device and file
name specification. If you use the ISI:n switch, a file name must be present.
?Missing open parenthesis

You did not specify an open parenthesis [(] when one was expected.
%More than 16 speeds specified for <terminal>

While processing the TERSPD.SYS file, a terminal was found that had mor~
than 16 speeds specified. Only the first 16 speeds are used.
?MOUNT privilege is required

You attempted an operation that requires MOUNT privilege. For example, l
user without MOUNT privilege can mount only private disks.
?MOUNT privilege required to dismount a public disk

A user without MOUNT privilege tried to dismount a disk initialized as
public. A user without MOUNT privilege can dismount only private disks.
?MOUNT privilege required to rebuild a disk

This message occurs when a user without MOUNT privilege attempts
to rebuild a private disk with the !REBUILD qualifier of the MOUNT
command. Have your system manager rebuild the disk.
?Name or account now exists

You attempted either to COPY to or RENAME an existing file. This error
can occur with RENAME if you do not specify !REPLACE and the output
file already exists. It can also occur with COPY if you specify INOREPLAC]
and the output file already exists.
?Network node names must be the same

Different node names were specified on input file specifications.
?No buffer space available

The system is overloaded and cannot complete your command because sma1
buffers are currently unavailable. Try the command again later.
?No channels available

An 110 or DCL command file channel is not available. You may have issuec
the at (@) command or an F$SEARCH or OPEN command, but a channel il
not available.
?No default {Print,Batch} queue

You did not specify a queue name with the PRINT or SUBMIT command,
and no default queue exists.
?No default print queue for log file

The /LOG_QUEUE qualifier in a SUBMIT command was specified without
an explicit queue name; however~ no default print queue exists. Use an
explicit queue name for the log file.

(continued on next page

General RSTS/E Messages

0-1

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?No file name or type permitted
A device:ppn syntax was expected and you supplied a full file specification
which included either a file name or type.
?Non-executable file
This error occurs if the file you nre trying to run is a source file; for example,
.BAS. You need to compile and link the file before you run it. (Note that an
executable file includes the value 64 in its protection code.)
?Non-printable character
You typed a control character.
??Non-res run-time system
This message generally indicates hardware problems. For example, the
run-time system referenced has not been loaded into memory and cannot be
loaded for some reason, and is therefore nonrosident.
%No owner rights to {print,batch} entry <entry-spec>
The entry shown is not your own and you do not have sufficient privilege to
delete it or modify it.
?NO prefix not allowed
You used the INO prefix improperly.
%No {Print,Batch} entry <entry-spec> foull1d
No entries matching the entry-spec you specified in a SHOW ENTRY or SET
ENTRY command were found.
%No {Print,Batch} queue <queue-name> found
No queues matching the queue name you specified in a SHOW QUEUE or
SET QUEUE command were found.
%No {Print,Batch} server <server-name> f'ound
No servers matching the server name you specified in a SHOW SERVER or
SET SERVER command were found.
?No qualifiers allowed
The user specified a qualifier on a command that does not allow qualifiers.
%No qualifiers are valid for RESTORE
You attempted to enter a qualifier to the RESTORE mount qualifier. No
qualifiers are valid.
?No (read,write,or read/write} access to fille <fIle-spec>- [by owner]
The file specifications in a PRINT or SUBMIT command requires access that
you or the specified owner does not have.

(continued on next page)

0-20

General RSTS/E Messages

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?No room for user on device

In attempting to create a file on a device, or write information to a device,
you exceeded some size limit. If it was a nondisk device, this error indicates
that the device was full. If it was a disk device, this error indicates one of
three conditions:
•
•

The disk as a whole is full.
You attempted to create a contiguous file, and there is not enough spaCE
on the disk. (However, the /CONTIGUOUS qualifier for the commands
COPY and CREATE produce only a warning if there is not enough
contiguous space.)
If you attempted to create a contiguous file but were unable to, try to
create a noncontiguous file of the same size. If you are able to create a
noncontiguous file t.hen you can conclude the problem is lack of contiguous
space.
•
If you have eliminated the first two conditions, then the disk directory
has reached its capacity. This capacity is independent of your disk
quota. The number varies, depending on the directory cluster size that
the system manager assigned when creating the directory and the sizes
of the files in it. Large files fill up a directory faster than small files,
especially if the files have small cluster sizes. Also, if the directory itselJ
has a large cluster size, it can hold more files and larger files.
?No run-time system

This error can occur if the program you are trying to run requires a run-timE
system that is not installed.
?Not a valid device

The device name that you gave is invalid for any of the following reasons:
•
•

It is not assigned as a system-wide or user logical name.
It is not a physical device name of any device that is installed on your

•

system.
The device name is valid, but not with this command. For example,
"INITIALIZE 'IT:" (you cannot initialize a terminal).

?Not enough available memory

An attempt was made to load a executable program that is too large to
run, given the job's private maximum memory size. Either the program
must be allowed to expand above a private maximum memory size, or the
system manager must increase the job's private memory size maximum to
accommodate the program.
?Not enough free memory to create dynamic region

There is not enough memory on the system to create a dynamic region
of sufficient size for the operation. If this is a BACKUP, reducing the
BLOCKSIZE or BUFFERSIZE parameters, or both, may correct the problem.
?Number not in range <low> or <high>

You typed a number where one is allowed, but the number is not in the valid
range. The valid range is from <low> to <high>.

(continued on next page)

General RSTS/E Messages

0-21

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Number too big

You typed a number where one is allowed J but the number is too large.
Refer to the command description to find out the largest acceptable value.
?Number too small

You typed a number where one is allowed J but the number is too small.
Refer to the command description to find out the smallest acceptable value.
?Pack-id labels don't match

The identification code for the specified disk pack does not match the
identification code already on the pack. This message can occur when you
try to mount or dismount a disk.
?Page limit exceeds queue's maximum

You specified a !PAGE_LIMIT vnlue in a PRINT command larger than the
maximum allowed for the queue.
?Parameter or argument too long

This message occurs if a file specification J text string J or qualifier argument
exceeds 255 characters.
%Please answer ''YES'' or ''NO''

Valid responses to the given prompt are YES and NO only.
%Please specify only a device name

You specified something other than a device name when only a device name
was valid.
?PPN does not exist

The project-programmer number (PPN) you specified as part of the job
specification does not exist.
?PPNneeded

The command you typed requirE~s that a PPN be specified.
Previous logical name assignment replaced

An informational message to indicate that an ASSIGN command redefined a
previously defined user logical.
{Print,Batch} entry <entry-number><entry-spec> created

Acknowledgment message to PRINT or SUBMIT command.
{Print,Batch} entry <entry-spec> - <error text>

For SET ENTRY indicates that an entry was not modified for the reason
stated in the error text.
J

?Print/Batch Services already started

You issued the START/QUEUEIMANAGER command; howeverJ the
PrintlBatch Services (PBS) package is already started.
?Print/Batch services not running

You issued a command that requires action by PBS; howeverJ the package is
not running.
{Print,Batch} queue <queue-name> assigIlled to server <server-name>

Acknowledgment message for ASSIGN/QUEUE command.

(continued on next page)

D-22

General RSTS/E Messages

Table D-1 (Cont.):
Message

General RSTS/E Messages

Meaning

{Print,Batch} queue <queue-name> deassigned from server <server-name>

Acknowledgment message for DEASSIGN/QUEUE command.
{Print,Batch} queue <queue-name> closed

Acknowledgment message for CLOSE/QUEUE command.
{Print,Batch} queue <queue-name> deleted

Acknowledgment message for DELETE/QUEUE command.
%{Print,Batch} queue <queue-name> - <error text:>

For SET QUEUE, indicates that a queue was not modified for the reason
stated in the error text.
{Print,Batch} queue <queue-name> marked for deletion

Acknowledgment message for DELETE/QUEUE command when the queue
you specified still has entries on it. The queue is deleted as soon as it
becomes empty.
{Print,Batch} queue <queue-name> opened

Acknowledgment message for OPEN/QUEUE command.
{Print,Batch} queue <queue-name> started

Acknowledgment message for START/QUEUE command.
{Print,Batch} queue <queue-name> stopped

Acknowledgment message for STOP/QUEUE command.
{Print,:Batch} server <server-name> {deleted,marked for deletion}

Acknowledgment message for DELETE/SERVER command.
%{Print,Batch} server <server-name> - <error text>

For SET SERVER, indicates that a server was not modified for the reason
stated in the error text.
{Print,Hatch} server <server-name> initialized [non-]shareable [with forms
<form-name> ]

Acknowledgment message for INITIALIZE/SERVER command.
{Print,Hatch} server <server-name> modified

Acknowledgment message for SET SERVER command.
{Print,Hatch} server <server-name> started

Acknowledgment message for START/SERVER command.
{Print, Batch} server <server-name> stopped

Acknowledgment message for STOP/SERVER command.
Print/Uatch Services started at <time>

Acknowledgment message when you issue the START/QUEUEIMANAGER
command to begin PrintlBatch Services.
PrintllJatch Services stopped at <time>

Acknowledgment message for STOP/QUEUEIMANAGER command if no
jobs are being processed.

(continued on next page)

General RSTS/E Messages

0-23

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

PrintlBatch Services will stop after {completing,aborting} <n> job[s]

Acknowledgment message for STOP/QUEUEIMANAGER command if any
jobs are currently in progress.
?Priority exceeds queue's maximum

You specified a priority for a PRINT or SUBMIT entry larger than the
maximum allowed for the queue.
? <privilege> privilege required

You typed a command that requires some privilege, and you do not have it.
??Program failure in <program-name>

This message reports a problem in the software. It is followed on the next
line by an explanation of the p"roblem. You should verify that the failing
program is correctly installed. If necessary, you should then submit an SPR.
The SPR should show the dialogue that preceded the message, the exact
text of the message, and a list of patches thBlt have been installed in the
failing program.
?Program PBS$:PBS.TSK does not exist

The PrintlBatch Services (PBS) program was not found when you issued the
START/QUEUEIMANAGER command.
?Protection violation

This error can occur for reasom~ similar to the following:
•

You typed a CCL command that your system manager has protected
against general use.
• You tried to run a program to which you do not have execute access.
• You tried to read a file to which you do not have read access.
• You tried to write to or delete a file to which you do not have write
access.
If this message occurs because of a protection violation with one of your
own files, you can use the SET PROTECTION command to change the
file's protection code. (However, this error can also occur because of other
conditions.)
%Public disk mounted as private

A user with MOUNT privilege mounted a diuk initialized as public without
specifying /PRIVATE, /PUBLIC, ISHARE, or INOSHARE. The system
mounts the disk as private.
?Qualifier confticts with file type

You specified the IAPPEND qualifier and the file has RMS attributes.
?Qualifier confticts with parameter

You typed a parameter and a qualifier that should not be present together. For example, with the DEASSIGN command, you specified the IALL
qualifier and a logical name.
?Queue already assigned to server

You issued an ASSIGN/QUEU]~ for a queue already assigned to the server
you specified. Use the SHOW QUEUEIFULL command to see which servers
are assigned to a queue.

(continued on next page)

0-24

General RSTS/E Messages

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Queue does not exist
You specified the name of a queue that does not exist.
?Queue is ~closed,deleted}
A PRINT or SUBMIT command entry could not be created on a queue
because the queue is closed or marked for deletion. Use a different queue.
?Queue is privileged
The queue you specified in a PRINT or SUBMIT command requires one
or more privileges that you do not have. Use the SHOW QUEUEIFULL
command to display the privileges that the queue requires.
?Queue manager message queue full
There is no room for additional messages to be sent to PBS at this time.
Reissue the request.,
?Quoted string expected
A quoted string or character was expected and was either missing or was
not enclosed in quotes.
%Record too long, line truncated
The record on a READ command was too long. The assignment occurred;
however, only the first 255 characters were processed.
?Reserved symbol name
The user attempted to define a local or global symbol which began with a
reserved prefix of $, F$ or f$, or delete a reserved global symbol from the
symbol table that began with the dollar sign ($).
?Backup set does not contain a, valid ANSI label
The Backup set to restore is not a valid ANSI tape.
%Backup set volume label is not <volume-name>
The Backup set volume name does not match the volume name specified by
the user.
?Server already exists
You issued the INI'rIALIZE/SERVER command to define a server that
already exists.
?Server does not exist
You specified the name of a print or batch server that does not exist. Use
the SHOW SERVER command to list the defined servers.
?Single character expected
A single character inside quotes was expected and more than one character
was supplied (for example, SET DATAlEND_OF_DATA = "$$").
?Some :file data in :file <:filename> may be missing
The Backup set did not contain the number of file blocks which it was
supposed to contain. If there were other errors, then this indicates that the
file is missing data. If there are no other errors, submit a SPR with a copy
of the Backup set.

(continued on next page)

General RSTS/E Messages

0-25

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Source device must be the same
Two different source devices were specified for the operation.
?Speed is not allowed for <terminal>
The speed specified is not one of the speeds allowed for this terminal in the
TERSPD.SYS file.
??Stack overHow
This message indicates a system problem. The system manager should send
in an SPR, giving the dialogue that preceded the message, the text of the
message, and a list of patches that have been installed.
?String too long
While in an expression, a string became too long to fit in the 255 available
bytes.
?Substitution too complex
The user's request for apostrophe substitution was rejected because the substitution was too complex (it reached the maximum number of substitution
iterations allowed on one command).
?Summary record not first record of Backup set
This problem indicates that the Backup set was not written correctly.
Submit a SPR including a copy of the Backup set.
?Symbol name conflicts with <symbol-name>
The user specified an assignment whose symhol name or abbreviation point
conflicted with an existing symhol definition in the same symbol table.
?Symbol name too long
The symbol name specified by the user exceeded 255 characters.
??Symbol table full
You attempted to define a label or symbol when the local or global symbol
table was already full. Note thnt if you are in a command procedure when
your symbol table gets full, the command procedure aborts.
?Syntax error
The command has improper syntax. This occurs when there is not a more
specific message describing the syntax error.
?THEN clause required
The action clause was missing on the IF <expression> THEN command or
on the ON <severity-level> THEN command.
?THEN keyword required
The THEN keyword was missing on the IF <expression> THEN command
or on the ON <severity-level> THEN command.
?This is not the correct output volume
The volume ID of the mounted output volume does not match the volume
name specified in the lOUT_LABEL qualifier'.
?Time limit exceeds queue's maximum
You specified a !TIME_LIMIT value in a SUBMIT command larger than the
maximum allowed for the queue.

(continued on next page)

0-26

General RSTS/E Messages

Table D-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Too many arguments

YQU used the nQtation Iqualifier=(arg,arg, ... ) with a qualifier that accepts
Qnly a single argument.
?Too many elements in list

In a list Qf file specificatiQns Qr Qther items (separated by CQmmas Qr plus
signs), yQU indicated mOore file specificatiQns than are allQwed. FQr example,
yQU exceeded Qne Qf the fQllQwing limits:
•
•
•

The DIBOL, RENAME, DELETE, and SET PROTECTION cQmmands
allQw six file specificatiQns.
The COPY cQmmand allQws six input file specificatiQns and Qne Qutput
file specificatiQn.
The PRINT and SUBMIT cQmmands allQw up tOo 11 file specificatiQns.

?Too many files or parameters

YQU specified tQQ many file specificatiQns in a PRINT Qr SUBMIT cQmmand,
Qr a parameter string in a SUBMIT cQmmand is too IQng to be handled by
PBS. Reduce the number Qf file specificatiQns Qr use a shQrter parameter
string.
?Too many logical names assigned

With the ASSIGN cQmmand, yQU exceeded the maximum number QflQgical
names. YQU can Qnly assign up tOo fQur IQgical names (Qnly three IQgical
names if any Qf the logical assignments includes a PPN).
'1Too many open files on unit

YQU specified the same magnetic tape Qr DECtape drive both as input and
Qutput files Qn COPY Qr APPEND.
?Too many parameters

This message Qccurs if yQU specify mOore cQmmand parameters than the CQmmand can accept. FQr example, yQU specified mOore than eight parameters
with the !PARAMETERS qualifier.
?Too many printers initialized

The device yQU specified CQuld nQt be initialized because the maximum
number Qf sPQQling devices has already been initialized.
?Unable to copy tape command file to disk

The user specified an indirect cQmmand file Qn magnetic tape, and SQme
errQr Qccurred when attempting to CQPy it to a temporary disk file fQr
prQcessing.
??Unable to create DCL work file
DCUs wQrk file fQr stQring the user's symbQI tables CQuld nQt be created Qn

disk. The user may be Qver qUQta. CQrrective actiQn shQuld be taken befQre
the user cQntinues prQcessing.
??Unable to read DCL work file

The user's symbQI tables CQuld nQt be read in frQm the DCL wQrk file. If
RSTSIE is prQcessing an indirect cQmmand file, then cQntrQI immediately
returns to the interactive level.

(continued on next page)

General RSTS/E Messages

D-27

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

?Unable to recover part of Backup set

This message indicates that BACKUP was unable to recover part of the
Backup set, after trying to use error recovery data. There will also be other
messages indicating what data. (if any) was detected to be missing.
?Unable to start PrintlBatch Services
<error message>

Some external condition, as the error message on the second line describes,
prevented PBS from being started.
??U~able to write DeL work file

The user's symbol tables could not be written out to the DCL work file. The
user may be over quota. Corrective action should be taken before the user
continues processing. If RSTSIE is processing an indirect command file,
then control immediately returns to the interactive level.
?Unbalanced parentheses

Parentheses do not match up (for example, A=«5)+6).
?Undefined label <label>

The label specified on the GOrrO command does not exist in the indirect
command procedure being executed.
?Undefined symbol

The symbol name specified is not defined in the global or local symbol tables.
?Unexpected character

A character was encountered in a qualifier or a parameter that was not
expected. For example, in SE'r TERMINALIWIDTH=80FOO, the "F" of
"FOO" would be an unexpected character.
?Unexpected error verifying file data foll." file <filenam.e>

The RSTSIE monitor reported an error in response to BACKUP's request to
verify data. Another message describing thEl problem follows this message.
?Unexpected error writing to output filE~ <filename>

The RSTSIE monitor reported an error when writing to the specified file.
Another message describing the problem follows this message.
?Unit number required

This message occurs when you specify a device-name without a devicenumber (for example, DM: instead of DMO:).
?Unsupported record structure level- record skipped

This message indicates that the version number of a record was higher than
the highest version known to BACKUP. If this error occurs trying to restore
a Backup set written by a newer version of BACKUP on an older version of
BACKUP, the operation may not work properly. If the version of BACKUP
writing the tape is older or the same as the one doing the RESTORE, submit
an SPR with a copy of the Backup set that generated the error.
?User macros nested too deep

There were more than 50 nested user macros or a user macro called itself.

(continued on next page)

0-28

General RSTS/E Messages

Table 0-1 (Cont.):
Message

General RSTS/E Messages

Meaning

%Valid qualifiers are /INITIALIZE and INOINITIALIZE

You entered a qualifier other than /INITIALIZE and INOINITIALIZE and
those were the only valid qualifiers.
'?Verification failure for directory information for file <filename>
There was a difference between the directory infonnation (file size, file flags,
etc) stored on the Backup set and in the directory structure.
'?Verification failure for file attributes for file <filename>
There was a difference between the file attributes stored on the Backup set
and in the directory structure.
?Verification failure for file data for file <filename>
The file data stored on the Backup set is different from the file data stored
on disk.
%Volume has not been initialized

The initialization request failed, so the output device is not in an initialized
state. BACKUP reprompts for an input device.
?Wildcard entry name not allowed
The command you specified does not allow wildcard entry names. Specify
the name of a single entry instead.
"lWildcards not allowed
You included a wildcard in a file specification, where wildcards are not
allowed.
?Wildcard PPN not allowed
The command you specified does not allow wildcard PPNs. Specify the name
of a single PPN instead.
'lWildcard queue name not allowed
The command you specified does not allow wildcard queue names. Specify
the name of a single queue instead.
%XOR block missing from Backup set
If this error occurs without any other messages, submit an SPR with a copy
of the Backup set. If there are other messages, this has probably occurred
because a block of the Backup set was unreadable.

%XOR verification failed· group may be unreliable
This message indicates that BACKUP, during a verify operation, concluded
that the recovery data for a group of data blocks does not match the data
found in the blocks. If this was not accompanied by other messages, submit
a SPR. It it was, consider the data in the Backup set to be unreliable.

General RSTS/E Messages

0-29

Appendix E

SAVRES System Program
The RSTSIE SAVRES system program is a disk backup and copy utility that has
opera tional functions:

E.,1

•

RESTORE recreates a disk from an archived SAVRES tape or disk.

•

IMAGE makes a copy of a disk.

•

IDENTIFY extracts label infonnation from a SAVRES volume or RSTS/E
disk.

When to Use SAYRES
SAVRES provides a nonselective, fast-volume image capability that requires few
operator responses during the operational dialogue. Unlike BACKUp, SAVRES
processes entire volumes only and does not allow selective file transfers. When
SAVRES finds bad blocks, it requires no operator intervention. For these reasons!
use the SAVRES program when you need to:
•

Crea te a fast, reliable image copy of an entire RSTS/E disk

•

Restore a previously archived SAVRES set.

E.2 Definitions of SAYRES Terms
To understand this appendix, you should know these tenns:
•

SAVE fonnat: The fonnat of the output written by a previous SAVE operation
and read by a RESTORE operation. These could only be created with versions
prior to V9.7.

•

SAVE Set: The set of magnetic tapes or disks created by a previous SAVE
operation. A SAVE Set must be composed entirely of disks or tapes; it cannot
be a combination of the two device types. However, you can mix different
drive types within one set, for example, two RK05's and one RK06.

•

SAVE volume: One of the magnetic tapes or disks of a SAVE Set.

SAYRES System Program

E-1

•

SAVE Set Name: One- to six-alphanmneric characters used to identify a
SAVE set. By default, the SAVE set name is the same as the Pack ID from
which it was created. However, you can specify another name for the SAVE
Set.

•

LIKE Disks: Like disks are units of the same device size (that is, the same
number of data blocks). SAVRES considers the following devices alike:
Two RP06 disks
An RM02 and an RM03
An RP04 and an RP05
An RK05 (RK05J) and one unit of an RK05F

Therefore, if you copied an RM02 to a SAVE Set, the SAVE Set can later be
restored either to an RM03 or another RM02.

E.3 Running SAVRES
You use the SAVRES program off line while running t.he system initialization
code INIT.SYS. In reply to the INIT OPTION prompt, type SAYRES. SAVRES
responds with the SAVIRES FUNCTION prompt. For example:
Option: SAVRES

DD-MMM-YY? 14-MAY-85
HH:MM? 12: 15

SAV IRE::; ./:!'unction:

NOTE

You must run SAVRES off line to copy (IMAGE) the system disk with
which you are currently running on line.
You can also use SAVRES on line, during timesharing. When you do, you must
not logically mount the RSTSIE disk(s) on which SAVRES operates. This ensures
the integrity of the data on the disk(s). 1b run SAVRES on line, type RUN
$SAVRES. SAVRES then prints its program prompt:
$ RUN $SAVRES<RET>
SAv/RES Function:

Mter the SAVIRES FUNCTION prompt appears, as a result of running SAVRES
either off line or on line, you are under the control of SAVRES and are ready for
restore processing.
Next, select from Table E-l one of the three operational functions:

E-2

•

RE[STORE]

•

IM[AGE]

•

ID[ENTIFY]

SAVRES System Program

Unless you need to exit SAYRES (by pressing Line Feed or Return, or entering
CtrllZ) or need further help (by typing HE[LP] or pressing the Return key),
type one of the three functions on your terminal in response to the SAVIRES
FUNCTION prompt. For example:
SAV/RES Function: RESTORE<RET>

This starts the SAVRES program, which immediately prints the first of several
dialogue questions. The section "SAVRES Dialogue" contains sample dialogues
for each of the operational functions.
Table F.,--llists the possible responses to the SAVRES prompt.
Table E-1:

SAYRES Functions

Function

Description

RE[STORE]

Recreates a RSTSIE file-structured disk from a SAVE Set.

IM[AGE:]

Copies a RSTSIE file-structured disk to a like disk.

ID[ENTIFY]

Prints label information and other volume characteristics of a
SAVE volume or a RSTSIE file-structured disk.

HE[LP]

Prints a table of SAVRES functions.

Exits the SAVRES program. If you run SAVRES with a RUN
command, pressing Return or Line Feed returns you to your
keyboard monitor. If you are running SAVRES off line with
INIT.SYS, pressing Return or Line Feed returns you to the INIT
OPTION prompt.

Displays the "Type HELP for help" message on your terminal.

Ctr1lC

Produces the same result as pressing the Line Feed key.

Ctr1lZ

Produces the same result as pressing the Line Feed key.

other

Produces the same result as pressing the Return key.

Your response to the SAVIRESFUNCTION prompt determines the type of
dialogue SAVRES performs. You can respond to the prompt in several ways:
•

With the function name, such as IMAGE or RESTORE, or by pressing a key
such as Return, as shown in Table E-l. If you enter only a function, the full
dialogue follows. See the section "SAVRES Dialogue."

•

With the function name and various switches. For example:
SAV/RES Function: RESTORE/VERIFY:l0-MAY-86

If you include switches with the function, the dialogue questions follow only
for the devices to be used and the unspecified switches. See the following
section.
•

With a full function command line. For example:
SAV/RES Function: IMAGE DM1:=DMO:MYPAK/VER/NOERR

If you enter a complete command line, SAVRES does not ask any dialogue
questions. See the section "Full Function Command Line."
SAVRES prints messages on your terminal to tell you its status. Some messages
are for information only; other messages are error conditions that may require
your attention during the dialogue, mount, transfer, or verification phases of
SAVRES.

SAYRES System Program

E-3

When the program finds an error during a dialogue or mount operation, it asks
you to correct the problem and to retype ,the appropriate command. Such errors
never affect the integrity of the data being transferred.
However, errors that occur while SAVRES is transferring or verifying data can
jeopardize data. In this case, SAYRES resolves the problem, aborts the run, or
asks you whether to continue or to abort the program. The program gives you
information to help you protect your data by printing informational messages and
error messages on your terminal.
Finally, SAVRES prints an optional summary report about RESTORE or IMAGE
operations at the end of the run. This report describes:
•

The operation performed

•

Device information

•

Start date and time statistics

•

Run total statistics, which includes the number of errors found during the
operation.

E.4 SAVRES Switches
You can include SAVRES switches in a function response or in a full function
command line. There are two types of switches:
•

The verification switch corresponds to questions in the SAVRES dialogue and
has default settings. If you do not include the switch in a function response,
the dialogue asks you for that infonnation.
If you include one of the switches in a function response, SAVRES does not
ask the corresponding question but does ask the question for the switch you
did not specify. For example, if you include the !EXPIRATION switch but not
NERIFY, SAVRES does not prompt you for an expiration date but does ask
the verification question.
If you do not include the switches in the full function command line, SAVRES

assumes the default settings. For example, if you do not include the NERIFY
switch, SAVRES assumes INOVERIFY. This means if you do not explicitly
include the NERIFY switch in a full function cornmand line, SAVRES does
not compare the data transfer.
•

E-4

The /STATS and /ERROR switches have no corresponding dialogue questions
but do have default settings. Therefore, if you do not specify these switches in
a function response or command line, SAVRES assumes the default settings.

SAVRES System Program

Table E-2 lists the allowable SAVRES switches
Table E-2:

SAVRES Switches

Switch

Default and Description

VE[RIFY]
NOVE [RIFY]

NOVERIFY
You can use this switch only for a RESTORE or IMAGE operation. If you specify the NE[RIFY] switch, SAVRES compares the
volume(s) to make sure that the transfer was accurate.

ST[ATS]
NOST[ATS]

STATS
SAVRES does not automatically print a summary report at the
end of a run, unless you specify the IST[ATS] switch with either a
function response or a full function command line. See the section
"SAVRES Summary Report."

ER[ROR]
NOER[ROR]

ERROR
If you do not include the !ERROR switch and a nonfatal error
occurs during a SAVRES run, SAVRES prints an error message an
continues processing. This is the default condition.
You must specify explicitly the INOERROR switch if you want
SAVRES to print an error message and terminate as a result of an
of the errors described in Table E-12.

E.S SAVRES Dialogue
The SAVRES dialogue begins after you run the SAVRES program and respond tc
the SAVIRES FUNCTION prompt. If you type RESTORE, IMAGE, or IDENTIF:
the program asks a set of dialogue questions. To give the default answers to thee
questions (printed in angle brackets), just press Return. If you press Return in
response to a question that has no default, SAVRES repeats the question.
There are two valid responses to a SAVRES program that can change its
operating status:
•

CtrVZ

•

CtrVC

If you enter CtrVZ at any point in the dialogue, SAVRES returns to the previous
question. A CtrVC response aborts the run immediately and returns you to
the INIT OPTION prompt if you are running off line under the control of the
INIT.SYS program, or to your keyboard monitor prompt if you are running
SAVRES on line. SAVRES automatically returns to the SAVIRES FUNCTION
prompt after the execution of a specified function when you are operating either
on line or offline. Exit from the INIT version of SAVRES to the OPrION promp
by pressing Line Feed or Return. Type CtrVZ to the SAVIRES FUNCTION
prompt to exit the online version of SAYRES.

SAVRES System Program

E-!

E.S.1

Device Specifications
SAVRES requires you to use device specifications as responses to several dialogue
questions. The general format is:
<device>[:[ <id> ]][ <switch(es»]
Device is a two-character device type (for example DU) followed by a one- or
two-digit unit number in the range 0 to 15. The identification can be either a
SAVE Set Name or a disk pack identification; the switehes are any of those from
Table E-3.
Table E-3:

E.S.2

SAVRES Device Specification Switches

Switch

Description

ISCR[ATCH]

The ISCRATCH switch is legal only on an output volume device
specification. If you use the /SCRATCH switch, SAVRES bypasses most volume label checking. The program always checks
the label to see if an output volume contains a SAVE Set written
at the current density It checks destination disks for bad block
information.

IDEN[SITY):[ 800]
[1600]
[6250]

The !DENSITY switch is legal only for magnetic tape. With this
switch, you can specify t.he density at which the tape is to be read.
If SAVRES cannot read the tape at the specified density, it tries to
read the tape at the other legal density setting.

Privileges Required
You need RDNFS, WRTNFS, and SYSMOD privileges to gain read/write nonfile-structured access to a disk. If a device is restricted, you also need DEVICE
privilege to gain access to that device.

E.S.3

Checking the Input Volume
SAVRES checks the input volume before transferring any actual data and warns
you of any problem it foresees in restoring or copying t.he volume. All of the input
volume error messages that are printed at this point are warning messages only;
you can recover from them by following appropriate procedures.

E-6

SAVRES System Program

Table E-4 lists and describes the error messages that result from a problem wit]
an input volume.
Table E-4:

SAVRES Input Volume Error Messages
Message and Meaning

%%% Input disk has only nn% free clusters.
Mount it anyway <NO>?

The input disk has very few clusters that are not allocated. There must
be at least as many good pack clusters on an output RSTSIE disk as there
are clusters to be transferred from the original input disk. This message
indicates that you may encounter problems copying the disk. Type NO or
press RETURN to return to the previous device prompt. If you type YES,
SAVRES proceeds with the IMAGE operation. The program also checks thE
output volume to see if the transfer can be made (see the section "Checkin~
the Output Volume"). At this point, you may want to free up space on the
input disk by deleting unnecessary files. This increases your chances of
completing a successful IMAGE.
%%% This is not the correct volume.

The pack ID or SAVE Set Name you included with the input device specification does not match the name already written on the volume. If SAVREE
finds that the names do not match, it prints the warning message and then
issues the input device prompt again. A dismount request follows the warn
ing message and lists useful information from the volume label of a SAVE
Set or a RSTSIE file-structured disk.
%%% This volume has no label.

The input volume is not a RSTSIE disk or a SAVE Set. SAVRES prints the
input device prompt again after issuing the warning message.
%%% Input disk should be rebuilt.

You removed a disk from a drive without logically dismounting it and then
attempted to remount it. Perform the clean operation with the ONLCLN
program or the REFRESH option of INIT.SYS. See the RSTS / E System
Installation and Update Guide for information on REFRESH. See the
RSTS / E Utilities Reference Manual for information on ONLCLN.

E"S.4

Checking the Output Volume
SAVRES checks the output volume before restoring or copying a RSTS/E disk
to make sure that each operation ends with the least chance of error. Mter
checking the output volume, SAVRES displays messages about the expiration
date, the label, and the available data space on the destination device. Some of
the messages are only for your information; the rest warn you of serious output
device problems.
When the output volume you are using is labeled a RSTS/E disk, SAVRES prints
the message:

*** 'I'he volume on dev: is a RSTS disk
*** Pack ID is <xxxxxx>
*** Pack will be reinitialized
Mount it anyway <NO>?

SAYRES System Program

E-j

If you type NO or press Line Feed or Return, SAVRES returns to the previous
output prompt. A YES response indicates you want to use the present volume.
If you restore (RESTORE) or copy (IMAGE) a disk, SAVRES checks whether
the total nwnber of clusters on the output disk minus the known bad blocks is
greater than or equal to the nwnber of allocated clusters on the original source
disk. If there is not enough space and the transfer cannot be made, it prints the
following message and then returns to the previous device prompt:
%%% Too many bad blocks on output disk.

However, if the program decides there are very few free blocks, it prints:
%%% Only nn% of the output disk clust,ers are available for relocation.
Mount i t anyway <NO>?

You must decide whether the nwnber of free clusters is large enough to permit
a successful SAVRES operation. If you decide it is not, type NO or press Line
Feed or Return to return to the output prompt. Type YES if you want SAVRES to
continue the operation.
NOTE

Whenever you get this warning message, DIGITAL recommends you
use another disk for output. If you do not, there is a greater than
normal chance the operation will not succeed.
SAVRES normally transfers each allocated pack cluster from the original source
disk to the same cluster on the output RS'TSIE disk. If SAVRES finds the output
cluster is bad, it tries to relocate the data. If the clust.er size of the item being
moved is larger than the pack cluster size, there must be enough contiguous free
pack clusters on the output disk to accommodate the entire source cluster. If
there is not, SAVRES aborts the operation.
If the specified output is not a RSTSIE disk, the program prints the new SAVE
Set Name and proceeds without further notice. It does not check the output
volwne label if you specified the /SCRATCH switch. It always attempts to recover
bad block information from a RSTSIE file-structured and SAVE disk.

E.S.S

Restoring a RSTS/E Disk using the RESTORE Dialogue
To recreate a RSTSIE file-structured disk from a SAVE Set, you must perform the
RESTORE operation. Because SAVRES writes the SAVE Set in the SAVE fonnat,
no other RSTSIE software can perform this rebuild process. Before transferring
any data, SAVRES tries to extract bad block information from the destination
disk and, if it finds any bad blocks, incorporates them into the new BADB.SYS
file.
The SAVRES program may find bad blocks during a RESTORE run. If it does,
it scans the SATT.SYS file to determine if there is a place where the block
corresponding to a bad block can be relocated. If a large enough area exists on
the destination disk, the program simply 'moves the data to the new location and
modifies the directory information accordingly.

E-8

SAVRES System Program

The relocation of blocks may require RESTORE to change a file's characteristics.
For example, this can occur if the program finds a bad cluster while transferring
a contiguous file. 1b relocate the cluster that falls on the bad block, SAVRES
must make the file noncontiguous. Changes of this type can cause problems, but
RESTORE preserves the data and informs you of these adjustments.
The operation ends if there is no way RESTORE can relocate the data cluster.
This can occur if:
•

No unallocated pack cluster remains on the disk

•

The file cluster size is larger than the pack cluster size and there is not an
equivalent number of contiguous pack clusters left on the disk

Normally, SAVRES uses the name of the input SAVE Set as the Pack ID of the
output volume. Mter you answer the From Device? question, SAVRES prints a
message showing you the name it plans to use. The general format is:

*** SAVE Set namelDefault Pack ID is xxxxxx
Mter you successfully respond to the TO RSTS DISK question, SAVRES prints
the message:

*** Pack 10 is xxxxxx

SAYRES System Program

E-9

Table E-5 lists the RESTORE dialogue questions. An example of a RESTORE
procedure follows the dialogue description.
Table E-S:

RESTORE Dialogue Questions

Question

Response and Description

From Device?

<devspec>
Data is transferred from the specified device. If you include
a SAVE Set Name in the device specification (see section
"Device Specifications"), SAVEIRESTORE checks whether it
matches the name on the device. If it does not, the program
prints a warning message (se,e section "Checking the Input
Volume") and prints the From Device? question again.
Type 2 characters to specify t.he input device type, followed
by the unit number. Include a SAVE Set Name if you
want SAVEIRESTORE to make sure it matches the one on
the volume. The device must be physically mounted and
write-locked.

To RSTS Dx: Disk?

<devspec>
Data from the previously spedfied SAVE Set is restored to
this disk. If you include a Pack ID in the device specification,
it becomes the Pack ID of the restored disk. Otherwise,
SAVEIRESTORE uses the SAVE Set Name (which may also
be the Pack ID of the original source disk). Physically mount
and write-enable the disk.
Type 2 characters to specify the output device type, followed
by the unit number. Include a Pack ID if you want to
override the default. This drive should have a SCRATCH
volume mounted, and must be write-enabled.

Verify(Yes or No)<NO>?

Y[ES] or N[O]
This is the dialogue form of the !VERIFY switch. If you type
YES, SAVEIREBTORE compares the output with the input at
the end of each input volume to make sure the volumes are
equivalent. Type YES if you want SAVRES to compare the
input and output volumes at the end of each input volume.
Type NO or press Line Feed lOr Return if you do not want
SAVRES to perform this verification.

Proceed(Yes or No)?

Y[ES] or N[O]
This question lets you either double check your dialogue
responses and abort the operation if you have made errors, or
proceed with the RESTORE operation. Type YES to proceed
with the operation. Type NO to abort and return to the
SAVIRES Function: prompt.

------------------------

NOTE

If you are using the output of a RESTORE operation as a system disk,
you must use the INIT.SYS INSTALL option to reinstall the desired
monitor Save Image Library (SIL). You must aIeJo reenter any necessary
HARDWR suboptions (such as changing Hertz) because the INSTALL
option clears various items set up by the HARDWR suboptions. This
procedure is necessary because a disk can be saved with SAVE from
one system disk and restored with RESTORE on a system with a
different hardware configuration.
The following RESTORE example shows the restoration of an RM03 disk from
the tape SAVE Set created in the previolls example. The letters are keyed to the
explanation that follows this sample.
E-10

SAVRES System Program

The RESTORE steps are:
1.

Mount the two SAVE Set volumes on drive MM1: and MM2:, write-protected.

2. Mount the destination disk on DR1:, write-enabled.
3. Type RUN $SAVRES and press Return.

o $ RUN $SAVRES<RET>
8

SAV/RES Function: RESTORE<RET>

From device? MMl:<RET>

*** SAVE Set Name/default Pack ID is SAVSET
C)

To RSTS DR: Disk? DR1:TSTPAK<RET>

*** The volume on DR1: is a RSTS disk
*** Pack ID is MYPAK
*** Pack will be reinitialized
Mount it anyway <NO>? Y<:R1fl'>
*** Pack ID is TSTPAK

Verify (Yes or No) <NO>? <RET>

Proceed (Yes or No)? Y<RET>

*** Begin RESTORE from MMl: to DR1: at 08:28 PM
Dismount Device:
Set Name:
Seq :fI::
Density:
Creation date:
Expiration date:

MMl:
SAVSET
1

800
Friday, 10-May-85
Saturday, 10-May-86

Mount volume :fI: 2 of SAVE Set SAVSET
Device? MM2:<RET>
Proceed (Yes or No)? Y<RET>

*** Begin RESTORE from MM2: to DR1: at 08:32 PM
Dismount Device:
Set Name:
Seq :fI::
Density:
Creation date:
Ex.piration date:

MM2:
SAVSET

Dismount Device:
Pack ID:
Pack Clustersize:
Creation date:

DR1:
'rSTPAK

800
Friday, 10-May-85
Saturday, 10-May-86

Monday, 15-Apr-85

Please label this volume!
RESTORE operation completed at 08:37 PM

SAVRES System Program

E-11

Summary Report
RESTORE of SAVE Set SAVSET to DR1:TSTPAK
Input Device:
Set Name:
# of volumes:
Density:
Creation date:
Expiration date:

Magtape
SAVSET

Output Device:
Pack ID:
Pack Clustersize:
Creation date:

DR1:
TSTPAK

800
Friday, 10-May-85
Saturday, 10-May-86

Monday, 15-Apr-85

RESTORE started on Friday, 10-May-85, at 08:28 PM
Run Statistics
Transfer Totals:
Total of 54184 blocks transferred
Error Totals:
Total of

o new bad blocks encountered on destination.

o files structurally altered.
Timing Totals:

Total elapsed time:

o hrs.,

9 mins.,

36 sees.

Total wait

time:

o hrs.,

o mins.,

16 sees.

Total process time:

o hrs.,

9 mins.,

20 sees.

SAV/RES Function: <Ctrl/Z>
$

The explanation of the RESTORE example follows:

o Type RUN $SAVRES, and then press Return to run the SAVRES program.
@ Begin the RESTORE operation by typing RES1'ORE and pressing Return

in response to the initial SAVRES prompt. Because you need to recreate an
RM03 from a tape SAVE Set, you select the RES'rORE function.
@) The RESTORE dialogue begins when it prints the From Device? question. In

response, you specify the device narn.e and the unit number of the SAVE Set
medium and press Return. Because the input medium is tape, mounted on
drive number 1, you enter MM1:. If you had included a SAVE Set Name in
the device specification, RESTORE would have checked it against the name
already on the set. If they had not nlatched, RESTORE would have issued a
warning message informing you of the discrepancy. After you terminate your
response by pressing Return, RESTORE reads the SAVE Set Name from the
input medium and displays it on your terminal.

e Because in this RESTORE operation. you are recreating an RM03 disk, you
respond to the To RSTS DR: Disk? question by typing DRl:TSTPAK. By your
response, you indicate that the device mnemonic of the RM03 is DR, that it
is mounted on drive number 1, and that you want it to have a Pack ID of
TSTPAK. Mter pressing Return, RESTORE checks the Pack ID of the output
disk. It then prints a message showing you that the disk isa RSTSIE disk,
has a Pack ID of MYPAK, and that RESTORE will reinitialize it. RESTORE
issues a Mount It Anyway? question that gives you the option to mount the
disk or not. Because the data on the disk is no longer valuable, answer Y and
press Return. RESTORE then initializes the disk with the new Pack ID.

E-·12

SAVRES System Program

o The comparison of the input data to the data already transferred to the
destination disk is not essential during this particular run; therefore, you
prel3s Return to accept the No default. When you ask RESTORE to verify
the transferred data, you are essentially requesting that RESTORE perfonn
two operations rather than one. That is, you make the program perform two
passes on the data, once for the transfer and once for the comparison.

o No mistakes were made during the dialogue phase; therefore, RESTORE proceeds with the operation in reply to your Y response. Mter you press Return,
RESTORE displays a message to indicate it has begun the transfer. After
RESTORE completes the first volume, it prints input medium information,
asks you to mount the next volume, and asks if you want to continue the
operation. Again, because there are no errors in the dialogue, RESTORE
proceeds. RESTORE moves the remaining data to the second volume, issues
the input and output dismount messages, and informs you that the operation
is completed.

o The Summary Report prints the SAVE Set Name, the Pack ID, and label
information for both the input and output medium, in addition to the date
and time of the RESTORE operation.
€) The Run Statistics Report prints the number of blocks transferred, the

number of bad blocks encountered, and the times required to perform various
operations.

o RESTORE returns to the SAVIRES FUNCTION prompt when the operation
has ended. Enter CtrllZ to return to your default keyboard monitor (DCL in
this example).

E.5,,6

Copying a RSTS/E Disk using the IMAGE Dialogue
'rhe IMAGE function lets you make an equivalent copy of a RSTS/E filestructured disk; however, only·disk-to-LIKE-disk transfers are legal when
using the IMAGE operation. LIKE disks are units that have exactly the same
device size. The SAVRES program handles bad block recovery during an IMAGE
operation in the same way as RESTORE.
Table E-6 lists the IMAGE dialogue questions. An example of the IMAGE
operation follows the dialogue description.
NOTE

If you are using the output of an IMAGE operation as a system disk,
you must use the INIT.SYS INSTALL option to reinstall the desired
monitor Save Image Library (SIL). You must also reenter any necessary
HARDWR suboptions (such as changing Hertz) because the INSTALL
option clears various items set up by the HARDWR suboptions. This
procedure is necessary because you can make a copy of a disk on one
system and use the copied disk on another.

SAVRES System Program

E-13

Table E-6:

IMAGE Dialogue Questions

Question

Response and. Description

From RSTS Disk?

<devspec>
SAVRES transfers all data from the disk specified in the
device specification response. If you include a Pack ID in the
device specification, SAVRES checks to see if it matches the
one actually on the disk. If it does not, the program prints a
warning message (see section "Checking the Input Volume")
and prints the :From RSTS Disk? question again.
Type 2 characters to specify 1~he input device type, followed
by the unit number. Include a Pack ID if you want SAYRES
to make sure it. matches the one on the volume. This device
must. be physic;ally mounted and write-locked.

To RSTS Dx: Disk?

<devspec>
SAVRES transfers all information from the previously
specified source disk to this disk. If you include a Pack ID in
the device specification, it becomes the Pack ID of the output
disk. Otherwise, SAVRES uses the Pack ID of the source
disk.
Type 2 characters to specify the output device type, followed
by the unit number. Include a Pack ID if you want to
override the default. This drive should have a SCRATCH
volume mounted, and must be write-enabled.
Note that the device mnemonic you specify in response to the
From RSTS Disk? question appears in place of the Dx in the
To RSTS Dx: Disk? question. In some cases, in your device
specification response to the To RSTS Dx: Disk? question,
you may specify a device mnemonic other than the one that
a ppeared for Dx. See the discussion of LIKE disks in the
section "Definitions of SAVRES Terms."

Verify<Yes or No)<NO>?

Y[es] or N[o]
This is the dialogue form of the VER[IFY] switch. If you type
YES, SAVRES compares the output with the input at the
end of each output volume to ensure that the volumes are
equivalent. If you type NO, SAVRES moves on to the next
question without performing a verify pass.
Type YES if you want SAVRES to compare the input and
output volumes at the end of the IMAGE copy operation.
Type NO or priess Line Feed or Return if you do not want
SAVRES to pelform the verification.

Proceed(Yes or No)?

Y[ES] or N[O]
This question lets you either double check your dialogue
responses and abort the operation if you have made any
errors or proceed with the IMAGE operation.
Type YES to proceed with the operation. Type NO to abort
and return to the SAVIRES Function: prompt.

Mter you answer the From RSTS Disk? question, the program prints a message
in the following general format to tell you the Pack ID it intends to use:

*** Input Pack IO/default Output Pack IO is XXXXXX
If you do not want the output disk to have the same name as the input volwne,
specify a different Pack ID in the <devElpec> response to the To RSTS DX: Disk?
question. Mter you select the Pack ID and answer the question, SAVRES prints:

*** Output Pack IO is XXXXXX

E-14

SAVRES System Program

rrhe following procedure describes an IMAGE operation that copies an RA80 to
an RA80 and verifies the transfer in the same pass. The letters are keyed to the
explanation that follows this sanlple.
The IMAGE function steps are:
1.

Mou.nt the source disk on drive DUO:, write-locked.

Mou.nt the destination disk on drive DUl:, write-enabled.

3. Type RUN $SAVRES and press Return.

€t

$ RUN $SAVRES<RET>

SAV/RES Function: IMAGE/NOERROR<RET>

From RSTS disk? DUO:<RET>

*** Input Pack ID/default Output Pack 10 is SOURCE
()

To RSTS DU: Disk? DU1:DESTIN<RET>

*** The volume on DU1: is a RSTS disk
*** Pack 10 is OUTPUT
*** Pack will be reinitialized
Mount it anyway <NO>? YES<RET>
*** Output Pack 10 is DESTIN
~

Verify (Yes or No) <NO>? YES<RET>

o Proceed (Yes or No)? YES<RET>
*** Begin IMAGE copy from DUO: to DU1: at 11:02 AM
*** Begin VERIFY pass from DUO: to DU1: at 11:04 AM
*** 0 differences found
Dismount Device:
Pack 10:
Pack Clustersize:
Creation date:

DU1:
DESTIN
8

Friday, 10-May-85

Please label this volume!
IMAGE copy operation completed at 11:07 AM
Summary Report
IMAGE copy of DUO:SOURCE to DU1:DESTIN
Input Device:
Pack 10 :
Pack Clustersize:
C:r:eation date:

DUO:
SOURCE
8
Friday, 10-May-85

Output Device:
Pack 10 :
Pack Clustersize:
Creation date:

DU1:
DESTIN
8
Friday, 10-May-85

IMAGE copy started on Friday, 10-May-85, at 11:02 AM
Run Statistics
Transfer Totals:
Total of 24400 blocks transferred
Error Totals:
Total of
Total of
Total of

o bad ce,mpares.
o new bad blocks encountered on source.
o new bad blocks encountered on destination.
SAVRES System Program

E-15

o files structurally altered.
Timing Totals:
Total elapsed time:

o hrB.,

5 mins.,

22 sees.

time:

o hrs.,

o mins.,

o sees.

Total process time:

o hrB.,

5 mins.,

22 sees.

Total wait

SAV/RES Function: <Ctrl/Z>

The explanation of the IMAGE example follows:

o Type RUN $SAVRES and press Return to run the SAVRES program.
f:) SAVRES prints the SAVIRES FUNCTION prom.pt to determine which
function you want to perform. 'lb create a copy of an RA80, select the

IMAGE function by typing IMINOERROR and pressing Return. Attach the
INOERROR switch to insure that no data is lost. The INOERROR switch
causes SAVRES to end the operation whenever the program finds a run-time
error. Before the program aborts, it. prints an elTor message indicating the
source of the problem.
@) In reply to the From RSTS Disk? question, specify the disk from which you

are to extract the data. Enter the device name and unit number of the input
disk and end the response by pressing Return. After this response, IMAGE
reads the Pack ID from the input disk and displays it on your terminal.

o IMAGE requests the device name and unit num.ber of the output disk. Type
DBl:DESTIN to indicate that the M80 is on drive 1 and you want DESTIN
to be the new Pack ID. As soon as you press Return, SAVRES tells you that
the output medium is a RSTSIE disk, its Pack ID is OUTPUT, and it will be
initialized. SAVRES initializes the disk when you type Y and press Return in
response to the Mount it anyway? question. A message that SAVRES prints
informs you that the disk was initialized to the specified Pack ID.

o 'lb insure against losing data, type YES to have IMAGE determine if the
input data equals the data that was transferred to the output disk. If IMAGE
finds that the data is not equivalent, it informs you of these differences.

o You are satisfied that you made no errors in the dialogue sequence and feel
that the operation should proceed. Mter you priess Return, IMAGE informs
you when it begins to transfer data and when it. begins the verify pass. Mter
completing the data comparison, IMAGE prints the number of differences
found between the two devices. It then issues the output device dismount
message. Immediately after IMAG]~ prints the dismount request, it prints a
m.essage to show you the copy operation ended EJUccessfully.
fj IMAGE prints the Summary Report unless you suppress it by using the

INOSTATS switch. The report includes the Paclt ID and label information of
each disk, in addition to the date and time that. the IMAGE operation was
run.

CD The Run Statistics Report provides a total figure for the number of blocks
transferred, the errors encountered, and the tiDles required to perform the
IMAGE operation.
~

E-16

SAVRES returns you to the SAVIR~~S FUNCTION prompt when the operation
ends. Enter CtrllZ to return to your keyboard DlOnitor (DeL in this example).

SAVRES System Program

E.S.7

IDENTIFY Dialogue
If you want label infonnation from a SAVE Set volume or a RSTSIE filestructured disk, use the IDENTIFY function. SAVRES asks only one question
before the IDENTIFY operation begins.
Table E-7 lists the IDENTIFY dialogue question. Examples of the IDENTIFY
operation appear after the table.
Table E-7:

IDENTIFY Dialogue Question

Question

Response and Description

From Device?

<devspec>
SAVRES prints the device label characteristics of the SAVE Set or
RSTSIE file-structured disk for the device specified in the dialogue
response. If you include a SAVE Set Name or Pack ID, SAVRES
checks to see if it matches the name on the device. If it does not,
the program prints a warning message (see the section "Checking
the Input Volume") and then prints the From Device? question
again.
Type 2 characters to specify the input device, followed by the unit
number. Include a SAVE Set Name or Pack ID if you want SAVRES
to make sure it matches the one on the volume. You must mount
and write-lock the device (for protection) before responding to the
From Devic:e? question.

The following IDENTIFY example shows the label characteristics of both a SAVE
Set and a RSTSIE disk. The letters are keyed to the explanation that follows this
sample.
The IDENTIFY procedures are:
1.

Mount a SAVE Set volume on MMl:, write-locked.

2. Mount the RSTSIE disk on drive DR1:, write-locked.
3. Type RUN $SAVRES and press Return.

«t

$ RUN $SAVRES<RET>

SAV/RES Function: ID MMl:<RET>

Device:
Set Name:
Seq :fI::
Density:
Creation date:
Expiration date:

()

1600
Friday, 10-May-85
Saturday, 10-May-86

SAV/RES Function: ID DR1:<RET>
Device:
Pack ID:
Pack Clustersize:
Creation date:

MMl:
SAVSET

DR1:
TSTPAK
8

Monday, 15-Apr-85

SAV/RES Function: <Ctrl/Z>
$

SAVRES System Program

E-17

o Type RUN $SAVRES and press Return to run the SAVRES program.
8 The SAVRES program prints the SAVIRES FUNC~rION prompt. You respond
to it with the first two letters of the IDENTIFY function and with the device
name and unit number of the mediunl you want to identify. Include a Pack
ID if you want SAVRES to compare the ID you specify with the one on the
input volume.

o IDENTIFY prints label information for the tape SAVE Set.
o SAVRES returns to the SAVIRES FlJNCTION prompt. In response, you ask
SAVRES to identify the disk pack mounted on DR1:. SAVRES prints the
information as soon as you end your response.

o Enter CtrllZ to return to your keyboard monitor (DCL in this example).
E.S.8

Full Function Command Line
Use the full function command line format if you do not want to answer each
dialogue question separately. The general format is:
SAVIRES Function: <function><outdevspec>=<indevsp1ec>[<switch(es» ]
Function can be RE[STORE], 1M[AGE] , or ID[ENTIFY] and the switches can be
any of those listed as valid with the associated function (see section "SAVRES
Switches"). See the section "Device Specifications" for an explanation of the
device specifications.
If you specify a full command in response to the SAY/RES FUNCTION prompt,
SAVRES prints only the PROCEED question (except during an IDENTIFY
question):
Proceed (Yes or No)?

You must type either YES or NO to this question. Type YES to have SAVRES
begin the operation; type NO to abort.
SAVRES does the same standard volume label checking during the full function
command line operation as it does during the longer dialogue procedure. If a
discrepancy occurs, SAVRES prints the error messagE~ and then returns to the
SAVIRES FUNCTION prompt. At that point, you must type the full function
command line again.

E.6 SAVRES and Booting
SAVRES always attempts to create bootable media. The output of a RESTORE or
IMAGE copy operation is bootable if the original disk contained an INIT.SYS file
in [0,1] (even if the original disk was not bootable).
If a SAVE volume is booted, a portion of INIT.SYS is loaded, allowing you to

perform the following INIT options:

E-18

•

SAVRES

•

DSKINT

•

BOOT

•

HARDWR option specifying the LIST suboption

SAYRES System Program

The types of SAVRES operations you can perform depend on the type of device
you boot from. You can run all three SAVRES options if you boot from a RSTSIE
disk or a RSTSIE magnetic tape or SAVE volume. If you boot a SAVE volume and
then perform a RESTORE operation, SAYRES prints a message at completion of
the run:

*** Please boot from the system disk
Boot device?

You must specify the device name and unit number of a bootable device.
Table E-8 sumrn.arizes the SAVRES operations you can perform while booting
various devices.
Table E-8:

Booting RSTS/E and SAVE Set Volumes

Device Booted
From

SAVRES Operations
You Can Perform.

SAVRES Operations You Can
Perform. Afterwards

RSTSIE Disk or
Magnetic Tape

IDENTIFY, IMAGE, RESTORE

SAVE 8et Volume
(Disk or Tape)

IDENTIFY, IMAGE

IDENTIFY, IMAGE, RESTORE

SAVE Set Volume
(Disk or Tape)

RESTORE

None--You receive a message
that states you must boot from
the system disk.

NOTE

You cannot use the console bootstrap to boot a magnetic tape SAVE Set
if:
•

The tape has a density of 1600 bpi.

•

The drive the tape is lnounted on has a TM02 formatter. (Use
the HARDWR LIST suboption of INIT to determine the type of
formatters on your drives.)

However, you can boot a 1600 bpi tape on a drive with a TM02
formatter if you use the BOOT OPTION prompt of INIT.SYS. That
is, you can boot under these circumstances using software but not
hardware boot procedures. As an additional restriction, you can boot
a magnetic tape SAVE Set only on the following drives: TS11, TU45 ,
TU80, TU16, TU77, and TE16.

E.J Operator Interface During Processing
SAVRES requires operator intervention at certain points during processing.
The operator must mount and dismount devices, reset hung devices, and abort
operations when necessary. The following sections describe these procedures.

SAVRES System Program

E-19

E.7.1

Mounting and Dismounting Volumes
A SAVE Set can contain more than a single volume. Because you may be
processing multivolume SAVE Sets, the SAVRES program contains procedures
that allow you to mount the additional volumes in a set. When SAVRES needs
an additional volume to complete a RESTORE operation, it prompts you with the
message:
Mount volume #,nn of SAVE Set <setnam>
Device?

Proceed (Yes or No)?

The letters nn represent the volume sequence number and <setnam> indicates
the SAVE Set Name and unit number of the drive on. which the next volume has
been readied. If you mount a device of the same type, you can respond to this
prompt only with a unit number. Mter you press Return to terminate your device
specification, SAVRES performs volume label checking (see section "Checking
the Input Volume"). At this point, if the additional volume you mount violates
any SAVRES rule, the program prints the error mesBage immediately after the
Device? question.
SAVRES prints a dismount message when:
•

SAVRES completes the data transfer to the output volume in an IMAGE copy
operation

•

You mount an incorrect volume

Note that the dismount message SAVRES generates for a SAVE Set (Format 1)
differs from a dismount message printed for a RSTS/E disk (Format 2):

Format 1
Dismount Device:
Set Name:
Seq -#:
Density:
Creation Date:
Expiration Date:

Format 2
Dismount Device:
Pack ID:
Pack Clustersize:
Creation Date:

The <device> field represents a two character device mnemonic followed by a
single digit unit number. The SAVE Set Name replaces <setnam>, and the disk
pack identification replaces <packid> during a valid :SAVRES operation. Density
figures are printed only for magnetic tape SAVE Setsl.

E-20

SAVRES System Program

E.7.2

Re-Accessing Devices
During an operation you may find that SAVRES does not have access to a
previously accessible device. If a device does become hung or is write-protected,
the program notifies you in the warning message:
%%% Device hung or write-locked
Retry (Yes or No)?

Correct the condition causing the problem if possible, and then type YES to
attenlpt to gain access to the device. Type NO to abort the attempt.

E.7.3

Aborting SAVRES
1b end a SAVRES operation at any time, enter Ctrl/C. This returns control to you
keyboard monitor if you running SAVRES on line. If you are running under the
control of the INIT.SYS program and the operation ends, INIT.SYS returns you
to its OPTION prompt. An abort during a RESTORE or IMAGE operation leavef
the output volume unusable. However, you do not jeopardize any IDENTIFY
volumes if you choose to abort this operation.

E.8 SAYRES Summary Report
SAVRES prints a Summary Report at the end of RESTORE, and IMAGE
operations. The report identifies the operation that was run, lists characteristics
of the input and output devices, and generates a list of data transfer, error, and
timing statistics. It also documents the time and date of the operation.

SAYRES System Program

E-~

E.8.1

Summary Report Format and Example
The format of the Summary Report varies with the operation. At times the
Input Device characteristics (see Figure E-l) identify a SAVE Set, as in the
case of a RESTORE operation.' Thus, depending on the operation, the SAVE Set
format may be represented 'under either the Input or Output Device category.
The RSTSIE disk format may be printed also under the Input or Output Device
categories. Figure E-l shows the general format of the SAVRES Summary
Report.

Figure E-1:

Summary Report Format

IMAGE
RESTORE

disk<packid>
SAVE Set <setnam>

Input Device:
Pack ID:
Pack Clustersize:
Creation date:
Output Device:
Set Name:
#" of volumes:
Density:
Creation date:
Expiration date:

disk<packid>
SAVE Set<setname>

<device>
<packid>
nn
*RSTS/E disk format
<day of week, dd-mmm-yy>
[magnetic tape]
[disk]
<setnam>
*SAVE Set format
nn
[ 800]
[1600]
<day of week, dd-mmm-yy>
<day of week, dd-mmm-yy>

Run Statistics
Transfer Totals:
Total of nnnnnnn blocks transferred
Error Totals:
Total of nnn bad compares.
Total of nnn new bad blocks encountered on source.
Total of nnn new bad blocks on destination disk.
Total of nnn files structurally altered.
Timing Totals:
Total elapsed time:
time:
Total wait
Total process time:

E-22

SAVRES System Program

nn hrs.,
nn hrs.,
nn hrs. ,

nn mins. ,
nn mins. ,
nn mins. ,

nn sees.
nn sees.
nn sees.

E.8.2

Summary Report Run Statistics
Table E-9 lists and describes the run statistics that appear at the end of a
RESTORE or IMAGE operation. Because each SAVRES operation prints its
own set of run statistics, Table E-9 also identifies the operations in which each
run statistic is printed. The operations are identified according to the following
conven tions:
R = ]RESTORE operation
I = IMAGE operation
Table E-9:

Summary Report Run Totals

Type of Statistic

Operation and Description

Transfer Statistics
Total of blocks transferred

This number refers to the total number of blocks
transferred (directory blocks, file blocks). In a SAVE
operation an extra set of directory blocks is stored at
the end of a SAVE Set. The total number of blocks
transferred includes these extra blocks. [RI]

Error Statistics
Total of bad compares

The total number of errors that occurred during the
VERIFY phase. [RI]

Total of new bad block errors

The number of previously unreported bad blocks
encountered on the source disk. [RI]

Total of bad block errors on
destination disk

The number of previously unreported bad blocks
encountered on the destination disk. [RI]

Total of files structurally altered

The number of files that were structurally altered to
enable relocation on the destination disk. (This meanl
contiguous files were made noncontiguous or placed
files were moved.) [RI]

Time Statistics
Total elapsed time (TE)

The total amount of real clock time that elapsed
during the run. [RI]

Total wait time (TW)

The amount of real clock time that elapsed while
waiting for operator responses and magnetic tape
rewinds. [RI]

Total process time (TP)

The amount of real clock time that was used for
processing (TE-TW=TP). [RI]

E.g SAVRES Error Handling
SAVRES issues dialogue, mount, transfer, and verification error messages. When
you find an error message during a dialogue or a mount, SAVRES asks you
to correct the condition, where appropriate, and then asks you to reenter the
command. Errors you encounter during either of these phases never affect the
integrity of the data being transferred.

SAVRES System Program

E-2~

NOTE:

When SAVRES reports an error during a transfer or verification
phase, it has detected a problem that might corrupt the data being
transferred. Depending on the nature of the errOl, a single file, or
perhaps an entire volume, might be affected. Gencerally, you encounter
fewer critical errors during a dialogue or mount phase than in a
transfer or verification operation.
SAVRES does not report an error when it finds a bad block on the output device,
unless it is unable to reallocate that block in a way that is transparent to you.
For instance, the one-block file SHIFT.LES is moved without notice if a free
cluster can be found. On the other hand, the program issues an error message if
it must move the placed file, MOVEME.NOT.
The INOERROR switch causes SAVRES to abort the run if it finds any condition
that warrants an error message. If you do not specify the INOERROR switch, the
program takes whatever action necessary. SAYRES always informs you of any
corrective action taken so you have the option of aborting the operation.
The most important problem that SAVRES encounters during a RESTORE
or IMAGE copy operation is finding a bad block on a RSTS/E file-structured
disk when it wants to write to that block. The program attempts to correct the
problem by performing the following actions in the ord,er listed:
1. Marks a contiguous file as noncontiguous and relocates the bad cluster but
not the entire file.

2. Marks the file as nonplaced if the bad cluster occurs at the beginning of a
placed file.
3. Finds enough contiguous free pack clusters on the output disk to hold the
number of clusters that must be moved. In other words, if the cluster size of a
file is 16, SAVRES finds 16 free contiguous blocks for the cluster. It proceeds
if the reallocation worked and aborts if it has not worked.
SAVRES always tells you about the changes it makes to a file that are caused by
this bad block processing. If there is no way to relocate a cluster, the program
aborts the operation. Unless you specify the INOERROR, SAVRES never
terminates an operation when it finds a bad block that results in a nonfatal error
(see the section "SAVRES Switches").

E-24

SAVRES System Program

EJ~.1

General SAVRES Error Messages
Table :E-IO describes the errors that can occur during SAVRES operations.
Table E-10:

General SAVRES Error Messages

Message

Meaning

111 Bad block in MFD
SAVRES aborts the operation when it finds a bad block in the Master File
Directory (MFD).
%%% Bad block in SATT.SYS

There is a bad block in the input SATT.SYS file. SAVRES does not use an
input disk that has a corrupt SATT.SYS file.
%%% Bad data:
in [MFD]
[<[P.PN] FILENAME.TYPE>]
[P,PN]]
[[GFD, nun]

A cluster contained in the described account or file could not be relocated
due to a fatal read error on the output volume.
1?? Bad directory for device
SAVRES discovered a bad link or bad block in the input or output directory
structure.
%%% Bad INIT.SYS on system disk

SAVRES found a bad block in INIT.SYS while transferring it from the
system disk to a SAVE Set volume.
%%% Can't mix device types in SAVE Set

A SAVE Set cannot be composed of both tapes and disks. Only all disk or all
tape SAVE Sets are legal.
%%% Can't read bad block file

SAVRES could not read the bad block file on a RSTSIE disk or a SAVE Set
volume.
??? Can't read SAVE Set volume SATT
A nonrecoverable 110 error occurred while SAVRES attempted to read the
Storage Allocation table on the SAVE Set volume.
%%% Can't write boot

The bootstrap could not be written to the first block of the output disk.
%%% Device hung or write-locked
Retry (Yes or No)?

The output device is off line or is write-locked. Correct the condition and try
again. Type YES to continue or type NO to abort.
%%% Device mnemonic must be specified

Enter the device name along with the unit number.

(continued on next page)

SAVRES System Program

E-25

Table E-10 (Cont.):
Message

General SAYRES Error Messages

Meaning

%%% Disk must be DSKINTed

When you are running SAVRES on line, the output disk must be RSTSIE
file-structured or a previous SAVE Set volume. Use the DSKINT option of
INIT.SYS to reinitialize the disk. See the RSTS/E System Installation and
Update Guide for a description of the DSKINT option.
%%% Duplicate switches

You specified the same switch more than once.
111 Error in BADB.SYS allocation

SAVRES was unable to rebuild the bad block file (BADB.SYS) after completing the current operation.
111 Fatal output bad block

A bad block was found at a critical position preventing SAVRES from
completing the current operation.
ll

%%% megal command

Your response to a prompt was in an unacceptable format.
%%% megal command:
no output device specified

The full function line command you specified. did not include an output
device specification.
%%% Illegal density

The density you specified was in an unacceptable format, was included
where none was acceptable, or specified an illegal density value.
%%% Illegal option

When booting from a SAVE Set volume, you can use only the RESTORE,
IMAGE, and IDENTIFY options.
%%% Illegal switch

You used a switch in an unacceptable way.
%%% Illegal switch combination

You specified conflicting switches.
%%% Input disk has only nn% free clusters.
Mount it anyway <No>1

SAVRES discovers the input disk has very fEtw clusters that are not allocated
and may find problems copying or restoring the disk. You may want to
delete some unnecessary file to increase the chances that SAVRES can
complete a successful IMAGE.
%%% Input disk should be rebuilt

The disk you mounted is "dirty." The SATr.SYS file may be corrupt.
SAVRES requires a valid SATT.SYS; therefore, rebuild the disk using
the REFRESH option of INIT or the ONLCLN program. See the RSTS / E
System Installation and Update Guide for a description of REFRESH. See
the RSTSIE Utilities Reference Manual for information about ONLCLN.

(continued on next page)

E-26

SAVRES System Program

Table E-10 (Cont.):
Message

General SAVRES Error Messages

Meaning

%%% Magtape select error
Retry (Yes or No)?
When access to a magnetic tape drive was attempted the selected unit was
offline.
%%% No default-type <cr> for .more information
You typed a LINE FEED or RETURN in response to a question which has
no default.
11? No INIT.SYS on system disk
There is no INIT.SYS in account [0,1] on the system disk.
%%% No previous question - type <cr> for more information
You attempted to return to the previous question but it no longer exists.
%%% Non-existent or hung device
You specified a disk or magnetic tape that does not exist on your system or
is not on line. Check your configuration or the device type again and retry.
11? Non-recoverable 110 error
SAVRES found a fatal error condition while trying to recover from a
DEVICE HUNG OR WRITE LOCK error. SAVRES aborts the current
operation and returns you to the SAVIRES FUNCTION prompt.
%%% Not a valid device

You specified an unacceptable device type in response to the question. This
may occur if you specify:
•
•
•
•

A device other than magnetic tape or disk
A tape rather than a disk (or vice versa)
A wrong disk type on an IMAGE
A device without including a unit number, for example, DK: instead of

DKO:
%%% Only nn% of the output disk clusters are available for relocation
Mount it anyway <No>?
SAVRES finds there are few free blocks available for relocation on the
output disk.
11? Operation aborted at user request
You specified the /NOERROR switch and one of the errors described in Table
E-12 occurred, or you typed NO to a RETRY (Yes or No)? question.
%%% Same device specified
You designated the same device for both input and output. Specify the
correct output device.
%%% Tape won't respond to density XXXX
SAVRES attempted to set the density of the specified magnetic tape to

x:xxx but the hardware rejected the request. Specify a density setting that
is legal for that device, or use a different device.

(continued on next page)

SAVRES System Program

E-27

Table E-10 (Cont.):
Message

General SAVRES Error Messages

Meaning

%%% Tape won't respond to any density

The input volume cannot be resid at any legal density setting.
%%% This is not the correct volume

The SAVE Set or Pack ID you specified does not match that on the mounted
device. A DISMOUNT DEVICE message follows the error message and
identifies the volume if it is a SAVE Set or a RSTSIE file-structured disk.
This message also occurs if SAVRES requests the next volume of a SAVE Set
and you specify a device containing a volume that is not part of the current
SAVE Set or is not the correct sequential volume of the current SAVE Set.
Mount the correct volume.
%%% This volume has no label

SAVRES did not find a valid SAVE Set label or Pack ID on the specified
input volume.
%%% Too many bad blocks on input disk

The number of bad clusters on the input disk exceeds the RSTSIE limit.
%%% Too many bad blocks on output disk

The number of allocated clusters on the original source disk exceeds the
number of available clusters on the output disk or, the number of bad
clusters on the output disk exceeds the RSTSIE limit (161 bad clusters). Use
a different disk pack.
%%% Unrecognized switch

You have specified an invalid switch.
111 Volume is full

There was not enough room for the relocation of data clusters due to the
blocks found on the output volume.
%%% Wrong size disk

The output volume of an IMAGE or RESTORE must be the same size as the
original source. See the discussion of LIKE disks in the section "Definitions
of SAVRES Terms."
%%% You will be writing to the booted device

If you write to this device, you need to reboot when you are finished with
SAVRES.

E.9.2 Transfer Errors-Fatal and Nonfatal
Transfer errors you may find in a SAVRI~S operation can be fatal or nonfatal.
Fatal transfer errors cause an immediate termination of the current operation.
Nonfatal errors, if the INOERROR switch is not in effect, let the operation
continue despite the problem. On the other hand, if you specify the INOERROR
switch and one of the errors in Table E-l1 occurs, SAVRES terminates the
operation immediately (which indicates the output is not a valid medium). In
other words, the nonfatal errors listed and described in Table E-ll become fatal
if the INOERROR is attached to the function response.
Whether or not you specify INOERROR, the error message ???Volume is full
occurs if SAVRES cannot relocate a file due to a bad block on the output device.
The relocation does not take place because SAVRES cannot find any additional
clusters available or does not find as many free clusters as it needed. This results
from too many bad blocks on the output volume.
E-28

SAVRES System Program

'rable E-ll contains a description of all nonfatal transfer errors that can occur if
you do not specify INOERROR.
Table E-11 :

SAVRES Nonfatal Transfer Errors
Message and Meaning

%%% Bad block on input, PCN.innnnnn
A bad block was foun.d on an allocated cluster on the input volume. A copied
version of the data in that block might be corrupt.
%%% Bad block on input, output PCN's affected:
A bad block was found on the input SAVE Set volume, which will affect
certain pack clusters on the output disk. A copied version of the data in that
block might be corrupt.
%%% Bad block on input:
in MFD [PCN.nnnnnn]
in UFD [P,PN] [, PCN.nnnnnn]
in GFD, nnn [, PCN.nnnnnn]
in rue [P,PN]FILENAME.TYPE,[PCN.nnnnnn]
A bad block was found on an allocated cluster on the input volume. A copied
version of the data in that block might be corrupt.
%%% Bad Compare, PCN=nnnnnn
The data on the input volume did not compare with the corresponding data
on the output volume. (See the explanation following this table.)
%%%Bad compare on input:
in MFD[PCN=nnnnnn]
in UFD [P,PN] [, PCN.nnnnnn]
in GFD, nnn [, PCN=nnnnnn]
in rue [P,PN] FILENAME. TYPE,[PCN.nnnnnn]
The data on the input volume did not compare with the corresponding data
on the output volume. (See the explanation following this table.)
%%% File x:xxxxx has been changed to noncontiguous
SAVRES found a badl block in the output area for file XXXXXX. 'lb complete
the transfer, one or more blocks of the file were relocated. Consider the
consequences of this change and proceed accordingly.
%%% Placed file x:xxxxx has been moved from aaaaaa to bbbbbb
A bad block was found in the output area for file xxxxxx. To complete
the transfer, one or more blocks of the file were relocated. Consider the
consequences of this change and proceed accordingly.

During a verification pass, SAVRES may detect differences between the input and
output. If there are differences, SAVRES informs you with two types of messages:
%%%Bad compare, PCN=nnnnn
%%%Bad compare in [MFO
[, PCN=nnnnn]]
[UFO [p,pn] [, PCN=nnnnn]]
[GFO, nnn [, PCN=nnnnnn]]
[file [p,pn]filenam.typ [, PCN=nnnnnn]]

SAVRES first prints a general error message and then, after some additional
processing, prints the second error message telling you more specifically the
accounts and/or files in which the bad comparisons occurred.

SAVRES System Program

E-29

If you request transfer verification, SAVRES checks (after the initial transfer
phase) whether the data it has written Inatches the corresponding data it read. If
there is a difference, SAVRES prints:
%%%Bad compare in [MFD

[, PCN=nnnnn]]

The SAVRES program transfers only clusters that are marked as allocated in the
Storage Allocation Table. During the actual transfer and verify pass, it can tell
only which Pack Cluster Number (PCN) it has completed. If INOERROR is in
effect and you are performing an IMAGE copy, SAVllES then tries to pinpoint
the problem by scanning the input direetory structure before aborting. If you
are doing a RESTORE, SAVRES simply aborts. Thi~3 difference exists because
SAVRES must have a complete directory to scan in order to find accounts and/or
files that caused the problems. In the case of a RESTORE, the output volume is
not yet complete and SAVRES cannot scan it.
If INOERROR is not in effect, SAVRES stores each PCN that caused a problem
in an internal table as it continues with the verify pass. If SAVRES finds more
entries than can be stored in the table, it prints the message:
"Affected file will not be reported"

This means there is no more room to store the information. At the end of the
verification pass, SAVRES prints the total number of variations found in the
form:
%%% nn differences found

Mter completing the transfer, SAVRES scans the input (IMAGE) or (RESTORE)
volume directories to find the accounts and/or files in which the bad comparisons
occurred. Then, it displays one or more of the following messages:
%%%Bad compare in [MFD
[, PCN=nnnnn]]
[UFD [p, pn] [, PCN=nnnnn]]
[GFD, nnn [, PCN=nnnnnn]]
[file [p,pn]filenam.typ [, l?CN=nnnnnn]]

E-30

SAVRES System Program

Appendix F

"Device Testing Package (DEVTST)

The Device Testing Package (DEVTST) exercises hardware on the RSTS/E
system during normal time~sharing operations. You can use DEVTST to simulate
lhardware errors on the system.

F.1

Introduction to DEVTST
DEVTST is a package of programs that provides the customer, System Manager,
or DIGITAL Field Service personnel with tests that verify the reliability of
RSTS/E~supported hardware. Execution of any DEVTST program is limited to
users with WREAD privilege. As distributed, DEVTST is in TEST$: (by default,
[0,5]).
'rhis appendix explains operational data, conventions, complete device testing
information, and error messages. Mter execution of individual DEVTST
exercisers, you should check the System Error Log for additional errors; for more
information on the System Error Log, see Chapter 10.
'rhe DEVTST exercisers are tests that can be initiated whenever you suspect
hardware trouble. These device exercisers generate excess activity in an attempt
to make the hardware fail during operation.
Use the DEVTST package in conjunction with the RSTS/E error package (see
Chapter 10). The Error Logging facility provides the means for determining
detailed information on hardware errors. In addition, the DEVTST programs
return error messages that may be useful in determining the failure; however,
these messages are not as specific as the information provided by the System
Error Log.

Device Testing Package (DEVTST)

F-1

The following conventions are used throughout the DEVTST package dialogue:
<>

Indicates the defaul'~ answer.

( )

Shows all possible r1eplies.

[ ]

Gives an optional reply; for example, ASC [II] means enter ASC,
ASCI, or ASCII.

Either key moves the program back one question. If you

or <ESC>

press either key as a response to l~he first question, DEVTST will
stop execution and return to the RSTSIE command level.

Pressing CtrllZ causes the program to stop execution and return to
RSTSIE command level. CtrllZ is detected only when the system
checks for new input.
Pressing CtrllC causes the program to stop execution immediately
and return to a command level.

<RET>

Use the Return key to terminate the input line.

Detach <NO>

All of the device exercisers have the ability to detach; this allows
you to run an exerciser while keeping your terminal free for other
operations.

-1

Negative one, when used as a value for number of iterations,
pages, or lines, tells the program to run indefinitely, or until you
press CtrllC.

F.2 DEVTST Operating Procedures
To run a DEVTST program, type:
$ RUN TEST$:program name

For example:
$ RUN TEST$:KBEXER

Table F-1 lists the program names and the hardware they test.

F-2

Table F-1:

DEVTST Prog rams

Program

Hardware Tested

CPEXER

CPU

CPUTST

CPU

DSKEXR

File-structured disk

DSKSEK

Non-file-structured disk

DXEXER

Diskette

KBEXER

Terminal

LPEXER

Line printer

MTEXER

Magnetic tape

Device Testing Package (DEVTST)

Table F-2 gives a complete list of RSTSIE supported hardware, including
designators and the associated! devices.
Table F-2:

RSTS/E Device Designators

Device Designator

Device

DK:, DL:, DM:, DP:,
DR:, DB:, DU:, SY:

RSTSIE public disk structure

SYO:

System disk (the unit that was bootstrapped)

DVO:

Virtual disk, unit 0 only

DKO: 1'00 DK7:

RK05IRK05F disk cartridge units 0 to 7

DLO: to DL3:

RLOllRL02 disk cartridge units 0 to 3

DMO: to DM7:

RK06IRK07 disk cartridge un.its 0 to 7

DRO: to DR7:

RM02IRM03IRM05IRM80 disk units 0 to 7

DBO: to DB7:

RP04IRP05IRP06 disk pack units 0 to 7

DUO: to DU15:

RA60, RA80, RA81, RC25, RD51, RD52, RD53,
RX50 disk units 0 to 15

CR:

CR11 punched or CM11 mark sense card reader

CD:

CD 11 punched card reader

MTO: to MT7:

TE101TU10trS03 magI?-etic tape units 0 to 7

MSO: to MS3:

TS11trSV05trK25trU80 magnetic tape units 0 to 3

MMO: toMM7:

TE16ITU16trU45trU77 magnetic tape units 0 to 7

MUO: to MU7:

TK50 magnetic tape units 0 to 7

LPO: to LP7:

Line printer units 0 to 7

Device Testing Package (DEVTST)

F-3

F.3 DEVTST Dialogue Errors
Table F-3 lists the error messages that can occur in DEVTST programs.
Table F-3:

DEVTST Error Messages
Message and Meaning

%Warning: <device name> is a logical device: <devic,e name> will be used.
DEVTST determines that the device name given is a logical namefor another
device. DEVTST performs a logical devicetranslation and the translated device
name is used.
?Illegal device: <device name> specified.,
The device name given by the useI' is either incorrect or is alogical device name
that cannot be translated (for example,incorrect characters or the wrong type of
device). Retry the operation with the correct device name.
?OPEN failure on <device name>: (Followed by an error message from Appendix
A of the RSTS / E System Installation and Update Guide.)
This error message can mean one of the following:
• The device specified does not exist
•
No volume was mounted on that device
• Another user has assigned the device
• The volume on the device has not been formatted
• The device is not on-line or the device is not write-enabled
To resolve: retry the operation using the correct option as determined by the
error message received.
?Illegal number of iterations selected.
Retry the operation using the correct number of iterations.

F.4 CPEXER
The central processor exerciser is designed to put a heavy load on the Central
Processing Unit (CPU). CPEXER runs compute bound for bursts and then sleeps
for five seconds so the average processor loading does not appreciably degrade
the performance of the system. CPEXER was specifically designed to test the
Floating Point Processor (FPP) option, but also serves to verify general CPU
integrity and the Extended Instruction Set (EIS) operation. There is also a test of
the PEEK function for kernel addresses 0 through 22000.
The next section itemizes the tests performed by CPEXER. In most of the tests,
CPEXER compares the results to known correct values. Two heavily computebound tests are included to verify consistent results of duplicate calculations.
Finally, CPEXER includes tests that verify FPP divide by zero trap, integer divide
by zero trap, integer conversion error, and EIS operation.

F-4

Device Testing Package (DEVTST)

F.4.1

CPEXER Tests
The CPEXER program includes the following 13 CPEXER tests:

F.4.2

Test

Description

SIN{X)

Uses SIN(X) extended function

SIN{X)

Uses polynomial approximation to SINE function

LOG(X)

Uses LOG(X) extended function

EXP{X)

Uses EXP(X) extended function

SQR(X)

Uses SQR{X) extended function

SQR(X)

Uses Newton-Raphson method to determine the square
root

LOG(EXP{X»

Compute-bound test

ATN(TAN(X»

Compute-bound test

A=1.0/0.0

Verifies FPP divide by zero trap

A%=1%/0%

Verifies integer divide by zero trap

A%=60000

Verifies integer conversion error

INT( 40.6621* 100+0. 5)1100=40.66

Verifies EIS operation

PEEK(X)

Verifies operation of PEEK function

CPEXER Dialogue
The foHowing is an example of the CPEXER dialogue:
$ RUN TEST$:CPEXER
CPEXER V9.0 RSTS V9.0 EDERE
How many minutes to run <15>? 1
Detach <No>? NO
11.9 seconds of CPU time used in 1 pass.
CPEXER finished at 03:36 PM
$

NOTE

The CPEXER program will build correctly only if the version of BASICPLUS or BASIC-PLUS-2 includes TRIG and LOG. These mathematical
fl.IDctions must be included during system installation of BASIC-PLUS
or the installation of BASIC-PLUS-2.

Device Testing Package (DEVTST)

F-5

F.4.3

CPEXER Errors
This program reports four types of errorEI:
1.

Instructions failed to execute.

Errors failed to trap.

Computation errors.

Computation consistency checks failed.

Check the System Error Log for reported errors. These errors may indicate
hardware problems.
When type one errors (instruction errors) occur, CPEXER prints a two-line
message: the first line gives the test that was being performed and the exact
error message, and the second line indicates that the test is continuing. For
example:
?CPEXER PEEK failed at memory location m.
-(Error Message)
Continuing ..... .

where:
(Error Message)

is the RSTSIE error message from Appendix A of the RSTS / E
System Installation and Update Guide.

is the memory location at which the PEEK function failed.

Examples of type two errors (trap failure) are:
?CPEXER Floating point divide by 0 didn't trap
Continuing .....
?CPEXER Integer divide by 0 didn't trap
Continuing .....
?CPEXER Forced integer conversion didn't trap
Continuing .....
?CPEXER (INT(40,6621*100=0.5))/100< >40.66
Continuing .....

Type three errors (computation errors) give the type of test (SIN,PLYNOM, LOG,
EXP, SQR, or NEWTON SQR), the error in the computed value, and the location
of the possible error. For example:
?CPEXER - SIN test failed for the following cases:
?CPEXER - Computed value of SIN(40) was .643, :3hould be .64
?CPEXER CPU,EIS,FIS,FPP,MUL,DIV ()r SHIFT E.rror
Continuing ...
?CPEXER - PLYNOM SIN test failed for the following cases:
?CPEXER - Computed value of PLYNOM SIN (40) wa:3 .643, should be .64
?CPEXER - CPU,EIS,FIS,FPP,MUL,DIV or SHIFT Error
Continuing ...
?CPEXER - LOG test failed for the following cases:
?CPEXER - Computed value of LOG (140) was 4.942, should be 4.939
?CPEXER - CPU,EIS,FIS,FPP,MUL,DIV or SHIFT Error
Continuing ...
?CPEXER - EXP test failed for the following cases:
?CPEXER - Computed value of EXP (4) was 54.598, should be 54.599
?CPEXER - CPU,EIS,FIS,FPP,MUL,DIV or SHIFT Error
Continuing ...

F-6

Device Testing Package (DEVTST)

?CPEXER - SQR test failed for the following cases:
?CPEXER - Computed value of SQR (SOO) was 22.361, should be 22.362
?CPEXER - CPU,EIS,FIS,FPP,MUL,DIV or SHIFT Error
Continuing ...
?CPEXER - NEWTON SQR test failed for the following cases:
?CPEXER - Computed value of NEWTON SQR (SOO) was 22.361, should be 22.:
?CPEXER - CPU,EIS,FIS,FPP,MUL,DIV or SHIFT Error
Continuing ...

Type four errors (consistency failure) in the CPEXER are shown in the following
examples:
?CPEXER - EXP/LOG Consistency check failed for the following cases:
?CPEXER - EXP (LOG ( ... ( 1 »»»»»
<> EXP (LOG ( ... ( 1»»»»»
?CPEXER - EXP (LOG ( ... (
<> EXP (LOG ( ... (
?CPEXER - CPU,EIS,FIS,FPP,MUL,DIV or SHIFT Error
Continuing ...

4 »»»»»

4»»»»»

?CPEXER - ATN/TAN Consistency check failed for the following cases:
?CPEXER - ATN (TAN ( ... (-1. S »»»»»
<> ATN (TAN ( ... (-1. S »»»»»
?CPEXER - CPU,EIS,FIS,FPP,MUL,DIV or SHIFT Error
Continuing ...

FilS CPUTST
The CPUTST dialogue first asks how many minutes you want the test to run.
The program then:
1.

Checks for expiration of this request time at eight strategic points during
each run sequence

Clears and/or sets three 2-dimensional arrays to specified values

3. Transposes Matrix A and places the result in Matrix B
4.

If no errors are found, returns to Step 2

5. If an error is found, reports the error and aborts the test
The program repeats this process until the run time expires, at which time it
prints the CPU time in seconds of expended job time.

F..5.1

CPUTST Dialogue
The following is an example of CPUTST dialogue:
$ RUN TEST$:CPUTST
CPUTST V9.0 RSTS V9.0 EDERE
How many minutes to run <lS>? 1
Detach <No>? NO
6.9 seconds of CPU time used.
CPUTST Finished at 10:S9 p~
$

Device Testing Package (DEVTST)

F-

F.S.2

CPUTST Errors
This device exerciser reports errors as shown in the following examples:
?CPUTST-Error in matrix transposition routine.
? (Error message) at line mm in CPUTS'I' V9. o.

where:
(Error Message)

is the RSTSIE message in Appendix A of the RSTS / E System
Installation and Update Guide for the error that occurred.

is the BASIC-PLUS line number in CPUTST where the error
occurred.

Check the System Error Log for reported errors. These errors may indicate
hardware problems.
If your hardware does not include a floating point processor (FPP), submit a
Software Performance Report (SPR) to DIGITAL whenever errors appear.

F.6 DSKEXR
This disk exerciser tests the normal operation of all fiRe-structured disks. It
is possible to run several copies of DSKEXR in order to test several drives
simultaneously or to put a heavier load on any single drive. The disk must
be physically and logically mounted. DSKEXR begins by asking questions to
determine the disk drive to be tested and the number of test iterations to be
performed.
Mter this dialogue, the exerciser:
1.

Opens and extends a temporary file in the current user's account on the
specified disk to a predetermined size that is dependent upon disk size and
the number of free blocks

Loads a pattern buffer with one of four patterns (all 1's and O's or a pattern of
alternating 1's and O's)

3. Writes the file
4.

Reads each block and compares it for each of the patterns.

Upon completion of all iterations for each drive you specify, DSKEXR prints a
status report.
NOTE

This write operation does not use CUITent disk files to read and write.
The temporary file will always be deleted after use, regardless of how
the DSKEXR program is terminated. Be sure to turn off data caching
before you run this test.

F-8

Device Testing Package (DEVTST)

F.6.1

DSKEXR Dialogue
The following is an example of DSKEXR dialogue:
$ RUNTEST$: DSKEXR
DSKEXR V9.0 RSTS V9.0 EDERE
Disk drive to test? DB1:
Number of test iterations <8>? 1
Detach <No>? NO
No errors detected on

OBI:

F.6.2

DSKEXR Errors
DSKEXR reports errors as shown in the following examples:
?(Error message) error at line mm in DSKEXR V9.0
?nn errors detected on DBn.

where:
(Error message)

is the RSTSIE message in Appendix A of the RSTS / E System
Installation and Update Guide for the error that occurred.

is the line number where the error occurred in the DSKEXR

is the number of errors detected by DSKEXR.

program.

When DSKEXR detects errors, do the following:
1.

Check to make sure that the volume is physically mounted and write-enabled

Check to make sure that the volume is logically mounted. This may be
verified by the use of the SHOW DISKS command. If the disk is not mounted
use the MOUNT command (see Chapter 15).
CAUTION

Note the following to avoid damage to the system:
.,

Packs mounted as unlocked public disks will be available to other
system users for creation and storage of their files. Subsequent
removal of a public disk may result in the elimination of these user
files from the system . For this reason, DIGITAL recommends the
use of a private pack if a nonsystem disk is to be tested .

N ever mount a potentially bad disk on another drive until you are
sure that the drive and pack are free of physical damage.

Persi~ltent errors may be due to bad blocks on the pack. Any potentially bad
blocks will be flagged in the System Error Log following unsuccessful access, to
prevent further access to the INIT option.

Device Testing Package (DEVTST)

F-~

F.7 DSKSEK
This disk exerciser tests the normal operation of non-file-structured disks by
randomly reading blocks from the disk. It is possible to run several copies of
DSKSEK in order to test several drives si.multaneously or to put a heavier load
on any single drive. If the disk being tested is a mounted RSTSIE file-structured
disk, DSKSEK will not read the known bad blocks. If an error occurs on a read
operation, DSKSEK reports it will be reported by DSKSEK and logs it into the
System Error Log.
NOTE

This is a read-only operation. Turn off data caching before you run this
test.

F.7.1

DSKSEK Dialogue
The following is an example of DSKSEK dialogue:
$ RUN TEST$:DSKSEK
DSKSEK V9.0 RSTS V9.0 ED ERE
Disk drive to test? DBO:
Number of test iterations <30>? 1
Detach <No>? NO
No bad block detections in 1000 sets from

DBO:

F.7.2

DSKSEK Errors
Errors detected by DSKSEK are displayed as follows:
?(Error message) error at line rom in DSKSEK V9.0.

where:
(Error Message)

is the RSTSIE message in Appendix A of the RSTS / E System
Installation and Update Guide.

is the BASIC-PLUS line number where the error occurred in
the DSKSEK program.

When DSKSEK detects errors, do the following:
1.

Check to see if the volume on the testing device is! physically mounted.

2. If it is, mount another scratch volume on the driv'B and retry the operation.
CAUTION

Never mount a potentially bad disk on another drive until you are
sure that the drive and pack are free of physical damage.
3. If the test continues to fail, check the System EITor Log to see if bad blocks
are recorded.

F-10

Device Testing Package (DEVTST)

If lbad blocks are recorded, add them to the bad block list on the disk using
the REFRESH BADS suboption of INIT. This can only be done if the disk
has been initialized (DSKINTed) as a RSTSIE disk. (See the RSTS / E System
Installation and Update Guide for information on BADS.)

5. If the problems continue, get a complete listing of the System Error Log
(using the ERRDIS prograrn).

F.B DXEXER
This exerciser tests the normal operation of the diskette controller and the
specified diskette drives. DXEXER starts by opening the floppy and writing
420 blocks of floating point numbers to it. These are written on the drive being
tested, then read back and verified. The program then displays the number of
words written and read, and the number of errors detected.
CAUTION

DXEXER destroys any information previously contained on the diskette
used for this test.

F.8·.1

DXEXER Dialogue
The foHowing is an example of DXEXER dialogue:
$ RUN TEST$:DXEXER
DXEXER V9.0 RSTS V9.0 EDERE
Warning: This exerciser will destroy data on the tested flexible
diskette.
Flexible drive to test? DXO:
Number of iterations <20>? 2
Detach <No>? NO
Flexible diskette DXEXER I/O data
Device
DXO:

Reads

Words
300

Writes

Words
76.8 K

300

76.8 K

DXEXER Finished at 10:53 AM
$

F.8.2

DXEXER Errors
DXEXER shows the number of errors detected during operation as shown in this
example:
?mm errors on DXn during iteration nn.

where:
mm

is the number of errors detected by the DXEXER program.

DXn

is the name of the device being tested.

is the number of the iteration during which the error occurred.

Device Testing Package (DEVTST)

F-11

When DXEXER detects errors, do the following:
1.

Check to see if the flexible diskette on the testing device is physically mounted
and the door closed.

2. If it is, mount another scratch flexible diskette on the drive and retry the
operation.

F.9 KBEXER
The keyboard exerciser (KBEXER) is used to test lo(:al and remote terminals
connected to the RSTS/E system. The three terminal tests are as follows:
•

The Space Test-Verifies that the terminal carriage will return reliably from
any position

•

The ASCII Pattern Test-Verifies that the terminal will print the standard
ASCII character set in all print positions

•

The Repeat Test-Repeats everything typed, one line at a time
NOTE

You can terminate the tests at any time by pressing CtrllC. You can
obtain HELP by typing HELP in answer to any dialogue question.

F.9.1

KBEXER Dialogue
The following is an example of KBEXER dialogue:
$ RUN TEST$: KBEXER
KBEXER V9.0 RSTS V9.0 EDERE
Keyboard to test <_KB68:>?
Test

<RET>

(SPA[CE], ASC[II], WOR[ST], REP [EAT],

Number of test iterations <8>?
Detach <No>?

HEI.[P])?

ASC

*** ROTATING ASCII CHARACTERS TEST ***
#$%&' () *+, -. /0123456789:; <=>? @ABCDEFGHIJKLMN'OPQRSTUVWXYZ [\] '"
! "#$%&
"#$%&' () *+, -. /0123456789: ;<=>? @ABCDEFGHIJKLMNOPQRSTUVWXYZ [\] A
! "#$%
!"#$%&' O*+,-./0123456789:;<=>? @ABCDEFGHIJKLt-1NOPQRSTUVWXYZ[\]A
!"#$
! "#$%&' () *+, -. /0123456789: ;<=>? @ABCDEFGHIJKIJMNOPQRSTUVWXYZ[\]A
! "#
! "#$%&' 0 *+, -. /0123456789:; <=>? @ABCDEFGHIJICLMN'OPQRSTUVWXYZ [\]A
!"
A
!" #$%&' () *+, -. /0123456789: ; <=>? @ABCDEFGHI~TKLMNOPQRSTUVWXYZ [\] ~
] '"
! "#$%&' () *+, -. /0123456789:; <=>? @ABCDEFGHIJKLMN'OPQRSTUVWXYZ [\]-;;:"
! "#$ %&' 0 *+, -. /0123456789: ; <=>? @ABCDEFGHIJKLMNOPQRSTUVWXYZ [\ ]"';;"
[\]-;;:"
! "#$%&' () *+, - ./0123456789:; <==>? @ABCDEFGHIJKLMN'OPQRSTUVWXYZ [\]-;;:"
Z [\]A
! "#$%&' () *+, -. /0123456789: ;<=>? @ABCDEE'GHIJKLMNOPQRSTUVWXYZ [\]
YZ [\] A_ ! "#$%&' () *+, -. /0123456789: i <=>? @ABCDE:FGHIJKLMN'OPQRSTUVWXYZ [\

\ fA

KBEXER Finished at 10:33 AM
$

F-12

Device Testing Package (DEVTST)

F.9.2

KBEXER Errors
Errors detected by KBEXER appear as errors in the keyboard test patterns.
Visually check the patterns for errors.

F.10 LPEXER
This exerciser prints a rotating pattern of ASCII characters on a specified number
of pages and is used to test any line printer.

F.10.1

LPEXER Dialogue
The following is an example of LPEXER dialogue:
$ RUN TEST$:LPEXER
LPEXER V9.0 RSTS V9.0 EDERE
Line printer to test? LPO:
Number of pages to output <4>? 1
Detach <No>? NO
LPEXER finished at 11:38 AM
$

F.10.2

LPEXER Errors
Problems found on the line printer appear as errors in the print-out of ASCII
characters. Visually check the line printer listing and if an error message is
printed by LPEXER, check to make sure that the line printer selected is on-line
and not out of paper.

F.111

MTEXER
The magnetic tape exerciser checks normal operation of the TMll or TSll
magnetic tape controller, the RHllIRH70ITM02ITM03 controller, and the
specified number of 9-track TElO, TUlO, TS03, TSll, TUl6, TEl6, TU45, TU77,
TUBO, TUBl, TK25, or TK50 drives. MTEXER allows you to select the drive to
be tested, the length of tape to be written and the number of iterations to be
perfonned. For each iteration, MTEXER initializes the tape, sets the drive to its
maximum legal density, opens a file and writes data until the specified length of
tape has been used. MTEXER then rewinds the tape, opens the file for input,
reads and verifies the data. If errors are detected, MTEXER prints a count of the
number of bytes found to be incorrect and then proceeds to the next iteration.
The data pattern used is a worst case NRZI pattern for 9-track drives. The
pattern is loaded into a 5l2-byte buffer, and successively greater numbers of
records are written to the magtape until the requested length of tape has been
written. Furthermore, the pattern base varies with the iteration number so the
contents of the pattern buffer also varies on successive iterations.
CAUTION

This exerciser overwrites data contained on the magnetic tape used for
the test.
Device Testing Package (DEVTST)

F-13

F.11.1

MTEXER Dialogue
The following is an example of MTEXER dialogue:
$ RUN TEST$:MTEXER
MTEXER V9.0 RSTS V9.0 EDERE
Warning: This exerciser will destroy data on the tested magtape.
Magtape drive to test? MT2:
Drive type is TE16/TU45/TU77
Density (800 or 1600) <800>? <RET>
Number of feet to test <100>? <RET>
Number of iterations <2>? 1
Detach <No>? NO
Magtape MTEXER I/O Data
Device
MT2:

Reads

Words Writes
387

No errors detected on

Words
99 .. 072 K

387

99.072 K

MT2:

The details of the dialogue vary with the type of drive used. For instance, if you
use a drive with only one possible density, the dialogue does not include a density
question like the one in the example. Also, the question about the number of feet
does not reflect the actual nwnber offee"t on a TK50 tape.

F.11.2

MTEXER Errors
Magtape errors are reported in these formats:
?Magtape select error at line nn in MTEXER V9.0.
?MTEXER-Selection error occurred.

where:
nn

is the line number in MTEXER at which the error occurred.

When MTEXER detects errors, do the following:
1.

Check to see if the volume on the testing device is physically mounted,
write· enabled, and assigned to you.

2. If it is, mount another scratch volwne on the drive and retry the operation.
3.

F-14

Check to make sure that the tape path and heads are free of oxide.

Device Testing Package (DEVTST)

Index
A
ABORT command
OPSER, A-45
SPOOL, A-26 , A-35
/A8:0RT qualifier
STOP/QUEUE/MANAGER command, 9-45
Accessing entries
entry number, 9-7
entry specification, 9-7
Acc:ount
associated with system logical name, 12-8
nonuser, 7-1
programming error in privileged, 2-5
user, 7-1
Acc:ount attribute
captive flag, 7-4
dial-up flag, 7·-4
network flag, '7-4
/ACCOUNTING qualifier
RESTORE command, '16-2
/AGCOUNTING_DATA qualifier
SHOW ACCOUNT command, 7-19
Account management
GACNT privilege, 6-3
WACNT privilege, 6-7
Account management, DCL commands, 7-H
/AGCOUNT_DATA qualifier
BACKUP command, 16-5
RESTORE command, '16-13
Action requests
BATCH, A-41
OPSER, A-10
table of, A-10
/ ADDRESS qualifier
CREATEIVIRTUAL_DISK command, 11-15
INSTALULIBRARY command, 12-23
INSTALURUNTIME_SYSTEM command, 12-26
/ADVANCED_VIDEO qualifier
SET TERMINAL command, 8-16
AFrER state, processing entries, 9-6
/ ALIGN switch, SPOOL, A-37
ALLOCATE command, 11-13
/ ALLOCATED qualifier
SHOW DEVICE command, 11-28
/ ALL qualifier
BROADCAST command, 13-9
DEALLOCATE command, 11-16
DEASSIGN/QUEUE command, 9-22
DELETE/COMMAND/SYSTEM command, 12-20
FORCE command, 13-12
LOAD/OVERLAY command, 14-23

/ALL qualifier (cont'd.)
SET CACHE command, 11-20
SET QUEUE command, 9-32
SET SERVER command, 9-35
SHOW ACCOUNT command, 7-19
SHOW DEVICE command, 11-28
SHOW FILE/SYSTEM command, 12-38
SHOW LOGICAUSYSTEM command, 12-41
SHOW LOGICALS command, 12-40
SHOW PRINTER command, 11-30
SHOW SERVER command, 9-37
SHOW TERMINAL command, 8-30
UNLOAD/OVERLAY command, 14-32
/ALT_MODE qualifier
SET TERMINAL command, 8-16
ANALYS
CRASH.SYS, 12-16
crash dump data, 10-23
crash error code, 10-23t
create error logging file, 10-1
dialogue questions, 10-22t
output, 10-23
run, 10-21
running after crash, 3-9
START.COM commands, 10-23
use, 10-21
/ANSI qualifier
SET TERMINAL command, 8-16
/ANSWERBACK qualifier, 14-21
SET SYSTEM command, 14-25
SET TERMINAL command, 8-16
ANSWER command
BATCH, A-41
OPSER, A-11
ASSIGN/QUEUE command, 9-20
ASSIGN/SYSTEM command, 12-9, 12-17
ASSIGN option, SPOOL, A-24
ATTACH command, 13-8
Attributes, printer forms, 9-13
/AUTOBAUD qualifier
SET TERMINAL command, 8-1, 8-16
TERSPD.SYS file, 8-2
Automatic restart
after power fail, 2-6
procedures, 2-2
RSTS/E system, 2-2
system initialized, 2-5
Auxiliary run-time system, 12-3
adding, 12-4
creation of, 12-4
removing, 12-4
START. COM, 12-4

Index-1

8
IBACKSPACE qualifier
SET PRINTER command, 11-23
START/SERVER command, 9-42
BACKUp, comparison with SAVRES, E-1
BACKUP command, 16-3 to 16-11
quick reference table, 16-10t
Backup sets
restrictions, 16-4
BADB.SYS file
adding bad blocks, 10-8
Bad blocks
adding, caution, 10-8
adding to BADB.SYS, 10-8
criteria, 10-7
during RESTORE, E-8
ERRDIS example, 10-8
ERRDIS prints list, 10-8
ERRDIS report, 10-7
list of potential, 10-4
report content, 10-7
SAVRES file changes, E-24
SAVRES output, E-24
Base priority
SET JOB/PRIORITY command, 6-6
TUNE privilege, 6-6
UU.PRI system directive, 6-6
BASIC-PLUS
keyboard monitor, 1-2
run-time system, 12-3
BATCH, A--1
ANSWER command, A-41
answering action requests, A-41
communication paths, A-5
CONTINUE command, A-42
default conditions for, A-38
default receiver identification, A-38
device type designators, A-41t
dismount a volume, A-42
error checking, A-38
interrupt commands, A-40t
logical device name, A-38
modules, A-37
mount requests, A-41
OPSER communicate with, A-40
OPSER package, A-37
overview, A-5
processing command file, A-41
processor, A-37
pseudo keyboard, A-38
requests for operator action, A-41
run, A-37
startup options, A-39t
startup procedures, A-41
BATCH.DCD file
command decoding, A-5
Batch jobs, exit status messages, 9-12
IBATCH qualifier
INITIALIZE/QUEUE command, 9-26
SET SERVER command, 9-35
SHOW SERVER command, 9-37
Batch servers
CPU limits, 9-12
Print/Batch Services (PBS), 9-2,9-11
pseudo keyboards, 9-12
time limits, 9-12

Index-2

Batch servers (cont'd.)
wildcard names, 9-11
IBEll qualifier
BROADCAST command, 13-9
Bit 0, setting, 2-5
IBlOCK_SIZE qualifier
BACKUP command, 16--5
Bootable medium, create with SAVRES, E-18
Bootstrap
RSTS/E after system hd, 2-2
RSTS/E into memory, 2-1
Bootstrap procedure, 2-1
IBREAK qualifier
SET TEHMINAl command, 8-17
IBRIEF qualifier
BACKUP command, 16--5
RESTOFIE command, 16-13
SHOW ACCOUNT command, 7-20
SHOW SERVER command, 9-37
SHOW TERMINAL command, 8-30
BROADCAST command, 13-3,13-9
SEND privilege, 6-5
IBROADCAST qualifier
SET TEHMINAl command, 8-17
IBROADCAST qualifier, permits receiving broadcasts.
8-6
Bucket size
cache, 11-10
RMS, 11-11
Buffer status report, SHOW BUFFERS command,
12-15
IBUFFER_QUOTA qualifier
SET TEBMINAl command, 8-17
IBUFFER_SIZE qualifier
BACKUP command, 16--5
RESTORE command, 16-13
Burst pages, A-25
BYE command, CCl, 3-8

c
Cache
clusters in, 11-8
data block, 11-7
size of, 11-8
space for, 11-8
updated by the monitor, 11-8
Cache cluster
eligible for replacement, 11-9
last block in, 11-9
size of, 11-8,11-19
specify size of, 11-20
ICACHE qualifier
SET FilE command, 12-35
Caching
checks by monitor, 11-"191
directory, 11-7
efficiency of, 11-10
guidelines, 11-10, 11-11
MODE checks, 11-10
optimize directory, 11-11
override UFO entry, 11-·10
random mode, 11-8
RMS indexed file, 11-"1"1
sequential mode, 11-9
types, 11-7
UFO entry marked for, '11-9

CatChing Control, DCl commands, 11-H
Captive account flag, 7-4
/CAPTIVE qualifier
CREATE/ACCOUNT command, 7-6
SET ACCOUNT command, 7-14
Catastrophic error, 2-2, 2-5
automatic recovery, 2-4,2-5
cause, 2-4, 2-5
handling, 2-5
recovery, 2-5
CCl (Concise Command language)
adding CCl commands, 12-11
adding in sequence, 12-11
BYE command, 3-8
command definition, 3-·8
defined at start of time sharing, 12-10
definition, 1-1
error checking, 12-11
HEllO command, 3-8
listing, 12-12
PLEASE, A-44
redefining, 3-8
removing CCl commands, 12-13
run RSTS/E programs, 12-10
use of small buffers, 3-·8
CCIL command, definition, 12-10
CHANGECONSOL.E command
OPSER, A-13
CHARS.QUE file, A-35
,character generation, A-4
CLOSE/QUEUE command, 9--21
/ClOSE qualifier
DETACH command, 13-11
INITIALIZE/QUEUE command, 9--26
Cluster size
Icache and pack, relationship, 11-9
Icache and RMS bucket Isize, 11-10
disk type, 15-8t
/ClUSTERSIZE qualifier
:SET CACHE command, 11-20
/ClUSTER_SIZE qualifier
CREATE/ACCOUNT command, 7-7
IINITIALIZE command, '15-7
/13:2_COlUMNS qualifier
SET TERMINAL command, 8-16
Command file
aUEMAN startup, A-19
:system startup, 3-2
Command level
Isystem, 1-11
jterminal at, 1-·11
Command line, full function (RESTORE), E-18
/CONFIRM qualifier
IREMOVE/JOB command, 13-16
Console terminal
lability to log on to, 13-2
I::hange, 13-2
CONTINUE command
BATCH, A-42
Controlled access
privilege, 1-12
/CONTROl qualifier
SET TERMINAL command, 8-17
/CONTROlS qualifier
IINITIALIZE/SERVER command, 9--29
SET PRINTER command, 11-23
SET SERVER command, 9--35

CPEXER program
CPU exerciser, short, F-4
CPU integrity test, F-4
EIS operation test, F-4
example dialogue, F-5
J=pp test, F-4
lOG feature requirement, F-5
PEEK function test, F-4
tests performed, F-5
TRIG feature requirement, F-5
type four errors, F-7
type one errors, F-6
type three errors, F-6
type two errors, F-6
CPU (Central Processing Unit)
determine the state of, 2-3
CPU exerciser, F-4
CPU integrity test
CPEXER program, F-4
CPUTST program, F-7
errors, F-8
example dialogue, F-7
/CPU_L1MIT qualifier
INITIALIZE/QUEUE command, 9--26
SET QUEUE command, 9--33
Crash
analyzing system, 10-21
causes of system, 2-5
documenting system, 10-21
error code, 10-23
handling of system, 2-5
memory dump, annotated version, 10-22
OPSER starts after system, A-14
random system, 2-5
running QUEMAN after system, A-20
system, 2-2, 2-5
CRASH.SYS file, 10-1, 10-21
crash dump, 12-16
Crash file
CRASHSYS, 12-16
information from, 10-1
/CRC qualifier
BACKUP command, 16-6
CREATE/ACCOUNT command, 7-5
CREATE/SERVICE/LAT command, 8-9
CREATENIRTUAl_DISK command, 11-12, 11-14
Create access
GWRITE privilege, 6-3
WWRITE privilege, 6-7
/CREATED qualifier
BACKUP command, 16-5
RESTORE command, 16-13
/CRFlll qualifier
SET TERMINAL command, 8-17
/CR_PREFIX qualifier
SET PRINTER command, 11-24
CtrllC
do not use in QUEMAN, A-16
FORCE, 13-4
SAVRES dialogue, E-5
typed at terminal, 1-11
CtrllZ
FORCE command, 13-4
SAVRES dialogue, E-5
CUSP (Commonly Used System Program)
definition, 1-2

Index-3

D
Data access, cached, 11-8
Data block
cache, 11-7
retrieval pointer, 11-2
Data caching
cache clusters used, 11-20
control, 11-7
disabling, 11-20
enable on the system, 11-18
guidelines, 11-9
memory access, 11-7
minimum residency time, 11-11
MODE values, 11-7
monitor directives, 11-7
read operations, 11-7
restriction with DSKSEK program, F-10
RMS flies, 11-11
support for, 11-7
write operations, 11-7
XBUF allocation, 11-8
/DATA_MAXIMUM qualifier
SET CACHE command, 11-20
/DATE qualifier
INITIALIZE command, 15-7
DATES privilege
system clock, 6-3
/DATE_FORMAT qualifier
SET SYSTEM command, 14-26
DCL (Digital Command Lat"guage)
definition, 1-1
keyboard monitor, 1-2
DCL (DIGITAL Command Language)
caution when defining, 3-8
CCL commands, 3-8
DCL command descriptions
account management, 7-5 to 7-20
BACKUP, 16-3 to 16-11
caching control, 11-12 to 11-30
device management, 11-12 to 11-30
disk handling, 15-3 to 16-1
job management, 13-7 to 14-1
LAT management, 8-9 to 9-1
managing logical names, 12-17 to 13-1
managing system files, 12-17 to 13-1
PrintiBatch Services Package (PBS), 9-19 to
9-47
RESTORE, 16-11 to 16-18
system functions, 14-22 to 15-1
tape handling, 15-3 to 16-1
terminal management, 8-9 to 8-29,9-1
DCL commands
account management, 7-H
disk handling, 15-H
job management, 13-1t
logical names, managing, 12-1t
PrintiBatch Services Package (PBS), 9-1t
system flies, managing, 12-1t
system functions, 14-1t
tape handling, 15-1t
DEALLOCATE command, 11-16
DEASSIGN/QUEUE command, 9-22
DEASSIGN/SYSTEM command, 12-18
logical name, 12-10
DECnetiE
crash analysis, 10-23

Index-4

DECnetiE (cc:mt'd.)
disabling
during shutdown, 4--4, 4-6
logins, 4-3
EVTLOG
program, 4-6
shutdown phase, 4·-8
interrupt shutdown, 4-6
Network Services ProtocoII file, 12-15
resetting shutdown status, 4-6
shutdown, 4-6
example, 4-12
phase, 4-5
use of NCP, 4-6
/DEFAULT qualifier
INITIALIZE/QUEUE command, 9-26
SET QUEUE command, 9-33
Default queues
PrintiBatch Services, 9-3
DEFINE/COMMAND/SYSTEM command, 12-10,
12-11, 12-19
/DELETABLE qualifier
SET FILE command, 12-35
DELETE/ACCOUNT comma!l1d, 7-10
DELETE/COMMAND/SYSTEM command, 12-13,
12-20
DELETE/QUEUE command, 9-23
DELETE/SERVER command, 9-24
DElETE/SERVICE/LAT command, 8-11
DELETENIFlTUAL_DISK command, 11-12,11-17
Delimiter
keyboard, 13-6
/DELIMITER qualifier
SET TEF~MINAl command, 8-17
/DENSITY qualifier
BACKUP command, 16--6
INITIALIZE command, 15-12
SET SYSTEM command, 14-26
/DENSITY switch, SAVRES, E-6
Description block, run-time system, 12-4, 12-5
DETACH command, 13-7, 13-11
OPSER, A-13
QUEMAN, A-16
Detached job
check with SYSTAT, 14·-8
SHOW JOBS/DETACHED command, 14-8
Detached-job quota, 5-3
/DETACHED_JOB_QUOTA qualifier
CREATE/ACCOUNT command, 7-7
SET ACCOUNT command, 7-14
Device designators, F-3t
table of BATCH, A-41t
Device management, DCl commands, 11-1t
DEVICE privilege
restricted device, 6-3
/DEVICE qualifier
INSTALL/SWAP_FilE command, 12-28
SHOW COUNTERS/LAT command, 8-24
SHOW SESSIONS/LAT c:::ommand, 8-27
Device sizes, disk type, C-.2t
Device testing package
definition, F-1
Device Testing Programs
see DEVTST programs
/DEVICE_TYPE qualifier
SET PRINTER command, 11-24
SET TERMINAL command, 8-18

DEVTST programs, F-2t
Iconventions, F-2
lerror messages, F-4t
how to run, F--2
/DFLENGTH switch
,effect on line printer, A--26
,effect on terminals, A-26
SPOOL, A-25
Diall-up flag
:account attribute, 7-4
dial-up line
Iring characteristics, 8-5
,set characteristics, 8-5
/DIALUP qualifier
CREATEIACCOUNT command, 7-7
SET ACCOUNT command, 7-14
SET TERMINAL command, 8-18
DirElctory
fragmentation, 1-11
minimize search overhead, 1-10
reducing accesses, 11-2
restructure with REORDR, 11-2
DirElctory caching, 11-7
4~ache clusters used, 11-20
43nable on the system, '11-18
GFD, 11-7
MFD, 11-7
4)ptimize, 11-11, 12-13
4)verlay code, 11-7
UFO, 11-7
/DIHECTORY qualifier
BACKUP command, 16-6
RESTORE command, 16-14
DirElctory structure
4)ptimizing on disk, 11-2
prevent a damaged, 11·-4
/DIHECTORY_MAXIMUM qualifier
SET CACHE command, 11-20
Dirty disk
,'ebuilding with MOUNT, 15-16
when dirty bit is set, 15-14
/DISABLE qualifier
SET CACHE command, 11-20
SET DEVICE command, 11-22
SET NODEILAT command, 8-12
Dislt
4::opying with IMAGE, E·-13
c::reate fast copy, E-1
differences between Level 0 and Level 1, 1-7
directory entries scattered, 11-2
dismount under SHUTUP, 4-10
t:ormatting with DSKINT option of INIT.SYS, 15-2
t:ree space, 14-8
IMAGE output as system, E-13
legal with IMAGE, E-13
MFD on system, 1-11
optimizing, 1-" 0
directory structure, 11-2
organization, 1-3
options, 1-3
private, 1-3, 1-10
rebuilding a dirty disk with MOUNT, 15-14
rebuilding with MOUNT command, 15-16
reorder disk structure, 11-4
reorder in public structure, 11-4
restore with SAVRES, E-1
restoring, E-8

Disk (cont'd.)
restructure directories, 11-2
SAVE Set, E-1
statistics (DISPLY), 14-16
status statistics, 14-16
stopping an RC25, 14-2
system, 1-2, 1-3
types, 1-3
virtual, 11-12
Disk characteristics, INITIALIZE display, 15-11
Diskette controller test
DXEXER program, F-11
Disk exerciser test
DSKEXR program, F-8
Disk handling, DCL commands, 15-1t
Disk quotas, 5-1
attribute block, 5-1
logged-in, 5-2
logged-out, 5-2
operations, 5-2
Disk size, disk type, 15-8t
Disk type, device sizes, C-2t
DISMOUNT command
disks, 15-3
tapes, 15-5
DISPLY
abbreviations in DISPLY report, 14-14
account location, 14-9
buffer status report, 12-15
busy device statistics, 14-16
commands, 14-11t to 14-12t
detached, 14-9
disk status abbreviations, 14-17t
disk structure statistics, 14-16
example, 14-13
free buffers, 14-18
information displayed, 14-10
INTERVAL question, 14-9
job status statistics, 14-14
memory status, 14-20
abbreviations, 14-21 t
message receivers, 14-17
abbreviations, 14-18t
resident libraries, 14-20
abbreviations, 14-19t
restrict use, 14-9
running, 14-9
run-time systems, 14-18
abbreviations, 14-19t
screen layout, 14-13
STATE column abbreviations, 14-15t
stop execution, 14-10
switches, 14-9t
titles used in report, 14-14
use, 14-7
WHY column abbreviations, 14-16t
DSKCVT program, 1-7 to 1-10
DSKEXR program
disk exerciser test, F-8
errors, F-9
example dialogue, F-9
DSKSEK program
data caching restriction, F-10
error messages, F-10
example dialogue, F-10
non-file-structured disks test, F-10
DUMP/SYSTEM command, 12-16, 12-21

Index-5

DXEXER program
diskette controller test, F-11
error messages, F-11
example dialogue, F-11
Dynamic regions, 5-4
IDYNAMIC_REGION_L1MIT qualifier
SET SYSTEM command, 14-26

E
lEIGHT_BIT qualifier
SET PRINTER command, 11-24
SET TERMINAL command, 8--18
EIS operation test
CPEXER program, F-4
Electronic messaging systems, 14-21 to 14-22
EMT
definition, 14-4
that cannot be logged, 14-5
EMT logging, 14-4
data contained in packet, 14-6
data returned, 14-6
how to program, 14-4
packet, 14-5
reasons for using, 14-4
shutdown under SHUTUP, 4-9
system security, 14-7
IEMT_LOGGING qualifier
SET SYSTEM command, 14-26
SHOW SYSTEM command, 14-30
IENABLE qualifier
SET CACHE command, 11-21
SET DEVICE command, 11-22
SET NODEILAT command, 8--12
END command, OPSER, A-44
lEND qualifier
BACKUP command, 16-7
RESTORE command, 16-14
Entry number, accessing entries, 9-7
Entry specification, accessing entries, 9-7
IEOT qualifier
SET PRINTER command, 11-24
IERASE qualifier
INITIALIZE command, 15-7
ERR.SYS file
allocate and position, 12-13
ERRCPY
activating, 10-1
active during time sharing, 10-1
error message from SHUTUP, 10-3
messages queued, 10-3
minimize size, 10-1
number of messages queued, 10-1
processing error messages, 10-3
shutdown procedures, 4-9
shutdown under SHUTUP, 4-9
total errors received, 10-10
use, 10-1
ERRCRS.FIL
error logging file, 10-1
ERRDET
criteria for bad blocks, 10-7
report content, 10-7
ERRDET module
ERRDIS, 10-4
ERRDIS
bad blocks, 10-8

Index-6

ERRDIS
bad blocl(s (cont'd.)
example, 10-8
potentially, 10-4
rep()rt, 10-7
categories of errors, 10--9
detailed report, 10-4
dialogue explanations, 10-5t
disk error description, 10-13t
error code mnemonic, 10-9
error file, zero contents, 10-4
error sequence number', 10-9
error title line, 10-9
full report, 10-9
examples, 10-10 to 10-19
functions, 10-4
modes, ()ptional, 10-5
modules, 10-4
nondisk peripheral device format, 10-17,10-18t
nonperipheral error
description, 10-19
format, 10-20t
running, 10-4
summary report, 10-4, 10-6
user description data, 10-10, 10-11t
zeroing the log, 10-20
ERRINT
chains to ERRCPY, 10--3
change the maximum size, 10-2
command fife, 10-1
example of dialogue, 10-3
location of program, 10--3
run, 10--1
run at startup, 3-7
validates error file, 10-"
ERRLOG.FIL file, 10-1
stores error messages, 10-1
Error file
adding, 12-16
initialization, 10-1
removing, 12-16
SHOW F:ILEISYSTEM c()mmand, 12-15
validation, 10-1
zero, 10-4
Error logging, 2-5
creation of file, 10-1
ERRCPY, 10-3
hardware errors, 10-1
initial conditions for system, 3-7
initialization, 10-1
software errors, 10-1
use of programs, 10-1
Error messages
processing, 10-1
processing with ERRCPY, 10-3
SAVRES, E-25t
nonfatal transfer, E-29
SPOOL, A-36
syntax, A-28 , A--29t
Error package
programs in System, 10-1
Errors
catastrophic, 2-2,2-5
automatic recovery, 2-5
causes, 2-5
handling, 2-5
categork~s of reported (ERRCPY), 10-9

Err'ors (cont'd.)
checking ceL commandi, 12-11
detailed report, 10-4
detection in RSTS/E monitor, 10-1
disk (ERROIS). 10--9
displaying, 10-4
example of MSCP in full report (ERROIS), 10-15
to 10--17
logging, 10--1
missed (ERRCPY), 10--9
MSCP in full report (ERROIS), 10-14
nondisk peripheral device, 10--9
nonperipheral, 10-9, 10-19
privileged job programming, 2-4
recovery from line printer, A-34
SAVRES, E-24
SAVRES input volume, E-7t
summary report, 10-4
summary report (ERROIS), 10-6
total received by ERRCPY, 10-10
Em)r trap
infinite loop, 2-5
two categories, 2-5
unexpected, 2-5
IERROR_FILE qualifier
SHOW FILE/SYSTEM command, 12-38
IEAIROR_LOGGING qualifier
INSTALULIBRARY command, 12-23
INSTALURUNTIME_SYSTEM command, 12-26
IESCAPE_SEQUENCE qualifier
SET TERMINAL command, 8--18
Ethernet, 8--6
EVEmt logger
OECnetlE, 4-6
EVrLOG program
OECnetlE example, 4-12
OECnetlE shutdown, 4·-6
resetting shutdown status, 4-6
shutdown phase, 4-8
IE)(CLUDE qualifier
BACKUP command, 16-7
RESTORE command, '16-14
EXElcute access
GREAD privilege, 6-3
WREAO privilege, 6-7
EXElcutive
definition, 1-2
IE)(ERCISE qualifier
INITIALIZE command, 15--8
ExiR status, Batch jobs, 9-'12
IE)(PIRE qualifier
CREATE/ACCOUNT command, 7-7
SET ACCOUNT command, 7-14
EXCITA privilege
logged-in quota, G-3
logged-out quota, 6-3
PrintlBatch Services (PBS), 9-5
Extended Instruction Set
see EIS

F
FOF
form characteristics, 9-,13
lFORMS qualifier, 9-13
PrlntlBatch Services (PBS), 9-14

File
accounting information, A-35
add swap file, 12-13
bad block, 10-8
bad cluster while transferring contiguous, E-9
caching, 11-11
contiguous, 11-11
random mode, 11-8
sequential mode, 11-9
cannot remove or add NSP, '12-15
commands
startup, A-13
directory entries, 11-2
disk structure, 1-3, 1-11
high access (definition), 11-11
minimize access overhead in data, 1-11
Network Services Protocol, 12-15
read from sequentially cached, 11-9
remove swap file, 12-13, 12-15
run request with no file type, 12-5
SIL, 12-7
structures, 1-4
swap file, 1-12
system control, 12-13
File access
SYSIO privilege, 6-5
IFILE qualifier
SET CACHE command, 11-21
IFILE_DATA qualifier
BACKUP command, 16-7
RESTORE command, 16--14
IFILE_ENO qualifier
STOP/SERVER command, 9-47
IFILL qualifier
SET PRINTER command, 11-24
FI P small buffer
OISPLY report, 14-18
use for CCL, 3-8
FMS-11 , 8--9
FORCE
CtrVC, 13-4
example, 13-4
FORCE command, 13-3, 13-12
IFORMAT qualifier
INITIALIZE command, 15--13
Formatting
disks requiring system shutdown, 15--2
Form characteristics
Forms Definition File (FDF), 9-13
FORM command
SPOOL, A-36
Form length
line printer, A-26
FORM option
SPOOL, A-24
Forms alignment
changing, A-36
during time sharing, A-25
procedures, A-25
Forms control, A-25
Forms Management System
See FMS-11
/FORMS qualifier
Forms Definition File (FDF), 9-13
INITIALIZE/QUEUE command, 9-26
INITIALIZE/SERVER command, 9-29
SET QUEUE command, 9-33

Index-7

IFORMS qualifier (cont'd.)
SET SERVER command, 9-35
IFORM_FEED qualifier
SET TERMINAL command, 8-18
IFORWARDSPACE qualifier
START/QUEUEIMANAGER command, 9-42
FPP test
CPEXER program, F-4
IFULL qualifier
BACKUP command, 16-7
RESTORE command, 16-14
SHOW ACCOUNT command, 7-20
SHOW SERVER command, 9-38
SHOW TERMINAL command, 8-30

G
GACNT privilege
account management, 6-3,6-7
General small buffer
DISPLY report, 14-18
missed errors, 10-10
SPOOL error, A-29
use for CCL, 3-8
GFD (Group File Directory)
with MFD and UFDs in Level 1, 1-6f
GREAD privilege
execute access, 6-3
read access, 6-3
IGROUP_SIZE qualifier
BACKUP command, 16-7
GWRITE privilege
create access, 6-3
rename access, 6-3
write access, 6-3

H
HANGUP command, 13-7, 13-13
IHANGUP qualifier
SET SYSTEM command, 14-26
IHARDCOPY qualifier
SET TERMINAL command, 8-18
Hardware
malfunctions, 2-4
Hardware controller
terminal device designator, 8-2
Hardware errors
logging, 10--1
IHEADER qualifier
BROADCAST command, 13-10
Heading burst pages, A-25
HELLO command, CCL, 3-8
Help report
ERRDIS (example), 10-6
IHOLD qualifier
SET JOB command, 13-17
SET SYSTEM command, 14-26
HOLD state
processing entries, 9-6
IHOSTSYNC qualifier
SET TERMINAL command, 8-18
HWCFG privilege
SET TERMINAL command, 8-3
system manager function, 6-3
terminal characteristics, 8-3

Index-8

HWCTL privilege
system control, G-4

IIDENTIFICATION qualifier
CREATEJSERVICEILAT ,command, 8-9
SET NODEILAT command, 8-12
SET SERVICEILAT command, 8-13
IDENTIFY, E-3t
dialogue, E-17t
example, E-17
IMAGE, E--3t
dialogue, E-13t
example, E-15 to E-16
extracting pack ID, E--17t
output from, as system disk, E-13
run statistics, E-23
SAVRES, E-13
Immediate mode
shutting down OPSER, 4-8
IINCLUDE qualifier
BACKUP command, 16-7
RESTOHE command, 16-14
Indexed file
caching, 11-11
IINDEX qualifier
INITIALIZE command, il5--9
INIT
command, 2-5
start OPSER, A-13
INIT.BAS program
OPSER shutdown conditions, 4-8
INIT.SYS fIIo
copies to disk, E-18
INIT.SYS program
control file, 3-1
DSKINT option to format disks, 15--2
use, 2-1
Initialization code
bootstrapped, 2-1
RSTS/E, 2-2
START option, 3-1
Initialization options
discussion, 1-2
requesting, 2-3
summary, 2-3t
INITIALlZElQUEUE command, 9-25
INITIALIZE/SERVER command, 9-29
INITIALIZE command
IDENSITY qualifier, 15-·12
disks, 15-6 to 15--11
display olf disk characteristics, 15--11
IFORMAT qualifier, 15-13
QUEMAN, A-15
tapes, 15--12
IINITIALIZE qualifier
BACKUP command, 16-8
RESTORE command, 16-14
Initializing disks
INITIALIZE command, 15-6 to 15--11
overview, 15--1
Initializing tapes
INITIALIZE command, 15--12
overview, 15--1
IINQUIRE qualifier
SET TEHMINAL command, 8-19

INSTALUERROR_FILE command, 12-13, 12-16,
12-22
INSTALULIBRARY
IDCL command, 12-7, 12-8
lLOCK qualifier, 12-8
l[NO]ADDRESS[=n] qualifier, 12-7
INSTALULIBRARY command, 12-23
INSTALUOVERLAY_FILE command, 12-13,12-16,
12-25
INSTALURUNTIME_SYSTEM
lADDRESS=n qualifier, 12-4
lLOCK qualifier, 12-6
lMAXIMUM:=s qualifier, 12-6
lMINIMUM=s qualifier, '12-6
/[NO]ERROR_LOGGING qualifier, 12-4
/[NO]KEYBOARD_MONITOR qualifier, 12-4
/[NO]READ_ONLY qualifier, 12-4
/[NO]SHAREABLE qualifier, 12-4
/[NO]UNLOAD qualifier, 12-4
/POSITION qualifier, 12-5
/PREFIX=n qualifier, 12-4
INSTALURUNTIME_SYSTEM command, 3-7,12-4,
12-26
,errors, 12--5
INSTALURUNTIME_SYSYTEM
lTYPE qualifier, 12-6
INSTALUSWAP_FILE
,errors, 12-15
lSIZE=n qualifier, 12-14
INSTALUSWAP_FILE command, 12-13, 12-14,
12-28,13-2
INSTAL privilege
defining CCLs, 12-1 0
system commands, 6-4
system flies, 6--4
system logical names, 6--4
/INTERACTIVE qualifier
CREATE/ACCOUNT command, 7-7
SET ACCOUNT command, 7-14
SET TERMINAL command, 8-19
INTERRUPT command
OPSER, A-11 , A-12, A-17t, A-44
response to, A-12
IntElrrupt commands
BATCH, A-40t
SPOOL, A-30
/IN__LABEL qualifier
BACKUP command, 16-8
RESTORE command, '16-14
/IN__QUOTA qualifier
CREATE/ACCOUNT command, 7-7
SET ACCOUNT command, 7-14

J
Job
accounting information, A-35
check detached, 14-8
check number of, 14-8
compute bound, 13-6
definition, 1-1
executing a read request, 11-7
header, A~5
keyboard monitor, 1-1
limit of spooling, A-20
maximum (JOB MAX), 12-14
maximum assigned size, 13-7

Job (cont'd.)
number allowed, 12-14
OPSER checks online, A-5
OPSER table, A-2
priority, 13-6
run burst, A-32
set maximum size, 13-5
setting priority, 13-5
size, 2-6
status statistics, 14-14
suspend, 13-5
Job area
size, 1-11
user, 1-11
Job control
SHOW USER command, 13-4
JOBCTL privilege
system control, 6-4
JOB MAX
restricts jobs, 12-15
IJOB qualifier
ALLOCATE command, 11-13
Job quota, 5-3
detached-job, 5-3
log-in request, 5-3
Job quotas, 5-1
Job run burst, 13-5
Job status, SYSTAT, 14-8
Job table, SHUTUP, 4-6
/JOB_END qualifier
STOP/SERVER command, 9-47
/JOB_QUOTA qualifier
CREATE/ACCOUNT command, 7-7
SET ACCOUNT command, 7-14

K
/KATAKANA qualifier
SET TERMINAL command, 8-19
KBEXER program
errors, explanation, F-13
example dialogue, F-12
local terminals test, F-12
remote terminals test, F-12
terminal tests, F-12
/KEBOARD_MONITOR qualifier
SET JOB command, 13-17
IKEEP=n qualifier
minimum residency time, 11-11
IKEEP qualifier
SET CACHE command, 11-21
Keyboard
controlling, 13-7
default SPOOL values, A-28
delimiter, 13-6
monitor, 1-11
spooling, A-28
startup on LA 180, A-33
startup on LA36, A-32
Keyboard monitor, 3-8
change with SWITCH, 1-1
default (definition), 1-1
definition, 1-2
job, 1-1
process command, 1-11
IKEYBOARD_MONITOR qualifier
INSTALURUNTIME_SYSTEM command,

12-27

Index-~

KILL
UTILTY program, A-45

L
LA180
form length, A-26
startup keyboard on, A-33
LA36
startup keyboard on, A-32
Label, specified for tapes
DISMOUNT command, 15-5
INITIALIZE command, 15-12
MOUNT command, 15--19
/LABEL qualifier
SET SYSTEM command, 14-27
LAT, ~6 to ~
enabling at startup, 3-9
enabling LAT, ~7
LAT problems, 8-8
log ins, 13-2
/LAT qualifier
SET SYSTEM command, 8-14, 14-27
SHOW SESSIONS/LAT command, 8-28
Level 0, 1-5f
compared to Level 1, 1-7
Level 1
compared to Level 0, 1-7
three structures involved, 1-6
Library
remove resident, 12-8
resident, 12-7
LIBRARY keyword, 12-7
line printer
default SPOOL values, A-27
/DFLENGTH switch effect, A-26
distinction from terminal, A-27
form length, A-26
output, A-35
recovery from errors, A-34
requests for output, A-4
spooling, A-21, A-27
startup
all defaults, A-31
narrow width, A-31
line printer test
LPEXER program, F-13
/UNE_NUMBER qualifier
DEFINE/COMMAND/SYSTEM command, 12-19
LIST command
OPSER, A-12
LIST JOBS command
OPSER, A-12, A-43
LIST OPERATORS command
OPSER, A-12
JUST_FILE qualifier
BACKUP command, 16-8
RESTORE command, 16-15
LOADIINDEX command, 14-22
LOAD/OVERLAY command, 14-23
ILOADABLE_CHARACTERS qualifier
SET TERMINAL command, 8-19
Load address
run-time system, 12-4
Local Area Transport
See LAT

Index-10

Local terminal test
KBEXER program, F-12
/LOCAl-.ECHO qualifier
SET TERMINAL command, 8-19
/LOCK qualifier
INSTALUUBRARY command, 12-23
INSTAlIJRUNTIME_SYSTEM command, 12-27
Log file
closing
under SHUTUP, 4~10
LOG FILE command
OPSER, A-13
Logged-In quota, 5-2
EXOTA privilege, 6--3
Logged-out quota, 5-2
EXOTA privilege, 6--3
/LOGGED_IN qualifier
BROADCAST command, 13-10
FORCE command, 13-·12
Logical device
BATCH, A-38
SPOOL, A-22
Logical end mode
shutdown of OPSER, 4-8
Logical name
account number associated, 12-8
adding new, 12-9
ASSIGN/SYSTEM command, 12-9
DEASSIGN/SYSTEM command, 12-10
job related, 12-8
list, 12--10
system, 12-8
system defined, 12-8
unique, 12-9
Logical Names
managing with DCL commands, 12-H
LOGIN
set priority, 13-6
LOGIN.COM
LAT reporting, 8-8
login command file, 6-8
LOGIN command, 13-14
LAT reporting, 8-8
Login command file
LOGIN.COM, 6-8
Logins
decrease the number of, 12-15
limit, 13-2
limit for DECnetlE, 4-4
limiting, in SHUTUP, 4--4
limit under SHUTUP, 4-4
over LA"r lines, 13-2
restrict, 13-1
restricted by swap space, 12-15
swap fllo space affects, 12-16
SWCTL privilege, 6-5
/LOGINS qualifier
SET SYSTEM command, 13-19, 14-27
LOGOUT command, 13--14
LAT reporting, 8-8
/LOG qualifier
CREATE/ACCOUNT command, 7-7
CREATE/SERVICE/LAT command, 8-10
CREATE/VIRTUAL_DISI< command, 11-16
DELETE/ACCOUNT command, 7-10
DELETEISERVICE/LAT command, 8-11
DELETENIRTUAL_DISKcommand, 11-17

ILOG qualifier (cont'd.)

INSTALUERROR_FILE command, 12-22
INSTALUOVERI..AY_FILE command, 12-25
INSTALUSWAP_FILE command, 12-28
SET ACCOUNT command, 7-14
SET FILE command, 12-36
SET NODEILAT command, 8-12
SET PASSWORD command, 7-17
SET SERVICElL.AT command, 8-13
SET SYSTEM command, 14-27
START/LAT command, 13-32
STOP/LAT command, 8--33
ILOOKUP qualifier
CREATEIACCOUNT command, 7-7
SET ACCOUNT command, 7-14
ILOWERCASE qualifier
SET PRINTER command, 11-24
SET TERMINAL. command, 8-19
LPEXER program
€lrrOrS, explanation, F-13
€lxample dialogue, F-13
Iline printer test, F-13

M
Maglnetic tape
density for Save Set, E--20
IDENSITY switch, E-6
SAVE Set, E-1, E-19
Ma~lnetic tape exerciser
See MTEXER program
MAt<SIL program
formats disk file, 12-7
Management
I=\STS/E system, 1-2
IMAXIMUM qualifier
SET CACHE command, 11--21
IMAXIMUM_SIZE qualifier
INSTALURUNTIME_SYSTEM command, 12-27
Memory
l!ldding resident library, 12-8
available range of, 12-15
~;lvoiding fragmentation, 12-4
dump, 10-21
l'ragmentation, 12-5
high, 12-5
Ilow, 12-5
istatus statistics, 14-20
Memory access
data caching, 11-7
Memory dump
,annotated version, 10-22
Icritical contents, 10-23
MESSAGE command
OPSER, A-10, A-13
Message quota, 5-4
Mei3sage receiver
OPSER, A-2
QUEMAN declares self, A-15
Me:ssage receivers
abbreviations (OISPLY), 14-18
statistics (DISPLY), 14--17
Me:ssages
? characters, 0-1
% characters, 0-1
general, 0-1, 0-2t to 0-29t
informational, 0-1

Messages (cont'd.)
severe error, 0-1
standard error, 0-1
warning, 0-1
IMESSAGE_QUOTA qualifier
CREATEIACCOUNT command, 7-8
SET ACCOUNT command, 7-14
Messaging systems, 14-21 to 14-22
M FD (Master File Directory)
and UFDs in Level 0, 1-5f
with GFDs and UFOs in Level 1, 1-6f
1M FD_ CLUSTER_SIZE qualifier
INITIALIZE command, 15-9
IMINIMUM_SIZE qualifier
INSTALURUNTIME_SYSTEM command, 12-27
Missed errors
reported to ERRCPY, 10-9
MODE values
data caching, 11-7
IMODIFIED qualifier
BACKUP command, 16-5
RESTORE command, 16-13
Monitor
caching checks, 11-191
changing monitors, 2-6
control of run-time system, 12-3
customizing, 2-6
default keyboard, 1-1
definition, 1-2
directives (data caching), 11-7
error detection, 10-1
job keyboard, 1-1
keyboard, 1-2
symbols (list), 10-23
take online dump, 12-16
update cache, 11-8
IMONITOR_NAME qualifier
SET SYSTEM command, 14-27
MOUNT command (disks), 15-14
I[NO]QUOTA qualifier, 15-16
I[NO]REBUILD qualifier, 15-16
I[NO]RESTRICT qualifier, 15-17
I[NO]SHARE qualifier, 15-16t, 15-18
I[NO]WRITE qualifier, 15-18
IOVERRIDE qualifier, 15-15
IPRIVATE qualifier, 15-15, 15-16t
IPUBLIC qualifier, 15-15,15-16t
MOUNT command (tapes), 15-19
IDENSITY qualifier, 15-19
IFORMAT qualifier, 15-20
lOVER RIDE qualifier, 15-20
IWRITE qualifier, 15-20
Mounting disks
DISMOUNT (disks), 15-3
MOUNT command (disks), 15-14 to 15-19
overview, 15-2
Mounting tapes
DISMOUNT (tapes), 15-5
MOUNT (tapes), 15-19
overview, 15-2
MOUNT privilege
public disk, 15-14
shared disks, 6-4
MSCP
ERRDIS full report, 10-14
example, 10-15 to 10-17

Index-11

MTEXER program
error messages, F-14
example dialogue, F-14
magnetic tape controllers test, F-13
magnetic tape drives test, F-13
/MULTICAST_TIMER qualifier
SET NODEILAT command, 8-12

N
NAME option
SPOOL, A-24
/NAME qualifier
CREATE/ACCOUNT command, 7-8
SET ACCOUNT command, 7-14
SET SYSTEM command, 14-27
NCP
network shutdown, 4-6
Network
shutdown of DECnetlE, 4-5
Network management
SWCTL privilege, &-5
/NETWORK qualifier
CREATE/ACCOUNT command, 7-8
SET ACCOUNT command, 7-14
/NETWORK_FILE qualifier
SHOW FILE/SYSTEM command, 12--38
/NEW_FILES qualifier
INITIALIZE command, 15-10
NEXT command
error in, A-19
error text, A-19t
QUEMAN, A-19
/NEXT_JOB qualifier
START/QUEUE/MANAGER command, 9--43
/NOBROADCAST qualifier, prevents receiving
broadcasts, 8-6
/NOCONTIGUOUS qualifier
SET FILE command, 12--35
/NODE qualifier
SHOW COUNTERS/LAT command, 8-24
Non-file-structured 110
RDNFS privilege, 6-4
WRTNFS privilege, &-7
Non-file-structure disk test
DSKSEK program, F-10
Nonuser account, 7-1
NORMAL option
changing in SPOOL, A-24
SPOOL, A-24
/[NO]FMS qualifier
SET SYSTEM command, 8-14,14-26
/[NO]STATISTICS qualifier
SET SYSTEM command, 14-28
NSP (Network Services Protocol)
cannot remove or add, 12-15
Number conversion, bit position, B-1

o
/OCCURRENCE qualifier
RESTORE command, 1&-15
ODT (Octal Debugging Tool)
setting breakpoints with, 12-4
/OMIT_CR qualifier
SET PRINTER command, 11-24

Index-12

ONLCLN
perform disk cleaning, E-7
OPEN/QUEUE command. 9--31
Operating system
definition, 1-2
OPERATOI=t command
OPSER, A-13
Operator communication prlogram, A-42
Operator services
BATCH, A-1
flowchal1, A--3f
OPSER, A-1, A-6
overview, A-1
QUEMAN, A-1
RESTOR, A-1
SPOOL, A-1
Operator table
OPSER, A-6
OPSER
ABORT command, A-45
action mquests, A-10
contents, A-11t
ANSWER command, A-11
cause INIT to start, A-13
CHANGECONSOLE command, A-13
changes OSC, A-13
checks for on line jobs, A-4, A-6
checks operator table, A-6
commands
from PLEASE, A-43t
from SHUTUP, A--44
START.COM, A--13
communicates
three ways, A-4
with BATCH, A-40
communicates with SHUITUP, 4-2
controlled program, A-2
controlling RESTOR, A-5, A-46
DETACH command, A--13
END command, A-44
immediate shutdown mode, 4-8
initial operating conditions for, A-13
interaction with action request, A-4
interaction with information line, A-4
interaction with messago, A-4
INTERRUPT command, A-11 , A-12, A-17t, A-44
LIST command, A-12
LIST JOBS command, A-12, A-43
LIST OPERATORS command, A-12
LOGFILE command, A-13
logical end mode, 4-8
message
contents, A-11 t
formats, A-10
levels, A-11
receiver identification, A-2
MESSAGE command, A-10, A-13
online job list, A-12t
OPERATOR command, A-13
operator commands, A-8t
operatoll' communicate with, A-43
operator table, A-6
overview, A-2
RESTOR commands, A-46t
restrictions of, A-2
RETYPE command, A--10
running with SHUTUP, 4-3

OIPSER (cont'd.)
send command to QUEMAN, A-18
shutdown, A·-5
shutdown in SHUTU P, 4-8
shutdown procedures, 4-8
start after system crash, A-14
starting, A-6
startup procedure, A-" 3
table of online jobs, A-1
terminating with SHUTLJP, A-5, A-44
valid operators, A-12, A-43
ways terminated, A-13
work file, A-6
OPSER.LOGfile
default log file, A-7
OPSER package
BATCH, A-37
OPSER Package Library
QUEMAN, A--14
IOIPTIMIZE qualifier
RESTORE command, 16-15
Options
INIT program, 2-5
OSC (Operator Sorvices Console)
INTERRUPT responses, A-12
OPSER broadcasts messages to, A-2
OPSER changes, A-13
output formats, A-10
send text to, A-42
IOUTPUT qualifier
BACKUP command, 16-9
RESTORE command, 16-16
SHOW ACCOUNT command, 7-20
SHOW COUNTERS/LAT command, 8-24
SHOW NODE/LAT command, 8-25
SHOW SERVICE/LAT command, 8-26
SHOW SESSIONS/LAT command, 8-28
SHOW TEBMINAL_SERVERILAT command,
8-31
lOUT_LABEL qualifier
BACKUP command, 16-8
RESTORE command, 16-16
lOUT_QUOTA qualifier
CREATE/ACCOUNT command, 7-8
OUT_QUOTA qualifier
SET ACCOUNT command, 7-14
OV4~rlay file
adding, 12-16
removing, 12-·16
SHOW FILE/SYSTEM command, 12-15
IOVERLAY_FILE qualifier
SHOW FILE/SYSTEM command, 12-38
IOVERRIDE qualifier
LOGIN command, 13-14
OVRSYS file
allocate and position, 12-13

p
Pac:k identification
19xtracting with SAVRES, E-17t
IPAGE qualifier
START/QUEUE/MANAGER command, 9-43
IPAGE_EJECT switch
SPOOL, A-25
IPAGE_END qualifier
STOP/SERVER command, 9-47

IPAGE_LENGTH qualifier
SET PRINTER command, 11-24
IPAGE_LlMIT qualifier
INITIALlZE/QUEUE command, 9-27
SET QUEUE command, 9-33
IPARITY qualifier
SET TERMINAL command, 8-19
IPASSWORD_PROMPT qualifier
CREATE/ACCOUNT command, 7-8
SET ACCOUNT command, 7-15
SET SYSTEM command, 14-27
PBS
batch servers, 9-2
default queues, 9-3
Forms Definition File (FDF), 9-14
print servers, 9-2, 9-7
queue manager, 9-2,9-3
queue names, 9-3
queue priorities, 9-5
queue types, 9-3
treatment by SHUTUP, 4-6
user intertace, 9-2
PBSCTL privilege
Print/Batch Services (PBS), 6-4
PDP-11 word
decimal values, B-1
octal values, B-1
PEEK function test
CPEXER program, F-4
IPERMANENT qualifier
SET TERMINAL command, 8-5,8-20
SHOW TERMINAL command, 8-30
Physical device names, 12-8
PHYSICAL option
SPOOL, A-24
IPLACED qualifier
SET FILE command, 12-36
IPLACED_POSITION qualifier
RESTORE command, 16-16
.PLAS system directive
RDMEM privilege, 6-4
PLEASE
CCL command, A-44
commands through OPSER, A-43t
commands to OPSER, A-43
message broadcast, A-43
OPSER commands through, A-43
restrict use, A-42
run, A-42
talks to OPSER, A-3
terminate, A-42
/POSITION qualifier
CREATE/ACCOUNT command, 7-8
INSTALUERROR_FILE command, 12-22
INSTALUOVERLAY_FILE command, 12-25
INSTALURUNTIME_SYSTEM command, 12-27
INSTALUSWAP_FILE command, 12-29
Power fail
automatic restart after, 2-6
IPOWERFAIL_DELAY qualifier
SET SYSTEM command, 14-27
/PREFIX qualifier
INSTALURUNTIME_SYSTEM command, 12-27
Primary run-time system, 12-3
position, 12-5
system disk, 1-3

Index-13

Print/Batch Service (PBS)
ASSIGN/QUEUE command, 9-20
INITIALIZE/QUEUE command, 9-25
Print/Batch Services
PBSCTL privilege, 6-4
Print/Batch services (PBS)
STOP/SERVER command, 9-46
Print/Batch Services (PBS)
CLOSE/QUEUE command, 9-21
DEASSIGN/QUEUE command, 9-22
DELETE/QUEUE command, 9-23
DELETE/SERVER command, 9-24
INITIALIZE/SERVER command, 9-29
many-to-one configuration, 9-4
one-to-many configuration, 9-4
one-to-one configuration, 9-4f
OPEN/QUEUE command, 9-31
SET QUEUE command, 9-32
SET SERVER command, 9-35
START/QUEUE/MANAGER command, 9-40
START/QUEUE command, 9-39
START/SERVER command, 9-42
STOP/QUEUE/MANAGER command, 9-45
STOP/QUEUE command, 9-44
Print/Batch Services Package (PBS)
DCL commands, 9-1t
managing, 9-2 to 9-12
Printer servers
ownership, 9-8
IPRINTER_PORT qualifier
SET TERMINAL command, 8-20
Printing device
adjustable top of form, A-26
/PRINT qualifier
INITIALIZE/QUEUE command, 9-:-27
SET QUEUE command, 9-33
SET SERVER command, 9-36
SHOW SERVER command, 9-38
Print servers
Print/Batch Services (PBS), 9-2,9-7
wildcard names, 9-8
Priority
basis, 13-5
byte, 13-6
format, 13-5f
decrease for job, 13-6
description, 13-5
LOGIN, 13-6
processing entries, 9-6
raise for SPOOL, A-32
setting job, 13-5
/PRIORITY qualifier
INITIALIZE/QUEUE command, 9-27
INITIALIZE/SERVER command, 9-30
SET JOB command, 13-17
SET QUEUE command, 9-34
SET SERVER command, 9-36
START/QUEUE/MANAGER command, 9-41
/PRIORITY switch
QUEMAN, A-16t
Private disk, 1-3, 1-10
/PRIVATE qualifier
INITIALIZE command, 15-10
Privilege
controlled access, 1-12
system damage, 1-12

Index-14

Privilege bit
TMPPRV privilege, 6-6
Privileged job
programming error, 2-4
Privilege mask
authorized, 6-8
current, 6-8
saved, 0-8
/PRIVILEGE qualifier
CREATE/ACCOUNT command, 7-8
SET ACCOUNT command, 7-15
SET JOB command, 13·-17
/PRIVILEGES qualifier
DEFINE/COMMAND/SYSTEM command, 12-19
INITIALIZE/QUEUE command, 9-27
SET QUEUE command, 9-34
Processing entries
AFTER state, 9-6
HOLD state, 9-6
priority, 9-6
READY state, 9-6
Program
system, 1-2
Program development, 1--2
Programming errors
causing crash, 2-5
/PROMPT qualifier
BACKUP command, 16-9
RESTORE command, 16-16
IPROTECTION qualifier
INSTALL/LIBRARY command, 12-23
SET FILE command, 12-36
Pseudo keyboard
BATCH, A-38
use with BATCH, A-5
Pseudo keyboards, 8-8
batch servers, 9-12
/PSEUDO_KEYBOARDS qualifier
SET SYSTEM command, 8-14, 14-27
Public disk structure, 1-3
/PUBLIC qualifier
DISMOUNT command, 15-3
INITIALIZE command, '15-10
Public structure, 1-2
/PURGE qualifier
DELETE/ACCOUNT command, 7-10

Q
QUE, A-4
QUE command
SPOOL, A-35
QUEMAN
command received through OPSER, A-18
commands
startup, A-19
consistency checks, A-·15, A-20
DETACH command, A--16
do not CtrVC, A-16
entry in AFTER queue, A-20
error in NEXT command, A-19
initial conditions for spooling queues, A-15
INITIALIZE command, A-16
Integrity checks, A-15
message receiver declaration, A-15
modules, A-14
OPSER Package Library, A-14

QlIEMAN (cont'd.)
requirements for running, A-14
resets to initial states, A-19
running
after SHUTUP, A--20
after system crash, A-20
running detached, A-1S, A-19
start, A-14
startup
command file, A-'19
commands, A-16t
switches, A-16t
STATUS command, A-·18
status printout, A-18
table of online spooling programs, A-4
/QUERY qualifier
BACKUP command, 16-9
DELETE/ACCOUNT command, 7-11
INITIALIZE command, 15-11
IREMOVE/JOB command, 13-16
IRESTORE command, 16-16
QUIEUE.SYS fUe, A-4
4::lear all entries from, A-15
4::onsistency checking on, A-20
4::orrupt, A-l 6
nag value, A-15
initial conditions, A-15
number of queue reques~s, A-20
QUElue limits, 9-7
QUElue management, A-14
QUEllue manager
Print/Batch Services (PBS), 9-2,9-3
QUEllue names
Print/Batch Services (PBS), 9-3
Queue priorities
Print/Batch Services (PBS), 9-5
Queues
l:ltarting, 9-5
s:topping, 9-5
Queue types
f)rint/Batch Services (PBS), 9-3
Quotas
Disk, 5-1
dlynamic regions, 5-4
Job, 5-1
message, 5-4
FIIB (Receiver 10 Block), 5-4
send/receive, 5-1

R
Random data caching, 11-8
RC25
sltopping, 14-2
RDMEM privilege
memory mapping, G-4
mad memory, 6-4
RDNFS privilege
nlon-flle-structured I/O, 6-4
RDS (RSTS/E Disk Structure)
levels, 1-4
SYSTAT, 1-4
Read access
GREAD privilege, 6-3
WREAD privilege, 6-7
READY state
processing entries, 9-G

/READ_ONLY qualifier
INSTALULIBRARY command, 12-24
INSTALURUNTIME_SYSTEM command, 12-27
Rebuilding dirty disks with MOUNT, 15-16
Receiver identification, A-24
RECOVR.COM file, 16-2
/REGIS qualifier
SET TERMINAL command, 8-20
/RELEASE qualifier
SET JOB command, 13-18
SET SYSTEM command, 14-27
Remote line, controlling, 13-7
Remote terminal test
KBEXER program, F-12
REMOVE/ERROR_FILE command, 12-16,12-30
REMOVE/JOB command, 13-16
example, 13-4
REMOVE/LIBRARY command, 12-31
REMOVE/OVERLAY_FILE command, 12-16,12-32
REMOVE/RUNTIME_SYSTEM command, 12-6,
12-33
REMOVE/SWAP_FILE command, 12-15,12-34
Rename access
GWRITE privilege, G-3
WWRITE privilege, 6-7
REORDR
account designators, 11-4
dialogue, 11-3
questions, 11-3t
directory restructuring functions, 11-2
error messages, 11-5
example, 11-6
fatal errors, 11-6
messages, 11-6t
precautions on use, 11-5
processing messages, 11-5
restructure directories, 11-2
run, 11-3
/REPLACE qualifier
ASSIGN/SYSTEM command, 12-17
RESTORE command, 16-2, 16-17
REQUE command
SPOOL, A-34
/RESET qualifier
DELETE/ACCOUNT command, 7-11
SET TERMINAL command, 8-20
SHOW ACCOUNT command, 7-20
Resident libraries
DISPLY program status, 14-20
Resident library, 12-7
memory location, 12-7
permanently resident, 12-8
remove, 12-8
removing
under SHUTUP, 4-9
RMS, 11-10
XBUF allocation, 11-10
RESTART command
SPOOL, A-34
/RESTART qualifier
START/QUEUE/MANAGER command, 9-43
RESTOR
attached to a terminal, A-46
commands through OPSER, A-46t
running detached, A-5, A-46
RESTORE, E-3t
aborting, E-21

Index-15

RESTORE (cont'd.)
bad blocks, E-8
dialogue questions, E-10t
example, E-11 to E-13
output used for system disk, E-10
relocation of blocks, E-9
restoring a RSTS/E disk, E-8
run statistics, E-23
SAVE Set format, E-1
RESTORE command, 16-11 to 16-18
quick reference table, 16-18t
restrictions, 16-11
system recovery, 16-2
Restricted device
DEVICE privilege, 6-3
Restricted receiver
SEND privilege, 6-5
Restrictions
Backup sets, 16-4
RESTORE command, 16-11
IRESTRICT qualifier
SET DEVICE command, 11-22
SET TERMINAL command, 8-20
IRESUME qualifier
SET TERMINAL command, 8-20
IRETAIN qualifier
INITIALIZE command, 15-11
Retrieval pointers
information for data blocks, 11-2
RETYPE command, OPSER, A-10
IREWIND qualifier
BACKUP command, 16-9
RESTORE command, 16-17
RIB (Receiver 10 Block), 6-4
quota, 5-4
IRIB_QUOTA qualifier
CREAT8ACCOUNT command, 7-8
SET ACCOUNT command, 7-15
Ring characteristics, 8-1
RJ2780 program
communication paths, A-5
shutdown, A-f,
RMS (Record Management System)
caching flies, 11-11
indexed file bucket sizes, 11-11
resident library, 11-1 0
RSTS/E
bootstrapping after system halt, 2-2
bootstrapping procedures, 2-1
disk structures, 1-3
file structure, 1-4
halting, 2-1
initialization code, 2-1
initialization options, 2-1
management, 1-2
monitor error detection, 10-1
run-time systems, 12-3
shutdown procedures, 4-1
starting up, 2-1
RSTS/E Error Package
use with DEVTST package, F-1
RSTS/E Privileges, 6-2t
RSX, keyboard monitor, 1-2
RT11, keyboard monitor, 1-2
RUn bUrst
decrease, A-32
default value, 13-6

Index-16

Run burst (cont'd.)
definition, 13-6
raise on SPOOL, A-32
set, 13-·5
system operation, 13-6
IRUNBURST switch
QUEMAN, A-16t
RUN command
how system treats, 12-5
no file type specified, 12-5
Run-time system, 1-2
auxiliary, 12-3, 12-4
BASIC-PLUS, 12-3
change default file type, 12-6
control, 12-3, 12-4
DCl command, 12-4
definition, 12-3
descripti<on block, 12-4
DISPlY report, 14-18
installing, 3-7
languago interface, 12-3
loading a, 12-5
permanently resident, 12-6
placement of, 12-5
position of primary, 12-5
primary, 12-3
removill!~

auxiliary, 12-4
from system, 12-6
under SHUTUP, 4-9
.RTS file type, 12-3
specifying load address, 12-4
storage of, 12-3
temporarily resident, 12-6
use, 12-3
IRUNTIME__SYSTEM qualifier
SET FILE command, 12-36
IRUN_BURST qualifier
INITIAUZE/SERVER command, 9-30
SET JOB command, 13-18
SET SERVER command, 9-36
START/QUEUE/MANAGER command, 9-41

s
SAV/RES
full function command line, E-18
SAV/RES FUNCTION
prompt, E-2, E-3
SAVE
run statistics, E-23
SAVE Set
default name, E-2
dismount message, E-20
extract label information from, E-17t
format, E-1, E-8
identify, E-2
magnetic tape, E-19
magnetic tape density, E-20
proces!;ing mUlti-volume, E-20
r~reate disk, E-8
volume
booting RSTS/E. E-19t
by-pass label checking, E-6
SAVRES, E-1
SAVRES
bad blc.cks

Sf\VRES
bad blocks (cont'd.)
output device, E-24
processing. E-24
changes caused by bad blocks, E-24
checking
input volume, E-7
comparison with BACKUP, E-1
create bootable medium, E-18
c:trVC in dialogue, E-5, E-21
c:trVZ in dialogue, E-5
definition of terms, E-1
device specification switches, E-6t
dialogue, E-3
defaults, E:-5
disks, like, E-2
dismounting volumes, E--20
dismount message, E-20
errors, E-3
dialogue, E-24
handling, E-24
messages, E-25t
transfer phase, E-24
verification phase, E-24
/E:RROR switch, E-4, E--5t
e:dt, E-3
feltal transfer errors, E-28
fllill function command line, E-4
function response, E-3
functions, E-1, E-3t
IDENTIFY
dialogue, E-17t
example, E-17
IMAGE
dialogue, E-13t
example, E-15 to E-16
function, E-13
input volume errors, E-7t
message types, E-3
mounting volumes, E-20
/NOERROR switch, E-5t, E-24 , E-28
nonfatal transfer errors, E-28, E-29t
nonselective backup, E-~
/NIOSTATS switch, E-5t
/N10VERIFY swit(~h, E-4, E-5t
operator intervention during processing,
prompt, E-3
reaccessing devic~es, E-~!1
RI:STORE
dialogue questions, E-10
example, E-11 to E-13
output as system disk, E-10
processing, E-2
re:store archived sets, E-1
RI::STORE function, E-8, E-9
run, E-2
off line, E-2
on line, E-2
SAVE set, E-1
S~\vE Set
name, E-2
volume, E-1
/S'TATS, E-5t
/S'TATS switch, E-4
summary information, E--4
summary report, E-22
format, E-22f

E-19

SAVRES
summary report (cont'd.)
totals, E-23t
switch types, E-4
transfer errors, E-28
use, E-1
/VERIFY switch, E-4, E-5t
volume label checking, E-18
/SCOPE qualifier
SET TERMINAL command, 8-20
/SC RATCH switch
SAVRES, E-6
/SEIZE qualifier
ALLOCATE command, 11-13
/SELECT qualifier
BACKUP command, 16-9
RESTORE command, 16-17
/SELECT_ERASE qualifier
SET TERMINAL command, 8-20
Send/receive quota, 5-4
message, 5-4
RIB (Receiver ID Block), 5-4
Send/Receive quotas, 5-1
SEND privilege
BROADCAST command, 6-5
restricted receiver, 6-5
Sequential data caching, 11-8, 11-9
example, 11-9
Servers
assigned to queues, 9-4
SET ACCOUNT command, 7-13
SET CACHE command
/ALL qualifier, 11-18
/CLUSTERSIZE qualifier, 11-19
/DATA_MAXIMUM qualifier, 11-20
/DIRECTORY_MAXIMUM qualifier, 11-20
/DISABLE qualifier, 11-20
/FILE qualifier, 11-18
/KEEP qualifier, 11-20
/MAXIMUM qualifier, 11-19
/NOFILE qualifier, 11-18
qualifiers, 11-18
TUNE privilege, 11-18
SET DATE command, 14-3, 14-24
SET DEVICE command
/DISABLE qualifier, 11-1
/ENABLE qualifier, 11-1
/NORESTRICT qualifier, 11-1
/RESTRICT qualifier, 11-1
SET FILElRUNTIME_SYSTEM=name command,
12-6
SET FILE command, 12-35
SET JOB/HOLD command, 13-5
SET JOB/RELEASE, 13-5
SET JOB/PRIORITY command, 13-6
SET JOB/RUN_BURST command, 13-6
SET JOB/SIZE command, 13-7
SET JOB command, 13-17
SET NODEILAT command, 8-12
SET PAS privilege
SET PASSWORD command, 6-5
SET PASSWORD command, 7-16
SETPAS privilege, 6-5
SET PRIVILEGE command, 6-1
SET QUEUE command, 9-32
SET SERVER command, 9-35
SET SERVICE/LAT command, 8-13

Index-17

SET SYSTEMtANSWERBACK command, 14-21
SET SYSTEM/DYNAMIC_REGION_L1MIT command,
5-4
SET SYSTEM/EMT_LOGGING command, 14-4,
14-26
SET SYSTEMtHOLD command, 14-2
SET SYSTEM/LOG INS command, 12-15,13-1,13-2SET SYSTEMtMONITOR_NAME command, 2-6
SET SYSTEMtNOLOGINS command, 13-1
SET SYSTEMtRELEASE command, 14-2
SET SYSTEM command, 8-14,13-19,14-25
SET TERMINAUANSWERBACKIPERMANENT
command, 14-21
SET TERMINAL command, 8-15
/AUTOBAUD qualifier, 8-1
/BROADCAST qualifier, 8-6
dial-up lines, 8-1
HWCFG privilege, 8-3
/INQUIRE, 8-8
/NOBROADCAST qualifier, 8-6
/PERMANENT qualifier, 8-5
START.COM, 8-1, 8-5
SET TIME command, 14-3, 14-29
Setting job priority, 13-5
Setting job size, 13-5
/SETUP qualifier
SET TERMINAL command, 8-20
/SHAREABLE qualifier
INITIALIZEISERVER command, 9-30
INSTALULIBRARY command, 12-24
INSTALURUNTIME_SYSTEM command, 12-27
SET SERVER command, 9-36
SHOW ACCOUNT command, 7-18
sample output, 7-18
SHOW BUFFERS command, buffer status report,
12-15
SHOW CACHE command, 11-26
SHOW COMMAND/SYSTEM command, 12-12,
12-37
SHOW COUNTERS/LAT command, 8-23
SHOW DEVICE command, 11-28
SHOW DEVICES/ALLOCATED
use for shutdown, 4-1
SHOW DISKS
use for shutdown, 4-1
SHOW DISKS command
free space, 14-8
SHOW FILE/SYSTEM command, 12-38
SHOW JOB/PRIVILEGE command, 13-20
SHOW JOBS
use for shutdown, 4-1
SHOW JOBS/DETACHED command, 14-8
SHOW JOBS command
job status, 14-8
SHOW LIBRARY command, 12-39
SHOW LOGICAUSYSTEM command, 12-10, 12-41
SHOW LOGICALS command, 12-40
SHOW NODEILAT command, 8-25
SHOW RUNTIME_SYSTEM command, 12-42
SHOW SERVER command, 9-37
SHOW SERVICEILAT command, 8-26
SHOW SESSIONS command, 8-27
SHOW SYSTEM/EMT_LOGGING command, 14-5,
14-30
SHOW SYSTEM command, 14-8, 14-30
monitor name, 2-6
swap size, 2-6

Index-18

SHOW TERMINAL command, 8-6, 8-29
sample output, 8-29
SHOW TERMINAL_SERVERS/LAT command, 8-31
SHOW USER command
job control, 13-4
SHOW USERS command
job status, 14-8
Shutdown
disabling DECnetiE during, 4-4
ERRCPY, 4-9
EVTLOG program, 4-8
final, under SHUTUP. 4-10
formula for waiting period, 4-5
manual procedure, A-44
normal (example), 4-10
OPSEB, 4-8, A-5
performing system, 4-1
RSTS/E system, 4-1
specify waiting period, 4-3
SHUTUP
attached jobs, 4-6
automatic restart, 4-4
communicate with OPSER, 4-2
detached jobs, 4-6
disable DECnetlE, 4-4
disabling logins, 4-4
disk dismount phase, 4-10
EMT logging shutdown phase, 4-9
ERRCPY shutdown phase, 4-9
error message to ERRCPY, 10-3
example
DECnetiE shutdown, 4-12
final job killing phase, 4-9
final shutdown phase, 4-10
formula for waiting period, 4-5
initial job killing phase, 4-6
limit logins, 4-4
limit network activity, 4-4
operational phases, 4-3
OPSER shutdown phase, 4-8
perform shutdown with, 4-1
phases of, 4-2
removing swapping flies, 12-15
resident library removal phase, 4-9
run from console terminal, 4-2
running QUEMAN after, A-20
run-time system removal phase, 4-9
run with OPSER, 4-3
run without OPSER, 4-3
set up dialogue phase, 4-2
shutdolwn of DECnetlE, 4-4
specify shutdown wait period, 4-3
swap file removal phase, 4-10
system recovery, 16-2
terminating OPSER, A-5, A-44
treatment of PBS, 4-6
unexpected errors, 4-10
use of, 2-3
warning message, 4-5
SHUTUP example
large BSTS/E system, 4-10
small RSTS/E system, 4-12
SHUTUP privilege
system control, 6-5
SIL (Save Image Library)
format, 12-7
MAKSIL program, 12-7

-,-

Simull!ltion of hardware errors
device testing package, F-1
/SIXEIL qualifier
SE:T TERMINAL command, 8-21
/SIZE qualifier
CFIEATE/ACCOUNT command, 7-9
INSTALUERROR._FILE command, 12-22
INSTALUOVERLAY_FILE (~ommand, 12-25
INSTALUSWAP_FILE command, 12-29
SET JOB command, 13-18
Small buffers
DISPLY report, 14-18
general (SYSTAT), 14-8
SHOW BUFFERS command, 14-8
Software errors
logging, 10-1
/SPEGIAL_CHARACTER qualifier
SET PRINTER command, 11-24
/SPEE:D qualifier
SET TERMINAL command, 8-21
SPOOL, A-1
ABORT command, A-26, A-35
/ ALIGN switch, A-37
ASSIGN option, A-24
chl!lnging NORMAL form, A-24
compiled version of, A-21
delfault form, A-24
de~ault keyboard values, A-28
delfault line printer values, A-27
/DIFLENGTH switch, A-25
error handling. A-34
error messages, A-36
error text in user output, A-36t
flle, header, A-35
FORM command, A-36
FORM option, A-24
forms control characters, A-25
int'Brrupt commands, A-30, A-30t
job header, A-35
IinlB printer errors, A-34
1000/icai device name, A-22
mc)dules, A-21
NAME option, A-24
NORMAL option, A-24
/PAGE_EJECT switch, A-·25
PHYSICAL option, A-24
QUE command, A-35
raise priority, A-32
raise run burst, A-32
RE:QUE command, A-34
RESTART command, A-34
re~;tart queued request, A-34
run, A-21
stalrtup
error processing, A-28
examples, A-31
options, A-23t
syntax error messages, A-28 , A-29t
terminate
queued request, A-35
without shutting down, A--44
Spooling
determining number of jobs, A-18
iniltial conditions for, queues, A-14
ke:vboard, A--28
Iinl9 printer, A-27
Spooling Package Library, A-21

Spooling programs
commands, A-12
controlling, A-12
line printer, A-21
messages, A-12
monitoring, A-12
SPOOL, A-21
SPOOL and BATCH, A-1
table of in QUEMAN, A--4
START.COM
automatic restart, 2--4
memory image, 2-5
sample, 3-2
SET TERMINAL command, 8-1, 8-5
START.COM file
ANALYS, 10-23
commands auxiliary run-time system, 12--4
contents, 3-1
logical names, 12-9
replace, 3-1
START/LAT command, 8-32
START/QUEUE/MANAGER command, 9--40
START/QUEUE command, 9-39
START/SERVER command, 9--42
Startup
BATCH options, A-39t
BATCH procedures, A--41
CCL command definitions, 3-8
controlling system, 3-1
SPOOL options, A-23t
terminal characteristics, 3-8
Startup control file
order of operations, 3-2
/STATIC_RATING qualifier
CREATE/SERVICE/LAT command, 8-10
SET SERVICE/LAT command, 8-13
Statistics
abbreviations in DISPLY report, 14-14
busy device, 14-16
job status, 14-14
SAVRES run total, E-23t
STATUS command
QUEMAN, A-18
STOP/LAT command, 8-33
STOP/QUEUE/MANAGER command, 9--45
STOP/QUEUE command, 9--44
STOP/SERVER command, 9--46
Summary report
ERRDIS, 10-6
run totals, E-23t
SWAPO.SYS file
allocate and position, 12-13
SWAP1.SYS file
allocate and position, 12-13
SWAP3.SYS file
allocate and position, 12-13
Swap file, 1-12
add, 12-13
affects log ins, 12-15
online creation, 12-14
removing, 12-13,12-15
under SHUTUP, 4-10, 12-15
SHOW FILE/SYSTEM command, 12-15
Swapping, 1-12
Swapping mode
TUNE privilege, 6-6
Swap space, 13-2

Index-19

swap space (cont'd.)
restricting logins, 12-15
ISWAP_FILE qualifier
SHOW FILE/SYSTEM command, 12-38
ISWAP_MAXIMUM qualifier
SET SYSTEM command, 14-28
SWCFG privilege
system manager function, 6-5
SWCTL privilege
EMT logging, 14-5, 14-26
Logins, 6-5
Network management, 6-5
SWITCH program
change keyboard monitor, 1-1
SYO:
system disk, 1-2
SYS functions
programming, 2-5
SYSIO privilege
file access, 6-5
write access, 6-7
SYSMOD privilege
system modification, 6-6
SYSTAT
free disk space, 14-8
guidelines, 14-8
use, 14-7
System
automatic restart RSTS/E, 2-2
commands, 1-11
EMT logging to record activities, 14-4
enable data caching, 11-18
enable directory caching, 11-18
halt, after bootstrapping RSTS/E, 2-2
halting the RSTS/E, 2-3
initialization options, 1-2
initialized in restart mode, 2-5
monitoring status, 14-8
operational concepts, 1-11
shutdown, 4-1
startup, 3-1
suspending operations, 14-2
System clock
DATES privilege, 6-3
System commands
INSTAL privilege, 6-4
System control
HWCTL privilege, 6-4
JOBCTL privilege, 6-4
SHUTUP privilege, 6-5
System crash, 2-2
ANALYS program, 3-9
causes, 2-4
handling, 2-5
randomly occurring, 2-5
System disk, 1-2
access, 1-3
contents, 1-3
recovery, 16-2
System error log, F-1
System Error Package
programs, 10-1
System errors
compilation, 1~1
extraction, 1~1
formatting, 10-1
retention, 10-1

Index-20

System fllEIS
INSTAl. privilege, 6-4
System Files
managing with DCL commands, 12-1t
System logical names
INSTAL privilege, 6-4
System manager function
HWCFG privilege, 6-3
SWCFG privilege, S-5
System mcxtiflcation
SYSMOD privilege, 6-6
System programs, 1-2
ISYSTEM qualifier
SET PASSWORD command, 7-17
System restart
START. COM, 2-5
System security
EMT logging, 14-7
System startup
command file, 3-2
conditic>ns at, 3-1
controlling, 3-1
crash recovery, 3-9

T
!TAB qualifier
SET PRINTER command, 11-24
SET TERMINAL command, 8-21
Tape handling, DCL commands, 15-1t
!TEMPLATE qualifier
CREATE/ACCOUNT command, 7-9
TERDFL.SYS
terminal characteristics, 8-4
TERMAC.SYS
terminal characteristics, 8-4
Terminal
at syst,em command level, 1-11
IBROADCAST command qualifier, 8-6
CtrllC typed at, 1-1 'I
IDFLENGTH switch effect, A-26
distinction from line printer, A-27
FORCE command, 13-3
hardware top of form, A-28
line speed characteristics file, 8-2
INOBROADCAST command qualifier, 8-6
OSC, A-13
setting characteristics, 8-1
top of 'form capability, A-26
Terminal characteristics
HWCFG privilege, 6-3
Startup, 3-8
TERDI=L.SYS, 8-4
TERMAC.SYS, 8-4
Terminal device designator, 8-2
Terminal line
dial-up lines, 8-1
local, 8-1
Terminal servers, 3-9, 8-6
!TERMINAL_SERVERS qualifier
SHOW COUNTERS/LAT command, 8-24
TERSPD.SYS file
account location, 8--2
I AUTOBAU 0 qualifier, 8-2
why create, 8-2
Timesharing, start, 2-3

ITllvtE_FORMAT qualifier
SET SYSTEM command, 14-28
ITIME_LlMIT qualifier
INITIALIZE/QUEUE command, 9-27
SET QUEUE command, 9-34
TMPPRV privilege
set privilege bit, 6-6
ITOP_OF_FILE qualifier
START/QUEUE/MANAGER command, 9-43
In'SYNC qualifier
SET TERMINAL command, 8-21
TUNE privilege
alter base priority, 6-6
SET CACHE command, 11-18
swapping mode, 6-6
swap size, 2--6
ITYPE qualifier
INSTALURUNTIME_SYSTEM command, 12-27
SET TERMINAL command, 8-21

u
UFD (User Filo Directory)
and MFD in Level 0, 1--5f
mark file for caching, 11-8
reorder, 11-5
with MFD and GFDs in Level 1, 1-6f
UNLOAD/INDEX command, 14-31
UNLOAD/LIBRARY command, 12-8, 12-43
UNLOAD/OVERLAY command, 14-32
UNLOAD/RUNTIME_SYSTEM command, 12-6,
12-44
/UNLOAD qualifier
DISMOUNT command, 1S-5
INSTALUUBRARY command, 12-24
INSTALURUNTIME_SYSTEM command, 12-27
/UPPERCASE qualifier
SET PRINTER command, 11-25
SET TERMINAL command, 8-21
/UP_ARROW qualifier
SET TERMINAL command, 8-21
USER1-8 privileges
User-defined privileges, 6-8
USler account, 7--1
USler-deflned privileges
USER1-8 privileges, 6--8
USler interface
Print/Batch Sel'Vices (PBS), 9-2
USler job
area, 1-1'1
/USER qualifier
CREATE/ACCOUNT command, 7-9
DELETE/ACCOUNT command, 7-11
SET ACCOUNT command, 7-15
SHOW ACCOUNT command, 7-20
/USER_DEFINED._KEYS qualifier
SET TERMINAL command, 8-22
UTILTY program
KILL, A-45

Volumes
dismounting SAVRES, E-20
mounting SAVRES, E-20

w
WACNT privilege
account management, 6-7
/WIDTH qualifier
SET PRINTER command, 11-25
SET TERMINAL command, 8-22
Window turning
reducing, 11-11
Work file
OPSER stores, A-2
WR EAD privilege
execute access, 6-7
read access, 6-7
Write access
GWRITE privilege, 6-3
SYSIO privilege, 6-7
WWRITE privilege, 6-7
/WRITE qualifier
INITIALIZE command, 1S-11
WRTNFS privilege
non-file-structured I/O, 6-7
WWRITE privilege
create access, 6-7
rename access, 6-7
write access, 6-7

x
XBUF
caching allocation, 11-7, 11-9, 11-10
resident library allocation, 11-10

z
Zeroing the ERRDIS log, 10-20
/ZERO qualifier
SHOW COUNTERS/LAT command, 8-24

v
/VERIFY qualifier
BACKUP command, 16-9
RESTORE command, 16-17
Vir'tual disks, 11--12

Index-2-

Reader's Comments

RSTS/E
System Manager's Guide
AA-2762G-TC

Please use this postage--paid form to comment on this manual. If you require a written reply to a software
problem a.nd are eligible to receive one under Software Performance Report (SPR) service, submit your
comments on an SPR form.
Thank you for your assistance.

I rate this manual's:
Accuracy (software works as manual says)
Completeness (enough information)
Clarity (e2lsy to understand)
Organizatilon (structure of subject matter)
Figures (useful)
Examples (useful)
Index (ability to find topic)
Page layout (easy to find information)

Excellent

Good

Fair

Poor

0
0
0
0
0
0
0
0

I would like to see more/less

What I like best about this manual is

What I like least about this manual is

I found the following errors in this manual:
Page

Description

Additional comments or suggestions to improve this manual:

I am using Version _ _ _ of the software this manual describes_

Name/Title

Dept.
Date

Company
Mailing Address
Phone

~;;~;;~ld Here and Tape· -------------------~lllr-------~¥~~y~-United States

~R~T~~~S~~~MI~;!~~A~!!g
POST AGE WILL BE PAID BY ADDRESSEE

Digital Equipment Corporation
Corporate User Publications
Continental Boulevard
MK01-2/E12
Merrimack, New Hampshire 03054

11111 ••• 11.1111 •• 1111 •• 1.11.1111.1 •• 1•• 1.1 ••• 1.11 •• 1
o Not Tear - Fold Here - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -