Digital PDFs

AA-FP67A-TM

April 1985

216 pages

Original

11MB

view download

OCR Version

10MB

view download

Document:	TOPS20 sysMgr
Order Number:	AA-FP67A-TM
Revision:	0
Pages:	216
Original Filename:	AA-FP67A-TM_TOPS20_sysMgr.pdf

OCR Text

TOPS-20
System Manager’s Guide
AA-FP67A-TM

September 1985
This document is intended for the person who is responsible for
making final decisions for setting up and maintaining the

efficient operation of a TOPS-20 installation.
Change bars in margins indicate material that has been added
or changed since the previous printing of this manual. Bullets
indicate that material has been deleted.

OPERATING SYSTEM:

TOPS-20 (KL Model B) V6.1

Software and manuals should be ordered by title and order number In the United States. send orders

to the nearest distribution center. Outside the United States. orders should be directed to the nearest

DIGITAL Field Sales Office or representative.

Northeast/Mid-Atlantic Region

Central Region

Western Region

Digital EQuipment Corporation

Digital Equipment Corporation

PO Box CS2008
Nashua, New Hampshire 03061

Digital Equipment Corporation

Accessories and Supplies Center Accessories and Supplies Center
1050 East Remington Road
632 Caribbean Drive

Telephone:(603)884—-6660

Schaumburg, lllinois 60195

Telephone:(312)640-5612

Sunnyvale, California 94086
Telephone:(408)734-4915

digital equipment corporatione marlboro. massachusetts

First Printing, October 1976
Revised, May 1977

Revised, January 1978
Revised, October 1978
Revised, January 1980
Updated, December 1980
Updated, April 1982
Revised, September 1985

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 furnished under a license and may
only be used or copied 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 or its affiliated companies.

The following are trademarks of Digital Equipment Corporation:

alilgliltlali
DEC
DECmate

MASSBUS
PDP

RSX
RT

DECsystem—10
DECSYSTEM-20

P/OS
Professional

UNIBUS
VAX

DECUS
DECwriter
DIBOL

Q-BUS
Rainbow

VMS
VT

RSTS

Work Processor

The postage-prepaid READER'S COMMENTS form on the last page of this
document requests the user’s critical evaluation to assist us in preparing future
documentation.

CONTENTS

PREFACE

DOCUMENTATION

2.5

SELECTING

AFTER SOFTWARE

3.1
3.2
3.2.1
3.2.2
3.2.3
3.2.4
3.2.5
3.2.6

OVERVIEW ¢ ¢ o o o o o ¢ 6 o o ¢ o
SPECIAL SYSTEM DIRECTORIES Py ° . °
<ROOT—DIRECTORY> . . [ [ [ . . [
<SYSTEM> e o o . . e o e o o o« o
Restoring the Directory <SYSTEM>
<SUBSYS> . . o o
. [
Restoring the Dlrectory <SUBSYS>
.
<NEW SYSTEM> and <NEW SUBSYS>

o
°
.
e
.
e
e
.

o
.
.
o
.
o
o
.

o
.
L)
°
«
o
e
.

o o
° °
L) L)
o o
« ¢
o .
o o
[ [

3.2.8
3.3
3.3.1
3.3.2
3.3.3
3.3.4
3-3.5
3.3.6
3.3-7
3.3.8
3.3.9
3.3.1%
3.4
3.5
3.6

<SYSTEM-ERROR> . . . L) . [
Other Useful Directories .
SYSTEM—LOGICAL NAMES . &« 4 o
SYSTEM:
. o e . e o o . .
SYS: ° [ . . e o o o ¢ o [
NEW: . . . . . o o e o [ .
OLD: . . . [ . . ) ) L] . [
HLP: ) ] . . ) e o [ . . .
& o o o o o o o o o
SERR:
DMP: e o o o o o o . e o .
DEFAULT-EXEC:
e o o o o o
POBOX. . . e o e o o e » .
NRT . . . .
. . . . [ .
CONSOLE FRONT- END FILES
&« ¢
TAILORING THE BATCH SYSTEM .
CHECKING THE SOFTWARE (UETP)

CREATING

4.1
4,2
4.2.1
4,2.2
4,3
4,4

OVERVIEW . .
e e o o o o s o o
THE PUBLIC STRUCTURE e e o s e o s o o
What Is the Public Structure?
. . ¢
The Contents of the Public Structure
ONE-STRUCTURE SYSTEMS
&« « o o o o o o
MOUNTABLE STRUCTURES &+ « & o o o o o o

SYSTEM

<ACCOUNTS>,

[}

L
*

[]
L]

L]
®

[]

[

[]

.
e
.

COMPUTER

[]

THE

ROOM

HANDLING USER REQUESTS
. .
ORDERING SUPPLIES
. .
SCHEDULING OPERATOR TASKS

FOR SOFTWARE

[]

SECURING

PREPARING

[]

[]
*

L]
L]

[]
[]

[]
L]

L]
[
L]

®
[}
[}

.
.
.

[]

.
.
.

Form
Form
Log

[]

Slgn—Up Log

2.2
2.3
2.4

3.2.7

CHAPTER

Structure

System Access Request
Operator Work Request
Operator Shift Change

2.1

CHAPTER 2

CHAPTER

Mountable

«3
.4
5

DOCUMENTS AVAILABLE FROM DIGITAL o o ¢ o
DOCUMENTS PREPARED AT YOUR INSTALLATION
System Log . . .
.
.
.
.
®

CHAPTER

o
.
.
)
o«
.
o
.

o
.
.
e
o
.
o
.

o
°
.
o
o
.

. L) L) .
o o o «
o o o o
o e o o
. o o e
o . ) .
] ) . o
) ) . )
o o o o
e e o .
o s o o
o e o o
L) . . .
o o o o
o o o .0
o o o o

)
o
«
o
e
.
o
.
o
.
o
.
.
o
o
o

3-13
3-14
3-15
3-16
3-16
3—16
3_17
3-17
3=17
3-17
3-18
3_18
3_19
3-19
3-22
3-22

o o
o s
¢¢«
. .
o o
o o

o
o
o«
.
o
o

o 4-1
o 4-2
o 4=2
« 4-2
o 4=3
o 4-4

INSTALLATION

FEATURES

INSTALLATION

<OPERATOR>,

<SPOOL>,

.
«
o
e
.
.
[
.
o
e
o
.
.
o«
o
.

[
«
o
o
e
.
.
[
o
o
o
e
[
o
o
o

[
¢
o
o
o
)
.
)
o
)
s
o
.
o
o
¢

and

[
¢
s
.
o
e
o
)
o
o
e
.
.
o
¢
¢

[
¢«
o
e«
o
o
o
.
s
o
o
e
.
s
o
o

3—1
3—1
3-2
3-2
3=5
3—5
3712
3—13

STRUCTURES

iii

o
o
¢
.
o
o

o
o
o«
.
o
o

4.5.,7
4.6

and
L

Public

)

¢ o ¢ o o
o o ¢ o o
Same Name

o
o
.

o
@
.

Publlc

Structures

MULTIPLE-STRUCTURE SYSTEMS .
Choosing Structure Names .
Mounting Structures Having

o ¢
« o«
the

Maximum Size of Structures . . o« ¢« ¢ o o o o
Increasing the Size of Structures
. . « « + &
Setting Up Structures for Maximum Availability
Taking Structures Off-Line . « o o ¢ o o o o @

Mounting
SHARING

Structures

STRUCTURES
L]

DETERMINING

Another

DRIVES)

SWAPPING
*

Swapping?

4,7.2

When

Increase

DETERMINING

from

(DISK

What

THE

Installation

BETWEEN

TWO

L[]

[

L]
L]

¢«

SPACE

¢«

Installation

Swapping

AVAILABLE

THE

SPACE

PUBLIC

Space

DISK

4.8.1

Determining

Disk

Space

4.8,.2

Determining

Disk

Space After

Before

Installation

CREATING

5.1

HAVING THE OPERATOR CREATE AND MAINTAIN ALL
DIRECTORIES (CENTRAL CONTROL)
e o o o o
o

5.2

DELEGATING

DIRECTORIES

CONTROL)

5.3
5.4
5.4.1
5.4.2
5.4.3
5.4.4

[]

THE

CREATION

PROJECT

ADMINISTRATORS

[

o
.
o
.

o
.+
o
.

o
.
&
.

e
.

o
.

e
.

o
.

o
.
o«

ALLOCATING
ENFORCING

5.7

PROTECTING

DISK
DISK

DIRECTORIES

Directory

and

File

o o 6 e
Project

STORAGE
STORAGE

QUOTAS

o o o o
Control
&«

Digits

File Protection
o o o o o o o o

.
o

QUOTAS

AND

FILES

Protection

5.7.2
5.8

Changing Directory
ESTABLISHING GROUPS

and
o o

CAPABILITIES

5.9

GIVING

5.10

PRINTING

DIRECTORY

CREATING

ACCOUNTS

6.1

SETTING

6.1.1
6.1.2

6.1.3
6.2
6.3

6.3.1
6.3.2

6.3.3
6.3.4
6.4

(PROJECT

‘e ®
and

e
.
«
.

5.6

MAINTENANCE
]

PrOjeCt Control
Combined Central

AND

COMBINING CENTRAL AND PROJECT CONTROL
c
CENTRAL AND PROJECT CONTROL DESCRIPTIONS
Central Control
. & o o ¢
o o ¢ o o
o
Central Control Using Subdirectories .

5.5

5.7.1

and

DIRECTORIES

CHAPTER

Mountable

STRUCTURE

CHAPTER

Between

4.7.1
4.8

Mountable
°

Similarities

SYSTEMS

4.7

USERS

SPECIAL

THE

INFORMATION

SYSTEM

USE

.«

ACCOUNTS

Enabling or Disabling Account
Setting up Account Validation

Validation . .
with Existing

Files
. & o o o o o
.
e o o o
Setting up the System for Accountlng
Changes
. .
e o & o o o
SELECTING AN ACCOUNTING SCHEME e 6 ¢ o
CREATING AN ACCOUNT DATA BASE
. & 4 ¢
Entering Accounting Data into Files
Sample Data FilesS
. ¢ ¢ ¢
¢ o o o o
Running the ACTGEN Program . « « o o
Data Base Failures/ReCOVErY
.« o o o
VALIDATING ACCOUNTS
¢ o o o o o o o o

o o &
Sshift

¢
.

o
.

»
.

o
o
o

¢
o
o

o
o
o

=EJAHhWN

4,5.3
4.5.4
4.5.5
4.5.6

4,5
4.5.1
4,5.2

4.4.2

Between
L]

Structures

Differences

[]

4.4,1

(and

ALLOCATION

Tape

Mounting

Policy
*
.

LJ
o«

*
o«

L]
«

L
+

L
&

L]
o

*
o

o«

RESTORING
A SINGLE DIRECTORY

RESTORING

<ROOT-DIRECTORY>

RESTORING

THE

ENTIRE

<ROOT—DIRECTORY>

Structure

Public

FILE

SYSTEM

Re-creating the File System on the Publlc
.

Re-creating
POWER

Mountable

FAILURES

Structures

o
o

SYSTEM

PERFORMANCE

THE

CLASS

19.1.1

Ooverview

10.1.2

Who

UNAVAILABLE

SCHEDULER

Should

Use

the Class

Scheduler’

o«

Noudwwwdhd
N

\1\1\1\171\1\1\1\1
W

8-16

THE

8-15

REMOTE DIAGNOSTIC LINK (KLINIK)
e o o o o o
MAKING THE CI UNAVAILABLE ON NON-CFS SYSTEMS
MAKING

8-13
8-15

the

8-12
8-13

Rebuilding

FILE

8-11
8-12

WO WY WO

SINGLE

8-11

Setting Up the System to Use Tape Labels o o
Initializing Tapes and Drives to Use Labels
SHARING TAPE DRIVES BETWEEN TWO SYSTEMS
. . .

VOWOIIBDHWNON

[]

.
.
L]
L]

*
L]

Tapes

When to Use Tape Drive Allocatlon
« o o
How to Enable/Disable Tape Drive Allocatlon

TAPE LABELING
* [
Why Tape Labels?

[

for Migrate o}

Archlve)
o«

[]

o
o

[]

o
o

)

Migration

o
o

o«
«

Requests

o
o

o«
«

«
«

«
.

Retrieval
o

Flle Mlgratlon

~NwhhhoH-

DRIVE

Program
Program

Use

W WOWWWYW

Recycling

REAPER
DUMPER

o
&

o«

o
«

.«

o
«

€0 €0 Co CO O
=
11

Files

Processing

&«

¢
«

the
the

o
.

loo]

Using
Using

System to

o«
.

[]

VWO OO

L]
[]

Up the

O
L
.
»
.
[

BDwWwh
-

OJO WD
®
N

WOWVWOVYW

(]

[

MIGRATION

Setting

RESTORING

19.1.3
19.1.4
16.1.5

FILE

SYSTEM PROBLEMS/CRASHES

1.1

o«

[)

o
o

W N

bdd

WwWHN

[]

When to Create Archive TapesS .«
Processing Retrieval Requests

TAPE

ARCHIVING

Setting Up the System to Use F11e Arch1v1ng
What Happens When Users Archive Files
. .
What Happens When Users Retrieve Files . .

[]

FILE

[]

STORAGE

SYSTE
TEM

W
-

BWN
AN
NN s
&
o
o
@
[]
[]
LJ L
Nd W+
W N
.
U

B wwwwWwN

TAPE

VWYLV

SAVING ALL FILES IN ALL DIRECTORIES
Full DUmPS . o o o o o ¢ o o o
Incremental Dumps
. « «+ &
Security of Backup Tapes .

Structure

CHAPTER

PROCEDURES

A COMMON BACKUP POLICY . . .
MAGNETIC TAPE REQUIREMENTS .
MAKING A SYSTEM CRASH TAPE . .
MAKING A CRASH TAPE USING BATCH
SAVING THE CONSOLE FRONT-END FILE

CHAPTER

BACKUP

]

SYSTEM

CHAPTER

19-2

19-4

.
How to Begin Using the Class Scheduler .
.
Procedures to Turn On the Class Scheduler
n
Changing Class Percentages During Timesharing

19-5
19-6
10-8

16.1.6

Disabling

the

Class

Timesharing
19.1.7

Getting

19.1.8

CHAPTER

About

During

e
Class

[

)

10-9

Scheduler

[

)

SeSSIioN o« o 4
¢ ¢
o ¢
o
Alternative to Using Accounts

10_9

10-10

10-12

., .
COMPUTE-BOUND

1l@-12
lg_l3

19.3

FAVORING

INTERACTIVE

PROGRAMS

LOW
.

PRIORITY

BATCH

VERSUS

[

)

JOBS

TIME

18-15

o«

1l0-16

18-16

CONTROL PROGRAM & ¢ 4 « o o
PASSWORD ENCRYPTION
3 ) L3 [} [] [3 [3
Moving Structures Among Systems

o
[

o
.

o
*

o
L)

1l1-1
11—4

e
System

11=5

10.4

IMPROVING

19.5

REINITIALIZING

1.6

DYNAMIC

ACCESS

11.1

ACCESS

11.2.3

PROGRAM

DUAL

STARTUP

DISK

PACKS

PORTING

&«

CONTROLS

Adding

Encryption

USing

DUMPER

)

Algorithms

)

the

llo3
11.4

[

)

PASSWORD MANAGEMENT
e
LAST LOGIN INFORMATION

[

11-6

11.5

e
.

o
&2

PREVENTING

&
o

*
o

&
o

FAST

6
o

o
o

e
o

LOGINS

o
o

e
o

@
o

11-7
11=7

©o

1l1-8

THE

FILE

SYSTEM

12.1

OVERVIEW

COMMON

12.1.1

[

)

CFS

HARDWARE

12-1

)

[

12—2

12.1.2

CFS

SOFTWARE

)

12'1.3

)

CFS

USERS

[

12-5

1201.4

[

CFS

)

and

12-6

12.1.5

CFS

and

DECnet e & 8 o e o o o
.
TIGHTLY-COUPLED SYSTEMS

[}

12-6

12=7

»
L3

o
[}

e
3

&
.

o
°

o
[

e
3

o
L]

12_7
12-7

L
o«

°
o

12-8

o«

12-8

12.1.6

12-2

Limitations

o
.

. ) . . L)
Served DiskS
FileS e o ¢ o o o

[
.

FileS

Mail

12.2.4

Sharing

12.3.2
12 5
12,6

12.6.1
12.7
12.8
12.9
12.lfl
12.11

12.11.1

o
[}

o
[

Files

LOAD

o
[

Update

12.4

e o
FILES

12.2.l

12.3

PLACEMENT

12.2.2
12.2.3

12.3.1

System

Assigning

o o
o
Systems

o
[

[]

[

[]

12_8

o«

12-8

[}

o«

12=10

[

[}

[

[}

[

12_lfl

12-10

)

12—11

FileS

BALANCING

Dedicating

e
3

Users

to SysStems . ¢«
) . .
. [
.
.
SYSTEM LOGICAL NAMES . [
.
.
. .
SHARING STRUCTURES AMONG SYSTEMS
STRUCTURE

NAMES

¢«

o«

)

o
.
.

o
)
[

12_11

12=12

o
o

o
&

o«
o

o
o

12-12
12=12

Sharing Public Structures
. « « o« o
RESTRICTING STRUCTURES TO ONE SYSTEM .
DISMOUNTING STRUCTURES . . .
* o o o
MAKING THE CI UNAVAILABLE TO A SYSTEM
USING DUMPER 3
3 o [ [3 [} )
[
[ °
[ [3 3
ERRORS .
.
.
.
. . . [ . .
[
[
. [ L) [
Communication

12-13

.
L)

.
[3

.
.

12-13

[

«
L3

)

12—14

o o«
o o
Drives

12-14
12-15

12=16

12,12

Problems . . « + &
¢« v
o
Massbus Problems with Dual-Ported Disk
SHUTTING DOWN A CFS SYSTEM « &
«
o
o o o

LAT

1301

OVERVIEW

13.2

LAT

13.3

DECNET

12.11.2

CHAPTER

Sample

SCHEDULING

11.2.2

CHAPTER

10.2

11.2
11.2.1

Scheduler

Information

StatUS

10.1.9

13.4
13.5

TERMINAL

12-14

SERVERS

]

SOFTWARE

.
. . .
CONTROLLING LAT FROM THE HOST
STARTING AND STOPPING LAT
. ¢

.
e

]
o

.
o

e
o

¢
o

o
o

.
&

13-3
13-3

13-7

13_1
13-2

APPENDIX

THE

[}

[]

[]
[}
LJ

L]
[]

L[]
[]
L]

13-8

[]

L]
*

Service Ratings
.
.
MONITORING LAT FROM THE HOST
Displaying User Information
Displaying Host Parameters .
Displaying Server Information
Displaying LAT Counters
. . .

[]

[

o«

[]

SERVICES

[]

13.8.4

GROUPS

HOST

13.8.1
13.8.2
13.8.3

LAT

13.6
13.7
13.7.1
13.8

13-9
13-9
13-10
13-19
13-11
13-12

COMMAND

BUILD

INDEX

<SYSTEM> Files * L ] ® *® *
STR:<SUBSYS> FileS ¢ « ¢
Console Front-End Files

[]
o
.

*
o
.

[]
o
.

*
o

[
o
o

[]
o
&

StruCtUreS

o«

vii

L]
[
[
L]
[

.
. .
Directory Protection Digits
.
File Protection Digits . . « « &
Special Capabilities « ¢« « ¢« ¢« o« + &
d
Summary of Account Data File Commands
.+ « « « « &
Tape Drive Allocation

L
LJ

.
.
.
.

[]

o o o
. o« =«
. . &
Space

[}

U
UTUT D B DD
t
11
HHWNMDHOUD
W

.+ « «
Sample Device NamesS
.
Maximum Size Structures
Determining Swapping Space
Calculating Available Disk

[

o«

Similarities Between Mountable and Public

Structures

Differences Between Mountable and Public

*®
=
®

&
o
o

[]
.
*
[]
L]
.
L]
[]
[]
[]
[
L]
L]
[] L]
L]
[]
L
L] []
[}
Q..

[}
L
L]
[
[]
*
[]
[]
L]
[
[3
£l
L]
[} []
[
L]
[]
[
L]
L]
N e

o
o
o
o
o
o
&

[]
o
o

TABLES

[]

Network

[]

LAT

Two Systems with Massbus Disks

(Yo

[

o
&

DiSks

¢«

Tape

. . ¢« « « &
Bias Control 'Knob'
Two Systems with Massbus Disks a n

¢«

TX@2

.«

.
&
.

Subsystem

Correct-Data Accounting Files
Unionbank Accounting Files . .
Organization of Labeled Tapes

o
o

.
.

1
2

Scheme
Scheme

Accounting
Accounting

Library Group
. « « &
Teacher-Student Group

.
.

Swapping Concept .
File-Sharing Group

Structures
¢« ¢ ¢ « &

Domestic and Foreign
Shared Disk Drive
.

[]

2 Structu r e

O,e

System

Three-Structure

.
3 Disk Drives and

System with

.
.
.

Operator Shift Change
Special System Directories

.
.

. .
Log

Request
Request

System Access
Operator Work

.C.O‘...O.........'m....‘.l

Sample System Log (Hardware Maintenance)
Sample System Log (Problem Report) . .
Sample Mountable Structure Sign-Up Log

HEROOONOOO
VMU
O DD WHEFEHEFEFEKE

FIGURES

9-1

<ROOT-DIRECTORY>

11-1
12-1

DUMPER

BUGHLTS

Directory Restorations
Comparison of CFS and DECnet .

viii

PREFACE

The TOPS-2f System Manager's Guide is written for the
person
who
is
responsible for establishing policies and procedures for a timesharing
and/or batch processing
installation,
wusing
the
TOPS-20
Operating
System.,
Usually,
this
person
1is
responsible
for
setting up and
maintaining
both
the
system
hardware
and
software.
The
Site
Management Guide and the TOPS—-2@ Operator's Guide provide you and your
operations people with the
necessary
information
to
maintain
vyour
system
hardware.
These
two
manuals are referenced throughout this
guide.

This guide deals primarily with your
system
software.
It
contains
general suggestions for planning the installation of your software and
for setting up your computer room
to
begin
operations.
The
guide
contains
hints and suggestions for your system's operation, including
when, and many times why, particular functions or procedures should be
considered.
It
assumes that your system operator is responsible for
implementing many of the decisions you make.
In
most
cases,
where
lengthy
implementation
procedures
are
required,
the
appropriate
reference

Chapters

noted.

and

describe the documentation,
system
logs,
and
special forms that you should have available to
you,
and
in
some
cases,
to
system
users.
Chapter 2 also
includes
preliminary
planning
functions that you can do before
the
software
is

Chapter

Chapter 4

installed.

describes

the system directories and

files

that

your
system
contains
immediately
after
you
install
the
software.
It
also describes the
mechanisms you can use to change the
installed
TOPS-20 batch system and to test the
integrity
of your newly installed or updated system.

describes using your disk-pack
and
disk-drive
resources
to
set
up disk structures in a way
that
best suits your installation's needs. It
also includes guidelines
for
determining
the
available disk space
user directories.

that

you

have

«create

CHAPTER

DOCUMENTATION

Section

1.1

describes

recommends

the

manuals

documentation

with

which

you

provided
should

DIGITAL

familiar

manage
documentation, for

your system,
Section 1.2 describes adding your own
example, special forms, to the documentation you receive
Be

sure

you

have

all

available

documentation

convenient

from
to

DIGITAL.

your

system

users.

l.1

DOCUMENTS

All

AVAILABLE

documentation

TOPS-20

Software

for

FROM

the

DIGITAL

TOPS-20

Notebook

Set.

cross

Operating

This

pertaining to
functionally

and

System

notebook

set

contained

contains

the

information

the most recent version of
TOPS-20.
It
1is
organized
to facilitate referencing manuals.
Each manual contains
references to other manuals within the set that further
explain

subject.,

This manual assumes that you are familiar with some of the manuals
in
the
notebook
set.
In
particular,
you should be familiar with the
information in the TOPS-2@ Operator's Guide, the TOPS-28 User's Guide,
the
DECSYSTEM-20
Technical
Summary,
and
the
~TOPS-2¢ KL1§ Model B
Installation

Guide.

Any

additional

your

system,

documents
For

that

you

example,

1if

termination
(DNxx),
Emulation-Termination,
procedures
your

and

system

the

connected

Monitor

your

Ethernet

provided

1local

with

LAT

the
the

Calls

system

various DECnet-2¢ manuals.
an

area

has

operator
ARPA

addition

the

Beware

configuration
IBM

user

Guide

and

you

should

If you are using
refer

terminal

servers,

TOPS-20

Software

Listing.

the
has

you

User's

hnhetwork,

File

and

network,

DEChet,

manual.

TOPS-20

system

software.

and

hints

and

1is

the

suggestions

familiar

installing

User's

you

receive
the

this

the

software

installing
Beware
have

that

with

the

servers

chapter

Set,

access

with

documentation

distributed

for

have

TOPS-2@AN

LAT terminal

addition

Notebook

interfaces.
should

installation and distribution magnetic tapes.
Before
version
of
the
software
on
vyour system, read the
contains last-minute changes
to
the
software
that

documented,

emulation

be
familiar
with
the
IBM
Manual.
It includes installation

DN65

depend

your

should

DN64

descriptions
1s

TOPS-2@AN

Guide.

you

need

new

File.
It
not
been

using

the

new

DOCUMENTATION

With each new system, you should also receive two stand-alone
documents, which are documents not included in the notebook set.
These manuals assist you in 1) preparing your site for the hardware
installation, the Site Preparation Guide, and 2) maintaining and
reporting problems about your system's software and hardware, the Site
Management

Guide.

NOTE

Site
the
delivers
Representative
Sales
Your
Field =@ Service
your
and
Guide,
Preparation
Representative delivers the Site Management Guide.

This manual (the TOPS-2@ System Manager's Guide) deals primarily with
installing and maintaining the software on your system. Therefore, it
is assumed that you have already used the Site Preparation Guide to
install

your

system hardware.

The Site Management Guide is designed for use by both you (along with
and your Field Service Representative. You
your operations people)
immediately after Yyou install vyour
should begin using this manual

It contains schedules, procedures, and logs for recording
hardware.
and evaluating all information pertinent to the operation and care of
The manual belongs to DIGITAL, but it is kept and
the system.

and
convenience
For added
maintained at your computer site.
system
important
their
all
Kkeep
managers
system
organization, many
For
information in the same binder as the Site Management Guide.
information
Change
Shift
Operator
and
Logs
System
example, they keep
Section 1.2
in the same binder, along with other special forms.
log book or,
system
a
in
include
may
you
that
forms
describes several
as suggested here, in the Site Management Guide.

DIGITAL places a major emphasis on the documentation provided to its
The Software Publications Department continues to solicit
customers.
suggestions for improvement and corrections from the users of its
Encourage users to comment on the manuals you receive
documentation.
with your system. For convenience, a Reader Comment Form is 1located
at the back of each manual.

1.2

DOCUMENTS PREPARED AT YOUR INSTALLATION

Sections 1.2.1 through 1.2.5 describe some forms that may be useful at
your installation. A sample form is provided in each section.

DOCUMENTATION

l.2.1

System

Every

system

procedures
system

the

must

have

relating

programmers

should

with

system

both

the

1log

hardware

record

time,

recording

software.

following

and

operators

and

activities

OSystem

Beginning and ending of timesharing (for example,
the
times
the system was started and stopped for preventive maintenance
or repair)
Problems

hardware

the

problems

(always

output

New

revised

New

users

New

structures

system

copy

problems

names:

procedures

correct

their

types

problems

All

along

backup

date,

the

for

and

log,

Most

Log

the

are

AND

the

actions

the

CTY

(operator

user

data

save

taken

terminal)

typescript)

software

changes

software

installed
existing

changes

easier

existing

solve

(and,

directories

structures
hence,

less

costly)

you
keep
an
accurate record of all activities.
The Site Management
Guide has a section
set
aside
for
system
log
information.
This
section
contains
preprinted
forms that you can use to record system
log information, or you can design your
own
forms.
You
<can
store
these forms in the Site Management Guide or in a separate binder.
You should
Remember,

design your log so
that
it
1is
easy
to
use
and
read.
you are likely to have the most problems when the system is
new, so NOW is the time to start using the
1log.
The
following
two
pages
‘contain
sample
1left- and right-hand pages of a log book.
The

left-hand page (Figure 1-1) contains
maintenance;
the
right-hand
page

information
(Figure 1-2)

concerning
hardware
is a problem report,

containing:
e

The

few words

describing

record

calls

e
e

time

"Y"

name

the

"N"

entry

answer

the

whether

person making

the

nature

Digital

Field

system

had

reloaded

entry
of

the

activity

Service

(F/S)

A description of the device or program causing the problem
Remarks

about

the

entry

DOCUMENTATION

SYSTEM LOG

MAINTENANCE PERFORMED

DATE

MR-5-526-80

Figure 1-1:

Sample System Log

(Hardware Maintenance)

DOCUMENTATION

PAGE

SYSTEM L.LOG

MONITOR OR

TIME | A
o

DATE

HARDWARE

|NAME

MAINTENANCE
ACTIVITY

N
A
T

DEC‘)’I'?CE

PROGRAM

ENTRY

MR--S-527-80

Figure

1-2:

Sample

System

Log

(Problem

Report)

DOCUMENTATION

1.2.2

Mountable

Structure

Sign-Up

Log

In addition to keeping the system log, you should also record requests
from
users
to
mount structures.
(Chapter 4 describes how to set up
and use structures.) Without a formal scheduling procedure, some users
may
monopolize
the use of a structure and frustrate other users, who
do not have the opportunity to mount and use their structures, usually
because
there are no disk drives available.
To avoid this situation,
set up a procedure whereby users inform the operator wheh they need to
use
a
structure.
The operator can then schedule the length of time
specified on the request log.
On a busy
day,
when
many
users
are
issuing
mount
requests
for
structures, the operator checks the log
before granting or denying the mount requests.
This scheduling allows
you
to
service
many
requests for mounting structures in a fair and

orderly manner.
Figure

1=-3

The

scheduled mounting

The

scheduled

The

actual

time

the

structure

was

mounted

The

actual

time

the

structure was

removed

The

name of

the

user

The

structure name

e¢

A column

Remember
best

1.2.3

The sample Mountable Structure Sign-Up Log

contains:

that

suits

this

your

System

for

needed

who

(or

time
to

is only a

use

the

initiated

pack

any special

log

own

time

shown

structure

the

request

ID)

instructions
or notes

sample;

you should design a

form that

requirements.

Access

Request

Form

Some installations have many users requesting access to the system for
the first time.
You need standard information from these users before
you can process their requests and create directories
for
them.
For
example,
you
must know which system they need to access (if you have
more than one system), their names, selected
passwords,
departments,
accounts,
etc.
You can organize these requests by providing a System
Access Request Form that is kept in an
easy-to-access
area,
perhaps
outside
the
computer room.
You can require signatures of department
managers on the access form to
ensure
that
prospective
users
have
approval to charge computer usage to accounts.
Figure 1-4 is a sample
of

system

access

request

form.

If you are using CFS-20 software, refer to Chapter 12, The Common
System, for further considerations in assigning users to systems.

File

DOCUMENTATION

MOUNTABLE STRUCTURE
SIGN - L.LOG

PAGE
DATE

SCHEDULED

ACTUAL

MOUNTING|

TIME

|MOUNTING|

TIME

USER

PACK

TIME

NEEDED

TIME

REMOVED

NAME

ID(s)

NOTES

MR-58-528-80

Figure

1-3:

Sample

Mountable

Structure

Sign-Up

Log

08-6¢5—S—HW
DOCUMENTATION

INIWIOVYNVW SNOILYHIJO A9 A3A0HddV 38 1SN,
i{

DOCUMENTATION

l.2.4

Operator

You may want

Work

form

Request

that

Form

allows

users

request

work

from

the

operator.
Examples of requests made to the operator are initializing
tapes, transferring files
between
systems,
and
making
changes
to
directories.
You
should
set
up
a
procedure
for
handling these
requests.

l.2.5

Figure

Operator

1-5

Shift

is a

sample of an operator work

Change

request

form.

Log

You may want to set up a
binder
to
contain
operator
shift
change
information.
Each
operator
records
new
procedures,
or
special
instructions that the incoming operator needs to know.
The
incoming

operator

reads

the operator shift log before starting the new shift.

For example, the
first
shift
operator
changes
the
procedure
for
storing
tapes, and records the new procedure in the shift change log.
The information in the shift change log should
not
concern
problems
with
the
system,
but should contain important information about the
system or the computer room.
The incoming operator
still
reads
the
system log book to determine the status of the system and any problems
that have occurred during the previous shift.
Figure 1-6 is a
sample
of

an operator

shift

change

log.

08-0£S-S—HIN

DOCUMENTATION

HOEE

I“N1OXH3d

‘HO1vYd3do ‘IW3LSAS

1-19

DOCUMENTATION

OPERATOR SHIFT CHANGE LOG

DATE

OPERATOR

SHIFT

COMMENTS

MR-5-531-80

Figure

1-6:

Operator

Shift

Change

Log

Chapter

describes creating and
It includes a detailed
methods

maintaining directories.
description of the three

administration

to
control
directories.

the
It

<creation
describes

you

can

choose

from

and
maintenance of
how to use directory

and file protection codes to
expand
or
1limit
the
type
of
access
users
can
have
to
directories and files, and how to
place
users
and directories in groups
so
that
wusers
can
share files.
Chapter

Chapter

describes the TOPS=-20 accounting facility. This
description
includes
how
to
choose
an
accounting scheme,
how
to
«create
accounting

files,
and
how
to
validating accounts.

set

describes

your

magnetic
It

backing

tape

recommends

procedures

soon

after
the

that

you

the

system

disk

structures

softwarle
follow

your directories and files on
a
and how to create a system crash
event of a major problem with the
Chapter

begin

onto

installation.

supplies

should

needed
to

and

save

all
daily
basis,
tape
in
the
file system.

describes how you
store
important

can
use
magnetic
tapes
files (file archiving) and

save

disk

valuable

infrequently

space

accessed

files

to
to

copying

tape

(file

migration).
It
also
describes
how
to
give
control
of
tape drive usage to the system and
the
operator
(tape drive allocation), and how
to

set

(tape

Chapter

your

system

use

labeled

tapes

labeling).

describes the procedures you must follow in the
event
that
you
have
a problem with the file
system
or
that a user has lost the files in a
directory. It describes using your system crash

tape

and

these

your daily

backup

tapes

resolve

problems.

Chapter

describes
the tuning mechanisms that allow you
to change the behavior
of
your
system.
Each
description
includes why you may want to use a
particular mechanism, how to use
1it,
and
the
effects it may have on your system.

Chapter

describes
you

can

the

use

access

increase

system
policy

control

alter

system

security

use. This
changes

mechanisms

policy

that

decisions

against

unauthorized

chapter includes the
type
of
you
may
want to make at your

installation.
Chapter

describes
feature

the
of

Common

TOPS—-20.

File

This

System,

chapter

rules,
options,
and
restrictions
with sharing files among systems.
Chapter

describes
software,
Ethernet

the

Local

for

wuse

local

area

Area
with

software
the

associated

Transport
(LAT)
servers in

terminal

networks.,

discusses

The

following

conventions

Convention/Symbol

and

symbols

used

throughout

this

guide:

Description
n
refers
file, for

n-—

are

to the latest version of
example, 6-1-CONFIG.CMD.

particular

representations,
must
be
entered

1indicates
exactly as

UPPERCASE

In
user
information
shown,

lowercase

In
user
input
representations,
indicates
variable information that is determined by you.

red

Indicates the information
your terminal.

input
that

that

you must

type

()

In user input representations,
encloses
guide
word
information.
Pressing
the
ESCAPE
or
ALTMODE
key on your terminal causes guidewords
to be printed by the computer.

<RET>

Indicates you
key
on
your
pressing

should press the
terminal. Unless

RETURN

terminates

all

RET
or
otherwise
command

RETURN
noted,
input

strings.
CTRL/

Indicates you should press
terminal.
The
CTRL
key
conjunction
with
another
CTRL/Z.

the CTRL key on your
1is
always
wused
in
key,
for
example,

CHAPTER

PREPARING

FOR

SOFTWARE

INSTALLATION

You can establish
many
of
the
policies
and
procedures
for
your
computer
site before you install the software.
It may help you later
if some of the preliminary decisions and preparations are done
before
you begin setting up the system and handling requests from users.
The
following
suggestions
for
©preparing
your
installation
are
not
all-inclusive.
Some TOPS-2@ installations have specific requirements

or restrictions that are not considered here.
as
a
guideline
stages of setting

2.1

You can use

for the types of decisions you can make
up your computer site.

this

1list

in the early

SECURING THE COMPUTER ROOM

Select the type of computer room security you need
and
a
method
of
enforcement.
Many system managers do not allow non-operations people
to enter the computer room.
Establish an open- or closed-door policy,
and
notify
users
of
your
policy.
If you decide on a closed-door
policy,

notify users of

you

(or the operator)

2.2

HANDLING

USER

the

and

procedures

that they should

to submit their job

requests.

use

contact

REQUESTS

Determine how user requests will be handled.
first-come
basis,
or
on
boxes outside the computer

You can handle jobs on a

a
priority basis.
You can set up request
room that the
operator
checks
regularly.

You
can
also
establish
a
location where users can leave disks and
tapes for the operator to mount.
Post a sign-up sheet so
that
users
can
specify
the
time they need the tape or disk mounted.
Chapter 1

describes

initial

sample

access

forms that can

completed

users

request

to the system and to request that work be done by the

operator.

2.3

ORDERING

SUPPLIES

Assign
someone
the
responsibility
for
ordering
paper
supplies,
ribbons, cards, and magnetic tapes.
Chapter 7 provides an estimate of
the number of tapes you
should
have
to
begin
a
backup
procedure
immediately
after
you install the software.
Be sure you have enough
CTY (operator terminal) and line printer paper to begin operations.

PREPARING

2.4

The

SCHEDULING

operator

as—-needed
regular

OPERATOR

performs

Dbasis,

tasks
Be

SOFTWARE

INSTALLATION

TASKS

Decide

schedule.

FOR

either

which

sure

operator

regular

tasks

include

regularly

The

hardware-

following
that

lists

your

are

samples

operator

may

Clean

tops

tasks

and

can

software-related

Tasks

Schedule

disk

software,

scheduled

and

performed

perform.

Hardware—Related

Regular

hardware,

documentation
related
tasks.
These
performed daily, weekly, or monthly.

tasks

basis

will

As—Needed

drives.

Replenish

Schedule

paper

the

1line

printer.

Clean magnetic

Vacuum
line
paper chad.

tape

drives.

printer

Remove
printer
to mail

remove

Load
mountable
structures
according to a schedule.

Replenish
console.

Run

scheduled

after

batch

As—-Needed

weekly

production

jobs.
Save

operator's

and
disk

unload
drives.

Bring

Maintain

system

the

Schedule

after

crash.

batch

system

for

disk

areas

users.
the

magnhetic

contents

disk

tape.

Create

for

1in

the line
(perhaps

Tasks

Schedule

Bring
up
system
maintenance.

paper

Physically
1load
magnetic tape and

Software-Related
Regular

reports
from
and distribute
boxes).

a
system
backup.

"crash"

Run

the SPEAR program
error analysis.

Submit a daily
accounting.

control

tape

Save

tape.

Restore

selected
needed.

areas
for

daily

file

for

Create
the
Message-of-the-Day
with the MAIL program.

special

magnetic

Interact

with

Create
and
directories.
Monitor

disk

user

disk

users.

update

user

space.

Establish a location for keeping the hard-copy output
from
the
Field
Service Representative needs this information if you
problems with your system.
Have the operator tear off
the
copy
store it daily.
Your

CTY.
have
and

PREPARING

Documentation-related
e

keeping

tasks

FOR

INSTALLATION

include:

hand-written

recording operator
Log)

SOFTWARE

log

shift

system

change

activities

information

(System

(Operator

Log)
©Shift

Change

coordinating
the
mounting
(Mountable Structure Sign-up

Chapter 1 describes creating a
log, and a mountable structure

2.5

SELECTING

Determine the
installation.

SYSTEM

and
Log)

dismounting

system log, an
sign-up log.

operator

structures

shift

change

FEATURES

system features
you
When you install the

want
to
software,

enable
during
software
you create a file called

n-CONFIG.CMD.
This file is read by a start-up program (n-SETSPD) when
the system is started for the first time and each subsequent time that
you reload and start the system.
The n-CONFIG.CMD
file
defines
the
line
speeds
for
your terminals and many system parameters.
Most of
the decisions you must make concerning the parameters in this file are
described
throughout
this manual.
As you read each chapter, you can
list the parameters that you want to place in the
n-CONFIG.CMD
file.
Many
system
managers
choose
to
introduce
new
pieces of software
slowly.
Therefore, you may want to disable
some
of
the
parameters

until
you
have
n-CONFIG.CMD file
should

edit

system.

Then,

run
the
new software for awhile.
You can edit
to add new software features
to
the
system.

the
when

file
the

system

convenient
restarts,

time

the

new

before

you

software

the
You

reload

the

features

are

enabled.

Chapters 3 through 13 describe setting up and
Read these chapters thoroughly.
They contain
help you
make
decisions
both
before
and
software.,

maintaining your system.
important information to
after
you
1install
the

CHAPTER

AFTER

3.1

SOFTWARE

INSTALLATION

OVERVIEW

After you install the TOPS-20 software, your system contains
all
the
directories
and
files
necessary
for
you
to
start
preparing for
timesharing
and
batch
processing.
This
chapter
describes
the
directories,
files,
and system logical names created during software
installation.
Also included are suggestions for
<creating
additional
directories and logical names to assist you and system users.

3.2

SPECIAL

SYSTEM

You initialize the
time,
the
system
that

you defined
Figure 3-1:

DIRECTORIES

file system during software
automatically creates nine

the

system

structure.

installation.
At
this
directories on the disk
These directories are shown

Public \

Structure

<Root-Directory >
<System>
<Subsys>>
<New-System>
<New-Subsys >
<Accounts>

MR-8-3718-84

Figure

3-1:

Sections
Section
they

are

Chapter
showing

Special

System

Directories

3.2.1
3.2.8

through 3.2.7 describe these directories
describes additional directories you can
useful.
5 also describes
the structure of

creating directories
directories.

and

and their use.
create and how

includes

diagrams

AFTER SOFTWARE INSTALLATION
3.2.1

<ROOT-DIRECTORY>

The <ROOT-DIRECTORY> contains a separate file
directory on the system structure as follows:

for

each

level

£first

STR: <ROOT=DIRECTORY>

STR:<SYSTEM> ... STR:<SUBSYS> ... STR:<DIRECTORY>
is the name of the structure).

(where STR:

The <ROOT-DIRECTORY> is the most important directory created. Without
it, directories and files cannot be accessed. You must NEVER modify
of
copy
backup
a
maintains
system
The
directory.
this
Iis
copy
original
the
if
accessed
be
can
<ROOT-DIRECTORY> that
destroyed. (Refer to Section 9.3, RESTORING <ROOT-DIRECTORY>.)

Each structure you create in addition to the system

structure

has

<ROOT-DIRECTORY>. The <ROOT-DIRECTORY> on any structure points to all
the first-level directories created under the <ROOT-DIRECTORY>.
install

After you

the

software,

give

the

DIRECTORY

command

for

The output on your terminal appears similar to the
<ROOT-DIRECTORY>.
the
in
file
that each directory is a
Note
example below.
The differences between this list and the one on
<ROOT-DIRECTORY>.
your terminal depend on the model system you have and the type of
unbundled software you have purchased.
(OF FILES)

$SDIRECTORY

STR:<ROOT-DIRECTORY><RET>

STR: <ROOT-DIRECTORY>
ACCOUNTS .DIRECTORY.1

BACKUP-COPY-OF-ROOT-DIRECTORY.IMAGE.,1
BOOTSTRAP.BIN.1
DSKBTTBL..1

1
FRONT-END-FILE-SYSTEM.BIN,

INDEX~-TABLE.BIN.1
NEW-SUBSYS.DIRECTORY.1
NEW-SYSTEM.DIRECTORY.1
1l
OPERATOR.DIRECTORY,
ROOT-DIRECTORY.DIRECTORY.1
SPOOL.DIRECTORY.1
SUBSYS.DIRECTORY.1
SYSTEM,DIRECTORY.1
SYSTEM—-ERROR.DIRECTORY.1
UETP.DIRECTORY.1

Total of

3.2.2

Files

The directory <SYSTEM> contains data and program files that the system

uses

appear

during normal operation.,
in

this directory.

Table 3-1 lists many of the files that

AFTER

Table

File

3-1:

Name

@gDUMP11 .BIN

SOFTWARE

INSTALLATION

Files

Explanation

Contains
after

the

dump

front

end

2060-MONBIG.EXE

The

smallest

2060-MONMAX.
EXE

The

largest

n=CONFIG.CMD

Contains

runnable

monitor.
line speeds,
printer
VFU

logical

files,

magnetic
DECnet
additional

tape
1logical
parameters,

numbers,

These

are

system

time
equals

every

The
value n
of TOPS—-240.

unit
and

system—-dependent

parameters.

set

memory

monitor.

definitions
of
names,

system

n-PTYCON,ATO

front-end
crashes.

the
the

parameters

system
latest

Contains the commands that
automatically
at
the

starts.
release

are
given
operator's

console every time the system
starts.
You may modify this file to suit
vyour
own installation.
n-SETSPD.
EXE

Program

file

the

that

runs

monitor

file

The

release

Program

n-SYSJOB.RUN

reads

sets

contains.

latest
n-SYSJOB.
EXE

that

and

the

n-CONFIG.CMD

parameters

value n

equals

that

the

TOPS-24.

process

takes

created

commands

from

n-SYSJOB.RUN.

Contains

commands

that

SYSJOB

processes.

ACCOUNTS-TABLE.BIN

Contains
validate

AN-MONBIG.
EXE

AN-MONDCN,
EXE

A monitor

large

the information
accounts,

ARPANET
that

necessary

timesharing
1includes

monitor.

ARPAnet

and

DECnet.
AN-MONMAX
. EXE

BUGS .MAC

The
largest
monitor.

ARPANET

Contains

and
CHECKD. EXE

Program

checks
DEVICE-STATUS.BIN

list

BUGCHK

BUGHLT,

BUGINF,

messages.

that
creates
structures
file-system consistency.

Contains

status

drives,

disk

structures.
MOUNTR.

all

timesharing

information

drives,
is

for

and

maintained

and

tape

disk
by

AFTER SOFTWARE

Table 3-1:
File

<SYSTEM> Files

Name

DUMP.CPY

(Cont.)
Explanation

the
Contains a copy of main memory at
1is
It
time of the last system crash.
a
maintain
to
DUMP.EXE
copied from
is
file
This
crashes.
of
history
when
written by the n-SETSPD program
the

DUMP.EXE

INSTALLATION

system

rebooted.

the

Contains a copy of main memory at

time
of
must have

the
this

dump after

ERRMES.BIN

Contains

EXEC, EXE

The

FEDDT.EXE

DDT

system crash. You
system
a
get
to

crash,

binary system

TOPS-20

front

1last
file

Processor,

Command

program

error messages.

used

for debugging

the

end.

HOSTS
. TXT

Defines ARPAnet host names
number translations.

INTERNET.GATEWAYS

Defines
reaching

and

their

the
network
gateways
for
remote
on
systems
host

networks.
IPALOD. EXE

KNILDR.EXE

Program that loads the CI20# microcode.
the
in
<contained
1is
(The microcode
file.)
After
the
1loading
has
completed, TOPS-2¢ starts the CI.
Program

that

loads.

the

NIA20

microcode. (The microcode is contained
in
the file.) It is run automatically
at system startup to start the NI.

MONITR.EXE

The

current monitor.

MONNAM. TXT

Contains the monitor
the beginning of the

at
printed
name
system
greeting

line.
PROGRAM-NAME-CACHE
. TXT

Contains
a
1list of the programs that
should be loaded into the program-name
cache. Read by the MAPPER program.

REAPER.,CMD

Contains a list of default commands to
REAPER. The REAPER program reads
this
file each time it 1is run.

RSX20F .MAP

Contains
front-end
FEDDT

SYSJOB.HLP

symbol
1locations
for
the
processor. It is used by the

program,

Contains

information about the SYSJOB

program.

SYSTEM.CMD

Contains
the

OPR

OPR
program

commands and
at

system

read

startup.

Table

3-1:

File

AFTER

SOFTWARE

Files

(Cont.)

Name

INSTALLATION

Explanation

TAPNAM, TXT

Text

file

installation
written

that

contains

identifier
VOL1
1labels

the

that
is
for
1labeled

tapes.

TGHA . EXE

Program
memory

TGHA.HLP

that

analyzes

and

corrects

MOS

problems.

Contains

information

about

the

TGHA

program.

TOPS-20.DOC

Text

file

that

information

about

3.2.3
If

Restoring

the

can

contents

restore

the
of

the

Directory

contains

the

summary

latest

release

TOPS-20.

are

directory

from

accidentally
the

TOPS-20

lost

destroyed,

Installation

Tape

you
your

latest system backup tape.
(Refer to Chapter 7 for information
about
creating
system
backup
tapes.)
Use
the procedure below to restore
<SYSTEM> directory.
If you have enabled tape
drive
allocation,
use

the

8.3

MOUNT
command
information

instead

for

Mount the
MTAQ:)
.

Give

the

@ENABLE

about

of
using

appropriate

following

the ASSIGN
tape drive

tape

(in

this

example,

your

terminal.

commands

(CAPABILITIES)

command.
(Refer
allocation.)
it

Section

drive

<RET>

SASSIGN (DEVICE) MTA@: <RET>
S$SKIP (DEVICE) MTA@: 4 FILES <RET>
$RUN (PROGRAM) MTA@: <RET>
DUMPER>

TAPE (DEVICE) MTA@: <RET>
DUMPER> RESTORE (TAPE FILES) DSK*:<*>*
*_*

DUMPER

TAPE

LOADING
END

The
that

INTO

FRIDAY

1-NOV-85

<RET>

330

SAVESET

gUMPER>EXIT

3.2,4

FILES

(TO)

<RET>

directory
the

user

<SUBSYS>
may

contains

want

run.

system
The

programs

directory

(and their help files)
protection code set for

prevents users from changing the
files
1in
this
directory.
the file protections require users to enable WHEEL or OPERATOR
capabilities to use the files.
(Refer to Chapter
5
for
information
about
directory and file protections and special capabilities.) Table
3-2 lists the programs and files
commonly
placed
in
<SUBSYS>.
An

Many

asterisk

precedes

all

unbundled

software.

AFTER

Table

3-2:

STR:<SUBSYS>

SOFTWARE

INSTALLATION

Files

Programs

Explanation

ACTGEN. EXE

Program that takes
files
and
creates

information
from
accounting
the
account validation data

base.

ACTGEN.HLP

Contains

information about

ACTSYM,.UNV

A file of universal

the ACTGEN program.

symbols

for USAGE

accounting

programs.

ANAUNV, UNV

A file of

ARPANET universal

symbols.

*B362LB.REL

BLISS
functions
needed
to
rebuild
the Record
Management
Services
facility
(RMS-20)
from
AUTOPATCH.

*BASIC.EXE

The BASIC compiler.

BATCON. EXE

Program

CDRIVE,.EXE

Program that controls card

CHECKD.EXE

Program

that

CHECKD, HLP

Contains

information about

CHKPNT.EXE

Program that makes accounting entries in the file

that controls batch jobs.

creates

file-system consistency

readers.

structures
(same

the

and

checks

in <SYSTEM>).

CHECKD program.

<ACCOUNTS>CHECKPOINT.BIN.

CHKPNT.HLP

Contains

CMD.REL

information

library file of

about the CHKPNT program.

routines

for the COMND monitor

symbols

for the COMND monitor

call.

CMD. UNV

A file of universal
call.

*COBDDT, HLP

Contains

information

*COBDDT.REL

The COBOL debugging

*COBOL. EXE

The

*COBOL.HLP

Contains

CREF. EXE

Program

that produces

., HLP
CREF

Contains

information

DIL.LIB

A library file
programs
that
(DIL) facility.

DIL.REL

The

DILV7.FOR

Contains

data

that

DIL.

COBOL

DIL

use

about

COBDDT.

program.

compiler.

information about the COBOL compiler.

of
use

cross—-reference

about
data

the

listing.

CREF program.

definitions
for
COBOL
Data Interchange Library

subroutines.
definitions

3-6

for

FORTRAN

programs

AFTER

Table

3-2:

STR:<SUBSYS>

SOFTWARE

Files

(Cont.)

Programs

Explanation

DITV7.FOR

Contains

data

that

the

DIXV7.FOR

DLUSER. EXE

use

data

that

the

Program

definitions

data

Contains

use

INSTALLATION

for

definitions

data

that

FORTRAN

transmission
for

conversion

saves

and

programs

component
FORTRAN

component

restores

DUMPER. EXE

Contains
Program

information
that

saves

and

magnetic

tape.

DUMPER. HLP

Contains

information

DX20LD,EXE

Program

DXMCA.ADX

Microcode

EDDT.REL

that

the

EDIT.HLP

Contains

*FAL, EXE

Program

FE.EXE

Program

directory

Contains

FEDDT
. EXE

The

FILCOM, EXE

Program

FILCOM.,HLP

Contains

editor.

The

and

from

program.

controller.

program

for

used

the

EDIT

for

DECnet

when

copying

the

program.

file

transfers.

files

TOPS-2¢

from

file

the

system

versa.
information

about

the

compares

the

contents

used

about
for

used

Contains

*FORO
.TS
EXE

The
FORTRAN
interface).

* FORTRA
. EXE

The

front

the

end.
of

FILCOM

examining

about

two

files.

program.

the

contents

FORDDT

program.

the

program.

format

RP@4/RPP6

timesharing

information

FORTRAN

program.

(DUMP.CPY).

debugging

FORMAT.HLP

for

information

system

the

program

information

FORTRAN

Program

about

system

program

* FORDDT.REL

GALGEN.EXE

text

'listens'

file

dumps

Contains

the

debugging

information

that

*FORDDT.HLP

FORMAT,EXE

subsystem

the

DUMPER

tape

debugging

DDT

the

DX2@

that

vice

system

about

program.

files

microcode.

that

FE.HLP

restores

DX2f

line-oriented

front-end

- FILDDT.EXE

DLUSER

DIL.

monitor.

EDIT.EXE

and

the

loads

for

component

TOPS-20

about

DIL.

programs

parameters,

DLUSER.HLP

object

about
time

disk

packs

while

mode.

the

system

FORMAT

program.

(operating

system

compiler.

Program

that

building

the

creates

batch

3-7

the

system.

©parameter

file

for

AFTER SOFTWARE INSTALLATION

Table

3-2:

STR:<SUBSYS> Files

Programs

(Cont.)

Explanation

GLOBS.UNV

A file of wuniversal symbols for the TOPS-ZQ

GLXLIB.EXE

Object-time system used by the GALAXY programs.

HELP.HLP

Contains information about the HELP command.

EXE
*IBMSPL.

Spooling program that sends

INFO. EXE

Program

monitor.

IBM-batch-job

to remote IBM host and retrieves the output.

files

that gives information to programs using

IPCF.

indexed

*ISAM,. EXE

Program that maintains COBOL

*ISAM.HLP

Contains information about the ISAM program.

KDDT.REL

single-key

files.

sequential

the
A component of
TOPS-2# monitor.

program

debugging

KNILDR. EXE

Program that loads the NIA2@ microcode.

LCPORN.REL

The

LCP subprocess to

LCPTAB.REL

The

LCP command

the

for

the OPR program.

table.

maintains, and lists the

creates,

Program

that

*LIBARY.HLP

Contains

information about the LIBARY program.

*LIBOl2.EXE

The COBOL object-time

*LIBARY.EXE

contents of COBOL library files.

(operating

system

interface).
*LIBOL.REL

Contains the COBOL library subroutines.

LINK.EXE

Program that loads relocatable binary programs.

LINK.HLP

Contains

LP64 .RAM

. RAM
LP96

information about the LINK program.

Translation

RAM

file

for

64-character line

Translation

RAM

file

for

96-character

printer.

Read

by n-SETSPD.

line

by n-SETSPD.

LPTSPL. EXE

Program that controls output to the line printer.

MACREL.REL

Run-time file

MACRO. EXE

The MACRO assembler.

MACRO.HLP

Contains information about the MACRO assembler.

MACSYM.UNV

Contains

MAIL.EXE

Program that

MAIL.HLP

Contains information about the MAIL program.

for macros

in MACSYM.

system macros.

sends messages

3-8

users.

AFTER

Table

3-2:

Programs
MAILER.EXE

STR:<SUBSYS>

Files

INSTALLATION

(Cont.)

Explanation
Program

and
MAKDMP . EXE

SOFTWARE

that

places

Program

receives

that

the

mail

from

the

appropriate

produces

MAIL

program

mailbox.

standard

DUMP.EXE

file

<SYSTEM>.
MAKLIB, EXE

Program

that

creates

relocatable

libraries.
MAKLIB,HLP

MAKRAM, EXE

Contains

Program

line

information

that

creates

Contains

information

MAKVFU, EXE

Program

that

(VFU)

Contains

MAPPER.EXE

Program

creates

Object
file
definitions.

MONSYM.UNV

Contains

MOUNTR.EXE

Program

MSCPAR.UNV

the

MAKRAM

vertical

about

the

that

symbol
that

file

MSCP

the

file

the

*NFT.HLP

Contains

*NMLT20

DECnet

program

functions.

file

Program

jobs

Contains

transfer

information
that

formatting
that

and

the

formatting

unit

program.

program

Program

that

MOUNTR,

LPTSPL,

Overlay manager

PAl1050.EXE

The

PAT, EXE

Version

for

(Refer
Time.)

for

the

symbols

about

calls.

used

build

the

monitor.

the

performs

unit

file

about

network

for

line

the

LINK

that

the

can

control

printers.

interface

with

OPR program.

messages

Compatibility

program.

system.

for

the

NFT.EXE

uses

the

EXEC,

3-9

monitor

symbol

program.

QUASAR,

PAl@58

call

structures.

TOPS-2#

processes

OVRLAY.REL

monitor

and

operator

devices

information

TOPS-1@

tapes

wuniversal

component

Vertical

contains

mounts

DECnet

all

for

program.

MAKVFU

debugging

definitions

*NFT.EXE

ORION. EXE

RAM

monitor.

MONSYM.REL

OPR.HLP

about

information

component

TOPS-20

OPR.EXE

translation

program.

that loads the program-name cache.
Section 10.4, Improving Program Startup

NORMAL.VFU

MAKLIB

file.

MAKVFU. HLP

the

subroutine

printers.

MAKRAM, HLP

MDDT.REL

about

sent

etc.

the

OPR,

programs.

program.

Package.
be

used

debugging.

AFTER SOFTWARE INSTALLATION

Table

3-2:

(Cont.)

STR:<SUBSYS> Files

Programs

Explanation

PHYPAR.UNV

PLEASE. EXE

file

Program

that

input/output

universal

symbols

establishes

programs.

TOPS-20

for

dialog

with

the

operator.

file

symbols used to build the

wuniversal

PROLOG. UNV

PTYCON,. EXE

Program that controls many

PTYCON. HLP

Contains information about the PTYCON program.

R
. EXE
QMANG

Program that manages the batch and print queues.

QUASAR, EXE

Program

RDMAIL. EXE

Program that allows a user to read mail sent with

TOPS-20 monitor.

scheduling

the
RDMAIL,.HLP
REAPER.EXE

Jobs

from

single

terminal.

MAIL

does

that

the

queuing

central

for the batch system.

and

program.

Contains information about the RDMAIL program.

Program

that

marks

files

for

migration

magnetic

tape.

REAPER, HLP

Contains

information about the REAPER program.

*RERUN. EXE

Restarts

COBOL

*RERUN. HLP

Contains information about the RERUN program.

RETRFB.SPE

Contains SPEAR report templates.

RFB.EYE

Contains internal definitions
function of

the SPEAR program.

RMS. EXE

RMS-2¢ used

in Section #

. EXE
RMSCOB

RMS-20 used

by COBOL V12B programs.

RMSINI.REL

Routine called by BLISS

RMSINT.R36

Unsupported BLISS interface file for RMS-20.

RMSINT.UNV

MACRO interface file for RMS-20.

RMSUTL. EXE

The RMS-20 file maintenance utility.

RSXFMT.EXE

Utility program used for converting TOPS-20 files
to a format used by the front end and vice versa.

RSXFMT.HLP

Contains

. EXE
RUNOFF

Program that helps with text preparation.

RUNOFF.HLP

Contains

initialize

programs.

for

the

RETRIEVE

of memory.

and

MACRO

programs

RMS-20.

information about the RSXFMT program.

information about the RUNOFF program.

AFTER

Table

3-2:

STR:<SUBSYS>

SOFTWARE

Files

Programs

Explanation

SCAPAR.UNV

(Cont.)

parameter

file

inter-system
SCSTST.EXE

SDDT.EXE

DDT

*SELOTS.EXE

program

that

debugger

that
and
(It

*SIX12.REL

The

*SORT.EXE

Program

*SORT.HLP

Contains

SPRINT.EXE

Program

that

Output

spooler

for

the

used

symbol

11.)
for

the
COBOL
object-time
system
versions of COBOL up to

the

SYSERR

error

that

sorts

files

information
creates
for

record

about

the

batch

jobs

card

punch,

record.

SORT program.
from

card

paper

tape

SPEAR.EXE

Segment

SPEAR

program.

SPRRET.EXE

Segment

SPEAR

program,

SPRSUM. EXE

Segment

SPEAR

program.

SYSJOB.HLP

Contains

SYSTAP.CTL

Control

TCX.EXE

TCX.HLP

Contains

information

about

the

TCX

TERMINAL,
HLP

Contains

information

about

the

TERMINAL

TOC.EXE

file

TOC.HLP

Contains

TV.EXE

UDDT.EXE

DDT

ULIST.EXE

information

DIGITAL

Standard

Runoff

information

Program

VERIFY, EXE

Program

for

SYSJOB

system

index

that

for

used

with

about

version

observing

3-11

program.

backup

tape.

utility.

for

TOC

command.

creating

utility.

symbol

table.

information

the

during

integrity

punch,

editor.

users,

the

and

text

the

input.

utility.

utility

printing

and

determine

Program

about

programs

information

checksums
WATCH,. EXE

Runoff

the

contents.

debugger

Contains

about

creates

Standard

directories

that

character-oriented

ULIST.HLP

codes.

debugger.

plotter.

table

table.

between

with

definitions

DIGITAL

SCA

routines.

without

COBOL

version

Contains

and

programs

interfaces

including

BLISS

symbols

SCA.

the

SERCOD, UNV

SPROUT. EXE

tests

for

Program
(LIBOL).

communication

language
and

INSTALLATION

ULIST

software

numbers
system

program.
installation

files.

.EXE

the

about

performance.

verifies

files.

AFTER SOFTWARE INSTALLATION

STR:<SUBSYS> Files (Cont.)

Table 3-2:
Programs

Explanation

WATCH,. HLP

Contains information about the WATCH program.

Library containing the BLISS

*XPORT.REL

memory, and string functions.

RMS-20¢ for use from memory

XRMS. EXE

Section

transportable

1/0,

other

than

sections

NOTE

All the .HLP files can be displayed wusing the HELP
command; for example, the command @HELP WATCH displays
the WATCH.HLP file.

Restoring the Directory <SUBSYS>

3.2.5

If the contents of <SUBSYS> are accidentally lost or destroyed, you
can restore the directory from the TOPS-20 Installation Tape or your
latest system backup tape. (Refer to Chapter 7 for information about
creating system backup tapes.) Use the procedure below to restore the
<SUBSYS> directory. If you have enabled tape drive allocation, use
the MOUNT command instead of the ASSIGN command. (Refer to Section
8.3 for information about using tape drive allocation.)

Mount the appropriate tape (in this example, it is

Give the following commands at your terminal.

drive

MTAQ:)

@ENABLE

(CAPABILITIES)

$ASSIGN (DEVICE) MTAQ:
$SKIP (DEVICE) MTA@: 4
(PROGRAM)

$RUN

DUMPER> TAPE

MTA@:

(DEVICE)

<RET>

<RET>
FILES

<RET>

MTA#:

<RET>

‘

|
DUMPER> SKIP (NUMBER OF SAVESETS) 1 <RET>
* % (70)
DUMPER> RESTORE (TAPE FILES) DSK*:<*>*
|

<RET>

DUMPER TAPE # 1 ,
LOADING FILES
END

, SATURDAY,

3-NOV-84 330

INTO STR:<SUBSYS>

SAVESET

DUMPER> EXIT <RET>
$

AFTER

and

<NEW-SYSTEM>

3.2.6

SOFTWARE

INSTALLATION

<NEW-SUBSYS>

program
DLUSER
the
The first time you install the TOPS-2¢ software,
creates
the
directories
<NEW-SYSTEM> and <KNEW-SUBSYS>.
They do not
release
new
a
when
directories
these
You can use
contain files.
When
updating the existing system.
are
you
and
available
becomes

Installation
TOPS-2@
the
on
DIGITAL distributes an updated monitor
this tape to the
from
savesets
two
first
the
restore
you
Tape,
You use these
directories <KNEW-SYSTEM> and <NEW-SUBSYS> respectively.
directories
until you feel comfortable with the new software.
Should
to
you have any problems with the new software, you can easily revert
using
the old software.
Appendixes A and B of the TOPS-2@ KL Model B
Installation Guide
detail
the
procedures
to
wupdate
one
software
release

another.

comfortable
If you have no problems with the new monitor, and you are
the files in the directory <NEW-SYSTEM> into the
copy all
1it,
with

directory <SYSTEM> and all

the

into

directory

<NEW-SYSTEM>
<KNEW-SUBSYS>
is

the files

<SUBSYS>.

You

the

<NEW-SUBSYS>

directory

cah now delete all the files in

and
<NEW-SYSTEM>
directories
The
<NEW-SUBSYS>.
and
remain empty until a new version of the TOPS-20 software

distributed.
NOTE

directories
the
into
files
new
After you copy the
vyou cannot revert to the old
<SUBSYS>»,
and
<SYSTEM>
using
system software unless you reinstall the system
the o0ld monitor or backup tapes.

3.2.7

<ACCOUNTS> - After installation, the directory <ACCOUNTS> contains one
This file contains all the accounting system
SYSTEM-DATA.BIN.
file,
the
If the directory <ACCOUNTS> is destroyed,
entries for each user.
accounting system creates a new SYSTEM-DATA.BIN file.
After

first

the

<ACCOUNTS>CHECKPOINT.BIN

the

file.

system,

This

the

system

creates

the

file stores accounting entries

user
a
After
in.
for each user during the time the user is logged
accounting data stored in CHECKPOINT.BIN is copied to
the
out,
logs
When the system comes up after a crash, the
the SYSTEM-DATA.BIN file.
to determine which users
<ACCOUNTS>CHECKPOINT.BIN
examines
monitor
in
the data
stores
and
crash,
the
were logged in at the time of
Therefore, users who did not log
in SYSTEM-DATA.BIN.
CHECKPOINT.BIN
for
charged
out in the normal fashion, because of a crash, are still
their

log—in

time,

file
<OPERATOR> - The directory <OPERATOR> normally contains the
This file usually contains a record of all the activities
PTYCON.LOG.
The
that occur under the operator jobs that are controlled by PTYCON.

directory
the

<OPERATOR> may also contain files the operator needs to run

system,

3-13

AFTER

SOFTWARE

INSTALLATION

<SPOOL> - The directory
<SPOOL>
contains
files
that
the
spooling
system
needs before performing any input or output.
They are kept in

this area until they can be output
to a slow-speed device
such
as
a
line
printer.
This
area
1is
also used for input of files from the
local card reader, if one is attached.
It may also be used for
input
of
files
from
IBM
remote
stations.
The
file
PRIMARY-MASTER-QUEUE-FILE.QUASAR is created
1in
this
directory.
It
contains
a copy of the input queues so that they are not destroyed if
the system crashes.
You must either delete this file or
process
all
entries
in
the
queues
before installing a new version of the batch
system that
has
a
different
queue
format.
The
GALAXY.DOC
file
describes
the
new
software
components
and
tells you if the queue
format

has

changed.

<SYSTEM-ERROR>
ERROR.SYS.
and

3.2.8

The

read

Other

The

directory

ERROR.SYS

by the

system

Useful

<SYSTEM-ERROR>

contains

the

file

file contains entries about system errors

error

recovery program,

SPEAR.,

Directories

You may want to create additional directories
for
storing
different
versions
of
programs
or
text.
Some useful directories are listed
below.
You should give these directories the proper protection number
and make them files-only directories.
Directory and

File

Protection

Directories and files that are executed or read
by
the
entire
user
community
should
not
be
given the default protection 7777660, which

allows no access.

777748

and

describes

<NEW>

the

They

file

directory and

file

given

777752

protections.)

the

directory

777712.

protection

(Section 5.7

The directory <KNEW> can contain versions of
your
software that are not completely tested
or
that
are
drastically
different
from
the
current
versions. If you create a directory <NEW>,
users
will
£find
it more convenient if you also create
the
system-logical
name
NEW:
defined
as
PUB : <NEW>,
SYs:,
where
PUB:
1is
the
public
structure. This logical name allows them
to
run
all
new
software
by merely typing NEW: and the
program name. If there is no file with the
given
name
in
<NEW>,
the
system
wuses
the
version
currently on <SUBSYS>. (Refer to Section 3.3
for
a

<OLD>

should

protection

description

logical

The directory <OLD>

of
If

software as
programs

names.)

can contain

the

o0ld

version

newer versions appear on <SUBSYS>.
or
data
do
not
work
with
new

software,
the
user
has a chance to correct the
problems before the older software is
no
1longer
available. Users will find it convenient
1f
you
also
define
the
system-logical
name
OLD:
as
PUB:<OLD>,SY¥S:,
where
PUB:
is
the
public
structure.

3-14

AFTER

creating

new

the

software

place

move

tape,
send

system

Store

time

you

apd

<OLD>,
a

software

has

<SUBSYS>,

and

place

<OLD>

version

on
the

introduce

becomes

available,

been

the

use

version

system

software,

back-up

you

should

users.
<HELP>

<NEW>,

list

gradually

the

contains

files
that
describe
different
versions
<SYSTEM>,

you

version

When

all

directory

new

<NEW>

version

change

message
The

1in

the

INSTALLATION

When

<KNEW>.

version

system-wide

<HELP>

<KNEW>

users.

directory

the

<OLD>.

Every
a

directories
to

the

awhile,

SOFTWARE

changes

and

documents

the
of

system
software

<OLD>,

vyou

incorporated

and

appear

should

the

help

software.
make

new

versions

it
in
the directory <HELP>. You can
files
with
the
file
type .HLP from
to
the
directory
<HELP>.
The
HELP

move

all
<SUBSYS>

command still works correctly if you
define
the
system-logical name HLP: to
be
PUB:<HELP>,SYS:,
where PUB: is the public structure.
<REMARKS>

The

directory

users

the

general
user

operator.

system

wants

contains

These

comments

send

does

not

require

send

message

the

MAIL

to
program.

the

MAIL

program

the

from

usually
When

message

response,

directory

(Refer
in

are

complaints.

operator

immediate

the

messages

that

can

using
description
of

the

TOPS-2@

User
Utilities
Guide.)
A
typical
message may be a request for
supplies,
for
example,
LA36
paper
or ribbon.
Creating the directory <REMARKS> avoids
constant

interruptions

requests.

messages

day,

3.3

PLEASE

the

simply

operator

The

when

from

operator

has

used

users
can

specified

issuing

read

the

time

each

time.

SYSTEM-LOGICAL NAMES

logical

locate

name

files.

usually

three

descriptive
can

six

word
to

establish

alphanumeric

characters;

characters.

Because

names

route

however,

logical

are used in place of device names, they always end with a colon.
Logical names tell the system where and in what order
to
search
for

files.
When
directories in
name,

a
wuser
types
a logical name, the system
the order they were defined or listed
by

Although

wusers

can

define

1logical

(refer

to the TOPS-20 User's Guide), the
can be used by all users of the system.
names in the n-CONFIG.CMD file.

names

logical
You can

for

searches the
the
1logical

their

own

use

names described
here
define system-logical

During installation,
the monitor, and may
S5YS:,

defined

several system-wide logical names are defined
be overridden in the n-CONFIG.CMD file.
They
PUB:<KNEW-SUBSYS>, PUB:<SUBSYS>; SYSTEM:,
defined

PUB:<NEW-SYSTEM>,

SYSTEM:EXEC.EXE;

aid

the

public

users

permanent,

PUB:<SYSTEM>;

and

accessing
place

<SYSTEM>n-CONFIG.CMD

SYSTEM:,

POBOX:,

structure.)

SY¥S:,

system-logical

You

files.

the
file

are

defined

the

may decide

you

want

definitions
on

DEFAULT-EXEC:,

names

DEFAULT-EXEC:,

the

POBOX:,

the

structure.

other

logical

1logical
an

defined

names

editor)

(PUB:

names

to
be

the

structure.

and

below.

3-15

add

(using

public

explained

public

by
are

some

other

frequently

used

AFTER SOFTWARE INSTALLATION

SYSTEM:

3.3.1

The logical name, SYSTEM:, defines a search list that contains all the
system programs and files that the system needs to operate. SYSTEM:
should always contain the directory <SYSTEM> on the public

structure.

contains the

directory

are updating the system with a new monitor, the definition of

you

SYSTEM:

<NEW-SYSTEM>.

n-CONFIG.CMD

the

example,

For

DEFINE SYSTEM:

also

file

STR:<NEW-SYSTEM>,STR:<SYSTEM>

where:

STR:

is the name of the public structure.

SY¥YS:

3.3.2

The logical name SYS:

defines a search list

that

all

the

file

also

(or

contains

system programs a user may want to run. SY¥S: should always contain
the directory <SUBSYS> and any other library directories that contain
If you are updating the system with a new
commonly wused programs.
definition

monitor, the

contains the directory <NEW-SUBSYS>.
DEFINE SYS:

the

n-CONFIG.CMD

For example,

STR:<NEW-SUBSYS>,STR:<SUBSYS>

where:

STR:

1is

the name of

the

public

structure.

Be sure to set the protection on the library
<NEW-SUBSYS>)

3.3.3

to 777748.

files

This protection allows access by all users.

NEW:

a directory
containing
defines a search list
The logical name NEW:
The
that has new software, followed by the system-logical name SY¥S:.
the
1is
in n-CONFIG.CMD,
put
definition for this, which you would
following:
DEFINE

NEW:

STR:<NEW>,SYS:

With this system-wide logical name, the user can give the command:
@DEFINE

(LOGICAL NAME)

5Y¥5:

(AS)

NEW:

<RET>

the
in
first
the system looks
Now, when the user runs a program,
and then in the normal system search list SYS:.
directory STR:<NEW>,
This way, the user always gets the most recent version of any program.

3-16

AFTER

you

OLD:

name

have

old

versions

may

helpful

OLD:

of
to

OLD:

programs,
users.

defining

The

usual

the

system-logical

definition

the

name

logical

©STR:<OLD>,SYS:

The definition OLD:
has
NEW:.
If the user gives

@DEFINE

3.3.5

INSTALLATION

is:

DEFINE

whenever

SOFTWARE

the
the

(LOGICAL NAME)

runs

same type
command:

5VYS:

program,

(AS)

will

effect

OLD:

get

the

definition

<RET>

oldest version

available.

HLP:

If
you
want
to
keep
programs
and
documentation
in
separate
directories,
you
should store the documentation in <HELP>.
The HELP
command searches the directories identified by the logical name
HLP:,
so you must define the logical name HLP:
to be the directory <HELP>.
The

definition
DEFINE

3.3.6

HLP:

n-CONFIG.CMD

should

be:

STR:<HELP>

SERR:

The logical name SERR:
defines a search list that contains the system
error
file
ERROR.SYS.
The
SERR:
1logical name tells the system to
search the directory <SYSTEM-ERROR> for the ERROR.SYS file.
This file
may

used

later

The

definition

DEFINE SERR:

produce

SERR:

reports.

nh-CONFIG.CMD

should

be:

STR: <SYSTEM-ERROR>

where:

STR:

1is

the

name

the

public

structure.

DMP:

3.3.7

When the system is re-booted after
a
crash,
the
file
DUMP.EXE
is
overwritten
with a copy of memory.
Upon system startup, the n-SETSPD
program copies the contents of DUMP.EXE
to
the
DMP:DUMP.CPY;P770000
file.
System
crashes
cause
DUMP.EXE
to
be
overwritten, but new
versions

DUMP.CPY

should

DUMP.CPY

files,

accumulate.

keep

the definition of

the

DMP:

public

structure

clear

the n-CONFIG.CMD file

be:

DEFINE

DMP:

STR:<DIRECTORY>

The structure and directory are your choice;
filename,
Versions
of
DUMP.CPY hereafter
area.

3-17

you should
accumulate

not specify
a
in the defined

AFTER

In
CFS
configurations,
DMP: definition, because
came from which system,

3.3.8

SOFTWARE

INSTALLATION

systems
this could

should
lead to

not
share
confusion about

a
common
which dump

DEFAULT-EXEC:

The

logical name DEFAULT-EXEC:
defines a search list that
points
to
the
TOPS-2@
Command Processor (EXEC).
When users log in or give the
PUSH command, the EXEC program is activated.
Some
experienced
users
may
choose
to
run
their
own
copies of the EXEC, not the standard
system

version.

name
after

for
their
giving the

level,
other

while

Such

program

users

can

private EXEC,
PUSH command.
batch

levels

may

define

interactive
invoke

the

Refer

the

3.3.9
The

for

Technical
complete

PUSH

commands

system

file

issued

version,

DEFAULT-EXEC:
if it
as SYSTEM:EXEC.EXE.

Summary

the

and

information

the.
on

the

from

unless

is defined.

TOPS-2

Commands

EXEC.

POBOX:

logical

files

DECSYSTEM-20

Manual

mode.

advantage of this feature
must be given at the EXEC

standard

the
program
has been written to use
By default, DEFAULT-EXEC:
1is defined

Reference

DEFAULT-EXEC:

and can take
This command

name

reside.

POBOX:
Mail

points

sent

users

the

structure

goes

the

where

MAIL.TXT

users'
files

mail
their

directories on the structure POBOX:.
Therefore, a directory
must
on
that
structure
for
each
wuser.
By default, POBOX:
defined as the public structure.
You
can
redefine
POBOX:
in
n-CONFIG.CMD file to be any structure you choose.

created

redefining
the public

POBOX:,

you
structure.

In CFS-20 configurations,
You
can
define
POBOX:
establishing
a
central

can

prevent

users'

mail

files

from

be
is
the

£filling

redefining

POBOX:
is
especially
wuseful.
the same structure for all systems,
for
all
mail
files
in
the

to
be
1location

configuration.
Then,
no matter what system users log onto, they are
automatically directed to this one area when
they
give
commands
to
access
their
mail.
They
do
not
have
to spend time logging onto
various systems to access mail that would otherwise have been sent
to
a
public structure.
To set up this central location, the same DEFINE
command

should

to
Chapter
CFS-20.

12,

entered

The

Common

each

system's

File

System,

3-18

n-CONFIG.CMD

for

further

file.

Refer

information

AFTER

3.3.10
The
your

SOFTWARE

INSTALLATION

NRT:

logical

name

system

NRT:

has

(Network

DECnet

Remote

Terminal)

communications

software.

applicable
When

the SET HOST command to connect to a remote system,
the
communications program is run by default.
If the remote
support

time

CTERM,

using

the

host

program

system

defined

tries

connect

the

user

only

a user

issues

CTERM-SERVER
node does not
again,

this

NRT:.

Examples
l.

For TOPS-20
definition:
DEFINE

TOPS-20

NRT:

communications,

give

the

following

SYS:SETHOST.EXE

For multi-operating system
DECnhet
communications,
you
can
specify the HOST program (located on the TOPS-20 tools tape):

DEFINE

3.4

CONSOLE

NRT:

FRONT-END

HOST.EXE

FILES

The console front-end computer consists of a PDP-1l1
with
28K
16-bit
words
of
memory.
When the system is brought up for timesharing, the
front—-end monitor, RSX20QF, is loaded in the PDP-11 memory and started.
The
TOPS-20 monitor is loaded in KL1# main memory and started.
Thus,
you have two computers working together.
Both
computers
have
their
own monitor and related software.
The

front—-end

file

system

consists

the

RSX28F monitor

and

programs
(tasks)
and
files.
During
software
installation, these
front-end files are transferred from floppy disks to a special area on
the
system
structure
unless
an
RP@7
1is
being used as the system
structure,
If an RP@7 is being used as the system structure, only the
files
on
the
TOPS-2@¢
Installation Tape will be placed on the RP#7.
The front—-end files,
on
the
floppy
disks,
must
be
placed
on
a
dual-ported
RP@P4 or RPP6 disk drive attached to the PDP-11 front end.
(Refer to the TOPS-20 KL Model B Installation Guide for the
procedure
for
<creating
the front-end file system when using an RP@7 disk drive
as the system structure.)
The area the
FILES
area,

front-end files are
or FILES-11 area.

placed on
Once this

is
area

<called
the
FRONT-END
has been set-up, there

is normally no need to get these files again from floppy
disks.
floppy
disks used to install the system become backup devices in
the

public

structure

destroyed,

the

case

where

RP@7

The
case
Iis

being
used as the system structure, they can be used to recreate your
front-end file structure.
It is a good idea to make an extra copy
of
your
installation floppies in the event one of your original floppies
is destroyed and you need to restore the FRONT-END FILES area.
Refer
to
Chapter
7, System Backup Procedures, for a description of the COP
program that is used to copy floppy disks.

As previously stated, the front—-end files must
dual-ported
RP@4
or
RPP6
attached
to
the
allows the front—end processor to access these
processor

accesses

TOPS-20

files

the

3-19

same

always be placed
on
a
PDP-11 front end.
This
files
while
the
main
or different disk packs.

AFTER SOFTWARE

INSTALLATION

The RSX20F monitor and its related programs do the following:
input/output and communications devices.

Control

Interface with the main computer.

Load the

TOPS-2@

monitor

TOPS-20
if

crash

system

startup,

system

and

reload

occurs.

Report

errors

TOPS-20.

Perform system diagnostic functions.

Table 3-3 lists the programs and files located

in the FRONT-END

FILES

Files with the file type TSK
with a brief description of each.,
area,
are programs that can be run under the front-end monitor RSX2@F; files
with the type MCB contain the microcode for the host processor (KL10);
the
1load
to
used
programs
bootstrap
are
files with the type EXB
monitor; and files with the type CMD are programs that record
TOPS-20¢

information about system errors.
This
information
1is
read
by
the
system
error recovery program, SPEAR.
Beginning with TOPS-2# Version
6,
console
front-end
filenames
include
edit-level numbers
that
indicate
the
versions
of
the
particular
programs,
for
example,
F11ACP.TSK;1505.

Table 3-3:

Console Front-End

Files

File

Contents or Function

BF16N1.All

MOS memory-timing RAM file.

BF64N1.All

Timing

B0OO.TSK

Used

BOOT.EXB

CLOCK.CMD

file

containing

file

is a nonexecutable

for 64K RAM chips.

boot

The central

boots

MOS memory data.

RSX20F.

processor disk bootstrap program that

TOPS-20

from disk.

Saves contents of memory locations for diagnosing
errors
that
are
purposely
induced
by
Field
Service.

COP.TSK

CRAM.CMD

Copies the contents of floppy disks.

Saves contents of memory locations fbr diagnosing

CRAM

parity

errors.

DEX.CMD

Saves contents of memory locations for diagnosing
Deposit Examine failures (PI level @ interrupt).

DMO.TSK

Dismounts a

DRAM.CMD

front—end device and allows a

reboot.

Saves contents of memory locations for diagnosing

DRAM

parity

errors.

EBUS.CMD

Saves contents of memory locations for diagnosing

FMPAR.CMD

Saves contents of memory locations for diagnosing

EBUS

parity

fast memory

errors,

parity

errors.

AFTER

Table

3-3:

Console

Front—-End

File

Contents

F11ACP.TSK

File

HALT.CMD

SOFTWARE

KLDISC.TSK

Provides

KLI.TSK

KLINIT

memory

front-end
KLINIK

program

Provides

KLX.MCB

KL1¢

KPALV.CMD

Saves

for

main
the

cease

Transfers

for

area.

disconnect

service.
the

microcode,

loads

line

diagnosing

central
configures

bootstrap.

ring

service.

file,

contents

alive

files.

initializing

memory,

microcode

files

Loads

KLINIK

disk

locations

line

(KL2@).

and

KLRING. TSK

keep

front-end

the

processor

LOOP.CMD

for

errors.

Initializes

cache

(Cont.)

Function

contents

halt

INI,TSK

LOGXFR. TSK

Files

handler

Saves

INSTALLATION

the

memory

locations

for

errors.
PARSER.LOG

error

file.

Saves

contents

errors

causing

the

memory

file

locations

hang

diagnosing

but

the

for

not

TOPS-~20

diagnosing

crash.

MIDNIT.TSK

Updates

MOU.TSK

Mounts

MTBOOT. EXB

Boots

PARSER.TSK

The
front-end command parser (prompts PAR>). The
primary means of access to front-end programs.

PIP,.TSK

Front-end

RED. TSK

Tells

the
a

day

through

for

use

with

the

monitor

from

magnetic

device
TOPS-20¢

program

for

file

the system where
device, SY@:.

files
RSX20F.MAP

Contains

RSX20F.SYS

Virgin

image

Saves

the

SAV. TSK

time

symbolic

disk.,

midnight.
front

tape.

transfer.

look

definitions

for

front-end monitor

front-end

end.

monitor

the

front-—end

RSX20F.

(RSX20F).

and

bootstrap

SB@.CMD

Helps
field
problems.

service

diagnose

SBUS-related

SB1.CMD

Helps
field
problems.

service

diagnose

SBUS~related

SETSPD,. TSK

Sets

TIMEO,.CMD

Saves

system

contents

protocol

TKTN. TSK

parameters,

Terminates

reloads.

timeout

tasks,

such

memory

line

locations

speeds.

for

errors.

reports

errors,

and

diagnosing

requests

AFTER SOFTWARE INSTALLATION

Table 3-3:

Console Front—-End Files (Cont.)

File

Contents or Function

Interfaces between

T2@8ACP,.TSK

front—-end

TOPS-28

and

user—-file

directories in the front-end

UFD, TSK

Sets

ZAP.TSK

Makes binary modifications to task images.

3.5

file

systems.

files

area.

TAILORING THE BATCH SYSTEM

Most installations use the parameters and defaults in the distributed
version of the batch system. However, you can modify some of these
parameters if required by the

batch

processing

procedures

your

installation.,.

DIGITAL distributes a program with the TOPS-2@ software that allows
you to tailor the standard batch system to the requirements of your
installation. This program, called GALGEN.EXE, is " located 1in the
directory <SUBSYS> on the system structure. You can run GALGEN at the
In either
time the system software is installed or at a later date.
case, you must have a working batch system before you can generate a
new one using GALGEN. This means if you are installing the systemnm,

you must first install the batch system that is distributed with every
new version of the TOPS-28 software (on the software installation
You can then run the GALGEN program and tailor the batch
tape).
system before it becomes available for general use.

If you tailor the batch system at a later date, you can run the GALGEN
program with users logged in. However, for safety reasons, the system

should be stand-alone during the critical phase of
system

batch

and

need not be empty.

starting

the new one.

stopping

the

old

The batch queues, however,

That is, batch jobs can be waiting to be processed

at the time you bring the system down.

The TOPS-28 KL Model B Installation Guide contains the procedures
s

running

3.6

—ed

the

GALGEN

for

program.

CHECKING THE SOFTWARE

(UETP)

After the system software is installed, you or the Software Specialist
UETP 1is a
can run the User Environment Test Package (UETP).
designed
files
control
batch
and
files,
data
collection of programs,

In
to allow you to test the integrity of various system elements.
installed,
properly
been
has
hardware
the
that
addition to testing
UETP ensures that the TOPS-2¢ Operating System is running and that the
languages you have selected for your operation are available.

UETP creates a moderate load on the system, consisting of various
defined procedures that closely resemble the 1load in an actual
operation. Later, you may want to tailor UETP to test a software load
that more closely resembles your particular system's use.

The TOPS-18/TOPS-28 User Environment Test Package Reference Manual
describes UETP, the individual component tests, typical message
information, and the procedures for adding new tests.

CHAPTER
CREATING

4,1
One

STRUCTURES

OVERVIEW
of

the

first

installation
needs.

BSome

decisions

you

what

the

type

must

disk

considerations

make

about

storage
that

determine

How

large

will

the

data

base

be?

How

many

users

will

using

the

How

experienced

Will

there

How

often

will

that

the

system

Must

all

files

system

Are

multiple

mountable

for

making

these

full-time
you

run

remain
be

users

new

(or

best

your

upgraded)

suits

decision

your

are:

system?

be?

operator?

diagnostics

available

and

how

critical

during

this

maintenance?

all

users

CFS

configuration?

all

times

during

operation?

Chapter

The

will

your

environment

systems

12,

The

structure

part

Common

File

Refer

also

System.

facility

of TOPS-20 provides
several
options
The
option
you
choose depends on the
answers to the previous questions and the number
of
disk
packs
and
drives
that
are
available.
For example, if your installation has a
number of disk packs and two or more drives, you can
store
data
and
program files on different structures.
A
or

this

decision.

structure is a collection of data and
more disk packs and referenced under

program
one

files

contained

one

name.

When you install your software, you create a structure
known
as
the
public
structure
(sometimes called the system structure).
All packs

this

structure

remain

on-line

all

times

during

If
your
public
structure
does
not encompass all
drives you can create and mount other structures.
Sections
best

4.2

utilize

through
your

4.8

disk

describe

resources

the
and

public
create

system

structure
and

use

operation.

your

and

other

available

how

you

can

structures.

CREATING STRUCTURES

4,2

THE PUBLIC STRUCTURE

the public
Sections 4.2.1 and 4.2.2 provide an overviewto ofthe what
and its
system
onship
relati
its
ng
includi
is,
structure
contents.

4.2.1

What Is the Public Structure?

re on your system.
The public structure is the most important structuinstall
ation when you
system
at
on-line
brought
and
It is created
(Refer
dialog.
ation
install
the
answer the appropriate questions in
public
the
of
name
The
Guide.)
ation
Install
B
Model
KL
to the TOPS-2¢
structure can be up to six characters.

g on the
The public structure can be one or more disk packs, dependin
You may
es.
resourc
drive
disk
your
and
system
your
of
configuration
e; the
structur
public
the
as
use one or more RP@4, RPP6, or RPP7 disks

max imum

number of disks per structure is given in Table 4-4.

NOT use an RA6@, RA81, or RP2@ as the public structure.

You may

d files to
While installing the software, you copy the console a front-en
rted drive
dual-po
on
mounted
Iis
that
pack
e
the public structur
(usually drive #). The dual port allows the front-end @ processor and
processor

access

the data on the public structure.

the central
However, if you are using an RP@7 as the public structure,diskyouthatmust
is
reserve space for the front-end files on an RP@4 or Rflfl6
ing
contain
e
structur
disk
dual-ported to the PDP-11 front end. If the
the front-end files has multiple packs, the first pack of the
structure must be permanently mounted on the dual-ported drive.

All disk packs in the public structure must be online at all times,
because this structure contains all the programs, files, and swapping
area that the system needs to operate. The structure also contains
the
all user directories necessary to support users logging into
system. If the file system is destroyed on the public structure, or
if a drive that contains all or part of the structure malfunctions,
the
the system halts. Refer to Chapter 9 in this manual and tor must
operato
the
and
you
that
steps
the
TOPS-2@ Operator's Guide for
follow if you have problems with the file system or if a drive goes
down,

as
You can have another structure online that is capable of being used
This structure must have a unique name,
the public structure.
(Section 4.5.2 provides
however, at least while it is mounted.
the same name.) Section
having
es
structur
g
mountin
about
ion
informat
4.5.5 discusses why you would have such a structure.

If you are using CFS-20 software, refer to Chapter 12, The Common File

System, for additional information on the public structure.
The Contents of the Public Structure

4.2.2

The following list provides an overview of the contents of the

public

structure:

The default TOPS-20 command processor, EXEC, which is usually

the
A <ROOT-DIRECTORY> (Section 3.2.1) that points to
public
the
on
ies
director
vel
first-le
all
of
location on disk

found

in <SYSTEM>

or <NEW-SYSTEM>.

structure, including the special system directories,
4-2

CREATING

All

the
files
(Sections 3.2.2

The

1in
and

directories

<OPERATOR>,

STRUCTURES

the

directorie
3.2.4).
:

<NEW-SYSTEM>,

<SYSTEM-ERROR>

and

<NEW-SUBSYS>,

and

<SPOOL>

<ACCOUNTS>,

(Sections

3.2.6

and

3.2.7).

The front—-end monitor
(RSX2@F)
files
(Section
3.4).
This
<ROOT-DIRECTORY>.

structure,

RP@4
6.

The
the

RPP6

required
TOPS-20

uses up to
to Section
7.

HOME

are

using

front-end

file

system

dual-ported

disk

drive.

the

swapping area.
monitor you are

the

RP@7

must

The size of
using.
For

the

reside

public

either

this area depends on
example, 2060-MONMAX

15,000 pages of disk space for
swapping.
(Refer
4.7 for a description of the swapping area.)

block

that

contains

the

structure's

the

number

which

this

the

address

and

file

system

the

physical

disk

pack

you

and
the
console
front—end
normally
appears
in
the

the

structure

number of
(usually 950)

pages

used

address

and

pages

set

the

address

the

serial

number
the

number

<ROOT-DIRECTORY>

parameters:

name

packs

following

the

CPU

for

the

aside

for

and

its

booted

front-end

swapping

backup

copy

from

the

structure

A directory

for

every

Users
must
1log
use the system.,

different

4.3

user

who

structure

and

access

the

system.

the public structure
mount and connect to

to
a

directory.

ONE-STRUCTURE SYSTEMS

A one-structure system consists of
a
structure,
which
1is always on-line.

requires

into a directory on
Afterwards, they can

on-line

for

the

system

operate.

Usually, a one-structure system has
Smaller TOPS-20 installations choose

files

one

structure

the

simplest

the

easiest

There is no
drives, for

The

majority of

All

files

access.

for

single
structure,
the
public
All packs in the structure must

some

only
one
or
two
disk
drives.
to keep all their directories and

the

following

reasons:

system.
system

to maintain.

requirement to physically remove
example, for security reasons.

are

users

are

available

packs

from

the

inexperienced.
all

times,

and

thus

are

easy

STRUCTURES

CREATING

A full-time operator may be

There

is only one disk drive

unnecessary.

(only one structure supported).

Chapter 5 describes the methods you can use

directories

MOUNTABLE

4.4

your

one-structure

create

and

maintain

system.

STRUCTURES

If the public structure does not encompass all available disk drives,
These
you can create and mount other structures on the unused drives.
The
CHECKD program.
the
using
"mountable" structures are created
TOPS-20 Operator's Guide describes creating structures with CHECKD.
NOTE

The public structure
time.
installation

at
is the only structure created
All other structures are created

(using the CHECKD program)

system

4.4.1

and brought on-line

during

operation.

Differences Between Mountable and Public Structures

and
Unlike the public structure, a mountable structure can be mounted
Also, it need not contain a front-end
during timesharing.
dismounted
Therefore, a mountable structure does not have to reside
file system.
Although a mountable structure has its
a dual-ported disk drive.
on
own <ROOT-DIRECTORY> and directory system, a user cannot
log
into
a
the public
wuser on
a
in as
log
but must
structure,
mountable
structure.

directories.
and

public

4.4.2

A user can then mount a different structure
and connect to

Table 4-1 summarizes the differences between a mountable

structure.

Similarities Between Mountable and Public Structures

There are, however, many similarities between the public structure and
A
user directories and files.
contain
Both
structures.
mountable
mountable structure can have a front-end file system, and can be
used
in
place
of the public structure to load the system for timesharing.
A mountable structure is created with the
eight
special
directories
(mentioned
in
Chapter
3)
as
for
a public structure.,
Likewise, a
mountable structure has a HOME block that contains information such as
the
name
of
the
structure
and
the
number
of
disk units in the
structure.
These and other similarities are summarized
in Table 4-2.

CREATING

Table

4-1:

Differences

Public

Structure

Always

Has

timesharing

front—-end

RP@7)

Resides

file

drive

with

computer

(unless

that

the

Can

system

dual

front-end

into

the

system

Belongs

Has

the

Public

Need

Structures

mounted

not

Structures

have

and
a

dismounted

front-end

file

system
<SUBSYS>,

and

belong

for

logging

private

into

user

Need

not have these directories
unless
the
structure
will be

<OPERATOR>,

<SYSTEM=-ERROR>,
directories

reside
disk drive

Cannot be used
the system
Can

<SYSTEM>,

<ACCOUNTS>,

Need
not
dual-ported

RP@7)

logging

for

and

system

ported

Used

Mountable

during

(unless

Between

STRUCTURES

<SPOOL>

used

can

the

public

structure;

deleted

from

the

structure

Must

contain

swapping

area

Need

not

area

unless

mounted

contain

the

swapping

structure

the

public

structure

Table

4-2:

Public
Has

Has

Similarities

Between

Mountable

Structure
HOME
a

Mountable

block

front-end

(unless

RP@7)

load

used

the

files

area

system

Has

Can

have

Has

system

user

files

All

packs

must

the

system

4.5

files

<ROOT-DIRECTORY>

Contains

MULTIPLE-STRUCTURE

Structures

HOME

block

front—-end

have

for

been

May

contain

Has

files

area

the system
areas
and
established

system

files

<ROOT-DIRECTORY>

Contains

be
on-line
operate

Public

Can
be used to load
if the proper
file

files
Contains

and

user

All

packs

on-line

files
the

use

structure

the

must
structure

SYSTEMS

A multiple-structure system consists of a public structure
and
one
or more additional structures.
Figure 4-1 illustrates a system with
three disk drives and two structures.
The two-pack public structure

MASTER:

must

structure

ADMIN:

one-pack

structure

online
can

can

during

timesharing.

removed
mounted in

during
its

The one-pack mountable
timesharing.
Another

place.

CREATING STRUCTURES

2
UNIT

UNITO

N
e

STRUCTURE ADMIN:

STRUCTURE MASTER:

MR-S-3719-84

Figure 4-1:

System with 3 Disk Drives and 2 Structures

to remain
Using Figure 4-1, suppose you want structure ADMIN:
The structure is
system operation.
times during
on-line at all
automatically mounted if you turn on the drive that contains the
The TOPS-2@ Operator's
structure before the system is brought up.
'
Guide describes mounting structures automatically.
permanent
another
perhaps
In addition to the public structure and
you may choose to keep one or more disk drives
structure,
on-line
packs during
available for users to mount and dismount "private"
timesharing.

three
Figure 4-2 illustrates a system with three disk drives and
ADMIN:, and
structure MASTER:,
public
the
structures,
one-pack
PROG:
.

1
UNIT

STRUCTURE
MASTER:

STRUCTURE

ADMIN:

PROG:
MR-8-3720-84

Figure 4-2:

Three-Structure System

contains all the directories necessary

In this example, MASTER:

contains the same, a superset, or a subset
ADMIN:
support log-ins.
at
online
and remains
of the same directories as those on MASTER:
PROG:
contains
that
drive
The
operation.
system
during
times
all
structures.

one-pack
different
is used for short-term mounting of
PROG:
remains online only for the time it is needed.

There can be up to 32 structures online at one time.

Several of the advantages
e

in a mountable structure environment are:

Some users or groups

exclusively
for

the

use

users

for their use.

certain

may

structures.

4-6

require

structure

They can 'own' or possibly pay

CREATING

Service

engineers

short-term
disturbing

drive
normal

can

STRUCTURES

mount

own

pack

diagnostics

the

without

Creating structures on mountable packs provides
additional
security
to
that already within the system.
For example,
you
can
create
a
structure
that
contains
highly
confidential data, remove it from the drive(s) when you are
done with it, and lock it in a security
cabinet
or
safe.
At
the
end
of
the
day,
confidential structures.

their

and
perform
some
system operation.

the

operator

1locks

wup

any

In this type of environment, you are
not
1limited
in
the
size
of
vyour
system's data base.
You can create as many
structures as you have disk packs to contain them, and
you
can mount as many at one time as your system can support.

The principal disadvantages of using mountable
structures
are
the
need for scheduling both access to the data and operator coverage to
install and remove packs on the
drives,
as
described
1in
Section
1.2.2,
Mountable
Structure
Sign-Up
Log.
There is also some risk
that packs will be damaged during handling.

After the system is operating and structures have been created,
the
operator
responds
to
requests
from
users
to mount and dismount
structures,
Section 4.5.5 describes how to place
wuser
directories
on
your
mountable
structures
to
obtain
maximum availability to
priority jobs.

4.5.1

Choosing

Structure Names

Each device on the system has a name,
called
the
physical
device
name,
which
is
wused when giving commands to the software.
Unlike
the generic device name that applies to
a
<class
of
devices,
for
example:
TTY:,
DSK:,
LPT:, the physical device name applies to a
particular device on the system; for example, TTY6:
and LPT@#:.
The
physical
device
names
for disks are structure names.
A structure
name can be from one to six alphanumeric characters of your
choice,
and,
like
other
device
names,
must be followed by a colon.
The
colon indicates to the software that a device is being used and not,
for

example,

file,

It is important to carefully assign a unique name to each
structure
that you create.
Section 4.5.2, Mounting Structures Having the Same
Name, explains why this precaution avoids confusion for users if
an
operator is unavailable during timesharing.

Because structure names are used in the device
field
(dev:)
of
a
file
specification,
you
should not create any structures with the
same name as a defined (or valid)
device
name.
Table
4-3
1lists
device names that may be defined in your system.

For the same reason, avoid naming a structure with a defined logical
name,
for
example,
SYS:, SYSTEM:, NEW:, OLD:, HLP:, etc., because
the system searches the list of
defined
systemwide
1logical
names
before
device
names.
(Refer
to Section 3.3 for a description of
logical

names.)

Refer to Chapter 12, The Common File
considerations in a CFS environment.

System,

for

structure-naming

CREATING STRUCTURES

Table 4-3:

Sample Device Names

DSK:

CDP:

MTAn:

FEn:

MTn:

TTY:

LPT:

TTYn:

LPTn:

PTYn:

PLPTn:

NUL:

CDR:

PLT:

PCDRn:

PLTn:

PCDPn:

DCN:

TCP:

SRV:

Where n

is the

unit number of the device.

To avoid duplication, you can get a list of all structure names
known
to
the
system by giving the operator command, SHOW STATUS STRUCTURE.
These structures do not have to be
online.
Their
presence
in
the
listing
indicates
that
they
were
previously
specified
in
a SET
STRUCTURE command,

4,5.2

Mounting

once mounted

Structures

Having

the

system.

Same Name

A situation may arise requiring you to mount a structure that has
the
same
nhame
as
a
structure
that is already online.
Perhaps another
installation has requested that you mount its public structure
(named
PUBLIC:)
for testing purposes, but you already have a structure named
PUBLIC:
online.
Because
the
system
notices
ambiguous
structure
names, you must mount the structure under a different name.
Each structure that is mounted is
identified
with
two
names:
the
physical
identification
and the alias.
Usually, these names are the
same.
The
physical
identification
is the
actual
|structure
name
written
in
the
HOME block of that structure.,
The alias is the name
that you use to reference the structure while it is mounted.
After
a

structure
1is
mounted,
it
1is
known
only
by its alias.
command is used to mount a structure and give it
an
alias

The MOUNT
different

from
the physical identification.
This allows two or more structures
with the same physical name to be mounted simultaneously.
The
system
distinguishes
them
by
their
different
aliases.
(The
TOPS-20
Operator's

Guide

describes

this

procedure.)

Note that the structures must be mounted one
at
a
time,
That
1is,
structures cannot be online simultaneously before the MOUNT command is
given.
One of the
structues
must
be
mounted
first.
It
may
be
necessary to power down disk drives and bring them up again.

CREATING

4,5.3

Maximum

Size

STRUCTURES

Structures

The

maximum size of a structure
structure of this size requires

is
approximately
3 RP2f disk drives

Structures of the maximum size, however, may not
installation.
Smaller
structures
enhance
availability of the system.
Remember that
you
structures on-line at one time.

Also, if a structure is contained
packs
and
drive
units for that
that is, either all RP@4s, RP@6s,
example,

805,680

pages.
spindles).

be practical for your
the
reliability
and
can
have
up
to
32

on more
than
one
disk
pack,
the
structure must all be the same type,
RPO7s, RP2@s, RA60s, or RA8ls.
For

SUPPORTED

NOT SUPPORTED

RP0O4

STRUCTURE
ADMIN:

RP0O4

STRUCTURE
ADMIN:

MR-§-534-80

Table 4-4 shows the maximum structure size
wusing
RPO4,
RPO6,
RP2@,
RA6#,
and RA81 disk drives.
For all drive types, there
12,000 directories per structure and 5,000 files per directory.
NOTE

The number of directories per structure and files
per
directory that can be created is approximate.
This is
because the disk space needed to create a directory or
file
varies
according
to
the
1lengths of the names
chosen for directories and files.

Table

4-4:

Maximum

Size

Structures

Type of
Disk Drive

Max .No.Packs
Per Structure

No.
Per

Pages
Pack

RP @4

38000

RP@6

76000

RP@7

216376

RP20

201420

RA60

90516

RA81

200928

RPO7,
can be

CREATING

4.5.4

Increasing

the

Size

STRUCTURES

Structures

You can add more disk packs
to
increase
the
size of
structure
(not the public structure) during timesharing.
you

mountable
To do this,

must:

Dump the
DUMPER

entire

file structure onto magnetic: tape

wusing

Run

the

Run the CHECKD program, specifying the new cbnfiguration,

DLUSER

Restore all

re-create

using

the

program

the

program

structure

the directories

DUMPER

program

and

files

from

magnetic

tape

IMPORTANT

If possible, re-create the structure and
restore
the
files
to
a different set of packs from the structure
that you dumped.
This precaution ensures that you
do
not
lose
any
valuable data should you have problems
reading the tape back to disk.
That
1is,
you
still
have
the
original
structure
intact
and
can rerun
DUMPER and copy the structure to another tape.
To

increase

system

and

structure

the

size

with

Problems/Crashes,

4.5.5

Setting

and

the

disk

public

structure,

installation
packs.

you must

procedure

Refer

shut

down

for bringing

Chapter

the

up this

System

the TOPS-2@ KL Model B Installation Guide.

Structures

for

Maximum Availability

Before you create structures and place user directories on
them,
you
should
determine
which
wusers
must
be
on the system at all times.
Place these users' directories and files on the public
structure,
or
on
another
structure
that
1is
always available during timesharing.
Divide the remaining users of the system by priority, and place
their
directories
and
files on the other structures.
Although these users
have log-in directories on the public structure, their
large
working
area
where
they
<create
and
store
files
1is
on the other on-line
structures.
You may want to help users set
wup
their
LOGIN.CMD
and
BATCH.CMD
files
so
that
they
can
mount,
connect, and access the
appropriate directory on the structure where their files are
located,
if it is not the public structure.
Dividing users into categories and
placing them on structures accordingly ensures that the failure of one
disk
drive
does
not prevent the most important users from using the
system.
For example, if the drive that contains
ADMIN:
goes
down,
you can remove the ADMIN:
pack from the broken drive, and mount it on
another drive that contains a less critical structure.
Also, on-line disk diagnostics can be
performed
during
timesharing.
Sometimes, the service engineer can dismount a non-critical structure,
mount the maintenance pack,
and
perform
preventive
maintenance
or
trouble-shooting with only a portion of the user community off-line.

CREATING

increase
system
for backup

structure

structure,

you

STRUCTURES

availability,
you
can
create
using the CHECKD program.
After

should

the

procedures

another
public
you create this
your

software

installation
manual
for
creating
the
front-end
files
area.
The
TOPS—-2@ Operator's Guide describes using the CHECKD program to
create
a
backup
public
structure.
If
you have problems with the primary
public structure, having a second public
structure
available
allows
you to resume timesharing without reinstalling the system.
The

backup

public

under

another

backup

public

the

structure

name,
regular

structure

can
it

public

1is

kept

structure periodically with
incremental
dumper
tapes.

be mounted
«can

structure

and

kept
is

storage,

the System
(Chapter

online
storage

destroyed.

the

at
and

all

times

mounted

the

operator must

backup

update

the

Backup
Tape
and
the
latest
7,
System
Backup,
describes

creating and using your System Backup
Tape
and
incremental
tapes.)
Occasionally
wupdating your backup public structure (in storage) keeps
it reasonably up-to-date.
If

you

keep

your

backup

public

structure

online

all

times,

and

vyou

have
important
files
that
are
constantly
accessed
by
the
user
community, you can improve your system performance
by
placing
these
files
on the backup public structure.
Now your swapping area and the
files that you access frequently are
not
on
the
same
disk.
This
procedure
is
useful
with any structure that you keep on-line at all
times.
If

multiple

12,
The
files and

4,5.6

systems

are

part

Common
File
System,
user directories.

Taking

Structures

CFS

for

configuration,

further

refer

discussion

Chapter

placement

Off-Line

When a structure must be taken off-line, the
operator
should
notify
users that it will be dismounted at a certain time.
Users should give
the DISMOUNT command for the structure before the specified time.
If
the
wusers
do
not cooperate, the operator can dismount the structure
(via the DISMOUNT command to OPR) without leaving files in an
unknown
state.
Files
that are open simply become inaccessible, and the user
who had the files open receives an error.
For information on
to Chapter 12, The

dismounting
Common File

structures
System.

CFS

environment,

refer

CREATING

4,.5.7

Mounting

Structures

STRUCTURES

from Another

Installation.

If you mount a structure
from
another
installation,
or
perhaps
a
structure
that
contains confidential data, some of the user names on
this structure may match the user names on your public structure.
You

must

mount this structure

the mishap of your

them.

The same

installation.

the

in what is called a FOREIGN state,

users accessing directories that do not

is true

to avoid

belong

if you bring one of your structures to another

You should

have the operator at

structure FOREIGN and then mount it.

the

installation

SET

Figure 4-3 illustrates this

concept.

YOUR INSTALLATION

—

O
PUBLIC

STR:

DOMESTIC

MR-5-3721-84

Figure 4-3:

Domestic

and

Foreign Structures

A structure

is brought online

1in

one

two

states,

DOMESTIC

FOREIGN, according to the setting that the operator last specified for
this structure with the SET STRUCTURE command.
The system
uses
the
FOREIGN state as the default if a SET STRUCTURE command has never been
given for this structure.
The
structure
remains
1in
the
specified
state
until
the operator changes the state with the SET STRUCTURE or
the UNDEFINE command.
Note
that
the
setting
1is unchanged
across
system

crashes

and

reloads.

You
should
bring
a
structure
online
as
DOMESTIC
only
|if
the
directories
on
that
structure
were
<created for the same people as
those on the public structure.
One can be a subset of the other,
but
a
given
directory
name
should
represent
the same person on both.
Conversely, you should bring a structure on-line
as FOREIGN
if
the
directories
on
that
structure
were not necessarily created for the

same people as those on the public structure.
This is because a
user
who is logged into a directory on the public structure is the owner of
an identically named directory on a DOMESTIC structure, and
can
give
the
CONNECT
or
ACCESS
command
to
that directory without giving a
password (provided the directory protection allows this type of access
for
the
owner,
which
1is the usual case).
However, a user who logs
into the public structure and gives the CONNECT or ACCESS command to a
directory
with an identical name on a FOREIGN structure must give the
associated

password.

CREATING

4,6

SHARING

STRUCTURES

(DISK

STRUCTURES

DRIVES)

BETWEEN

TWO

SYSTEMS

If you have two DECSYSTEM-20's and one or more structures that contain
data
common
to
both
of
these systems, you may want to set up your
system

disk drives

alternately.

For

example,

you could

allow

System
A
to
use
the
drive
that contains structure ADMIN:
in the
morning and allow System B to use this structure on the same drive
in
the
afternoon.
THE SYSTEMS CANNOT, HOWEVER,
SAME TIME.
(Refer to Chapter 12, The
Common
exception.)
Also, if one of your systems goes
the drive that is connected to both systems.

ACCESS

File
down,

THE DRIVE AT THE
System,
for
the
you can still use

A drive that is to be shared by two systems must be supported by
both
systems.
For
example,
you
cannot
connect an RM@3 disk drive to a
DECSYSTEM-20200 and a DECSYSTEM-2@65 because the 2065 system
does
not
support
RM@3s.
Also,
the
shared
drive must be dual-ported.
Your
field service representative must
make
the
appropriate
connections
from
each
DECSYSTEM-20 to a port on the disk drive.
Be sure to have
the field service representative tell you which system is connected to
which port on the drive.
Figure 4-4 illustrates this connection.

—

Duat-Ported

DRIVE

DECSYSTEM-20

FRONT-END

FRONT-END
MR-5-536-80

Figure 4-4:

Shared

Disk Drive

The port switch on this drive must be in either the A or
B
position,
unless
the
systems
are
part
of
a
CFS configuration.
Otherwise,
TOPS-20 will not permit either system to access the disk while
it
|is
in the A/B position.
Error messages will be generated.
To use
the
drive,
place
the
port
switch
in
the
position
that
corresponds to the first system that is using the drive (A or B).
The
operator mounts a structure using the
normal
procedure.
After
the
first system is no longer allowed to use the drive, the operator gives
the

DISMOUNT

command

for

the

structure,

To use the drive on the second system, the operator leaves the pack on
the
drive
(if
vyou
are
using
the same structure), turns the drive

off-line, changes the port switch to
the
corresponding
system,
and
turns
the
drive
back
on.
Then, the operator (or a user) gives the
MOUNT command to mount the
structure
on
this
system.
The
system

automatically
structure.

recognizes that another drive

is on-line and mounts the

CREATING

4.7

DETERMINING

Sections 4.7.1
determine
the
your

system.,

4.7.1

What

SWAPPING

SPACE

STRUCTURES

THE

PUBLIC

STRUCTURE

and 4.7.2 describe what swapping space is
amount
of swapping space that you should

and
how
allocate

¢to
for

Swapping?

The
number
of
wuser
processes
that
can
fit
into
main
memory
simultaneously
depends
on
the size of the individual processes, the
size of the memory-resident portion of the monitor, and
the
size
of
memory
physically
available.
(Only a portion of the TOPS-2# monitor
resides in main memory at one time.) If a user wishes
to run a process
that is not currently in memory, process space must be provided.
This
may necessitate moving some other process out of memory.
The
wuser's
program or data that is transferred out of memory is placed on disk in
the

swapping area.
The system
sets
storage space on the public structure

On
to

aside
a
portion
of
specifically for this

some timesharing systems, a program must be
execute.
Swapping then consists of moving

disk

and

memory.

Under

TOPS-20,

only

portions

the
disk
purpose.

entirely in main memory
entire programs between
of

program

(those

containing
the instructions and data currently being referenced) need
be in memory.
Other portions of the program are brought
into
memory
from
disk
as
they
are
needed.
1In this case, swapping consists of
moving portions of a program or data between
disk
and
memory.
The
monitor decides which portions of which programs to swap, and when.,
The size of a program is measured in
a
unit
called a
page.
When
swapping
occurs,
some
of
these pages are copied between memory and
disk.
PFiqure 4-5 illustrates the concept of swapping.

CORE MEMORY

MONITOR

USER 1

\\[j:)

DISK

USER 2

PAGES OUT

USER 3

PAGES IN

USER 4

N———"

USER
5
L]

Figure

4-5:

4.7.2

When

Swapping

one

time.

Concept

Increase

For the most part,
cumulative size of

Table

MR-8-537--80

Swapping

Space

the size of your
swapping
processes you estimate will

4-5

contains guidelines

for

space
be on

depends
on
the system at

estimating

the

amount

the
any

swapping
space
required for an approximate number of user jobs based
on typical requirements.
This amount is
given
1in
response
to
the
question,
"HOW MANY PAGES FOR SWAPPING?" in the software installation
procedures.

CREATING

The

actual

disk

space

used

for

STRUCTURES

swapping

depends

the

number

pages

you
give,
The
system rounds the number of pages given upward to an
integral number of cylinders.
The swapping space is
divided
equally
among
You

the

can

disk

allow

structure.

packs
for

the

swapping

However,

this

public
space

structure.
structures

necessary only

structure as the public structure in the
space
avoids
re-creating the structure

the

public

other

than

you plan

the

future.
Allocating
should you decide to

public

mount

the

swapping
mount it

structure.

All the monitors are
pages for swapping.
The guidelines in
editing
Jobs and
production jobs.
and
production
double the size

designed to default to
In most cases, you can

an appropriate
number
take this default.

Table 4-5 apply to systems whose users perform
many
an average or small amount of debugging programs and
If your users perform a great
number
of
debugging

Jjobs
and
only
of your swapping

a small
space.

amount of editing, you should
However, if
vyou
double
the

size
of your swapping space, check the maximum swapping space allowed
for the monitor you are running.
(The TOPS-2¢ KL Model B Installation
Guide lists the maximum number of swapping pages you can use with each
monitor.) You cannot exceed this maximum.
If you enter a number
that
is
larger than the maximum, the monitor uses the maximum allowed.
If
you must exceed the maximum,
you
<can
bring
up
a
1larger
existing
monitor,
or you can tailor your monitor by following the instructions
in the BUILD.MEM file.
This file
1is
1located
1in
the
documentation
files saveset on the TOPS-2@ Software Distribution tape.

Table

4-5:

Determining

Estimated
Number of Jobs
20

Swapping

Space

Recommended Number of
Pages for Swapping*

less

3000

4500

490

6000

5¢

7500

* For each additional 10 jobs, increase
swapping by approximately 1,5040.

the

number

pages

for

If disk space
1is
available,
it
1is
better
to
overestimate
swapping
space
needed.
If
not
enough
swapping
space has
reserved, system service may be disrupted.

the
been

CREATING

4.8

DETERMINING

4.8.1
To

Determining

determine

among

THE AVAILABLE

your

the

Disk

Space

available

users

swapping space required
insert
the
number you
shown in Table 4-6, and
Table 4-6 outlines
public
structure.

other

structures,

space

for

any

structure,

SPACE

Before

Installation

space

that

installing

you

the

will

system,

have

first

divide

calculate

by
your
system
(Section 4.7.2).
calculated for swapping space into the
perform the appropriate steps.

the

Second,
formula

how to calculate the available disk
space
on
the
If you are calculating the available disk space on
follow this same procedure but
eliminate
reserving

directories

any possibility

DISK

disk

before

STRUCTURES

exists

reserve

the

that

areas

that

are

not

structure may

swapping

that

used

space.

structure.

the

NOTE

Remember that as
pages
in
the

increases.
directory

your

system expands,
the
number
of
<SYSTEM>
and
<SUBSYS>
directories

Also, the number
of
pages
reserved
<SPOOL>
should
be
increased if:
(1)

maintain

large

numbers

operator

files

backlog

user

file

up much

the

<SPOOL>

log

files

large

retrieval
area.

(2)

files

users

to
requests

gopy

LPT:.

can

for
you

large

A large
‘also
use

1If

public

CREATING

Table

4-6:

Calculating

STRUCTURES

Available

Disk

Space

Total Disk Space:
Number of RP04 disk drives *

38000 pages
per drive

76000 pages
per drive

216376 pages
per drive

201420 pages
per spindle

90516 pages
per drive

200928 pages
per drive

TOTAL DISK SPACE

OR:

Number of RP06 disk drives *

TOTAL DISK SPACE

OR:

Number of RP07 disk drives *

TOTAL DISK SPACE

OR:

Number of RP20 disk drives *

TOTAL DISK SPACE

OR:

Number of RA60 disk drives *

TOTAL DISK SPACE

OR:

Number of RA81 disk drives *

TOTAL DISK SPACE

Reserved Disk Space:
Front-end file system

Swapping Space

950

(Enter number of pages
selected and allocated
for swapping)
<SYSTEM>

1876

1780

<NEW-SYSTEM>

<NEW-SUBSYS>

500

<UETP.*>

1701

<ROOT-DIRECTORY>

1000

TOTAL RESERVED SPACE

SUBTRACT
TOTAL RESERVED SPACE
FROM TOTAL DISK SPACE

AVAILABLE DISK SPACE

4.8.2

Determining

Disk

Space

After

Installation

Shortly after you have
installed
the
system,
you
can
1log
1in
as
OPERATOR
and
give
the
command INFORMATION (ABOUT) DISK-USAGE.
One
line of output tells you the actual number of system
pages
that
are
available
on
the public structure.
You can divide this disk storage
amohg
your
users.
(Refer
to
Chapter
5,
Creating
Directories.)
However,
be
careful
about
over—-allocating disk space on the public
structure.

free
up

the

entire

the

space

public

allowed

structure

public

after

structure.

user

directories

installation,
But

exceeds

then

TOPS-200

users

cannot

the

space

can
free

fill
up

adequate
space by
may be disrupted.

expunging the
Even if space

public
can be

structure, then system service
freed to allow the
system
to

keep

user

may

disrupted.

running,

some

programs

CHAPTER
CREATING

prospective

user

who

requires

DIRECTORIES

access

the

system must

assigned

a user name (normally the user's surname), a password, an account, and
disk storage quotas, and must have a directory created for him or
her
on
the
public
structure.
Optionally,
you
can
assign
certain
capabilities and/or make the user or the user's directory a member
of
onhe or more groups.
(Refer to Section 5.8 for a description of how to
establish group relationships
among
users
and
Section
5.9
for
a
description of the capabilities you can assign to users.) You can also
create additional directories for users on mountable structures.
This chapter describes
three
methods
you
maintain
directories.
Using
one
method,
maintains all the directories on the system.

can
use
to
<create
and
the operator creates and
A second
method
allows

you
to
delegate
the
responsibility
for
<creating
and maintaining
directories to project administrators.
The third method combines
the
first
two
methods,
thus providing additional flexibility.
Sections
5.1 through 5.3 explain these methods and the determining factors
for
choosing
one
of
them.
These
sections
also
include
some of the
decisions necessary to assign user names and capabilities, and how
to
allocate
disk
storage
according to the method you use to create and
maintain directories.
(Refer to Chapter 4 to determine the amount
of
disk space available to divide among the directories you create.)
Chapter

assign

and

allow

describes
users

to
valid
directories.

Refer

creating

5.1

how

wvalidate
to

log

into

accounts

Chapter

12,

directories

HAVING THE
CONTROL)

set

accounts.
the

You
system

immediately

The
in

Common File
a

CFS-20

accounting

should
and

read

charge

after

System,

scheme

and

Chapter 6

how

you

want

their

computer

usage

have

created

their

you

for

<considerations

1in

environment.

OPERATOR CREATE AND MAINTAIN ALL

DIRECTORIES

(CENTRAL

In this type of installation, the operator creates a directory on
the
public
structure
for
each
new
user, and specifies the appropriate
parameters.
The name of the directory is the
same
as
the
assigned
user
name.
Each
user
informs
the
operator
when a change to his
directory parameters is required.
This type of operation
allows
you
as
system
manager
to
have
central administrative control over all
directories and parameters.

CREATING

DIRECTORIES

and
create
Therefore, central control means that you or the operator
Central
your system users.
all
for
the directories
all
maintain

to
One scheme allows you
control has two types of directory schemes.
The other
create up to approximately 5,000 directories per structure.
scheme allows you to use subdirectories and create up to approximately
12,000

directories

per

structure.

NOTE

The number of directories

allowed

structure

per

approximate, because the disk space needed
directory or file varies.

5.2

to create

is
a

DELEGATING THE CREATION AND MAINTENANCE OF DIRECTORIES TO PROJECT

ADMINISTRATORS

(PROJECT CONTROL)

administrative
project
involves
An alternative type of installation
control.
Under this type of control, the operator creates directories
only for the users who have been designated as project administrators,
(e.g.,
the
representatives
of major
departments).
The
project
administrators, in turn, create subdirectories for users within
their
departments
or
projects
and
control the assignment of those users'
directory parameters.
This type of control allows you to delegate the
responsibility
for
creating
and
maintaining
directories for other
users and still maintain ultimate control over
your
system
and
its
resources,

Therefore, project control means that most of the directories
created
by
you
or
the operator are project directories (e.g., MATH might be
the assigned name of a project).
The system's resources, such as disk
space,
are
divided among these project directories either equally or
according to the expected size of
the
project.
Subdirectories
are
then
created
under project directories for users within the project.
The people who have been appointed project leaders

are responsible for creating,
the subdirectories within

administrators

assigning parameters to, and maintaining

their

project.

The

resources

that

vyou

allocate to the project directory are divided among its subdirectories

by the project administrators.

create

5.3

COMBINING

subdirectories)

per

Under

approximately

structure.

project control,

12,008

you are allowed

directories

(including

CENTRAL AND PROJECT CONTROL

A combination of central and project control can be used if
you
want
to
keep
the majority of the user directories at the management level
of control and separate only a portion of your
system
into
projects
and

administrative

control.

Therefore, combining
central
and
project
control
means
that
the
operator
creates
and
maintains
directories
for most of the system
users and creates
project
directories
for
special
projects.
The

project
administrators
create
directories
directories and are responsible for maintaining
two
types
of
control
still
allows
up
to
directories

per

structure.

under
the
project
them,
Combining
the
approximately
12,000

CREATING

5.4

CENTRAL AND

Sections

5,4.1

PROJECT CONTROL
through

control,

project

control.

Each

5.4.4

control,

description

DIRECTORIES

DESCRIPTIONS
describe

and

the

The

determining

The

format,

The

procedure

for

assigning

The

procedure

for

creating

user

The

procedure

for

creating

files-only

The

restrictions,

factors

including

for

any,

apply

5.4.1

description

contains

all

the

Central

DETERMINING

directory

types

central

and

project

general

particular

each

type

control

names
directories

apply

that

thoroughly.

very

user

that

Also, any additional considerations
description are included.
each

choosing

diagram,

control

description

two

includes:

Read

the

combination

apply

The

directories
using

headings

first

information

particular

within

central

and

each

control

suggestions

schemes.,

that

Control

FACTORS:

Your

business

therefore,

installation

relatively

uncompl icated;

there
1is no need to separate projects and assign
the
creation
and
control
of
directories
to
various
administrators.
All
the
directories
on
the
system
are

created

and

You

sure

than

are

maintained
that

approximately

the

you

number

or
of

the

system

directories

5,000.

operator.
you

need

1less

FORMAT:
<ROOT-DIRECTORY>

can

structure.

All

directories

under

point

approximately

<ROOT-DIRECTORY>

are

5,008

one

directories

level.

per

CREATING DIRECTORIES

DIAGRAM:

CENTRAL CONTROL (5,000 DIRECTORIES)

< ROOT-DIRECTORY >

< PORADA >

< FORTRAN-LIB>|

|<MCELMOYLE>

< HOLLAND >

—»- UP TO APPROXIMATELY 5,0 0 DIRECTORIES PER STRUCTURE

>
< T-SOURCES

MR-5-3722-84

USER NAMES:

ASSIGNING

The user name that you assign should include the user's last name.
This convention is true for any type of directory scheme that you use.
The system uses this name when recording the authors of files, sending

mail

users,

and

displaying

system

status.

If you follow this

convention, you can easily identify who is using the system

when

you

give a SYSTAT command. If just using last names will not result in
duplications, you can simply use the last names. Otherwise, you might

want to wuse the first and middle initials followed by the last name.
(If a user has no middle initial, you can use a dash in its place.) It
is most convenient for users when user names begin with unique
or
characters, allowing use of ‘“"recognition" when typing user
directory names. Use of leading initials often yields this result.
USER DIRECTORIES:

CREATING

All directories are created using the "ECREATE command.

who

have

(Only

users

WHEEL or OPERATOR capability enabled can use this command.)

In the next example, the operator is connected to the public structure

PUBLIC: and uses the “ECREATE command to create a new directory named
<BECKER> for the user who has been assigned the user name BECKER.

Also, the operator assigns the password MARTIN.
@ENABLE

(CAPABILITIES)<RET>

$"ECREATE

(NAME) PUBLIC:<BECKER><RET>

[NEW]

$SPASSWORD MARTINKRET>
$S<RET>
SDISABLE (CAPABILITIES)<RET>
a

This directory is called the user's logged-in directory and is

always

on the public structure. Whenever the user logs into the system, he
is connected to this directory. He can remain in this directory or
connect to and use files in another directory.

Refer to the TOPS-2@¢ Operator Command Language Reference Manual for

complete description of the ~ECREATE command that the operator uses to

create new directories, and the ULIST program that prints
about all

the directories on the system.

information

CREATING

After
to

creating

update

the

DLUSER,

DLUSER

Chapter

new

tape

directory

(either

containing

data,

System

the

DUMPER,

Backup

DIRECTORIES

files-only

monitor,

<SYSTEM>,

TOPS-2¢
and

user),
Command

<SUBSYS>.

Procedures.)

remember
Processor,
(Refer

CONSIDERATIONS:

two

try

users

have mistakenly

«create

the

second

been

prints

[OLD]
instead of [NEW].
a slightly different user
the new directory name.
A

user
with

assigned

directory

names

with

the

same

this

user

name

duplication,

and

the

you

system

Give

the ABORT subcommand, assign
the
name, and reissue the “ECREATE command
common practice
is
to
precede
such

with
the user's first initial.
This allows recognition on the
user or
directory
name
without
typing
the
entire
name
and
the
distinguishing
character.
For
instance,
if you have the two users

Stephanie

names

the

Sheldon

S-SHELDON

names

CREATING

and

Andrew

SHELDON~S

and

FILES-ONLY

structure

another

you

SSHELDON

should
and

assign

ASHELDON

them

the

rather

SHELDON-A.

user

than

use

DIRECTORIES:

If a user wants
directory,
you
or

Sheldon,

A-SHELDON

have a library area in addition
to
can
create
a
files-only
directory
structure.

to
this
directory
associated with the

giving
directory.

For

have

The

the

user

can

CONNECT

gain

command

his
on

logged-in
the public

owner

privileges

and

the

password
the
directory
is
located
on
a
regulated
mountable
structure,
the
user
must
also give the MOUNT
command to use the structure before he gives the CONNECT command.
The
user
cannot
give
the
ACCESS
command
for or log into a files-only
directory.

example,
regular

directory

you

basis,

and

keep

the

user

named

may

want

payroll

data

To
accomplish
this,
you
can
logged-in directory <BECKER> and

The

directory

can

ADMIN:<PAYROLL>).

BECKER

protection

restrictive

CONNECT

give

the

ACCESS

The next example
public structure
@ENABLE

command

now

the

protection

directory

password.

for

log

into

shows how
MAIN:.

or
to

create

the

public

on
his

directory.

structure

the
directory
<PAYROLL>.
other
structure, (e.g.,

some

its

payroll

programs

restricted

the
files-only

normal

the

processes

payroll

more
on

giving

(NAME)

his

<PAYROLL>
BECKER

directory,

and

cannot,

can

still

however,

<PAYROLL>.
directory

the

MAIN:<PAYROLL><RET>

[NEW]
$SPASSWORD MONIES<RET>
$SFILES
(ONLY)<RET>
$SPROTECTION
(OF DIRECTORY)
$$DEFAULT
(FILE) PROTECTION
$S<KRET>

who

the

(CAPABILITIES)<XRET>

$"ECREATE

SDISABLE

create

has

BECKER

develop

77400@<RET>
77020@<RET>

(CAPABILITIES)<RET>

Now

if user BECKER logs in and wants to
can give the following CONNECT command:

@CONNECT (TO DIRECTORY)
Password:
monies<RET>

use

the

files

<PAYROLL>,

CREATING DIRECTORIES

The TOPS-20 Operator Command Language Reference Manual ‘describes

all

the parameters you can give to directories and describes how to create

directories on mountable

structures.

CONSIDERATIONS:

When you create additional directories on mountable structures,
Some users will not
if files-only directories are suitable.
consider
time they require
each
directory
a
to
command
CONNECT
the
give
to
want
Also, files-only
in that directory.
owner access to the files
members and
group
directory
as
only
groups
of
members
are
directories
not user group members. Therefore, if you create ALL the directories
user
on a structure as files-only, you cannot establish any valid
to Section 5.8
(Refer
group relationships among those directories.
for a description of setting

up groups.)

Conversely, if you create user directories, users can give the ACCESS
and group
directory and gain owner
additional
their
to
command
other
use
can
they
and
directory,
the
to
connecting
without
privileges
Also, if the name of
the structure as group members.
directories on

1is the same as the wuser's logged-in
the directory you create
directory, and the structure is mounted as DOMESTIC, the user does not
have to specify a password when giving the CONNECT or ACCESS command
(Refer to Section 4.5.7.)
to the directory.
No password
user is submitting a batch job.
In
input; therefore, security is preserved.
directories on the structure, you can create
be

used

This is valuable when the
is required on the batch
addition to creating user
files-only directories to

library areas.

It is also possible to create only one user directory on

structure

In this case, all
to create all other directories as files-only.
and
structure could
the
users required to use a files-only directory on
give the ACCESS command to the user directory (gaining owner and group
to
privileges), and use the files in a files-only directory according
This would be useful if you have a
the group protection codes set.
private structure that contains several library areas that are

common

Each owner could give the
to the owners of the private disk pack(s).
ACCESS command for the one user directory and gain group privileges to
Therefore, these users would need only
all the 1library directories.
one password to gain access to all the information on the pack.
Note that defined groups may provide

better

security

controls

than

for the ADMIN:<PAYROLL> directory is
If the password
passwords.
be guessed, or user BECKER may write it down or tell
MONIES, it might
On the other hand, group membership can be
some other user.
it to

centrally controlled, and the access can be withdrawn, 1if necessary.
the directory structure provides a record of group memberships,
Also,
Wise use of directory
which can be displayed with the ULIST program.
protections can allow user members of a group to connect to a
files-only directory without giving a password.

Refer to the TOPS-20 User's Guide or the TOPS-2¢ Commands Reference
Manual for a complete description of the CONNECT and ACCESS commands.
RESTRICTIONS:

structure cannot exceed
The number of directories you create per
n because the disk
approximatio
an
is
number
This
5,80@.
y
approximatel
space that it takes to create a directory or file varies.

CREATING

5.4.2

Central

DETERMINING
®

Control

Using

DIRECTORIES

Subdirectories

FACTORS:

stated

does

not

However,

the

warrant

this

structure

directory

than

directory

segregation

the

scheme

the
operator
can
create
directories
per
structure
directory parameters.
®

You

can

adding

easily

project

management

(Refer

expand

level

into

over

the

Section

Control.)

central

control

installation

and

control.

directories

scheme.

approximately
assign and maintain

and

form

and

still

majority

5.4.4,

your

projects

allows

directories,

scheme,

project

maintain
the

Combined

per

You

control

user

12,000
all the

the

directories.

Central

and

Project

FORMAT:
<ROOT-DIRECTORY>

points

directories.

The

name

each

directory

is a letter of the alphabet, <A>, , <C>, <ee, <Z>,
The directories
point to all
wuser
and
files-only
directories.
The
single-letter
directories
are
on
one
level below <ROOT-DIRECTORY>.
The user and
files-only directories are on a second
level
below
<ROOT-DIRECTORY>
and

are

pointed

the

alphabetic

directories.

The

diagram below illustrates this flow.
Also,
although
it
1is
not
shown
1in the diagram, the directories pointed to by the single letter
directories, (e.g., <A.JONES>), can be allowed to
create
directories

under

them,

Perhaps

subdirectories
responsible
use

only

allow

for

the

directory.

store special
maintaining the

disk

Refer

some

user

quota

users

you

Section
to

create

JONES

wants

files,

create one or

such as
memos.
directory he created and

originally

5.4.3,

allocated

Project

directories

Control,

their

The
is
his

would

own.

CENTRAL CONTROL USING SUBDIRECTORIES

< ROOT-DIRECTORY >

_———

VAN

[

<A.JONES> |

[<<B,PARKER>|

| <C.BAKER>

< D.LAWES >

<E>

<D.HALL
>

<Z>

< Z.FORLIB >

|is
to

logged=-in

you

DIAGRAM:

<D>

two

user
allowed

< Z,TESTS
>

UP TO APPROXIMATELY 5,000 DIRECTORIES PER SINGLE LETTER DIRECTORY
UP TO APPROXIMATELY 12,000 DIRECTORIES PER STRUCTURE

MR-S-3723-84

CREATING DIRECTORIES

ASSIGNING USER NAMES:

possible to the
Each user name that you assign should be as close as add
a period. For
user's last name prefixed by a first initial
Under
C.BAKER.
name
user
the
assigned
be
would
example, Charles Baker
this type of directory scheme, you must follow the principle of
prefixing the name with the first initial and a period.
CREATING USER DIRECTORIES:

Have the operator create 26 directories using the “ECREATE command.
The name of each directory is a letter of the alphabet, that is, <A>
through

<zZ>,

The theory behind creating these alphabetic directories is the same as
described in Section 5.4.3. That is, you must create directories that
are allowed to have subdirectories. The directories <A>, , .oy

<Z> can have approximately 5,000 user and files-only directories rsunder
in
them. Therefore, you must include some of the same paramete
these directories, as you would in project directories.

Refer to the TOPS-2¢ Operator Command Language Reference Manual for a
complete description of the parameters that are defined when the
operator uses the “ECREATE command to create directories, and the

program

ULIST

that

system.

information about all directories on the

prints

The procedures for creating the alphabetic directories and the user
and files-only directories under them are described below. In the
1is the name of the public structure,

example, COMMON:

NOTE

The general considerations described in Section 5.4.1
for creating directories are also applicable to this
directory description.

users,
However, do not assign

Even though the alphabetic directories are not associated with

they

must

passwords.

directories.

@ENABLE

created as log-in directories.

This

prevents

users

from

gaining

access

these

(CAPABILITIES)<RET>

$"ECREATE

(NAME) COMMON:<A><RET>

[NEW]

Assign each directory a large number for creating subdirectories,
example,

for

400.

$$MAXIMUM SUBDIRECTORIES

(ALLOWED) 400<RET>

Because many user directories are created under each of these
alphabetic directories and the page quota (disk space) from these
the user
alphabetic directories is divided among (or passed on to)
large
very
a
directories
alphabetic
the
assign
must
directories, you
permanent and working page quota. Assigning them a sufficiently large
page quota prevents any of these alphabetic directories from exceeding

g to make a change to the quota at
you
their page quota, thus requirin
directory at least 500,000 pages
each
assign
Therefore,
a later time.

of permanent and working disk page quota.

$SPERMANENT (DISK STORAGE PAGE LIMIT) 590808@@8<RET>
$SWORKING (DISK STORAGE PAGE LIMIT) 5@00P00<RET>
5-8

CREATING

Assign a list of
The
list should
you use
groups,

DIRECTORIES

SUBDIRECTORY-USER-GROUP numbers
be the same for each directory.

depends on how many groups
10
groups,
40
groups)
.

examples

are

for

illustration;

you

The numbers
choose any

can

$$ SUBDIRECTORY-USER-GROUP (ALLOWED)
$$ SUBDIRECTORY ~-USER-GROUP (ALLOWED)
$$ SUBDIRECTORY~-USER-GROUP (ALLOWED)
$$ SUBDIRECTORY-USER~-GROUP (ALLOWED)
$S$ SUBDIRECTORY-USER-GROUP (ALLOWED)
Later,

when

group,

Yyou

you

create

must

user

enter

one

plan

The

the

following

@ ENABLE

COMMON:

example,

and

(CAPABILITIES)

$ “ECREATE (NAME)

the

creates

204<RET>

operator

first

that
this

<RET>

COMMON: <A><RET>

COMMON: <RET>

$$MAX 40¢ !You can use abbreviations<RET>
SSWORKING 500008 <RET>
$$ PERMANENT 500000<RET>
$$SUB 20B<RET>
$$SUB 201<RET>
$$ SUB 202<RET>
$$SUB 2@3<RET>

$$SUB 204<RET>
$$ <RET>
$ !Continue creating directories <C>
in

exactly the

same manner.

user in
list.
and

Section

5.8

conhected

two

[NEW]

lthrough <2Z>

following

sequence:

202<RET>
20 3<RET>

[NEW]

numbers

200<RET>

$$ MAXIMUM-SUBDIRECTORIES (ALLOWED) 400<RET>
SSWORKING 500000 <RET>
$$ PERMANENT 5000@@<RET>
$$ SUBDIRECTORY-USER~-GROUP (ALLOWED) 200<RET>
$$ SUBDIRECTORY~-USER=-GROUP (ALLOWED) 201<RET>
$S SUBDIRECTORY-USER~GROUP (ALLOWED) 202<RET>
$$ SUBDIRECTORY-USER-GROUP (ALLOWED) 203<RET>
$S SUBDIRECTORY-USER~-GROUP (ALLOWED) 204<RET>
§$<RET>
$ "ECREATE (NAME)

201<RET>

directory and place
of
the numbers in

the

directory.

range

establish,
used in the

(Refer to Section 5.4.3,
numbers will be valid.
description of why you are using this list of numbers,
for establishing valid group relationships.)

structure

each

directories,

<A>

and

:

CREATING

After all 26
command
to

DIRECTORIES

directories are
created,
you
can
give
the
DIRECTORY
see
all the directory files that have been created under

<ROOT-DIRECTORY>.

$DIR

COMMON:<ROOT-DIRECTORY>

COMMON: <ROOT-DIRECTORY>
A.DIRECTORY.1

B.DIRECTORY.1

C.DIRECTORY.1

Z.DIRECTORY.1

Next,

after

created,
the user
In

the

example

structure

user

all

alphabetic

the operator
directories.

who

below,

COMMON:

has

again

been

the

and

directories

uses

the

operator

creates

assigned

the

user

have

“ECREATE

1is

connected

directory

name

been

command

named

A.JONES.

successfully

and

creates

the

<A.JONES>

This

all

public
for

the

directory

A,
Jones'
log-in
directory
on the public structure.
Each time A,
Jones logs into the system, he is connected
to
directory
<A.JONES>.
In
this
example,
the
operator also assigns the password 2BY4.
The
operator gives the directory the system default of 250 pages for
both
working and permanent disk quota.
Because he is using the default, he
does

not

pages

have

given

make

this

any

entries

directory

for

are

these

taken

two

from

parameters.

the

superior

The

250

directory's

quota (directory <A>).
(The PRESERVE
subcommand
can
be
issued
to
avoid
having
the
disk
space
quota
subtracted
from
the superior
directory's allocation.) The operator also places user A.JONES in user
group

202

@ENABLE

and

places

directory

<A.JONES>

directory group

202.

(CAPABILITIES)<RET>

$"ECREATE

(NAME)

COMMON:<A.JONES><RET>

[NEW]
SSPASSWORD

2BY4<RET>
SSUSER-OF-GROUP (NUMBER) 202<RET>
$SDIRECTORY-GROUP (NUMBER) 202<RET>
S$<RET>
SDISABLE

(CAPABILITIES)<RET>

@
After

creating

should
Command

any

new directories

update
the
Processor,

<SUBSYS>.

(Refer

(either

files-only

backup
DLUSER,

tape
that
<contains the
DLUSER
data,
DUMPER,

Chapter

System

Backup

user),

you

monitor, TOPS-20
<SYSTEM>,
and

Procedures.,)

CREATING

DIRECTORIES

CONSIDERATIONS:

users

have

duplicate

middle

initials.
assign either

can

C.LBAKER.
<CL>

For
one

you

addition

first

initials

wuse

the

directory

<C>.

the
and

name

method),

C,LBAKER

(the

not

create

need

could

create

the

levels

directories

followed

their

last

above

you

using

additional

third-level
by

names,

you

not

can

use

C.BAKER,
CL.BAKER

you
or

a directory
create

this

wuser's first and middle initial, you
will
not
be
able
to
create
the
instead,
you assign user Baker the user

preferred

described

two

last

form CL.BAKER you must create

additional
directory
with
will receive error messages
directory
<CL.BAKER>,
If,
<C.LBAKER>

and

example, if two users have the name
of them a user name
in
the
form

on
for

create

standard

directory

"initial"

based
names,

can

the

<CL>.

directories,

the

directory

procedure.

You do
Alternatively,
you

and

their
first
and
example, C.L.BAKER.

assign

middle

users

initials

If
a
wuser
requires
special
capabilities
to
perform
privileged
functions,
the
operator can include the parameter for the capability
in the user's directory accordingly.
(Refer
to
Section
5.9
for
a
description
of
the
capabilities you can assign to certain users who
require them.)
CREATING

you

FILES-ONLY

user
can

wants

create

a
a

DIRECTORIES:

library

area

files-only

addition

the

Files-only directories

can

also

prefixed

Because

by a

initials

users

not

place

the

first

name

logged-in

directory,

directory.

letter
always

and

period.

encompass

every

letter in the alphabet, you may want to use
those
infrequently
used
letters
as the prefix to the files-only directories, e.g., <X.FORLIB>

<Z.TESTS>.

under

root

Alternatively,

special

prefix,

directory.

<X.FORLIB>

The
the

password

SQUASH,

default

for

directory
can

group

now

such

example

public

makes

working
this

below shows

permanent

how

and

files-only

them

282.
the

the

operator
takes

assigns

quotas,

and

(User

members

files

contains

under

the

directory

the

storage

protections.)

@ENABLE

directories

directly

create

The

files-only,

disk

number

directory

place

PUBLIC:.

directory

directory group

accesss

can

LIB.,,

structure

the

and

you

the

25@-page

places

the

groups

202

according

(CAPABILITIES)<RET>

$"ECREATE (NAME) PUBLIC:<X.FORLIB><RET>
[NEW]
SSPASSWORD SQUASHXRET>
SSFILES-ONLY<RET>

S$SDIRECTORY~-GROUP

SS<RET>
SDISABLE

(NUMBER)

2@2<RET>

(CAPABILITIES)<RET>

Follow the procedures in
directories on mountable

the TOPS-2@
structures.

Operator's

Guide

for

creating

CREATING DIRECTORIES

CONSIDERATIONS:

control
central
previous
The CONSIDERATIONS described in the
(Section 5.4.1) for files-only directories also apply to
description
'

this description.

If the number of files-only directories you want to create is small,
you can create them on the same level as the alphabetic directories.
That is, <X.FORLIB> can be created as <FORLIB>.

may look

Your directory scheme

like:

< ROOT-DIRECTORY >/

<A>

<J>

<X>

< A, JONES >

< X.FORLIB
>

OR:

< ROOT-DIRECTORY >

<A>

<J>

<Z>

< FORLIB>

< A.JONES >

MR-5-541-80

RESTRICTIONS:

The number of directories you can create per structure

approximately

The number of subdirectories
example,

<A>,

cannot

exceed

12,000.

under

single-letter directory,

cannot exceed approximately 5,000.

for

CREATING

the

you

reach

system

MAXIMUM

the

maximum

prints

number

the message:

DIRECTORY

NUMBER

DIRECTORIES

EXCEEDED;

If you reach the maximum number of
directory can point to, the system
SUPERIOR

You can
because

groups
a

DIRECTORY

define up
all
the

problem

directories

allowed

INDEX

NEEDS

TABLE

subdirectories

prints

the

per

structure,

EXPANDING

that

single

letter

message:

FULL

to only 40 user groups with
the
“ECREATE
command,
wuser
groups must be specified as subdirectory user

the superior
when
all

single-letter directory.
This is not
the
wuser
directories
are
directly

as
great
under the

ROOT-DIRECTORY.

If you make a superior directory FILES-ONLY, be sure to make
subdirectories
FILES-ONLY.
Otherwise, you will be unable to
the subdirectories (refer to Chapter 9) if the structure
is
until all the superior directories are made not FILES-ONLY.

5.4.3

Project

DETERMINING

The

Control

FACTORS:

complexity

warrants

and

perhaps

separating

geography

small

1large

vyour

groups

projects.
The responsibility for
creating
the
directories
within
a
project
can
administrator.

This

directories

wusers

have

especially
on

Even though
groups
of

you delegate
directories

helpful

the

the operator
from
spending
an
creating directories and changing
e

all
its
recreate
damaged,

organization
of

users

into

and
maintaining
be
given
to
an
if

system.

large

This

excessive
directory

number

method

frees

amount
of
parameters.

time

the task of
creating
and
to project administrators,

managing
you still

maintain ultimate control
of
the
overall
system
and
its
resources,
This means that you still determine and allocate
the disk space that each
project
uses.
The
administrator
distributes the disk space you allocate to directories within
the project.
Also, administrators can
create
and
maintain
directories
for
their
projects
without
having
WHEEL
or
OPERATOR capabilities by using
the
TOPS-2¢
BUILD
command.
Therefore,
you
do
not
weaken the security of your system.
Unless

you

give

administrator,

WHEEL

cannot

OPERATOR

assign

those

capabilities

other

users.

addition

for

allowing

project,

create

you

administrators

can

subdirectories.

files

These

users

place

12,000

created

their

per

directories

structure.

(including

create

users

can

subdirectory.
According to
subdirectories, they
can
with
other
wusers
without
1losing
the
superior
directory.
The
users
are
maintaining the directories they create.
e

allow other

directories

the

separate

system
and

store

the protection they
share
their
files
security
of
their
responsible
for

subdirectories)

<can

CREATING DIRECTORIES

FORMAT:

<ROOT-DIRECTORY> can point

5,008

approximately

directories

per

structure.

Each directory under <ROOT-DIRECTORY> can point to approximately 5,008
subdirectories.

Each subdirectory can also point to approximately 5,800 subdirectories
it.

directly under

The number of subdirectory levels is determined by a maximum length of

alphanumeric

characters,

because each subdirectory name contains

the
Using
1it,
the name or names of any superior directories above
under
who owns directory <PHYSICS>
user
the
below,
diagram
The new
<ROOT-DIRECTORY> creates the subdirectory <PHYSICS.LAB-12>.
name (LAB-12) has its superior directory's name (PHYSICS)
subdirectory
The period separates the different levels of the
as its prefix.
is counted as one of the characters in the
directory name and
directory name.

DIAGRAM:

PROJECT CONTROL (12,000 DIRECTORIES)

< ROOT-DIRECTORY >

|
< CHEM.

HESS >

< CHEM,

STUDENT >

< COMP,SCI>] e

< MILLER >

< PHYSICS >

< CHEM >

eeo | <HURLEY>

< CHEM.

STUDENT >

<PHYSICS. | | <PHYSICS.

<PHYSICS.

STUDENT >

TESTS >

ADMIN >

|
LAB-12>

< MILLER.

| 444

>
STUDENT

.. | <cowmp-sci. |

STUDENT >

ses|

< PHYSICS.
LAB-12,

STUDENT >

< PHYSICS.
LAB-12.

STUDENT >

< PHYSICS.
LAB—12,

>
STUDENT

|<comp-scl.

STUDENT >

|***

UP TO APPROXIMATELY 5,000 SUBDIRECTORIES PER DIRECTORY
UP TO APPROXIMATELY 12,000 DIRECTORIES PER STRUCTURE
MR-5-3724-84

|,,,

CREATING

ASSIGNING

USER

DIRECTORIES

NAMES:

The names that you assign to users should be as close as
possible
to
the
wuser's last name.
1In addition, the project names that you assign
and that will be used for project directory names
should
be
closely
related to the project, e.g., PHYS might be used for Physics and PHYED
for Physical Education.
When

you give

names
which

SYSTAT

make
it
project.

CREATING

command,

easier

PROJECT AND

USER

the

identify

user

who

surnames

using

and

the

obvious

system,

project

and

under

DIRECTORIES:

The user and
project
directories
that
(first-level directories) are created by
“ECREATE command.

<ROOT-DIRECTORY>
points
you or the operator using

The procedures you should use and the parameters
in these directories are described below.

that

Create all project directories
as
log-in
(user)
would not create a project directory as files-only,
directories
cannot
have
1log-in
However, log-in project (or user)
files-only subdirectories.

you must

to
the

include

directories,
You
because files-only

directories
<created
under
them.
directories can have both log-in and

Assign a disk storage quota to each
project
directory.
This
quota
must
be large enough to accommodate both the files that are contained
in the directory and the directories that are created under it.
Each
time
a
directory
1is
.created
wunder
a
project
directory,
that
directory's disk quota is taken
from
the
project
directory's
disk

quota.
The
total disk quota for directories
directory cannot exceed the quota
originally
directory.

created under a
given
to
the

In the example below,
the
operator
begins
to
directory
<CHEM>,
He creates the directory as
the public structure ORANGE:
and
assigns
the
procedure
allows
an
administrator to log into
its

password,

and

create

the

required

project
project

create
the
project
a log-in directory on
password
H24.
This
the directory, giving

subdirectories.

may

also

want
to
store
his
files in this directory.
The operator gives the
directory a 10,008-page
working
and
a
10,000-page
permanent
disk
storage

quota.

@ENABLE

(CAPABILITIES)<RET>

$“ECREATE
[NEW]

(NAME)

ORANGE:<CHEM><RET>

$SPASSWORD H2@<RET>
SSWORKING 10000<RET>
SSPERMANENT 10@@@<RET>
$$

CREATING

DIRECTORIES

Next, the operator enters the parameter that allows the owner
of
the
project
directory
to
<create subdirectories.
This parameter, called
MAXIMUM-SUBDIRECTORIES (ALLOWED), specifies how many
directories
can
be
created under the directory.
Unless you enter this parameter (the

default
1is
@),
the
owner
of
the
directory
cannot
create
subdirectories.
For
example,
all
wusers of the system can type the
BUILD command to the TOPS-20 Command Processor, but only
those
users
who
have
the
MAXIMUM-SUBDIRECTORIES
(ALLOWED)
parameter
in their
directory with a number greater than zero can actually use
the
BUILD
command

create

subdirectories.

The following entry in the sample project directory <CHEM> allows
administrator

to create

$SMAXIMUM-SUBDIRECTORIES

100

subdirectories.

(ALLOWED)

the

1@g@<RET>

The administrator who is responsible for
this
sample
project
might
create
6@
directories
wunder
the
project
directory
and give each
subdirectory approximately 50 pages
of
working
and @ permanent
disk
quota.
He
Kkeeps enough pages in the project directory to allow that
directory's files to grow and
to
create
additional subdirectories.

(Refer to the TOPS-2@ Commands Reference Manual

for the description of

the
BUILD
command,
including
distributing
working
storage quotas and maximum subdirectory quotas.)

Also,

some of the MAXIMUM-SUBDIRECTORIES

project

directory

(ALLOWED)

and

permanent

quota given

under it can be created.
The
quota
for
the
project
directory
decremented by the amount of quota given to the subdirectory.
For

The

example,

directory <CHEM>

administrator

the

can be given to a subdirectory so that directories

creates

is given a

subdirectory

dquota

Iis

104d.

the directory <CHEM.STUDENT> under <CHEM>

and gives the directory a subdirectory quota of
16.
The
number
of
subdirectories
that
can
now
be created under <CHEM) is 89,
If the
administrator
<creates
another
subdirectory
under
<CHEM>
called
<CHEM.STUDENT2>
and
gives
that directory a subdirectory quota of 6,
the number of subdirectories that can now be created under
<CHEM>
is
82.

If the administrator gives an

command,

the output line

MAXIMUM NUMBER

The two
created
the

INFORMATION

(ABOUT)

DIRECTORY

for maximum subdirectory quota

SUBDIRECTORIES

ALLOWED

is:

<CHEM>

directories
<CHEM.STUDENT>
and
under
<CHEM> account for the two

<CHEM.STUDENT2>
that
were
subdirectories not shown in

subtraction.

Next, the operator enters the parameter that allows the
administrator
for
this project to place users in groups.
The administrator can use
the group facility as described in
Section
5.8
to
set
up
library
directories and allow file sharing among members of the project.

CREATING

The SUBDIRECTORY-USER-GROUP
262143
as
1its
argument.

DIRECTORIES

parameter
You
can

accepts a number between
list a range of numbers that

administrator can use to establish groups within the
Be careful
you
must
‘enter each number separately.
to
project.
of numbers that is unique
that
directory <CHEM> may be given the range:
$$SUBDIRECTORY-USER-GROUP

(ALLOWED)

2600<RET>

$$SUBDIRECTORY~USER-GROUP

2601<RET>

$$SUBDIRECTORY-USER-GROUP

(ALLOWED)
(ALLOWED)

$$SUBDIRECTORY -USER-GROUP

(ALLOWED)

260 3<RET>

$$SUBDIRECTORY-USER~GROUP

(ALLOWED)

2604<RET>

Project directory
numbers different

2602<RET>

may
be
<PHYSICS>
from project CHEM.,

given

following

$$SUBDIRECTORY-USER-GROUP
$$SUBDIRECTORY-US ER-GROUP

(ALLOWED)

3001<RET>
3002<RET>

$$SUBDIRECTORY
-US ER-GROUP

(ALLOWED)

300 3<RET>

$$SUBDIRECTORY-USER-GROUP

(ALLOWED)
(ALLOWED)

3004<RET>
3005<RET>

$$SUBDIRECTORY
-US ER-GROUP

the
project; however,
to assign a range

you assign the same range of numbers to different projects, you can
For example, a user in group
a
security break among projects.

cause

2602

project

the directories

CHEM should

and

files

not

able

project

access,

group

PHYSICS.

directory's
in
placed
project
range of
numbers
imply
the
directory
or any of its
that
not
parameter
1list
does
It
subdirectories has access to those groups.
(or owner of the directory) can use those group numbers
administrator
among
that
directory
relationships
to
establish
group
subdirectories.
The

the
The following example shows
sample project directory <CHEM>:

@ENABLE
$

(CAPABILITIES)

"ECREATE

[NEW]
$$SPASSWORD

(NAME)

completed

<RET>

ORANGE:<CHEM><RET>

H20<RET>

$SWORKING 10000<RET>
$SPERMANENT 100Q@<RET>

$SMAXIMUM

SUBDIRECTORIES

(ALLOWED)

$SSUBDIRECTORY-USER-GROUP

(ALLOWED)

$$SUBDIRECTORY-USER-GROUP

(ALLOWED)
(ALLOWED)
(ALLOWED)
(ALLOWED)

$$SUBDIRECTORY-USER-GROUP
$$SSUBDIRECTORY-USER-GROUP

$$SUBDIRECTORY-USER-GROUP
SS<RET>

$ DISABLE

parameter

(CAPABILITIES)

1@0<RET>
2600<RET>
2601<RET>
2602<RET>
260 3<RET>
2604<RET>

<RET>

Refer to the TOPS-2@ Operator's Guide for a
complete
description
of
the “ECREATE command that the operator uses to create new directories,
and
the
ULIST
program
that
prints
information
about
all
the
directories

the

system.

After creating a new directory (either user or
files-only),
remember
to
wupdate
the backup tape that contains the monitor, TOPS-20 Command
Processor,
DLUSER,
DLUSER
data,
DUMPER,
<SYSTEM>,
and
<SUBSYS>.
(Refer to Chapter 7, System Backup Procedures.)

CREATING

DIRECTORIES

CONSIDERATIONS:

If two projects or users have mistakenly been assigned the same
name,
and
vyou try to create the second directory with this duplication, the
system prints [OLD] instead of
[NEW].
Give
the
ABORT
subcommand,
assign
the user or project a slightly different name, and reissue the
“ECREATE command

with

the

new directory name.

A subdirectory is just like any other directory.
It can
be
1logged
into
(if
it
1is
not specified as files-only), it can be a member of
user
and
directory
groups,
and
it
obeys
the
wusual
protection
mechanisms.
Therefore,
there
are
no
implied
rights
between
a
directory and its subdirectories, or between two subdirectories of the
same

directorye.

and world;
Refer
to

Files

have

three

protection

fields:

owner,

group,

and each directory has the same
three
protection
fields.
Section
5.7
for
a
description
of
directory
and
file

protections.

The only additional rights that the owner of a directory has over that
directory's
subdirectory is the power to change its parameters (e.g.,
directory protection, password, or group memberships), or to
use
the
KILL subcommand, which deletes the subdirectory.

you or another

delete
any
directory or

protection

user choose

to delete a directory,

you

must

first

subdirectories
under the directory.
You cannot delete a
subdirectory
that
has
existing
subdirectories.
This

insures

that

someone

(possibly

administrator

special

capabilities

of a

project) does not accidentally delete a directory
that
points
to
a
large
portion
of
the
database.
The operator or administrator must
connect
to
the
directory
immediately
above
the
lowest
level
subdirectory
to
begin
deleting any directories.
For example, using
the diagram in the FORMAT description, if the owner of
the
directory
<PHYSICS>
wants
to
delete
the
directory <PHYSICS.LAB-12>, he must
first connect to
directory
<PHYSICS.LAB-12>
and
delete
the
three
<PHYSICS.LAB-12.STUDENT>
directories.
Then, he connects to directory
<PHYSICS>
and
gives
the
KILL
subcommand
to
delete
directory
<PHYSICS.LAB-12>,
Note
that the operator is the only person who can
delete

the

you

user,

you

directory

the

<PHYSICS>.

administrator

can

include

the

choose

parameter

grant

for

the capability

in the

user's

directory.
(Refer
to
Section
5.9
for
a
description
of
the
capabilities you can assign to certain users.) You should instruct the
administrator to inform you when special capabilities are given
to
a
system
user.
You
are protected against users randomly giving other
users special capabilities, because the operator or the
administrator
who assigns special capabilities to a user must have (as a user) those
same capabilities.
A person with WHEEL or OPERATOR
capabilities
can
assigh any capability to another user.
Also, the user or operator who
is assigning the capabilities must have those capabilities enabled
at
the
time
the
©privileged
parameter
1is
entered
into
the
user's
directory.

Once a SUBDIRECTORY-USER-GROUP number has been allocated

to a

project

directory,
be
careful
about removing it.
If it is in use in any of
the
subdirectories,
either
as
a
USER-OF-GROUP
number
or
as
a
SUBDIRECTORY-USER-GROUP
number,
you
will
be unable
to recreate the
subdirectories (refer to Chapter 9) if the structure is damaged, until
you
manually
restore the SUBDIRECTORY-USER-GROUP number into all the
superiors.

CREATING

FILES-ONLY

DIRECTORIES

DIRECTORIES:

Administrators or users can have library areas in
addition
to
their
logged=-in
directories.
They
can
wuse
the BUILD command and create
files-only directories under their logged-in directories, provided you
have
given
them
the
capability
to
do
so
by
adding the MAXIMUM
SUBDIRECTORIES

the

(ALLOWED)

parameter

the

directory

that

will

contain

subdirectories.

CONSIDERATIONS:

Refer

to the CONSIDERATIONS
Control,
Section
5.4.1.
Administrative Control.

under

These

the
first
description
of
considerations also apply to

Central
Project

RESTRICTIONS:

You

cannot

exceed

approximately

12,000

directories

per

directory points

structure.

The number of
cannot exceed

If you reach
a structure,
MAXIMUM

the
the

directories that a superior
approximately 5,004d.

maximum number of directories
system prints the message:

DIRECTORY

NUMBER

If either
you
or
an
directories that can be
prints the message:
SUPERIOR

DIRECTORY

EXCEEDED;

INDEX

administrator
created under

that

TABLE

you

NEEDS

can

create

EXPANDING

reach
the
maximum
superior directory,

number
of
the system

FULL

Files-only

directories

cannot

have

log-in subdirectories.
1If
create
user
(log-in)
you
must
make
his

you
want
to
allow
a
wuser
to
subdirectories
under
his
directory,
directory a log—in directory.

5.4,4

Combined

DETERMINING

Central

and

Project

Control

FACTORS:

Only a portion of your organization warrants being
separated
into
projects.
The directories for the majority of the user
community
are
created
and
maintained
at
the
central
management
level.’
But,
where
project
administration
1is
appropriate, the task of creating
and
managing
directories
within a project is given to administrators.

For example,
1f
vyour
company
has
groups
of
users
with
terminals
in several distant locations, you may want to have
the administrator at the remote location create and
maintain

all

the directories

for

that

site.

You canh create

directory for the remote location, perhaps using the
the
site
as
the
project
directory
name
(for
<CHICAGO>

directories
operator.

<CHIC>,

the

<SEATTLE>,

central

«see¢)e

location

are

The

project

name
of
example,

remaining

created

the

user

system

CREATING

DIRECTORIES

FORMAT:

<ROOT-DIRECTORY>

alphabetically

points

all

directories,

named

the

<A>

project

through

directories

The

<Z>,

and

project

directories point to the
wuser
and
files-only
directories
that
an
administrator
creates
for
a
given
project.
The
directories <A>
through <Z> point to
wuser
and
files-only
directories
created
and
maintained
by the operator.
These directories can also be allowed to
have

subdirectories.

DIAGRAM:

COMBINED CENTRAL AND PROJECT CONTROQL (12,000 DIRECTORIES)

< ROOT-DIRECTORY
>

s | <CHICAGO
>

< BOSTON >

<A>

¢ | <A.JONES
>

< BOSTON.
ESTEY >

< ASMITH
>

<C>

|see

eee|

<Z>

< Z.TESTS>

>
MEMOS

.| <CHICAGO.
HESS >

< CHICAGO.

OVERHEAD >

<SEATTLE.

< SEATTLE.

SALES >

SMITH >

< SEATTLE.
BOSACK >

< CHICAGO. |
>
INV

|eoe-

<BOSTON. {_, . | <A.SMITH,
HALL >

<SEATTLE> | ¢ o »

< SEATTLE,

< GEATTLE.

SALES.

OVERHEAD >|

[PROJECTION >

UP TO APPROXIMATELY 5,000 SUBDIRECTORIES PER DIRECTORY

UP TO APPROXIMATELY 12,000 DIRECTORIES PER STRUCTURE
MR-5-3725-84

ASSIGNING

USER

NAMES:

If
you
create
any
user
directories
that
are
pointed
to
by
<ROOT-DIRECTORY>,
assign
project
names
and
user names in the same
manner as described under
Project
Control,
Section 5.4.3.
Again,
assign
the 26 directories that will point to the majority of the user
directories, the names <A>, , <C> ....
<2>.
The user
names
that
will
be
the directory names under the alphabetic directories should,
as previously stated, be
the
user's
surname
prefixed
by
a
first
initial
and a period.
(Refer to Section 5.4.2, Central Control Using
Subdirectories.)

CREATING

USER

AND

FILES-ONLY

DIRECTORIES:

Create the user,
files-only,
following the instructions in

Create

the

Section
(TOPS-2@

project

DIRECTORIES

and
<A>
through
Section 5.4.2,

directories

according

<KZ>

the

directories

instructions

5.4.3,
and
distribute
the description of the BUILD command
Commands Reference Manual)
to
the
administrators
who
are

responsible

for creating

the

user directories within their project.

CONSIDERATIONS:

All the considerations that apply to both Central
also apply to combining the two types of control.

and Project

Control

You may want to allow users
whose
directories
are
created
by
the
operator
to
create
several
directories
wunder
their
1logged-in
directories.,
The diagram
under
FORMAT
1illustrates
this
facility.
User
A.SMITH
has
created
the subdirectory <A.SMITH.MEMOS> to store
files that he wants to keep separate from his programming files.
This
user
uses
the
BUILD
command to create the number of subdirectories
that he is allowed to create and divides the quota for
his
1logged-in
directory among the directories he creates.,

In general, users can store files in these directories or, if they set
the
appropriate
protection, can share the files in these directories
with

other

users.

RESTRICTIONS:

Combined Central and Project Control
directories per structure.

The number of directories that
cannot

exceed

you reach

system

prints

MAXIMUM

If you reach

directory can
SUPERIOR

5.5

approximately

the maximum

allows

superior

up to

approximately

directory

can

12,000

point

5,000.

number

directories

per

structure,

the

the message:

DIRECTORY NUMBER

the

point

maximum
to,

DIRECTORY

ALLOCATING

the

EXCEEDED;

number

INDEX TABLE NEEDS

directories

system prints

that

the message:

EXPANDING

superior

FULL

DISK STORAGE

QUOTAS

In Chapter 4 you determined the amount of disk space that is available
on
the
public
structure
after
installation.
Once
you
know the
available disk space, you can decide how
to
allocate
it
among
the
directories you create.
Each directory is given a number of pages for
both
working-storage
and
permanent-storage
allocations.
Working
storage
refers to the disk space that a user can have during the time
he is logged-in.
Permanent storage refers to
the
total
disk
space
that a

user can have

store

files after

he has

logged-out.

CREATING

DIRECTORIES

The number of pages that you should assign to directories
depends
on
whether
vyou (or the operator) are creating all the directories on the
system (central control) or you are delegating the
task
of
«creating
and
maintaining
directories
to
project
administrators
(project
control).
When using central control, you may divide the
disk
space
equally
among
directories,
giving regard to special requirements of
certain users.
In the case in
which
the
operator
creates
project
directories,
you
should
allocate
a
disk
quota
1large
enough
to
accommodate the expected size of each project.
Remember that
project
directories
must
distribute
their
disk
space
to
'the directories
created

under

Several

important

discussed

Assign

who

for

them.

points about working

and

permanent

allocations

are

below.

a large

(2000-3000

page)

working-storage

perform considerable sorting

because the

this operation can occupy substantial

disk

allocation

users

temporary files required
space.

As the number of users on the system increases and your disk space
on
the public structure becomes low, you can decrease the working-storage
and permanent allocations on the public structure to
add
new
log-in

directories,
If
you
have
additional
disk
drives not used by the
public structure, you can accommodate the
directories with
many
or
large
files by creating other structures and directories.
Users will
log into their
directories
on
the
public
structure,
request
the
operator
to
mount
the proper structure using the MOUNT command, and
access their additional directories with
the
ACCESS
and/or
CONNECT
command.
Note that in setting up the system, it is easier to accustom

users

from the

structures

and

start to use other structures than
retrain

users

after

to reorganize

the

space has run out on the public

structure.

5.6

ENFORCING

DISK

STORAGE

QUOTAS

Working-storage
allocations
are
strictly
enforced.
Users
cannot
exceed
their
working-storage allocations unless they enable WHEEL or
OPERATOR capabilities.
(Refer to Section 5.9 for a description of the
special
capabilities that can be given to users who require them.) If
users request additional space, you can increase their allocations
as
required.
If a user exceeds
create
or
change

his
a

working-storage
allocation
and
attempts
to
file,
the
system
prints
the following error

message:

?2QUOTA

EXCEEDED

The user must decrease his disk usage to less than the working—-storage
allocation
for
the
directory (in which the file is being changed or
created) before he can create or change any more files.

The system
allocation
directory.
or

LOGOUT

informs a
user
if
he
1is
over
this
permanent
storage
when
he
1logs
off
the system or connects to a different
The system prints the following message after the
CONNECT
commands:

<directory>OVER

PERMANENT

STORAGE

ALLOCATION BY nn

PAGES

CREATING

DIRECTORIES

This message reminds users that although they may not
be
over
their
working~storage
allocation,
they
have exceeded their expected total
disk usage.,
Users should delete any files that
are
unnecessary
for
their
job.
Also,
because
permanent quotas are not enforced, it is
wise
to
instruct
the
operator
or
administrator
to
police
each
directory's
disk
usage.
The operator should run the CHKPNT program
daily to keep a record of each directory's disk
usage.
The
TOPS-20
Operator's
Guide
contains
the
description
of
running
the CHKPNT
program.

Chapter

8),

command

Every

than

time

500

you

are

you

force

the

users

the

want
to

file

stay

run

below

the

[CAUTION-DISK
IN

SPACE

[EXPUNGING

LOW

structure

Again,
users

the
are

this

message

when

5.7

the

name]

quotas.

[DELETED FILES

SWAPPING

operator
over

SPACE

Chapter

the

following message
and
starts
directories
on the structure

the

expunging

complete:

for

allocations.

more
disk
to the one

should
Also,

check

see

are

using

you

how
the

directory or

many
file

to migrate files on the
system
more
running low on systemwide disk space.

description

DIRECTORIES

file when there is no
error message similar

EXCEEDED

administrator

permanent

PROTECTING

when

less

WILL

file

migration.)

AND FILES

Every directory and file has a protection number associated
with
The
system
wuses
a
default protection number for each directory

file

TRIM

EXPUNGE COMPLETED]

migration facility, you may want
frequently if you are constantly
(Refer

permanent

(refer

with

the
all

If anyone tries to create or change a
space available, the system prints an
below:
?FILE

program

DELETED FILES]

prints

[SYSTEM

their

SECONDS]

30 seconds, the system prints
expunging
any
deleted
files
in
mentioned in the warning message:

system

facility

REAPER

disk space on the public
structure
is
system prints the following warning message:

After

The

migration

the

available

pages,

EXPUNGED

using

may

file

it.
and

created.

Whenever a user accesses a file, the system first checks the directory
protection.
If that protection allows the user the appropriate access
to the directory,
the
system
then
checks
the
protection
of
the
individual file.

CREATING DIRECTORIES

5.7.1

Directory and File Protection Digits

The directory and file protection numbers have three 2-digit fields.
The first field applies to the owner of the directory or file, the
second field to members of the same group as this directory, and the
(or world).

third field to all other users

Protection Code

The default

directory
access.

Owner

Group

World

protection

for

and

files is

7777090.

protection

allows the

owner

and

directories

protection

file

For example, the default

77 in any given field allows full

members of his group full access but all other users no access.
Protection Code

73"/

Owner

Group

World

Table 5-1 contains a list of the directory protection digits.
Table 5-1:

Directory Protection Digits

Digits

Privilege

Permits creating files in the directory.

Permits connecting to the directory without giving

password and changing the accounts and protection
numbers of the files therein. Thus it gives many of
the privileges the directory owner has. (Refer to the
TOPS-2@ Monitor Calls manual.)

Permits, subject to the protection on the

individual

files with the
the names of the
1listing
file,
e.g., via the
file,
the
reading
and
command
DIRECTORY

TYPE,

PRINT,

or LIST commands.

These protection codes are actually bits in a protection word.

To get

more than one protection, add the digits (octal) corresponding to the
protection you want. Thus, 44 allows listing the files and creating
new files. There are unused bits in the protection number; therefore,
to provide complete access to files, use 77. Useful digit pairs are:
access.

%17

Permits no

Permits the files to be listed and read.

Permits

full

(owner)

access.

A file protection number has the same format as a directory protection
number, but the meanings of the digits are different. Table 5-2
contains a list of file protection digits.

CREATING

Table

5-2:

File

Protection

DIRECTORIES

Digits

Privilege

Permits

wildcarding

Permits

appending

Permits

executing

the

Permits

writing

and

deleting

- Permits

reading

the

file,

the

file.

file,
the

file.

Obtain a protection number by adding the file protection digits of the
different protections you need.
For example, protection number 775200
allows the owner full
privileges;
the
members
of
the
same
group
reading,
executing,
and
directory listing privileges; and all other
users

privileges.,

Permits

file

Useful

1listing

the

specified

digit

file

pairs

with

explicitly

are:

the

and

DIRECTORY

Permits executihg and using the DIRECTORY
file

only.

This

protection

program

and

system

users

protection

file

agree
to

Permits
list the

Permits

when,

for

contract

write

into,

execute-~only

additional

file

command

example,

not

executing,

only

and

copy

771212.

using

you

allow
the

the

1list

the

purchase
any

file.

The

considerations

files.

reading,
file.

full

your

read,

provides

execute-only

useful
in

command

completely.

Set

TOPS-2¢

for

the

Beware

setting

DIRECTORY

your

command

access.

The system checks protection numbers starting with the
two
rightmost
digits.
Therefore,
users
do
not
restrict
members
of a group by
assigning the file protection 770052, because the group gets at
least
the
execute,
read,
and
directory
1list
access (52) granted to all
users.,

Also,

because

file
still

protection,
secure in a

For

example,

the

system

checks

suppose

the

user

directory <HESS>.
The
777700
and
the protection

and directory
<HESS>
protection
applies.
777708.
the

directory.

the

corresponding

though

the

file

KOHN were

KOHN

tries

User

KOHN

not

before

the

low file protection are
directory
protection.

type

on
the
EDIT.MAC

protection on
executed, and listed
access to files in the

read,

allowed

protection

are
not
in
the
same
First,
the
system
The last two digits
(#@¢)

access

even

directory

been given a
the
default

protection
on the file

protection,
to

the

files that have
directory with

the

file

EDIT.MAC

directory
is 777752.

group,
checks
apply

allowed

the

file

with

the

so
the
and

type

<HESS>
is
User KOHN

the
world
directory
permit
the

file,

(52) would
allow
DIRECTORY command

directory.

CREATING DIRECTORIES

Changing Directory and File Protection

5.7.2

Users can change file protection numbers via the SET

command or

FILE

PROTECTION

the RENAME command.

the SET DIRECTORY
You can, however, prevent users from
PROTECTION or BUILD command.
making changes to their directory protection numbers by including the
DISABLE DIRECTORY-PARAMETER-SETTING command in the system file called
<SYSTEM>n-CONFIG.CMD on the public structure. If you make this entry
in n-CONFIG.CMD, only users with WHEEL or OPERATOR capabilities can

Users can change directory protection numbers via

change

directory

the

(via

parameters

PROTECTION commands).

ENABLE

and

SET

DIRECTORY

NOTE

Make an entry in n-CONFIG.CMD only if you DO NOT want
allow users to change their directory protections;
to

otherwise, the system assumes that you want to use the

ENABLE
of
command
default
system
(Refer to the TOPS-20 KL
DIRECTORY-PARAMETER-SETTING.

Model B Installation Guide for a description of the
parameters that are placed in the n-CONFIG.CMD file.)

5.8

ESTABLISHING GROUPS

You can let users share files by placing users and directories in
Members of a group can access directories and files in that
groups.
group according to the middle digits of the directory and file
protection code

fields.

PROTECTION CODE FIELD

MR-5-644--80

USERS and
Each group that you establish has two types of members:
Each group is identified by a number. This number is
<DIRECTORIES>.
included as one of the directory parameters in each directory
to the group. Any directory (including subdirectories) or
belonging
user can belong to as many

groups.

You

can

set

group

the
using
Dby
directories
individual
the
in
relationships
and
“ECREATE
the
to
subcommands
UP
USER-OF-GRO
DIRECTORY-GROUP and
The following example shows that you have placed user
BUILD commands.
Smith in user group 268 and directory groups 268 and 418:
<RET>
@ENABLE (CAPABILITIES)
$ “ECREATE (NAME) MAIN:<SMITH><RET>
$$ PASSWORD SOAR<KRET>
SSWORKING 500<RET>
$$ PERMANENT 50@<RET>
$SUSER-OF-GROUP 268<RET>
$$SDIRECTORY~-GROUP 268,418<RET>

CREATING

DIRECTORIES

The

DIRECTORY-GROUP or USER-OF-GROUP parameter that you place
in
the
user's
directory
determines:
l)
if
this
wuser can access another
directory's files as a group member 2)
if the
files
in
this
user's
directory
can
be
accessed
by another user as a group member, or 3)
both.
The diagrams on the following pages illustrate
the
difference
between being a member of a group as a directory and/or as a user.

When
same

user

accesses

group,

the

file

system

first

directory

checks

that

see

this

member
user

of
the

the
owner

of the directory.
When, in this example, it finds that
the
user
|is
not
the
owner,
the
system then checks to see if the user is in the
same group as this directory.
1In this case, the
user
and
directory
are
1in
the same group; that is, the group numbers match.
The system
now checks the group protection code
field
of
the
directory
being
accessed.
If the group protection allows the type of access that the
user requested, the system proceeds to check the group
protection
on
the individual file.
If you are setting up groups on
different
structures,
correlation between a group number on one structure and

there
1is
no
the same group

number on another structure.
For example, group 202
on
not
necessarily have the same user and directory members
on another structure.

MAIN:

does

group

202

USER

Each directory has two lists of
group
numbers:
Directory Group Numbers and
User

Group

Numbers.

DIRECTORY GROUP NUMBER LIST

Directory Group

Numbers
groups
of

various

VSER GROUP NUMBER LIST

users

of
MR-8-545-80

which

the
this

member.

User Group Numbers are associated with

identify

and

which

identify

each

user

the

various

a member.

groups

CREATING

DIRECTORIES

The

USER

Directory

Group

Numbers

are

to users who require access
important
who
users
Those
to this directory.
group number in the
matching
a
have
User Group Number List can access this
group
1its
to
according
<DIRECTORY>

DIRECTORY GROUP NUMBER LIST

protection code.

The User Group Numbers

USER GROUP NUMBER LIST
n

are

important

This
the owner of this directory.
to
that
directory
any
owner can access
group number in its
matching
a
has

Note:
List.
Number
Directory Group
Because files-only directories are not
not
do
they
associated with a user,
contain User Group Numbers.

MR-5-546-80

There are three common types of groups:

A file-sharing group, whose users can access a set of library
directories and each other's logged-in directories.

A library group, whose users can access a set of library
directories and their own logged-in directories, but not each
other's logged-in directories.

A teacher-student group,

in which the teacher can access

the

and the students can access their own
directories
students'
directories
logged-in directories, but not their classmates'’
or

the

teacher's directory.

Figures 5-1 through 5-3 illustrate these three common groups

association between USER and <DIRECTORY> members of a group.

and

the

users share all their files
In a file-sharing group (Figure 5-1),
library
in the
both
the group protection field,
to
according
directories (here it is <MANUALS>) and in their logged-in directories.,

CREATING

DIRECTORIES

USER PORADA

USER HOLLAND

< HOLLAND
>

DIRECTORY GROUP NUMBER LIST

DIRECTORY GROUP NUMBER LIST
2,4

—

~~~~~~

USER GROUP NUMBER LIST
-

<MANUALS?

~—

/
14

DIRECTORY iROUP NUMBER LIST

V4
2,8

USER GROUP NUMBER LIST
NONE

MR-5-547-80

Figure

5-1:

Figure

same

File-Sharing

5-1,

dgroup

access

are

the

The

also

members

fields.
according to

the

access

PORADA

2).

own

can

users,

(group

their

code

two

Group

directory

and

HOLLAND,

directories

group

and

2.,

files

PORADA

are members

<PORADA>,

Users

PORADA

according

directories

and

according

group.

library

HOLLAND

can

owner

protection

can access directories
<HOLLAND>
and
<MANUALS>
group protection code fields, and conversely, HOLLAND

protection
code
fields.
The
other
numbers
shown
indicate that a user or directory can be a member
of

the
and

and

the

<HOLLAND>,

group

(Figure

members

but

5-2),
not

USER

members

can

each

other's

logged-in

This

library

group

their

in
more

the

access

group

figure
than
one

all

the

directories.

The

library directories are usually files~only directo
ries.
This
figure
illustrates
a
library
group
that
consists
of
the
files-only
directories:
<SUBROUTINES>,
<TAPE-TESTS>,
<MACROS>,
and
users:

ALUSIC,

because

BROPHY

and

you are a
directory need not

KOHN.

member
belong

of
a
to the

group
as
same group.
5-29

illustrates

user,

your

that

just

logged-in

CREATING DIRECTORIES

USER KOHN

<KOHN>

DIRECTORY GROUP NUMBER LIST
1

USER GROUP NUMBER LIST

DIRECTORY GROUP NUMBER LIST

USER GROUP NUMBER LIST

NONE

USER ALUSIC

USER BROPHY

DIRECTORY GROUP NUMBER LIST

DIRECTORY GROUP NUMBER LI{ST

USER GROUP NUMBER LIST

2
MR-5-548-80

Library Group

Figure 5-2:

In Figure 5-2,
members of the
the

same

according

group

<SUBROUTINES>,

directory

group

directory

<KOHN>

to the

group

users KOHN, ALUSIC and BROPHY are not

same group; however, they are all user group members in
(group 2).

the

owner

User

KOHN

can

access

protection field and can access directories

<TAPE-TESTS>,

and

according

KOHN can access <BROPHY> and <ALUSIC> according to
protection field.
the "world" protection field. Although the arrows have not been drawn
from users BROPHY and ALUSIC, their access privileges are the same as
KOHN's.

5-30

CREATING

DIRECTORIES

TEACHER WILEY

<WILEY>

DIRECTORY GROUP NUMBER LIST
NONE

USER GROUP NUMBER LIST
5

/)
STUDENT HURLEY

STUDENT HALL

<HALL>

DIRECTORY GROUP NUMBER LIST

DIRECTORY GROUP NUMBER LIST
5

USER GROUP NUMBER LIST

NONE

-_—
STUDENT MILLER

STUDENT RUSSELL

—

DIRECTORY GROUP NUMBER LIST

5‘———/

USER GROUP NUMBER LIST

NONE

MR-5-549-80

Figure

5-3:

Teacher-Student

Group

In a teacher-student group (Figure 5-3),
the
teacher,
WILEY,
1is
a
member of the group as a USER, while the directories <HURLEY>, <HALL>,
<MILLER>, and <RUSSELL> are <DIRECTORY> members.
The teacher,
WILEY,
can
access
the
files in the directories <HURLEY>, <HALL>, <MILLER>,
and <RUSSELL> according to the group protection.
The
students
whose
logged-in
directories
are
in
this group as <DIRECTORY> members can
access the files in <KWILEY> according to the protection
set
for
all
users,
because
only their directories are members of the group; they
are

not members

the

group

users.

5-31

CREATING

5.9
You

GIVING
can

USERS

give

OPERATOR,

SPECIAL

special

DIRECTORIES

CAPABILITIES

capabilities

CONFIDENTIAL,

certain

MAINTENANCE,

users;

IPCF,

they

ENQ-DEQ,

are

WHEEL,

ARPANET-WIZARD,

and

ABSOLUTE-ARPANET-SOCKETS.
Each capability that you give to a user
is
placed
in
the user's directory parameter list when you create or
change the directory.
The person
who
enters
the
capability
in
a
user's
directory
must
have
that
capability
himself,
and have it
enabled at the time the
capability
1is
entered
into
the
directory
parameter list.
You should grant these capabilities only to users who
absolutely need them,
Table 5-3 lists all the available
capabilities
and a brief description of their function.

Table

5-3:

Special

Capabilities

Capability

Description

WHEEL

Allows

the

user

modify

parameters
or
data.
WHEEL
capability
1is

wants

give

any

system

In
particular,
needed
1if
the

the

“EEDDT

the
user

“EQUIT

commands.

OPERATOR

Allows
to

the

user

control

however,

all

the

capabilities

system.

give

the

The

required

user

cannot,

"EQUIT

“EEDDT

commands.
CONFIDENTIAL

MAINTENANCE

Allows

the

wuser

information

for

Allows

user

the

obtain

another

user's

(usually

representative)
maintenance

the
IPCF

Allows

the

field

service

perform

functions,

but

certain

cannot

give

commands.
the

user

functions
of
Monitor Calls
ENQ-DEQ

accounting

job.

Allows

perform

the

privileged

IPCF.
(Refer to the
Reference Manual.)

the

user

TOPS-20

perform

global

ENQUEUE/DEQUEUE
functions.
(Refer
to
TOPS=20AN Monitor Calls User's Guide.)
ARPANET-WIZARD

Allows

the

privileged
TOPS=20AN

Monitor

Allows

the

SOCKETS

numbers

Allows

the

ARPANET

DECNET-ACCESS

DECNET

wuser

his

perform

User's

place

ARPANET

the

Guide.)

absolute

socket

programs.
owner

establish

connections,

directory

network

certain

(Refer

Calls

directory

network

Allows the

functions.

ABSOLUTE-ARPANET-

ARPANET-ACCESS

user

the

owner

connections.

CREATING

With the exception of WHEEL and
listed in a format where having

DIRECTORIES

OPERATOR, these capabilities
one capability means you also

capabilities

are
have

not
the

listed below it.
The user who has WHEEL capabilities can
the
OPERATOR,
CONFIDENTIAL,
MAINTENANCE, IPCF, and ENQ-DEQ
functions.
The user who has OPERATOR capabilities
can
also
perform
these privileged functions with the exception of the “EEDDT and "“EQUIT
perform

commands.
But the
perform
functions
cannot

perform

ABSOLUTE

user

SOCKETS

individual capability.
capabilities.
you are giving
Therefore,
if
wuser
connect

capabilities..

has
CONFIDENTIAL
are
allowed by this

MAINTENANCE,

ARPANET

Also,

who

that

The

IPCF,

capabilities

HALL

access

However,

the

ENQ-DEQ,

functions,

same

has

to a
WHEEL

directory

ARPANET

unless

principle

they log

capabilities
can
only
capability; that is, he

true

has

for

WIZARD,

and

been

given

the

remaining

the

user,

not the user's directory.
capabilities,
other users who

<HALL>

user

not

HALL,

obtain

they will

WHEEL

obtain

HALL's

capabilities.
For this reason, users with special capabilities
(especially
WHEEL
and
OPERATOR)
should
be
especially
careful in
selecting and protecting their passwords.
They should
be
encouraged
to
change
them
often,
and
to use passwords that cannot be readily
guessed.

5.180

PRINTING

DIRECTORY

INFORMATION

The

ULIST program prints information about directories on
the
system
and
1is
described
1in
the
TOPS-20 Operator's Guide.
1In addition to
listing information about each directory, the ULIST program
can
1list
information about groups, capabilities, and related information.

The
the

LIST subcommand of the “ECREATE command prints
information
about
directory or user name you are currently creating, and the “EPRINT
command also prints the information on an individual basis.

CHAPTER
CREATING

The

TOPS-2f

computer

adding
usage

accounting

usage

facility

valid

ACCOUNTS

allows

accounts.

you

provides

assign

you

with

and

charge

means

for

security to your system, 2) determining
charges
for
computer
and
billing
users by account, and 3) associating classes with

accounts

for

validation

wuse

for

One or more

each

records

CPU

all

time
time,

account,

reports

the

<class

scheduler.

these

and

they

are

structures

are

stored

used.
used,

You

can

wuse

account

reasons.

accounts can be assigned

validated

valid

one

user

All
and

usage

for specific

tasks

accounting

data,

peripherals

used

file

and

can

and

including
under

used

later

for

billing.

This chapter describes how to set up the system to
use
accounts
and
establish
an
accounting
data
base.
The
TOPS-200
USAGE
File
Specification describes how to
create
accounting
reports
from
the
Usage
file
and establish billing procedures.
The following sections
include:
¢

How

set

How

select

How

use

base
e¢

How

6.1

run

your

account

takes
data

accounting

and

the

these

system

account

accounts

scheme

subaccount

account

use

and

create

the

necessary

files

generator

data

files

program

and

(ACTGEN),

creates

the

which

accounting

base

What

the

work

properly

operator

can

How

initialize

your

SETTING

THE

SYSTEM

Disabling

the

system

accounting

start

data

base

validating

does

not

accounts

USE ACCOUNTS

6.1.1

Enabling

During
create
entry

software installation, you can
specify
whether
vyou
wish
to
the
account data base and validate accounts.
You can make an
in
the
n-CONFIG.CMD
file
that
specifies
either
DISABLE

ACCOUNT-VALIDATION

entry

ENABLE

the

Account

ENABLE

n-CONFIG.CMD

ACCOUNT-VALIDATION.

Validation

ACCOUNT-VALIDATION.

file

for

accounting,

the

you

system

not

make

assumes

CREATING

If you enter

ACCOUNTS

DISABLE ACCOUNT-VALIDATION,

use

the

account

each

account.

only for length.
maximum
number

validation

meaning
the

you do

system

not

checks

wish

each account

The purpose of the
check
1is
to
ensure
that
the
of
alphanumeric
characters has not been exceeded in

use or create
truncates the

facility,

other

checking

performed.

an account greater than 39 characters,
entry to the 39-character maximum.

user

the

attempts

¢to

system simply

If you have instructed the system
to
ENABLE
ACCOUNT-VALIDATION
but
have
not
vyet
created an account validation data base, you receive a

warning
on
operation.

the
<console
terminal
The message is:

(CTY)

when

the

system

starts

<SYSTEM>ACCOUNTS-TABLE.BIN NOT FOUND - ACCOUNT VALIDATION IS DISABLED
The system continues its normal operation; however,
no
accounts
are
validated
(except for length checking) until you create the necessary
account data files and run the account generator program
(ACTGEN)
to
create your account data base.

You should

your

not

recelve the

above warning message

account data base prior to bringing

Users can

6.1.2

log

into

Setting

the

system using

their valid

up Account Validation with

if you

have

created

the system up for operation.
accounts.

Existing

Files

If you are using account validation
on
a
system
that
already
has
files,
the accounts for these existing files should be updated before
account validation is enabled in the n-CONFIG.CMD
file.
Notify
the
users
who created these files to change the existing account on every
file to their new account(s).
This procedure ensures accurate billing
immediately after

the

system

brought

up and

that daily

DUMPER

tapes

contain files with valid
accounts,
This
means
that
1if
you
must
restore files from a backup tape, the correct account for each file is
properly restored; therefore, the disk file storage
continues
to
be
accurately
charged.
(Refer
to the TOPS-20 Operator's Guide for the
procedure

6.1.3

Setting

if all

up the

files do

System

not get

for Accounting

updated.)

Shift

Changes

The accounting facility also allows you to change your
billing
rates
for
system
wusage
at
selected times during the day.
This action is
called an accounting
shift
change.
Accounting
shift
changes
are
selected
by day-of-week

and

time-of-day.

You must enter the appropriate commands in the
n-CONFIG.CMD
file
to
initiate
accounting
shift changes.
The n-SETSPD program reads these
commands each time the system is reloaded.
The format
of the
command
placed

the

CHANGE

n-CONFIG.CMD

time

file

days-of-week

is:

CREATING

You

can

use

any

ACCOUNTS

format

for

the

3:0@pm,
MONDAY, MON.,
WEEKDAYS.
The default

Or,
for

you can use the
days—of-week is

commands

time

typical

set

that may

CHANGE

9:00

CHANGE

10:00

WEEKENDS,MONDAY

CHANGE

12:08

TUESDAY,THURSDAY,SAT

CHANGE

17:00

and

day,

appear

that

is,

1500,

15:00,

keywords ALL, WEEKENDS, and
ALL.
The
following
1is
a

the

n-CONFIG.CMD

file:

WEEKDAYS

The

CHANGE (ACCOUNTING
SHIFT
NOW)
command
to
the
CHKPNT
program
provides
you with a means of changing shifts during system operation.
This command causes an accounting session to end and a new
accounting
session
to
begin
for
all
active jobs on the system.
Refer to the
TOPS-20 Operator's Guide for a description of all
the
commands
that
can be given to the CHKPNT program.

6.2

SELECTING

AN ACCOUNTING

SCHEME

The first thing you must do before you create account
data
files
1is
set
up
an accounting scheme.
This procedure includes deciding which
accounts you wish to create, their expiration dates if you
are
going
to

open

and

accounts,

the

names

charge to)
those accounts and, if you are
the scheduling class associated with each
The TOPS-2f account
validation
project
administration
in
a
account.,

For

the

users

using the
account.

facility
allows
group of accounts

who

<c¢lass

can

use

(or

scheduler,

several
levels
of
having the same base

example:

DENVER

Base Account

Subaccount

|
CHEM

N
Subaccount ——
|

OVERHEAD

LAB-12

--«——— Subaccount

MR-8-551-80

The

accounts

you

would

create

using

the

above

example

are:

DENVER
DENVER.CHEM
DENVER.,CHEM,OVERHEAD

and
DENVER,CHEM,LAB-12

In this example, users at Denver University taking
course
in
chemistry
(e.g.,
12)
would
1log
in
assigned account, DENVER.CHEM.LAB-12.

a
and

particular
1lab
charge to their

CREATING

ACCOUNTS

All accounts that you assign to users can have a maximum length of
39
alphanumeric
characters.
The
system allows you to use a hyphen (-)
within
the
accounts
you
create
(e.g.,
LAB-12),
but
no
other
punctuation (including spaces) can be used.
Note that the system uses
the period (.) as a delimiter to separate
each
part
of
multi-level
accounts
and
the
period
1is
<counted
as
one of the 39 characters.
Therefore,

The

type of

DENVER.CHEM.OVERHEAD

accounting

administration

you

scheme

have

you

your

user

account with 20

use depends on the
installation.

form

characters.

Multi-level

project

accounts

are usually created through a form of project
administration
similar
to
that
used
when
allowing
certain
wusers
(perhaps
heads
of
departments) to create subdirectories.
(Refer to Chapter 5,
Creating
Directories.)
Remember
that
subdirectories
are just like any other
directory.
Therefore, users must have
accounts
to
1log
into
their

directory.

Generally, all files that contain data pertaining to base accounts are
created
by you or the operator, and all files that contain subaccount
data
are
created
by
one,
or
perhaps
more
than
one,
project
administrator.
A
project
administrator,
for example, might be the
head of the Chemistry Department.
(This could be the same person
who
handles
the
subdirectory
creation
for a group or groups of users.)
Allocating the subaccount file creation
to
a
project
administrator
allows
you
to
collect
or
budget
for
one
base
account
(e.qg.,
DENVER.CHEM) and not be directly concerned with the
subaccounts,
In
the
example,
the head of the Chemistry Department is responsible for
creating the subaccount files under DENVER.CHEM., that is, LAB-12
and
OVERHEAD.
Section
6.3
describes
how
to create these account data
files using a sample accounting scheme.
Figures 6-1 and 6-2 illustrate several ways that you can set
up
your
accounting
scheme.
Figure
6-1
is
a
simple
scheme
that a small
organization might use.
It also allows you,
as
system
manager,
to
have complete control over all accounts because you are aware of every
account assigned.

CREATING

Figure
to

6-1

set

each

shows

one

customer

that
base

using

the

manager

account

his

accounts.
All
the
charge their computer
Unionbank,

L
people

for

system.

ACCOUNTS

Correct

Data

Company has

Correct

Data

and

the

customer

people
who
usage to the

used

one

base

decided

account

name

for

the

use
the system at Correct Data can
Correct Data account, CORRECT-DATA.

& P Food, and Town Square Magazine submitted the names of
who will be using the system from their respective sites.
These
are
the only people who will be able to log in from their site
and charge to their assigned account.
The
manager
at
Correct
Data
also planned expiration dates for each customer account.

those

SAMPLE COMPANY:

Correct Data

TYPE OF BUSINESS.

Timesharing House
PRIMARY MODE OF OPERATION: Batch

ACCOUNT

CORRECT-DATA

USER

Hudson, Holland, Gerard, Gionet, King, Kelly, Kohn
(Note: These 7 people are all the users at Correct Data)

ACCOUNT

UNIONBANK

USER

Warriner, Bloomstran, Prest, Pendergast

EXPIRES

June 1, 1986

ACCOUNT

LP-FOOD

USER

Schied, Queeny, Smith

EXPIRES

July 1, 1986

|ACCOUNT

TOWN-SQUARE

USER

Markley, Gerhard, Dole

EXPIRES

July 15, 1986

MR-5-3726-84

Figure

6=1:

Using

the

Accounting

Scheme

same
sample
company,
Figure
6-2
shows
how
a
simple
scheme of this type can be expanded into a form of project
administration.
Here,
Correct
Data
and
one
of
its
customers,
Unionbank, broke down the base accounts into subaccounts.
accounting

Because

Correct

Data

account,

the

manager

bills

Unionbank

for

Correct

Data

subdivides its account,
subaccounts
at
Unionbank.
The
concerned with the computer usage
within
file

his
company.
that contains his

all

its

usage

concerned

onhe

with

how

and

is probably
not
aware
of
the
manager
at
Unionbank, however, is

costs incurred
supplies Correct Data
subaccount information.

computer

not

with

each

the

department

name

the

CREATING

ACCOUNTS

SAMPLE COMPANY: Correct Data
TYPE OF BUSINESS: Timesharing House
PRIMARY MODE OF OPERATION: Batch

ACCOUNT
USER

CORRECT-DATA
Hudson, Holland

SUBACCOUNT
USER

PAYROLL
Gerard, Kelly

SUBACCOUNT
USER

OVERHEAD
Gionet, Kohn, Kelly

EXPIRES

December 31, 1986

SUBACCOUNT
USER

PROGRAMMING
King, Carlson

UNIONBANK
Warriner

ACCOUNT
USER

SUBACCOUNT
USER

TRUST
Bloomstran

SUBACCOUNT
USER

LOANS
Prest

SUBACCOUNT
USER

MSTRCHG
Prest, Pendergast

SUBACCOUNT
USER

PAYROLL
Bloomstran

MR-§-3727-84

Figure 6-2:

Accounting

Scheme

Section 6.3 describes how to
6-2 into files that are used

6.3

CREATING

Sections

AN ACCOUNT

enter the
to create

information for
an account data

Figures
base.

6-1

and

DATA BASE

6.3.1
through
6.3.3
describe
how
to
wuse
your
selected
accounting
scheme
and create the necessary files for your data base,
Specifically, these sections include how to enter accounting data into
files, how to run the account generator program (ACTGEN) to create the
data base, and what to do if an error occurs.

CREATING

6.3.1

Entering

Accounting

Data

ACCOUNTS

into

Files

Base and subaccount files are created using a text editor.
The format
below
shows
the
combination
of
entries you can make in accounting
files
wusing
the
CREATE
command.
Each
file
you
or
a
project
administrator
creates can contain one or more accounts.
Each account
‘can point to one subaccount file, where additional account information
is
stored
pertaining
to
that
account.
Following the format is a
summary of the valid commands in an account file.
ACCOUNT

DATA

FILE

FORMAT

@CREATE

(FILE)

INPUT:

filename.type.l

00100
20200

ACCOUNT account/SUBACCOUNT:dev:<dir>filename.type/CLASS :n/ALLOW:n,n
USER name,name,name,
...
DIRECTORY dev:<directory>
GROUP (ON STRUCTURE) dev:/USER:user group number

go300

00400
ga500

00600

GROUP

00700
*EU

<ESC>

<directory>filename.type

(ON STRUCTURE)

dev:/DIRECTORY:directory group number

<filename.type.l>

In addition to the
expiration date in

above entries,
the form:

each entry

the

file can

have

/EXPIRES:dd=mm-yy hh:mm

This switch
that

entry

indicates when
in

the

file.

the account will

For

longer

wvalid

for

example:

USER namel,name2/EXPIRES:10-JAN-86,name3,named

In the above example, name2 can no longer use this
account
after
10
January
1986.
Namel, name3, and named4, however, can continue to use
the account beyond
that
date.
You
could
also
place
the
switch

immediately following

the USER entry.

USER/EXPIRES:14-JAN-86,

namel,

For example:

name2,

name3,

named

This format specifies that none of the users in the list can
wuse
account after a certain date.
The account, however, remains open,
you can place another list of users
in
the
£file
who
can
use

the
and
the

account.

Table 6-1 summarizes the account data file commands.
You can type the
entire
command,
or
Jjust the characters necessary to distinguish one
command from another.
For example, ACCOUNT can be typed as AC.

Because the ACCOUNT command has several modifiers,
you
may
have
to
continue
typing
the
modifiers
on the next line.
To do this, use a
hyphen at the
end
of
the
1line
and
continue
typing
the
ACCOUNT

modifiers

the

line.

@10@

ACCOUNT

0200

/CLASS:2/ALLOW:1,3

For

example,

TEST/SUBACCOUNT:SYSA:

ACCOUNT.TXT-

CREATING

Table

6~-1:

Summary

Account

ACCOUNTS

Data

File

Command
ACCOUNT

/SUBACCOUNT:

Commands

Description
Specifies

the

administrator

project

Note:

The

entry

name

the

account

wish

ACCOUNT command
an account
data

must

wup

Modifies

command.

the ACCOUNT

specification
data

for

Note:

the

The

the

file

account

can

ACCOUNT

file

subsequent
entries
entry are modifiers.

that

the
first
because
all

ACCOUNT

includes

where

you

assign.

the

additional

found.

command

accepts

only

one

/SUBACCOUNT:modifier.
Example:

One

contains
ACCOUNT

the

your

following

accounting

files

commands.

CORRECT-DATA/SUBACCOUNT:

<GERHARD>ACCT.TXT

ACCOUNT

UNIONBANK/SUBACCOUNT:

<WARRINER>ACCTG.TXT

/CLASS:n

ACTGEN looks in
account data for
and
it looks in

<GERHARD>ACCT.TXT
for
more
the
account.
CORRECT-DATA,
<WARRINER>ACCTG.TXT for more

data

UNIONBANK.

for

account

Modifies the ACCOUNT command and is
used
in
conjunction
with
the
<c¢lass
scheduler.
It
specifies
the scheduling class that is valid
for

this

ACCOUNT
means

account.

For

example,

CHEM=-2@7/CLASS:3
that

account

class

CHEM-207.

valid

when

ACTGEN

using

places

the
this

information in the system's
accounting
data
for
use
by
the
<class
scheduling
routines. Use the /CLASS:n switch only if you
have
selected to specify class scheduling by
base

account.

(Refer

complete

description

scheduler

When

have

system

Section
wusing

10.1
the

account.)

wuser
gives an account that
valid class associated
with
default

the
If

it,
CPU
1in

class

scheduling

the

appropriate

class,

not

it,

the account has a class associated
with
that class is used.
The
percentage
of
time that classes can receive is
defined
the n-CONFIG.CMD file.
use

the

does
@.

uses

for
a
class

<class

account,

you

have:
e

made

entries

the

must

CREATING

Table

6-1:

Summary of Account

ACCOUNTS

Data

File

(Cont.)

Description

Command

updated your ACCOUNTS.CMD file (and
subaccount files) to specify the classes
that are associated with each account.

run ACTGEN
update the

given

to
up

/ALLOW:n,n

Commands

with the INSTALL command
ACCOUNTS-TABLE.BIN file.

the ENABLE CLASS-SCHEDULER command
OPR or brought the system down and back
again to start class scheduling.

Modifies the ACCOUNT command and is
wused
in
conjunction
with
the
class
scheduler.
It
allows
you
to
delegate
the
assigning
of
classes
to
subaccounts
by
project
administrators. It
specifies
the
class
or
classes that can be used
by
subaccounts
of
this account. For example,

ACCOUNT CHEM/SUBACCOUNT:<ABC>MORE.TXT/CLASS:2/ALLOW:1,3

means
the
CHEM
account
is in class 2, and
that subaccounts created under CHEM can be in
either
class
1
or
<class
3.
If no /ALLOW
switch
is
given,
the
administrator is not
restricted
to
wusing
certain
classes
and,
therefore,
can
give
the
subaccounts
any
class. For
example,
the
administrator
can
give them the
same
<class
as
the
superior
directory.
The
/ALLOW switch is useful when

you
want
the
superior
account
to
be
1in
perhaps
a
higher
percentage class than its
inferior
accounts.
Remember
that
1if
the
administrator
does
not give a /CLASS switch
to the subaccount,
users
who
1log
into
or
change to this subaccount will be in class 8.
USER

Specifies
one
are allowed to

Specifies
users
on
argument

user or
use this

the list of users
account.

that an account is
a
system.
The
*
to

the

USER

wvalid
1is
a

who

for
all
special

command.

Example: One instance when you might use * is
if
you
have
not
established an accounting
scheme
but
would like to allow users to log
into the system.
and use the * to
use

that

You

could

ACCOUNT

USER:

You could set
indicate that

up one account
all users
can

account.

also

use

MATH-101
MATH.*

the

follows:

CREATING

Table

6-1:

Summary of

Account

ACCOUNTS

Data

File

Command

(Cont.)

Description

This
means
that
all
beginning
with
MATH

users with a user name
can
use
the MATH-101

account.

the

user

For

access
users

to
to

users

MATH.SMITH,

can

Specifies a
the account

or
to

example,

names

MATH.BROWN
DIRECTORY

Commands

all

use

assigned

this

and

account.

directory name. It indicates that
is valid for
anyone
with
write
the

directory.

create

store

This

feature allows
in
systemwide

files

groupwide
directories and to charge them
an "overhead" account different from their

own

account.

The

usage

charged

directory
could
be
absorbed
project
administration.
This
prevents

storage
Note:

users

from

The

DIRECTORY

the

wildcard

are:

*:{*>,

dev:<*>,

the

entry

being

that

may

this

charged

for

file

users

command also accepts a
entry. The
wvalid
forms
or *:<dir>. The asterisk

with

directories

certain

by
system or
command
also

files-only directories.

form

indicates

charge

write

access

matching

the

file

creation

their

This

PS:<FORLIB>

entry

access

means
that
anyone
directory
<FORLIB>

account

files

Specifies
certain

an account is valid for
and
directory
groups

(Refer

description

for

can be used
decimal
user

5.8

groups.)

the GROUP command
combination.
(nnn is a
directory group
number.)

and

any

account

use
on

Section

Modifies

after

when

there.

that
user

structure.

placed

with
write
can
use the

CORRECT-DATA.UNIONBANK.FUND

storing

data
has

entry:

DIRECTORY

/USER:nnn
/DIRECTORY :nnn

account.

The file that contains account
account
CORRECT-DATA.UNIONBANK.FUND

for
the

any

wildcard

Example:

GROUP

the

MATH.JONES,

either

becomes

/EXPIRES:

modifier
invalid

to
for

can

indicate
wuse

be
when
the

group.

Note: Using the GROUP command is
helpful
if
many people are eligible to use
an
account.
Specifying their group number (if they are in
a

group)

names

eliminates

incorrectly.

typing

1long

1list

CREATING

6.3.2

Sample

The

examples

6-1

and

file

6-2

for

and

the

file.,

should

when

you

run

ACTGEN

the

TAKE

(FILE)

ACCOUNTS.CMD.1

go600
#0783
Bo8a0
00900
21000
#1100

This

file

1is

1lines

in Figures
shows the data

beginning

with

in
in

your

directory.

the

information
example

the file for ease in reading
particular contains all the base

ACTGEN

(Refer
command.)

Input:

#9499
@500

within

for.

@CREATE

20300

comment

used

stored

the

first

and

ACCOUNTS.CMD

looks

enter
The

can

and

giving

02100
@02@30

Tabs

(!)

ACCOUNTS.CMD.

under

6-1.

point

formatting

easier

show how you could
data files.

account

Figure

accounts

Files

below
into

exclamation
or

Data

ACCOUNTS

you

default

Section

You

name

file

that

6.3.3

for

may

find

the

file

TAKE

command

running

ACTGEN

the

<HUDSON>ACCOUNTS.CMD<RET>

!This file contains definitions of top-level
ACCOUNT CORRECT-DATA<KRET>
USER Hudson,Holland,Gerard,Gionet<RET>
USER King,Kelly,Kohn<RET>
ACCOUNT UNIONBANK/EXPIRES;1-JUN-86<RET>

accounts<RET>

USER Warriner,Bloomstran,Prest,Pendergast<RET>
ACCOUNT LP-FOOD/EXPIRES:1-JUL-86<RET>

USER Schied,Queeny,Smith<RET>
ACCOUNT TOWN-SQUARE/EXPIRES:15-JUL-86<RET>
USER Markley,Gerhard,Dole<RET>
<ESC>

*EUKRET>
<ACCOUNTS.CMD
1>
.

NOTE

Lines

300

and

400

Data.

However,

because

would

including

the

Square

charge

contain

all

you

would

enable

all

users
to

the
not

the

users

Unionbank,

the

wusers
use

CORRECT-DATA

Correct

the

asterisk

the

Food,

(*%*)

system,

and

Town

account.

Figures 6-3 and 6-4 show what
information
to
enter
1into
base
subaccount
files
for
Figure 6~2.,
Each block in Figures 6-3 and
contains information to be entered into a separate file.
Some of
blocks
contain
containing more

subaccount

information

(/SUB:)

about

entries

that

particular

point

other

CORRECT-DATA

Figure 6-4 shows which entries to make for account UNIONBANK
subaccounts
(the
bottom
half of Figure 6-2.) Note that all
Data

and

the
files

account.

Figure 6-3 shows which entries to make for
account
its subaccounts (the top half of Figure 6-2.)

account entries for both Correct
the file <HUDSON>ACCOUNTS.CMD.

and
6-4

Unionbank

are

and

and
1its
the base
contained
in

Figure 6-3:

Correct-Data Accounting Files

< HUDSON > ACCOUNTS. CMD

ACCOUNT CORRECT—DATA/SUB: < GERARD > ACCOUNT. TXT
USER Hudson, Holland

ACCOUNT CORRECT-DATA/SUB: < GIONET > ACCT. TXT
ACCOUNT CORRECT-DATA/SUB: < KING > ACCTG.TXT

< GIONET > ACCT.TXT

< KING > ACCTG.TXT

ACCOUNT OVERHEAD/EXPIRES:31-DEC-86

ACCOUNT PROGRAMMING

USER Gerard, Kelly

USER Gionet, Kohn, Kelly

USER King, Carlson

¢t-9

ACCOUNT PAYROLL

SLNNODOV

< GERARD > ACCOUNT, TXT

ONILVIYTMO

NOTES

The accounts that will be created are;

CORRECT—-DATA.PAYROLL
CORRECT—DATA.OVERHEAD

CORRECT-DATA.PROGRAMMING

Note that users Hudson and Holland can use any account number
that begins with CORRECT—DATA, but user Gerard can use only
the account CORRECT—DATA.PAYROLL. User Kelly can use the

accounts CORRECT—DATA.PAYROLL and CORRECT—DATA,
OVERHEAD.

The file type for account data files is optional. Your project

administrators can use any file type, e. g., ,TXT, .CMD or .ABC.

MR-5-3728-84

< HUDSON > ACCOUNTS, CMD

ACCOUNT CORRECT-DATA/SUB:<WARRINER>ACCOUNT.TXT
USER Hudson

<WARRINER > ACCOUNT, TXT
ACCOUNT UNIONBANK/SUB: < BLOOMSTRAN > ACCT. TXT

USER Warriner

£€T-9

ACCOUNT UNIONBANK/SUB: < PENDERGAST > MCA. TXT

|
< BLOOMSTRAN > ACCT, TXT

<PREST > ACCO.TXT

< PENDERGAST > MCA.TXT

ACCOUNT TRUST

ACCOUNT LOANS/SUB:< THOMAS > LO.TXT

ACCOUNT MSTRCHG

USER Bloomstran

USER Prest

USER Pendergast, Prest

ACCOUNT PAYROLL

ACCOUNT LOANS/SUB:< MILLS > DATA, TXT

USER Bloomstran

|
<THOMAS > LO. TXT
ACCOUNT INSTAL
USER Thomas

ACCOUNT CORP
USER Thomas, Mills

<MILLS > DATA. TXT
NOTE
The accounts that will be created are:

ACCOUNT COMM

USER Mills

CORRECT—-DATA.UNIONBANK, TRUST

CORRECT—DATA, UNIONBANK. PAYROLL
CORRECT—DATA, UNIONBANK.MSTRCHG
CORRECT—DATA. UNIONBANK.LOANS. INSTAL

CORRECT—DATA. UNIONBANK. LOANS. CORP
CORRECT—DATA. UNIONBANK, LOANS, COMM

MR-S-555-80

SLNNODOY ONILVIYD

ACCOUNT UNIONBANK/SUB: < PREST > ACCO.TXT

CREATING ACCOUNTS

the ACTGEN Program

Running

6.3.3

After you create the base account files and the project

administrator

notifies vyou that all his subaccount files are complete, you can tell
the operator to run the ACTGEN program. ACTGEN takes the accounting
information in these files and creates an account validation data
base. It is through this data base that the monitor validates all
accounts entered by the users of your system. ACTGEN is a privileged
program, so you must enable WHEEL or OPERATOR capabilities before

giving the ACTGEN command.

The command appears as follows:

(CAPABILITIES)<RET>

@ENABLE

SACTGEN<KRET>

ACTGEN>

ACTGEN

are

HELP,

EXIT,

The HELP command lists information to

assist

you

when

running

Valid commands that can be given
CTRL/A,

TAKE,

INSTALL,

and

the

program.

ACTGEN

The EXIT command terminates the program and returns you to the TOPS-20¢

command

level

($).

The TAKE command accepts as its argument either a file specification
A carriage return defaults to your connected
or a carriage return.
If you do not name your base
ACCOUNTS.CMD.
filename
the
and
directory
account file ACCOUNTS.CMD, the file you specify should be the one that
contains your base account information and points to all the existing
The TAKE command tells ACTGEN to look in the
subaccount files.
It also tells
file for account information.
default)
specified (or
ACTGEN to look at any subaccount file specifications for additional
information pertaining to the account(s) in the base account file.
Using Figures 6-1 and 6-2, the manager, Hudson, would specify the TAKE
follows:

command

@ENABLE

(CAPABILITIES)<RET>

SACTGEN<LKRET>

ACTGEN> TAKE

(COMMANDS FROM FILE)<RET>

If the manager in these examples

had

named

his

base

MACCT.TXT, he would specify the TAKE command as follows:
ACTGEN> TAKE

(COMMANDS FROM FILE)

account

file

<HUDSON>MACCT.TXT<RET>

CAUTION

base
by your
to
If the data files that are pointed
file are located on structures other than the
account
public structure, be sure the required

structures

are

the ACTGEN program will fail on
Otherwise,
mounted.
those accounts that have subaccount files on unmounted
structures.

ACTGEN takes all the information specified in the account files, forms
file
and creates a new version of the
accounts,
valid
the
Each
running.
1is
ACCOUNTS-TABLE.BIN in the directory where ACTGEN

time

file

the

ACTGEN

created.

program

is run successfully, a new version of this

CREATING

While ACTGEN
entries,

is creating

e.g.,

two

the data

accounts

ACCOUNTS

base

the

file,
same

it checks

name,

and

for

the

duplicate

length

of the
accounts.
If an error occurs, an appropriate message
1is
printed
on
the
terminal
where
ACTGEN
1is running, but the program continues to
build the data base, using only the accurate data.
You should make
a
note
of
the
error
on an error log sheet that you have prepared and
later correct the appropriate files using an editor.

ACTGEN also checks expiration dates.
If two or more expiration
dates
are
given
for the same entry in a file, the system uses the earliest
date.
For example:
1if you specify May 15,
1985
as
the
expiration
date
for account MATH and your project administrator specifies August
30, 1985 for the account MATH,LAB-201, the system will stop validating
all

accounts

You can
return
version

beginning

with

MATH

May

15,

1985.

press CTRL/A while ACTGEN is running to stop the
program
and
to ACTGEN command level.
The data files are closed and no new
of the data base file ACCOUNTS-TABLE.BIN is created from
this

session.

The INSTALL command starts account validation.
When
you
enter
this
command,
ACTGEN
copies the ACCOUNTS-TABLE.BIN file in your connected
directory
(or
the
directory
where
ACTGEN
created
the
file)
to
<SYSTEM>ACCOUNTS-TABLE.BIN on the public structure and enables account
validation using this new data base.
Because the new version of ACCOUNTS-TABLE.BIN is kept in the directory
where
ACTGEN
was
run
and not in the directory <SYSTEM>, you have a
means of correcting any errors that might occur without disturbing the
version
currently
running
in the <SYSTEM>ACCOUNTS-TABLE.BIN file on
the public structure.
You can give the INSTALL command after you have
corrected any problens.
If you do not receive any errors while ACTGEN
base
file
(ACTGEN has successfully completed
you

receive

the

ACTGEN

prompt),

you

can

give

is
creating
the
data
the accounting file and
the

INSTALL

command

immediately.

You
should
keep
track
of
which
version
of
the
<SYSTEM>ACCOUNTS-TABLE.BIN
file
you
are
using.
A
log
book that
contains the date that ACTGEN was run and the version
number
of
the
data
base
file
1is helpful should a system problem occur and you are
not sure of which data base file you were using.
To
find
out
which
version
you
are
using,
enable
capabilities and give the DIRECTORY
command for <SYSTEM>ACCOUNTS-TABLE.BIN on the public
structure.
The
generation
number
indicates
the
version that is presently running.
The system looks in the current data base file each time it
wvalidates
a

given

6.3.4

account.

Data

Base

Failures/Recovery

If your accounting files were set
up
inaccurately,
or
you
entered
random incorrect data into the data base file, account validation will
not work proverly.
You are aware of this because users cannot log
in
and/or
use accounts that are normally valid for them.
Therefore, the
account OPERATOR is set up for the user OPERATOR and is always
valid.
The
operator
can
log into <OPERATOR> with the OPERATOR account, fix
the files that are in error, and run ACTGEN to get account
wvalidation
working

again.

CREATING ACCOUNTS

6.4

VALIDATING ACCOUNTS

An account is validated when a user gives any

one

the

following

commands.

TOPS-20

the

log

system

(TO)

argument

SET ACCOUNT

the
if
SUBMIT,
PRINT,
example,
for
commands,
Any queue
account is different from the currently validated account

SET FILE ACCOUNT

File creation with an explicit account

(OF FILES)

arguments

(TO)

argument

This
accounts.
The system records the computer time used for valid
time used per structure,[l] and peripherals used
includes CPU time,
on the line printer,
per job, that is, the number of pages printed
tape records read/written, card reader usage, and disk
tape mounts,
The computer usage incurred by each account is stored in the
storage.
This file is used for reports and billing.
accounting USAGE.OUT file.
information about
File Specification for
Usage
TOPS-20
the
(Refer to
reading

and

using

this

file).

on
depends
file
How often you run ACTGEN and create a new data base
vyou expect to have
If
accounts.
your
frequently vyou change
how
frequent changes (e.g., opening and closing accounts), you may want to
Your
run ACTGEN.,
to
week
each
time
standard
a
establish
to
made
are
changes
administrators should inform the operator when
their

subaccount

files.

NOTE

been
has
file
base
Once ACTGEN is run and the data
created,
vyou
can
dump
all
the
account
files
to
magnetic tape and conserve some of
your
disk space.
You must copy the files to disk the next time you need
to

[1]

run

the

ACTGEN

program.

To account for the time used on a structure, you must
set
the
Refer to the TOPS-20 Operator's Guide
REGULATED.
as
structure
for

a description of

REGULATED and

6-16

NONREGULATED structures.

CHAPTER
SYSTEM

BACKUP

7
PROCEDURES

All the disk packs on your system must be backed up on magnetic
tape.
This procedure provides both a permanent record of the contents of the
disk and a precautionary measure in the event a disk pack
and/or
its
contents are destroyed.
On the first day of operation, start a system
backup procedure that includes:
l.

Saving

all

the

Saving

the

directory

parameters

Saving

the

front—end

file

These procedures
duties.

should

files

become

all

the

directories
and

system

part

critical

(one

time

the

all

structures

system

programs

only)

operator's

scheduled

It is important to start
backing
up
the
system
immediately
after
installation.
If
you
follow
the
backup
procedures
as
they are
outlined here and in the TOPS-2@ Operator's Guide, you can restore the
file system quickly and easily should a mishap occur.
DUMPER

The following sections discuss using the DUMPER program to save files.
Make
sure
when
you
restore
these
files
with DUMPER that you are
running the correct version of DUMPER with your
TOPS-20
monitor
and
that
the
tape
version
1is compatible with the software.
Otherwise,
directory passwords could become unusable and you may have to manually
respecify
them
with
the
BUILD
command.
Refer
to
Section 11.2,
Password Encryption, for
details.
In
addition,
project-programmer
numbers,
supported
in
TOPS-20 version 6, may not be restored at all
with incompatible versions of the software.
In

tape

CFS

configuration,

drives

are

DUMPER must

run on

the

system

which

the

attached.

simplify backup

and

restore

operations,

you

can

<create

DUMPER

command
files
for the operator.
Rather than type a list of commands
to DUMPER, the operator can then just give the
TAKE
command
with
a
command
file
name
as an argument.
DUMPER will sequentially execute
commands

contained

The TOPS-2f)
discuss the

the

file.

User Utilities Guide
and
DUMPER program in detail.

the

TOPS-20

Operator's

Guide

SYSTEM

7.1

SAVING

Have

the

ALL

FILES

operator

run

the

ALL

BACKUP

PROCEDURES

DIRECTORIES

DUMPER

program

(with

the

/FULL-INCREMENTAL

switch
to
the
SAVE
command)
to save all files in all directories.
This procedure includes saving
all
the
directories
on
the
public
structure
and
all
the
directories on any additional structures you
have created.
You can save all the files (a full dump)
or
Jjust
the
files
that
have
<changed since the last time the operator ran DUMPER
(an incremental dump).

Start a library of the magnetic
tapes
from
the
DUMPER
operations.
Each
structure
should
be
copied
to a separate tape(s).
Each tape
should be identified
with
the
system
model
number
or
name,
for
example,
2068,
or
System-A,
the
date,
the
¢type of save (full or
incremental), the name of the structure, and the tape number.
(A tape
set
name
may
replace the tape number, if labeled tapes are used.) A
typical identification may look like:
SYSTEM-A (20680)
30-JANUARY-1985
Incremental

SYSTEM-A (20640)
30-JANUARY-1985
Full

OR:

ADMIN:

Tape

ADMIN:

Tape

In addition to keeping the tapes, keep the listing of their
contents.
(The operator includes a command to DUMPER to list the contents of the
magnetic tapes on the line printer.) These DUMPER
1log
files
can
be
conveniently
kept
in
a
binder with the most recent listing on top.
Identify each binder with the system model or name, for example,
2065
or System-A.,
(Chapter 9, System Problems/Crashes describes how to use
these log files to restore directories and files.)
Tell users that backup files do exist and
post
the
times
when
the
operator
normally
creates the backup tapes.
Many system managers do
not allow users to enter the computer room to mount and use the
tapes
themselves.
NOTE

The DUMPER program DOES NOT
save
console
front—-end
file
system.
files, you must restore them from
(Refer to Section 7.6)

7.1.1

Full

the
If
the

files
1in
the
you lose these
floppy
disks.

Dumps

Full dumps contain
all
the
information
on
the
system,
with
the
exception
of
the console front—-end files, and can be used to restore
many of the files that were on the system
to. their
previous
state.
Therefore,
full dumps contain a copy of every file in every directory
on

every

structure.

NOTE

Full dumps are known as FULL-INCREMENTAL dumps.
name
corresponds to the /FULL-INCREMENTAL switch
you give with the SAVE command to DUMPER.

This
that

SYSTEM

7.1.2

Incremental

Incremental
command)

PROCEDURES

Dumps

dumps

cause

BACKUP

(using

DUMPER

the

/INCREMENTAL

save

the

files

switch

that

have

with

the

SAVE

never

been

saved

(new files) and the files that have been updated since the
1last
time
an
incremental DUMPER operation was performed.
Many managers request
an incremental dump Monday through Thursday and a full dump on Friday.
The

File

Descriptor

necessary

saved
since

Block

determine

each

file

contains

the

file

has

been

updated

since

information
it

was

last

during an incremental DUMPER operation.,
A file that has changed
the last time it was saved is automatically saved again on tape;

otherwise,

passed

over.

The operator should give
each
structure
one
at
individually,

the

have

been

the INCREMENTAL switch to DUMPER and
specify
a time.
By running DUMPER for each structure

operator

tape.
After
running
presently on—-line, the

that

(FDB)

used

can

copy

DUMPER
operator

that

day

each

structure

and
copying
should mount

and

run

DUMPER

onto

separate

the
structures that are
any additional structures

for

them

also.

An incremental dump is faster than
a
full
dump
and
requires
1less
magnetic
tape.
By
specifying a value with the /INCREMENTAL switch,
you
can
cause
modified
files
to
be
written
to
more
than
one
incremental
backup
tape.
This is helpful if you want to be certain
you

can

7.1.3
It

recover

the

Security of

good

idea

files,

Backup
to

even

one

the

tapes

has

data

errors.

Tapes

protect

the

security of your
backup
to mount them, he or
such as other user's

you
allow
a
non-privileged
wuser
access to confidential information,

tapes.

she may gain
data
files.

If
the
unencrypted
passwords
were saved along with other directory
parameters, a technically sophisticated user can
retrieve
them
from
the
DUMPER
backup
tapes, and thus obtain unauthorized access to the

privileged

7.2
A

COMMON

common

backup
1.

accounts

BACKUP

backup

POLICY

policy

Each day, take
changed
from

incremental

can
the

the
time
3.

system.

outlined

below.

You

can

set

your

own

policy.

you
2.

the

saves

recycle
end

system.
you can

Every six
years,

an incremental
the
previous

until

the

full

incremental

week,

Keep
recycle

save of
the
day's
backup

the
the

take

full
tapes

save

files
tape.

is made,

that
have
Keep the

which

time

tapes.
full

save

all

the

files

saves for six months, at
into the backup system.

months, take a full save and keep
or if you choose, indefinitely.

for

which

number

SYSTEM BACKUP PROCEDURES

7.3

MAGNETIC TAPE REQUIREMENTS

You need a supply of magnetic tapes to start a system of backup
tapes
This section provides a guideline for the number
procedure.
It is assumed that
you should have on hand for your installation.
there are 2400 feet per reel of tape.

- Reels Per
Drive

Type of
Disk Drive

Bits Per
Inch

RPO6

l608

RA60

1600

RP@4

1600

RP2@ (one

1600

RP2@ (both

1600

RP2# (one

6250

RP2@ (both

6250

RP@7

6250

RPQ7
RA81
RA81

19
6
19

1600
6250
1600

spindle)

spindles)

spindle)

spindles)

Therefore, the number of tapes you stock depends on the type of edisks
It also depends on the backup procedur you
at your installation.
your dally incremental tape dumps for a
save
you
if
example,
use. For
longer time than wusual, it takes longer to recycle these tapes into
the backup system, and thus you need more tapes.

Generally, during the first month after installation, you may hneed
approximately 36 (2408 ft.) tapes for each RP@6 or RA6Q, 24 tapes for
each RP@4, 45 tapes (6250 bit/in) or 180 tapes (1600 bit/in) for each
RP2@ disk (2 spindles), and 72 tapes for each RP#7 or RA8l (1600
bit/in).

NOTE

These estimates assume a magnetic tape blocking factor
You can specify a higher blocking factor and
of 1.
Before doing this,
save much space on your tapes.
however, there are cautions that you must consider.

User
TOPS-20
in the
The description of DUMPER
Utilities Guide explains how and when you can increase
blocking

factors.

SYSTEM

7.4

MAKING

SYSTEM

CRASH

BACKUP

PROCEDURES

TAPE

As the name implies, the system crash tape is used
to
re-create
the
public
structure
should it become unusable.
This tape is created in
addition to the tapes that contain
FULL-INCREMENTAL
and
INCREMENTAL
saves
of
all
files
and
directories.
You should make a new system
crash
tape
whenever
you
add
a
new
user,
change
any
directory
parameters, or make a change (patch) to:
e

The monitor

you

are

The

TOPS-20

Command

The

DLUSER

The

DUMPER program

running
Processor

program

Therefore, the crash tape contains only the files necessary to recover
user
directory
parameters and important system programs,
User files
themselves are restored
from
the
FULL-INCREMENTAL
and
INCREMENTAL
saves

the

public

structure.

Label this tape SYSTEM BACKUP TAPE and include
the
public
structure
name;
DECSYSTEM-20
model
number
or
name,
for
example,
2060
or
System-B; and the date and time the
tape
was
created.
You
should
follow
this procedure once a day if users are allowed to change their
own directory parameters.
(Refer to
Section
5.7.2
for
information
about
allowing
users to change directory parameters.) If you are not
using password encryption (refer
to
Section
11.2),
be
careful
to
protect
the
backup
tapes
against
reading
by
unauthorized users,
because all the passwords for your users will be accessible.

If you are also using mountable structures,
you
should
periodically
run
DLUSER
to
<copy
their directory parameters to a file on another
structure, preferably the public structure.
Then,
if
the
mountable
structure is destroyed, you will be able to recreate the directories.
NOTE

Do not use a labeled tape when creating a System Crash
Tape.
The
reason
for this is that the installation
software that is used to create the
public
structure
cannot

The order of

read

tape

labels.

files on the crash

<SYSTEM>MONITR.
EXE

<SYSTEM>EXEC,EXE

<SYSTEM>DLUSER.EXE

DLUSER data

<SUBSYS>DUMPER.EXE

files

tape

is:

SYSTEM

BACKUP

PROCEDURES

DUMPER save sets containing the directories:
<SYSTEM>
<SUBSYS>
<NEW-SYSTEM>

<NEW-SUBSYS>
<UETP>

<GALAXY-SUBSYS>

Notice that the format of
this
Installation Tape that you used

7.5

MAKING A CRASH TAPE

USING

tape
is
the
same as
the
TOPS-20
to install the TOPS-20 software.

BATCH

You can create a batch job
to
make
your
<crash
tape
or
type
the
commands
at
the
operator's
terminal.
Example 1 shows the standard
control file that you can submit as a batch job to create
this
tape.

In the example, PUB:

EXAMPLE

1is the name of the public structure.

SYSTAP Control

@TYPE(FILE)

File

SYS:SYSTAP.CTL<RET>

! Obtain a tape drive
@MOUNT TAPE TAPE:/WRITE/LABEL:UNLABELED

I1Systems not using Tape Drive Allocation must replace the
IMOUNT TAPE command with @ASSIGN MTA@: and @DEFINE TAPE:
! (AS)

MTA@:

@ENABLE
@REWIND

commands.

(CAPABILITIES)
(DEVICE) TAPE:

the monitor

!Save

QGET (PROGRAM) PUB:<SYSTEM>MONITR.EXE
@SAVE (ON FILE) TAPE:

ISave the TOPS-2¢ Command Language
@GET (PROGRAM) SYSTEM:EXEC.EXE
@SAVE (ON FILE) TAPE:

Interpreter

{Save the DLUSER program
@GET (PROGRAM) SYS:DLUSER.EXE
@SAVE (ON FILE) TAPE:

!Run

the

lon

tape

same DLUSER program,

@START
*DUMP
*EXIT

(TO

FILE)

!Save

DUMPER

TAPE:

@GET (PROGRAM) SYS:DUMPER.EXE
@SAVE (ON FILE) TAPE:

saving

the directory structure

SYSTEM

!Run .the

same

DUMPER,

saving

BACKUP

SYSTEM:

PROCEDURES

and

SYS:

@START

*TAPE (DEVICE) TAPE:
*LIST (LOG INFORMATION ON FILE) SYSTAP.LPT
*SSNAME SYSTEM-FILES
*SAVE (DISK FILES) PUB:<NEW-SYSTEM>,PUB:<SYSTEM>
*SSNAME SUBSYS-FILES
*SAVE (DISK FILES) PUB:<NEW-SUBSYS>,PUB:<SUBSYS>
*EXIT

!Print

the

@PRINT

SYSTAP.LPT/NOTE:BACKUP

@DISMOUNT
@
!Systems

DUMPER

log

file

TAPE:

not

using

Tape

Drive

!DISMOUNT TAPE: command with
I@DEASSIGN TAPE: commands.
To

run

TAPE

SYSTAP,

submit

the

Allocation

@UNLOAD

batch

must

replace

(DEVICE)

control

file

TAPE:

using

the

and

the

TOPS-20

SUBMIT

command.

the

now

the event the public structure becomes unusable,
be
used
by following the instructions in the
Installation Guide.

TOPS-2¢ KL

crash

tape can
Model B

HINT:

Before you store
made
a
usable

the crash tape, verify that
tape.
That
1is, mount the

tape, follow the
B
Installation

instructions in
Guide
to
1load

DUMPER

7.6
The
file

SAVING
DUMPER

THE

get

CONSOLE

program

system.

You

listing

FRONT-END

does

not

can,

however,

save

tape's

FILE

SYSTEM

the

contents
a

have
crash

the TOPS-2¢ KL
Model
the monitor, and use

the

make

you
new

contents.

backup

the

copy

console

the

prints

operator's

console,

type

the

front-end

system
by copying your floppy disks using
the
front-end
program
COP
(for
copy) .
You
should
make
at
1least
one backup copy of your console
front—-end file system (refer to Section 3.4).
To run the COP program, follow the steps outlined
stop timesharing when following these steps.

below.

CTRL/backslash;

file

You

the

need

not

system

PAR>.

CTRL/backslash
PAR>

Type
COP

MCR

PAR>MCR

cop>

COP

and

prompt.
COP<KRET>

press

the

RETURN

key;

the

system

prints

the

SYSTEM BACKUP PROCEDURES

Place the floppy disk to be copied in drive # and the floppy
to contain the new files in drive 1. Be sure to mount the

floppy disks correctly; this includes checking that the paper

containing

the floppy directory is not accidentally attached

to the back of the floppy disk.

Type DX1:=DX@:

and press the RETURN key; the

system

starts

the copying, which takes a few minutes. After the copying is
complete, the system verifies the copy and prints a messagde.
Type CTRL/Z

to exit COP.

COP>DX1:=DX@:<RET>

COP - STARTING VERIFY
COP>"Z

To return to TOPS-28 Command Level, type a CTRL/backslash;
the system prints PAR>; type another CTRL/Z2, or type QUIT and
press the RETURN key.

CTRL/backslash

PAR>"Z

CHAPTER
TAPE

STORAGE

Chapters 1 through 6 deal primarily with setting
up
and
wusing
your
disk
resources.
Chapter 7, System Backup, describes how to save all
the data from your disk structures onto magnetic
tape.
These
tapes
are
the
backup tapes that you use to restore directories and perhaps
entire disk structures if something happens to
the
disks
(refer
to
Chapter
9).
In
addition
to using magnetic tapes for system backup
tapes, you can use magnetic tapes to store other
types
of
data
and
save valuable disk space.
This chapter
describes
two
other
wuses
for
magnetic
tapes,
File
Archiving and File Migration.
It also describes how you can allow the
system and the operator to control all tape
drive
assignments,
Tape
Drive
Allocation, and how you can set up some or all of your tapes to
contain labels, Tape
Labeling.
These
uses
are
described
1in
the
following order.
e

FILE

ARCHIVING

FILE

MIGRATION

TAPE

DRIVE

TAPE

LABELING

ALLOCATION

In addition, the last section of the chapter discusses
two DECSYSTEM-2@s to share a TX@2 tape subsystem.

File
way

archiving
to
move

provides you and
important
files

users
from

how

set

of the system with
a
voluntary
the
disk
to magnetic tape for

long—-term storage.
These tapes are stored separately from your system
backup
tapes.
Users
can
access these tape files as easily as they
access files on the disk.
When users want to restore
archived
files
to
disk,
they
give a command to the TOPS-20 command processor.
The
system then tells the operator which tapes to mount
and
proceeds
to

restore
the
files.
Section 8.1 describes why you would use the file
archiving facility, and how to set up your system to archive files
to
magnetic tape.

File migration provides you with a means of
controlling
the
use
of
disk space.
File migration is especially useful if your disk space is
very low on a particular structure, for example, the public structure.
This
type of disk space control is, for the most part, involuntary on
the

part of the user.
01d unused disk files
are
periodically
moved
(migrated)
to magnetic tape by the system operator.
Again, you should
store these tapes separately from your
system
backup
tapes.
Users
still
maintain
easy access to these files and retrieve them the same
way as they retrieve archived files.
Section 8.2
describes
why
and
when
you would use the file migration facility and how to set up your
system to migrate files to magnetic tape.

8-1

TAPE

If you use the file
remember that these

archiving
tapes are

They are not replacements.
program and create full and

or
in

STORAGE

file migration facilities,
or
addition to your system backup

You
must
incremental

continue
to
system backup

run
the
tapes.

Dboth,
tapes.
DUMPER

Tape drive allocation provides the system and computer
operator
with
complete
control
over tape drive usage.
This means that it prevents
users from issuing the ASSIGN command and reserving
tape
drives
for
their
jobs.
When
users
issue
the MOUNT command, the TOPS-20 Tape
Drive Allocation system and the operator
control
the
allocation
of
tape
drives.
You must use the tape drive allocation facility if you
use tape labeling; however,
using
tape
drive
allocation
does
not
restrict you to using labeled tapes.

Tape
tape

labeling
itself

provides a means of storing
1label
information
on
that
1identifies
the tape and describes the data on

the
the

tape.
This label information is in an industry standard format so you
can
read
and
write tapes to be used with different computers.
Tape
labels can also add more security and reliability to your tape system.,
Section
8.4
describes why you would use tape labels, and also how to
set up your system to begin labeling tapes.
There are no dependencies among file archiving,
file
migration,
and
tape
1labeling.
For example, you can use the file archiving facility
without using file migration or tape labeling.
Each tape facility can
be used separately.
There is, however, the dependency that tape drive
allocation

8.1

FILE

must

turned

use

tape

labels.

ARCHIVING

File archiving provides a means of storing data on magnetic
tape
and
freeing
wvaluable
disk
space.
This type of off-line storage allows
users to store (archive)
important files on tape, keeping
their
disk
space
below their permanent allocation, and still have easy access to
those files.
:
If

your

installation

has

than

one

computer

system,

the

directory

Restore

<ROOT-DIRECTORY>

Restore

the

entire

file

the

these

file

arises,

From

are:

nature

address

system

you

the

topics.

can

1least

seldom
perform

most

directory
(other

than

<ROOT-DIRECTORY>)

system

The

TOPS-20 Operator's Guide provides all
the
necessary
information
for
the
operator
to
correct these types of problems.
Sections 9.1
through 9.4 provide you with an overvi
ew of
how
these
problems
are
solved,

9.1

RESTORING A SINGLE

FILE

If you receive a request
following procedure.
l.

Look in the
log files.

Write
and

file
to

©Submit

file

and

the

file,

sure

structure

the

directory

and

operator

you

(including

restored,
and

user,

the listing
of
creating DUMPER

specification

structure

request

for

different from the
file was saved.
3.

that contains
(Chapter 7 describes

containing

destination

restore

binder

down the
directory)

tape

date

directory

restore

use

structure

number

the

both

are

indicate

one

from

the

the
DUMPER
log files.)

the

and

can

which

file,

the

SYSTEM PROBLEMS/CRASHES

9.2

RESTORING A SINGLE DIRECTORY

the
If you receive a request to restore a directory, you can use
following procedure.

Determine the structure that contains the directory.

in this directory
Make sure you have a copy of the files
)

on a

down the 1list of param
number, allocation, etc. You may need this information

when

DUMPER tape.

(Check the log files.

with the LIST subcommand.direcWrite
Give the "“ECREATE command eters
tory
, for example, the
you re-create the directory.

subcommand for the
Give the “ECREATE command with the KILL
tor's Guide provides additional
directory. (The TOPS-20 Opera
e directory 1if the “ECREATE
singl
a
ing
procedures for delet
mmand is unsuccessful.)
command with the KILL subco

, first restore the files from
Using your DUMPER backup tapes
R operation. Then, restore
DUMPE
L
MENTA
the last FULL-INCRE
the time when the
files from each INCREMENTAL tapethe until
operator gives the CREATE
files

were

lost.

sure

.
command to DUMPER to restore the directory parameters

1is not
roducible information and ittance
If the directory contains unrep
. It
assis
for
ces
Servi
are
Softw
al
backed up on tape, call Digitt the directory without losing the valid
may be possible to reconstruc
information in it.

9.3

RESTORING <ROOT-DIRECTORY>

-DIRECTORY> and a backuryp
Each structure has its is ownused <ROOT
by the system Iif the prima
<ROOT-DIRECTORY> that The directory
<ROOT-DIRECTORY> contains asa
<ROOT-DIRECTORY> is bad.
as well
l directory on a structure,trate
pointer to each first-levefiles
how
illus
5
ter
(Chap
.
several important system directories.) If <ROOT-DIRECTORY> sis lost
to
s
point
>
<ROOT-DIRECTORY
t access any files in the system.t
on the public structure, users canno
mountable structure, users canno
a
on
lost
1is
>
1f <ROOT-DIRECTORY
access files on that structure.

is bad
CTORY> on the public structureliste
You can tell that the <ROOT-DIRE
d in
Ts
BUGHL
the
of
one
any
s
print
le
conso
if the operator's
A
.
stops
m
syste
the
le,
conso
the
Table 9-1. When a BUGHLT appears on
BUGHLT appears in the form:
khkkkkkkkkk

*BUGHLT name AT dd-mm-yy hh:mm:ss
*JOB:n,

USER:

user—name

*ADDITIONAL DATA:
Ahkkkkkhkkk

data,data,data

The lines beginning with JOB:

or ADDITIONAL DATA:

may nhot appear.

SYSTEM

Table

9-1:

<ROOT-DIRECTORY>

PROBLEMS/CRASHES

BUGHLTS

BUGHLT

Meaning

BADROT

<ROOT-DIRECTORY>

FILIRD

The

not

FILMAP

system

The

could

system

invalid

could

initialize
not

map

<ROOT-DIRECTORY>

into

memory

BADXT1

The

<ROOT-DIRECTORY>
FND.
If
this

index

table

may also be bad
error
appears,

is missing

and

cannot

created

if you get the
BOOT
error
?2FIL NOT
be sure you have mounted all devices

correctly.

recover

structure

from

on
a
mountable
ROOT-DIRECTORY
Operator's Guide

and

contains

using

bad

<ROOT-DIRECTORY>,

the bad

directory.

first

the

bad

determine

which

<ROOT-DIRECTORY>

structure,
you
c¢an
run
CHECKD
with
RECONSTRUCT
and
specify
the
proper
structure.
The
TOPS-20
details the procedure for determining
the
structure

CHECKD

for

reconstructing

<ROOT-DIRECTORY>

on mountable

structures,

If the bad <ROOT--DIRECTORY>
instruct
the
system
to
<ROOT-DIRECTORY> and rebuild
this procedure.

1is

on
the
public
structure,
use
the
backup
public
this directory.
Section 9.3.1

you
can
structure
describes

NOTE

If your
first
attempt
fails, call your DIGITAL
NEVER

try

rebuild

the

Public

to
rebuild
<ROOT-DIRECTORY>
Field Service Representative,

this

directory

twice

any

structure.

9.3.1

Rebuilding

Structure <ROOT-DIRECTORY>

To rebuild <ROOT-DIRECTORY> on the public structure, halt the
central
processor
and
perform Steps 7 through 21 in Chapter 2 of the TOPS-20
KL1@ Model B Installation
Guide.
The
steps
are
shown
below
for
reference.

Type CTRL/backslash
prints PAR>.

Type SHUTDOWN and press
few message lines.

the

PAR>SHUTDOWN<KRET>

**HALTED**
¥DECSYSTEM-20

NOT

RUNNING

operator's

console;

RETURN

the

key;

system

the

system

prints

SYSTEM

PROBLEMS/CRASHES

Mount

System

Floppy A

in drive

(Step

Mount

System

Floppy B

in drive

(Step 8).

Mount

the

TOPS-2#

If the TOPS-20
system
backup
contains

DLUSER

the

data,

Installation

the

Tape on MTAQ:

Installation Tape
1s
not
tape,
mount
your
SYSTEM
monitor,

TOPS-20

DUMPER,

<SUBSYS>.
(Refer to
system directories.)

Place

7).

front—-end

and

Section

Command

(Step 9).

your
most
recent
BACKUP
TAPE that
Processor,

DLUSER,

save sets containing <SYSTEM>
3.2

HALT switch

for

and

a description of special

the

ENABLE position

(Step

000087

(Step

19).

Set the front-end switch register
11).

(octal)

Press the ENABLE and SWITCH REGISTER switches simultaneously
(Step

12).

RSX-2¢F

VB15-21

8:55

1-JAN-85

[SY®:
REDIRECTED TO DX@:]
[DX@#:
MOUNTED]
[DX1:
MOUNTED]
KLI -- VERSION VB15-15 RUNNING

KLI

ENTER

DIALOG

[NO,YES,EXIT,BOOT]?

KLI>
NOTE

The version and edit numbers in
this
manual
may
differ
from the numbers printed on your
console.
The numbers on your console must be
equal
to or greater than the numbers:
in this
manual.

Type

YES

and

press

the

RETURN

key

(Step

13).

KLI>YES<RET>

KLI

KL1¢

S/N:

KLI

KL10

HARDWARE

2136.,

MODEL

HERTZ

ENVIRONMENT

MOS

MASTER OSCILLATOR
EXTENDED ADDRESSING
INTERNAL

CHANNELS

CACHE

KLI
KLI>

RELOAD MICROCODE

[YES,VERIFY,FIX,NO]?

SYSTEM

1a.

Type

YES

KLX

press

the

RETURN

key

(Step

14).

KLX<RET>

KLI>YES

KLI

and

PROBLEMS/CRASHES

MICROCODE

VERSION

400

LOADED
NOTE

If
your
system
has
cache
memory,
following
question
appears
previous to
question in Step 11.
(Step 15.)
KLI

Type

RECONFIGURE

ALL

and

press

CACHE

the

the
the

(FILE,ALL,YES,NO)?

RETURN

key

(Step

16).

key

(Step

17).

KLI>ALL<RET>
KLI

11.

Type

KLI

ALL

~--

CACHES

and

ENABLED

press

CONFIGURE

the

MEMORY

RETURN

CFILEsALL >REVERSE»FORCE,YESyNOI7?

KLI>ALL<RET>
LLOGICAL

MEMORY

CONFIGURATION
CONTROLLER

ADDRESS
00000000
KLI -~ LOAR

SIZE
RQO
RQ1
RQ2
256K
00
01
00
EKOOTSTRAF CYES,NO,FILENAMEJ?

RQ3
01

CONTYFE
MB20

INT
4

12,

Type

MTBOOT and

press

KLI>MTBOOT<LRET>
-- CONFIGURATION

KLI

the

RETURN

FILE

ALTERED

KLI

-- WRITE CONFIGURATION FILE

KLI

(Steps

and

19).

[YES,NO]?

BOOTSTRAP

BOOT

key

LOADED

AND

STARTED

V11.0(306)

MTBOOT>

13.

Type

MTBOOT>
[BOOT:
[BOOT:

and

press

the

RETURN

key

(Step

20).

/L<RET>
STARTING CHN:n DX2@x:d MICROCODE
LOADING RESIDENT MONITOR] [OK]

Vn(n)] [OK]

MTBOOT>
NOTE

The message concerning the DX2f microcode
1is
printed
only
if
you
are
installing
the
TOPS-20 software on
a
DECSYSTEM-2@
with
a
DX20

tape

disk

controller.

SYSTEM

14,

Type /G143 and

PROBLEMS/CRASHES

press

the

RETURN key

(Step 21).

MTBOOT> /Gl143<RET>
[FOR ADDITIONAL INFORMATION TYPE "?" TO ANY OF THE
FOLLOWING QUESTIONS.]
DO YOU WANT TO REPLACE THE FILE SYSTEM ON THE SYSTEM
STRUCTURE?
NOTE

Read

Step

carefully before

answering

this

question.
15.

Type N and

all

the

all

the

press

the

RETURN key.

You DO NOT

information on the disk packs.
information

YOU WANT

THE

SYSTEM

REPLACE

STRUCTURE?

the

file

system,

THE

FILE

SYSTEM

l6.

always type

clear

retain

PASS1] [OK]

ROOT-DIRECTORY?

Type Y and press the RETURN key.
of <ROOT-DIRECTORY> to be used.
RECONSTRUCT

N<RET>

[PS MOUNTED]
[BOOT:
LOADING SWAPPABLE MONITOR,
RECONSTRUCT

want

If you want to

ROOT-DIRECTORY?

[RECONSTRUCTION

PHASE

This causes

the

backup copy

‘

Y<RET>

COMPLETED]

SETSPD

$¥NO

System
ENTER

restarting,
CURRENT

DATE

wait...
AND

TIME:

The system restarts and runs CHECKD to
reconstruct
the
bit
table.
The bit table contains one bit for every page in the
file system.
If the bit is on, the page is available; if the
bit is off, the page is in use.
‘
17.

Type the date and time and press the RETURN key.
Type
press the RETURN key to confirm the date and time.
ENTER

CURRENT

DATE

AND TIME:

1-JAN-81

@931<RET>

YOU HAVE ENTERED THURSDAY, @1-JANUARY-1981
IS THIS CORRECT (Y,N) Y<RET>
WHY

RELOAD?

9:31AM,
|

and

SYSTEM

18.

Type

OTHER

and

WHY RELOAD?
[REBUILDING

PROBLEMS/CRASHES

press

the

RETURN

key.

OTHER<RET>
BIT TABLE]
NOTE

You

should

type

response

WHY

RELOAD?,

that
reminds
you
of
why
vyou
did
this
procedure.
This response is
stored
in
the
<SYSTEM-ERROR>ERROR.SYS
file.
Refer to the
TOPS-20 KL1@¢ Model B Installation Guide for a
complete list of valid abbreviations.
19.

The system prints some standard
messages,
the
CHECKD, RUNNING DDMP, and the output from SYSJOB
[{REBUILDING

BIT

[WORKING

STRUCTURE

output

from

RUNNING

output

TABLE]

PS:]

CHECKD

DDMP

from

SYSJOB

and

PTYCON

Refer to the TOPS-2f Operator's
Guide
output from CHECKD, SYSJOB and PTYCON.
20.

Log

TOPS-2@

@LOGIN
Job

9.4

you

may

still

have

user

TTYl

TOPS-2@

OPERATOR

3-MAY-85

ENTIRE

FILE

receiving

for

samples

the

OPERATOR.

SYSTEM,

(USER)

RESTORING THE

you are

output
from
and PTYCON.

restore

Monitor

6.1 (6701)

(PASSWORD)
1@:33:32

——-

(ACCOUNT)

OPERATOR<KRET>

SYSTEM

random

the

errors

entire

and

cannot

file

use

system

the

system,

the public

structure.
Before
doing
this,
you
should
contact
your
software
specialist
to
ensure
that resorting to this procedure is necessary.
The procedure requires shutting down the system and
reinstalling
the
file

system.

9.4,1

Re-creating

The following
system on the
l.

Type

the

steps outline
the
public structure.
CTRL/backslash

CTRL/backslash
PAR>

File System on the

Public Structure

procedure

start

the

for

restoring

front—end

command

the

file

parser.

SYSTEM PROBLEMS/CRASHES

Type SHUTDOWN to
stop
prints a few messages.

the

<central

processor;
'

TOPS-20

the

system

PAR> SHUTDOWN<LRET>

**HALTED**
$DECSYSTEM-20

Start at

NOT

Step 9

RUNNING

Chapter

the

Model

Installation Guide and follow all the steps through Step 60.

In Step 9, instead of mounting the TOPS-2¢ Installation Tape,
mount
your public structure SYSTEM BACKUP TAPE that contains
the monitor, TOPS-20 Command Processor, DLUSER, DLUSER
data,
DUMPER, and save sets containing <SYSTEM> and <SUBSYS>.
4,

After performing Step 60, restore
your
entire
file
system
using
DUMPER.
First run DUMPER, then mount your first reel
of the most recent full DUMPER tape and follow the procedures
in the TOPS-2@ Operator's Guide.

Re-—create
the
front-end
directions
in
Chapter
Installation Guide.

Finally, restart the system by following
the
directions
Chapter 5 of the TOPS-2@¢ KL Model B Installation Guide.

file
system
by
following
the
4
of
the
TOPS-2¢
KL
Model
B

NOTE

Restore the
incremental
saves
to
obtain
the
most
recently
saved
files.
Before
restoring
each
incremental tape, type DUMPER>CREATE
to
restore
all
the
directories that were created since the last time
you

9.4.2

ran

the

Re-creating

DLUSER

program.

Mountable

Structures

If, after your efforts to correct errors on a structure other than the
public
structure
(using
the
command
RECONSTRUCT ROOT-DIRECTORY to
CHECKD), you still cannot use the structure, you
can re-create
that
structure
and restore all the directories and files.
You can restore
structures other than the public structure during timesharing.
To

restore

mountable

structure

you must:

Give the SET STRUCTURE IGNORED command to the OPR program
to
prevent other users from mounting the structure while you are
re-creating it.

Run

Run DLUSER to restore directory parameters
if you previously used the program to save

for
the

Run DUMPER using the backup tapes for this
the
CREATE and RESTORE commands to DUMPER
directories and files.

structure.
Give
to restore all the

The

TOPS-2@

and

DUMPER

CHECKD

create

the

structure.

Operator's Guide details
to

re-create

structure.

the

procedures

for

the
structure
parameters.

running

CHECKD

SYSTEM

PROBLEMS/CRASHES

POWER FAILURES

9.5

Unfortunately,

power

installations.

failures

and

brown-outs

You should be aware of the

sometimes

occur

immediate steps to perform

and transmit this information to your operations people.
The kind
of
attention
vyou
should
direct
to this type of problem depends on the
type of outage you have.
These steps can
protect
your
system
from
physical damage and perhaps unnecessary loss of files.
If

your

system

experiences

total

Immediately power-off

Inform

your

may ask

you expect

all

DIGITAL Field

resume

power

failure,

components
Service

power.

The

be present while

you should:
the

system.

representative

field

you bring

when

service representative

your

system

back

up.

If your system experiences a short instance
of
a
power
failure
or
brown-out,
the
system
may recover on its own.
You may not have any
problems and, usually, all your data remains intact.
If you notice
a
problem, call your Field Service representative.,
High-performance computer systems are sensitive to the quality of
the
electrical
power
supply.
An investment in a power conditioner or an
uninterruptible power supply may more than repay
itself
in
improved
system
reliability
and
availability.
Your
Field
Service
representative may be able to assist you in evaluating
the
need
for
such

equipment.

9.6

REMOTE

DIAGNOSTIC

LINK

(KLINIK)

You may occasionally have
a
problem
with
your
system
and
cannot
determine
the
cause.
The
remote diagnostic link, available on all
systems, allows a DIGITAL Field Service engineer to access your system
from
a
remote lccation and run diagnostic programs.
This capability
is called KLINIK.
The DIGITAL
engineer
accesses
KLINIK
through
a
terminal
and
telephone
1line
at
the
DIGITAL
Service Center.
The
TOPS-20 Operator's Guide describes when and
how
to
use
the
KLINIK
capability.

9.7

MAKING THE CI

UNAVAILABLE

ON NON-CFS SYSTEMS

The CI, a computer—-interconnect bus, is a key piece
of
hardware
for
connecting
systems
in
a
CFS-20
configuration
and
for connecting
HSC5@-based disks (RA8ls and RA6@s) to a system.
Ordinarily, you need
do
nothing
at
all
to
operate
the
CI,
However, you may need to
disengage

problems

instruct

PORT

system

with

the

from

operator

the

CI20

to make

UNAVAILABLE command.

for

details.)

When

the

1is

back

the

wunavailable

Field

Service

HSC5@0.

personnel

At those

can

times,

unavailable by means of

(Refer

¢to

system,

correct

you should
the

users

cannot

access

HSC5@-based
disks,
which
rely
on
the
CI
to
transmit data.
procedure calls for the operator to dismount any structures
that
system indicates are mounted on these disk drives.
To

put

the

in operation,

OPR>SET PORT CI AVAILABLE

SET

the TOPS-2# Operator's Guide

the operator gives

the command:

The
the

SYSTEM

Structures

that

were

affected

Refer to Chapter 12,
disengaging the CI on

9.8
A

MAKING

system

may

THE

need

THE
CFS

PROBLEMS/CRASHES

must

then

COMMON
systems.

FILE

This

command

servers

PORT NI

TCP/IP software is
between
systems.

the

SYSTEM,

for

information

disengaged

from

the

with the
operator

NI or
gives

that

the
the

users

any

other

the

system

software

that

from
uses

using
the

NI.

installed, it cannot use the NI for
To make the NI available again, the

command:

OPR>SET PORT NI

field

NIA2@.
To
command:

service

make

the

UNAVAILABLE

prevents
well

remounted.

UNAVAILABLE

personnel can diagnose problems
NI unavailable to a system, the

OPR>SET

AVAILABLE

9-10

LAT
If

terminal
DECnet

data
transfer
operator gives

CHAPTER
SYSTEM

PERFORMANCE

The configurations of systems running TOPS-20 and the mix of
jobs
on
these
systems
vary
from
installation
to installation.
TOPS-20 is
designed to provide better
response
to
interactive
users
than
to
provide
higher
throughput to computational tasks.
On systems with a
typical mix of interactive and batch jobs, this design
provides
both
good
response and adequate throughput.
Therefore, if your system has
many timesharing users and an average amount of batch processing jobs,
you
will
find that your system's response is good and most users are
satisfied.
If

you

have

system,

you

mix
may

jobs

want

configuration

change

the

response

different
of

your

from

system

typical

provide

better

service to specific users.
TOPS—-20
provides
several
tuning
that
allow you to experiment with and change the behavior
of your system.
One mechanism allows you to favor classes of users by
allocating
each
<c¢lass
a
specified
percentage of the CPU.
Another
mechanism allows you to
favor
computational
Jjobs
over
interactive
jobs.
A third mechanism allows you to disable features that normally
provide better performance, but that may not
be
applicable
at
vyour
installation.
mechanisms

This

chapter

describes

the

mechanisms

for

tuning

your

system's

response
and provides guidelines for using these mechanisms.
Because
the response of your system can be felt during actual
use
only,
vyou
can

expect

system

analyzing

the

inputs

from

tuning

statistics

your

wusers,

from

you

experimental

the

can

WATCH

and

program

determine

the

iterative
and

best

way

process.
gathering

tune

your

system.

The

tuning
e

mechanisms

The class
processor

scheduler, which allows you
(CPU)
resource among groups

Assigning

low

do
e

include:

not

Bias

use

control,

compute~bound
e

The

program

frequently

Each

there

priority

the

class
which

jobs

name
used

Reinitializing
improving file

batch

divide
users

for

the

central

installations

that

scheduler
allows

your

cache

for

you

favor

either

interactive

system
improving

the

startup

time

programs

disk packs
processing

in
time

mechanism

jobs

to
of

heavily

can be used
independently.
interrelationship among them.

19-1

used

Unless

structures

otherwise

for

noted,

SYSTEM

14.1

THE

CLASS

PERFORMANCE

SCHEDULER

Occasionally, certain jobs monopolize
the
central
processor's
time
while
other
more
critical
jobs
wait for the CPU.
You may want to
control the amount of CPU time a job receives.
You can use the
class
scheduler
to
provide
an even distribution of CPU time or to provide
more CPU time to critical jobs on the system.
Some
system
managers
may
want
to
allocate
a
1larger amount of processor time to special
users than to other system
users.
Sections
1¢.1.1
through
10.1.9
describe:
e

overview of

who

should

how

turning

changing

use

begin
on

the

class

scheduler

the

class

scheduler

using

the

class

the

class

scheduler

percentages

® disablihg the class scheduler

10.1.1
The

getting

sample

information
session

alternative

about

using

class

using

the

scheduler

scheduler
class

status

commands

scheduler

with

accounts.

Overview

class

scheduler

allows

you

allocate

percentages

the

processor's
time to individual classes of users.
Each job
receives a portion of the class percentage.,
Therefore, by
class
scheduler,
you
can provide a consistent service to
groups

central

in a class
using
the
predefined

users,

The diagram below illustrates the concept of classes
and
percentages
of CPU time allocated to each class.
Note that you can set up a class
to include all batch jobs.
This allows
you
to
control
batch
Jjobs
separately
from
timesharing
jobs.
The batch class is given a high
percentage or a low percentage,
Now, each time a user submits a batch
job,
the
scheduler
wuses
the
batch class and not the user's class.
Section 10.1.4, Step 5, describes how to create a batch class.

CLASS O

CLASS 1

STUDENTS
10%

CLASS 3

RESEARCH
40%

CLASS 2

MR-S-566-80

19-2

SYSTEM

You

can

define

class

PERFORMANCE

memberships

using

the

TOPS-28

accounting

facility
or
by
writing
your
own
access control program.
Section
10.1.9 provides an overview of using
an
access
control
program
to
define
classes.
The following description applies to using the class
scheduler

with

the

accounting

facility.

Using the accounting facility, you can associate a class
number
with
each
account
on
the
system.
Each
account
has
only
one
class
associated with it,
Therefore, only a user who
has
access
to
more
than
one
account
can
have access to more than one class.
The user
changes classes by changing accounts.
To use this
method,
you
must
enable
account
validation
so that users are required to use a valid
account,
Section 10.1.4 describes the accounting file entries.
The scheduler periodically computes
individual jobs within a class.
The
a class receive its share of the CPU
greater

its

percentage

target

time

the

time,

class

The

that

scheduler

the

and
that

gives

farthest

away

from

percentage
allocated
to
the
<class).
The
is that each job within a class receive its
the scheduler computes the
average
amount
of
CPU
time
that
a
Jjob has accrued.
This quantity determines the
job's priority within the class.
A job that is requesting the CPU and
is
farthest
way
from
receiving
1its
fair
share
within the class
receives a greater percentage of time within that class.
scheduler's
fair share.

Generally,

windfall.
jobs, or if
to

You

use

all

(the

the
time
wused
by
classes
scheduler's first concern is

total

second concern
Periodically,

the

CPU

has

wunused

its

class

This

unused

time

1is

<called

percentage.

can

handle windfall in
Allocating
windfall

it.

time,

Windfall
occurs
1f
one or more classes has no logged-in
one or more classes has an insufficient demand for the CPU

one

of
means

two

ways:
that the

allocate it
or
withhold
excess CPU time is awarded

proportionately to the active classes.
(In this discussion, the
term
active class refers to a class that has logged-in users requesting CPU
time.) Higher percentage classes receive proportionately more
of
the
available
windfall than lower percentage classes.
This means classes
can

receive

windfall
to

each

slightly

available.
class

and

not

than

their

Note

that

windfall

each

job

allowed

within

percentages

is distributed
a

when

Usually,

throw

better

away valuable

allocate

windfall,

computer

time.

19-3

class.

Withholding windfall means that the excess CPU time is idle
it
is
not
distributed
to
the
active classes.
It also
classes do not receive more than
the
percentage
allowed
don't

there

proportionately

and

not

withhold

time
and
means that
for
them.
it

you

10.1.2

Who

Should

Use

SYSTEM

PERFORMANCE

Class

Scheduler?

the

As mentioned earlier, the class scheduler
user
community
1into defined classes and
time to each class.
This
intended
use
practical for some systems.
Tradeoffs do
do

not

require

paragraphs

Several

class

and

instances
You

scheduling.

determine

can

that

can

elect

customer

can

Therefore,

your

benefit
sell

system

from

some

read

needs

using

portions

purchase

allows
vyou
to
divide
the
allocate a percentage of CPU
may
hot
be
beneficial
or
exist and some installations

this

the

CPU

the
type

class

control.

scheduler

time.

percentage

following
of

For

are:

example,

computer

time

proportional cost.
You should, in this case, invite customer
X
to
your
installation
and
provide
an
environment that
simulates the kind of response this customer
can
expect
at
this
percentage.
Because
it
1is
1impossible to create the
environment exactly as the customer will experience it at all
times,
customer X may decide later to change the purchase to
a different percentage.
If you have a natural
you
can divide these
time to these classes
system.

have

For

faculty

users

student

the

<class
of

If you
class
the
Conversely,

the

(perhaps

users

the

each

educational

(doing

working

(who

CPU

are

computer

vyou

time

can

environment,

payroll,

funded

research

science

establish

according

the

you

grades,

project),

majors).

three

may

etc.),
Using

classes

importance

and

pay

class.

wish to give
scheduler to

preference to batch jobs, you can use the
give the batch class a high percentage of

CPU,
you may

following
If

wusers

scheduler,

distribute
rate

example,

administrative

and

division among the users of the systen,
users into classes, then distribute CPU
with respect to their importance on the

not

want

have

not

scheduling,

you

the

realized
1if

you

immediate requirement
set it up.
Systems with
increase
in
any

use

class

scheduler

for

some

reasons:

for

need

have

class

1in

new

the

system

scheduling,

minimal amounts
of
memory
swapping.
This additional

advantage

gained

using

past
and

the

class

for

you

not

class
see

should

an
not

may
experience
an
overhead may offset

scheduler.

In
addition
to
other
tasks,
the
scheduler
constantly
maintains
usage
values
for each class and job.
Because of
this extra overhead, the overall system throughput decreases.
Therefore,
wuse the class scheduler if allocating percentages
to individually
defined
<classes
outweighs
the
throughput
depreciation.
'
To

give

have

placing
to

batch

jobs

use

the

place

low priority
<class

BATCH-BACKGROUND
all

batch

the

scheduler.

Jjobs

queue.,

19-4

command
in

low

system,

Section

the

you

16.2

n-CONFIG.CMD

priority

(or

not

describes

file

background)

SYSTEM

19.1.3

How

Begin

Using

the

PERFORMANCE

Class

Scheduler

If you elect to use the class scheduler, first divide the system users
into
classes.
Next,
determine
the
amount
of CPU time each class
should receive.
Percentages given to classes are typically
in
units
of
5,
that
1is,
5%, 10%, 15%, ....100%.
The sum of the percentages
given to all classes cannot exceed 10# percent.
The result
of
these
two
steps
depends
on
the
reason
you
elected
to
use
the class
scheduler, and how you plan to use it at your
installation,
Because
of
this
system " dependency,
step-by-step
procedures
for selecting

classes and
allocating
the
CPU
cannot
be
given.
The
discussion
provides
you
with
guidelines
for
setting up
scheduler to meet your system's needs.

following
the class

Start with as few classes as possible,
The
class
scheduler
allows
eight.
classes further, if necessary.

three or perhaps four.
Later, you can divide

Determine the number of logged-in users you
expect
in
each
class.
Be
sure that you do not overload a class, making it
impossible to give a sufficient percentage
of the CPU to each
job in the class.
Also, you can limit the number of users in
a class, but you
cannot
1limit
the
number
of
Jjobs.
For
example,
the
SUBMIT
command
creates
an
additional
Jjob.
Therefore, consider the type of work that users
in
a
class
perform.

Estimate the percentage of CPU time (or time purchased)
each
class should receive.
This is a difficult step; however, you
can experiment with
and
later
alter
the
percentages
you
choose.
(Sectiqn
10.1.5 describes how you can change class
percentages.) If a class consists of a large number of
users
who are generally logged in at the same time, be sure to give
this class sufficient
CPU
time.
. For
example,
using
the
diagram
below,
<class
2
has
60
members.
It also has 60
percent of the CPU.
This means that if approximately 68 jobs
in
class
2
are
demanding
the
CPU, each job will receive
approximately 1 percent.
(It may be slightly greater than
1
percent
if you allocate windfall.) This also means that on a
per-job basis, users in class 2, with the higher
percentage,
may
not
get as much of the CPU as users in class # or class
1.

WINDFALL

CLASS 1

CLASS O

CLASS 2
MR--S-567-80

The
198

above diagram
percent
of

percentage

for

all

illustrates that you can
the
CPU
to
classes.
classes

cannot

190-5

exceed

allocate
However,

100

less than
the total

percent.

SYSTEM

PERFORMANCE

The system uses class 0 as a default
for
any
account
that
does
not
have a valid class assigned to it.
Therefore, you
can divide a portion of the system into well-defined
classes
and
have all other users receive the percentage given to the
default

class.

You

can set up the class scheduler so that one class, perhaps
the
default
class,
receives
only
windfall.
For example,
suppose you allow some users to log into an account
that
Iis
not
yet
assigned
a class.
These users are assigned to the
default
class
4.
Or,
suppose
several
users
log
in
infrequently, read mail, and perform little computing.
These
users may log into an account that is not
assigned
a
class
(and,
therefore,
are
assigned the default class, 0), or an
account that is associated with class
#@.
You
subsequently
assign a zero percentage to class #.
These users may receive
enough windfall to get their job done.
However, they are not
allocated
CPU time and can have periods of extremely slow or
no response.
You may find,
after
experimenting
with
this
type
of
procedure,
that
it
is
better
to associate each
account with a class and give the class a very low percentage
of

the

CPU.

Situations may arise that affect the scheduler's
ability
to
give
a
<class
1its
percentage of the CPU.
For example, the
members of
a
class
cannot
use
the
amount
of
CPU
time
allocated
to
the
<class.
Also, a situation may arise where
the demands of the class exceed the percentage
of
CPU
time
given the class, and windfall is available but not allocated.
In either case, you should reevaluate
both
the
percentages
given
to
<classes,
and
whether or not you want to continue
withholding

10.1.4

Procedures

After dividing
time,
follow
scheduler.

users
these

allocating

Turn

the

windfall.

Class

into classes and
steps
to
set

‘

Scheduler

estimating
up classes

the
amount
of
CPU
and turn on the class

PROCEDURE

STEP

Edit the <ACCOUNTS>ACCOUNTS.CMD file on the public structure
(and
the
subaccount
files,
1if any, that contain all your
accounting
data).
Follow
the
procedure
in
Chapter
6,
Creating
Accounts,
for
modifying or creating each account
with the /CLASS:n switch,
For example,
ACCOUNT COMP1/CLASS:1
USERS

KOHN,HOLLAND,MILLER

means that when users Kohn, Holland, and Miller log into
or
change
their
account to COMPl, they are placed in class 1.
Each
account
has
only
one
class
associated
with
it.
Therefore,
only
the
user
who has access to more than one
account can have access to more than one class.
A
Jjob
can
be in only one class at any one time.
Run the ACTGEN program.
Give the TAKE command
and
specify
the
<ACCOUNTS>ACCOUNTS.CMD file (or the file containing the
accounting data).

19-6

SYSTEM

PERFORMANCE

When ACTGEN returns its
This
command
updates
system uses to validate

prompt, give
the
INSTALL
the ACCOUNTS-TABLE.BIN file
accounts.

command.
that the

Edit the n-CONFIG.CMD file to include
the
CREATE
commands
that
specify
the
classes.,
Be sure account validation is
enabled.
That is, check that the DISABLE ACCOUNT VALIDATION
command
is
NOT
in
the
file.
(The
system
default
Iis
ENABLED.,)
If account validation is disabled, users
can
use
any
account and therefore any class they choose.
following commands in the order shown below.

Enter

the

The

and

the

CREATE

command

percentage
example,
CREATE

CPU

defines
time

the

<class

number

should

receive.

For

.40

means

that the default class, @, should receive
of
the
CPU.
Enter
the
CREATE
command
for
classes that you defined
1in
the
ACCOUNTS.CMD

4@
percent
each of the
file.
For

example,
CREATE

.49

CREATE

.20

CREATE

.30
NOTE

Remember

that

user

whose

class

account

the

(this

default

includes

not associated with a specific class
Next, if
separate

you have decided
class, enter the

BATCH-CLASS

class.

subaccounts)

placed

to
place
all
batch
BATCH-CLASS command.

Any

Jjobs
1in
a
For example,

means that
all
batch
Jjobs
will
be
placed
in
class
2
regardless of the user's associated class.
You must use the
CREATE command to define the percentage of CPU time that the
batch class should receive.
Note that timesharing users can
be in the same class as well, if the
account
they
use
Iis
associated
with
that
<class.
The CREATE command in step 4
defines class 2 to receive 30 percent of the processor.
If
you
do
not place a BATCH command in the n-CONFIG.CMD file,
the scheduler treats all batch jobs the same as
timesharing
jobs.
Finally, enter the ENABLE CLASS-SCHEDULING command.
Be sure
this is the last command entered.
If you reverse the order,
that is, you enter ENABLE CLASS—-SCHEDULING before the CREATE

commands
percent

and
of

the

BATCH

command,

CPU.

19-7

all

wusers

will

receive

zero

PERFORMANCE

SYSTEM

The

ENABLE

scheduler

CLASS-SCHEDULING

command

the next system reload.

turns

the

class

It also specifies if

you are
using
accounting
or
an
access
control
program
(policy
program)
for
class
scheduling
and
if
you
are
allocating or withholding the windfall.
The format of
this
command

is:

ENABLE CLASS—-SCHEDULING ACCOUNTS

WITHHELD

ALLOCATED

POLICY-PROGRAM

For

example,

ENABLE

CLASS-SCHEDULING ACCOUNTS

means turn

method,
and
windfall.
Do
good
reason
scheme

The

the

class

ALLOCATED

scheduler,

use

accounting

works.

entry,

ENABLE

CLASS-SCHEDULING

POLICY-PROGRAM

ALLOCATED

means turn on the class scheduler, use
the
control) program, and allocate the windfall.

the

allocate
the
windfall.
Always allocate the
not withhold windfall unless you have
a
very
to
do
so,
and
you are sure that your class

policy

(access

At the next system reload (the system is
brought
down
and
back
up
again)
the
new commands in the n-CONFIG.CMD file
take

19.1.5
During

effect.

Changing

’

Class Percentages During

timesharing,

you

can

change

Timesharing

the

percentage

that

class

receives
by
using
the
SET command to OPR.
This change lasts until
either the next system reload, or you make another
change
using
the

SET command.

The

format of

SET SCHEDULER CLASS

n can be
the

To make
CREATE
the

from @-99

CREATE

command

the

(number)

inclusive.
is

not

SET command
m

Note

allowed

(to

that

appears:

percent)

this

the decimal point
command.

required

the percentage that
a
class
receives
permanent,
edit
the
commands
in the n-CONFIG.CMD file.
For example, you can edit

commands

CREATE

.60

CREATE

.30

CREATE

.19

CREATE

.10

CREATE
CREATE

1
2

.@5
.80

Next, reload the system to process the commands
in
the
n-CONFIG.CMD
file.
The classes that received a low percentage before you made the
change now receive a higher percentage of CPU time.

19-8

SYSTEM

PERFORMANCE

Changing class percentage may be useful if you decide
that
users
in
one
or
two
classes
should receive a greater percentage of CPU time
than other classes during the
day.
However,
these
high-percentage
classes
may
nhot
require this time during the evening shift.
If you
have a recurring need for such
changes,
you
may
want
to
put
the
appropriate commands into files for use with the TAKE command in OPR,

19.1.6

Disabling

the

Class

Scheduler

During

Timesharing

You can turn the class scheduler off and back on during timesharing by
using
the
DISABLE
and
ENABLE commands to OPR.
The class scheduler
remembers the class percentages that were in effect before the DISABLE
command
was
given.
For example, suppose you use the SET command to
change the percentage that class
1
should
receive
from
@5
to
15
percent,
Then,
you
give
the
DISABLE CLASS-SCHEDULER command, and
later give the ENABLE CLASS-SCHEDULER command.
The
class
scheduler
uses 15 percent for class 1.
If you reload the system after disabling
the class scheduler, the class scheduler is enabled again and uses the
percentages given in the n-CONFIG.CMD file.,

18.1.7

Getting

Information About

Several TOPS-20 commands
scheduler statistics.

The

INFORMATION
e

(ABOUT)

the

class

the

accounting

provide

Class

information

SYSTEM-STATUS

scheduler

Scheduler

command

enabled

method

allocated

withheld

Status

about

different

informs

you

<class

if:

disabled

access

control

program

1is

being

BATCH

CLASS

used

For

windfall

the system's batch jobs are in a

separate class.

example

@INFORMATION

CLASS

(ABOUT)

SCHEDULING

SYSTEM-STATUS<RET>

ACCOUNTS

ENABLED,

WINDFALL

ALLOCATED,

The INFORMATION (ABOUT) MONITOR-STATISTICS command
provides
a
table
that
shows
each active class, its target use of the CPU, its current
use of the CPU, and the load averages for that class.
For example,
@INFORMATION

Class

l
2

(ABOUT)

MONITOR-STATISTICS<RET>

Use

Loads

@0.80

0.79

6.56

4,36

¢.15

@6.21

5.46

1l.38

.95

0.05

0.00

2.00

.00

2.00

3.57

19-9

SYSTEM

PERFORMANCE

The SYSTAT command outputs load averages for either the entire
system
or
a
specific
class.
When
the class scheduler is disabled, these
averages represent the load of the entire system.
For example,

@SYSTAT<KRET>

Wed 11-Jul-84 10:17:09
Up
52416 Jobs
Load av
5.30

11:38:12
4.03
4.86

The last three numbers following Load av indicate the
average
number
of
runnable
processes over a period of one minute, five minutes, and
fifteen minutes.
These numbers start at zero.
The higher the numbers
the
1longer
a job has to wait for CPU time on the average.
Using the
example, over a fifteen minute period, a given job demanding
the
CPU
waits
approximately 4.86 times longer to run than it would if it were
the only job running on the system.
When the class
status
of the
For

scheduler
job doing

is enabled, these load
the SYSTAT command and

example,

@SYSTAT<KRET>
Wed

11-Jul-84 10:28:97
Up
52+1@ Jobs
Load av (class

The

averages represent the
not the entire system.

last

the

three

<class

command

following

Load

that

the

job giving

the

CLASS

argument

with

numbers

11:49:12
1)
1.79

2.36
av

2.88

indicate

the

SYSTAT command

provides

load

breakdown

the
system,
This
breakdown
includes the class each
average share of the class percentage that this job can
how much CPU time the job is currently using.
The

TOPS-20

Commands

Reference

Manual

describes

averages

in.

these

The

each

SYSTAT
Jjob

job is in,
receive,

the
and

commands

1i1n

detail.

10.1.8

The

Sample

following
e

The

Session

examples

show:

ACCOUNTS.CMD

file

after

associating

classes

with

accounts.

! The

STYPE

The
procedures
that
accounting files.

USERS

(FILE)

USERS

has

been

MYBANK/CLASS:2

TRUST/CLASS:1

OVERHEAD

SAMBERG,BERKOWITZ,
TAYLOR

19-10

after

commands

edited.

MAIN:<ACCOUNTS>ACCOUNTS.CMD<RET>

BRAITHWAITE,HURLEY,HALL,CRISS

ACCOUNT

USERS

file

BLOUNT, KONEN,ENGEL

ACCOUNT

sample of the class scheduling
the 6~-CONFIG.CMD file.

<ACCOUNTS>ACCOUNTS.CMD

ACCOUNT

you

editing

that

are

all

placed

your

SYSTEM

ACCOUNT
USERS

PERFORMANCE

PROG/CLASS:3

BLO'NT,KOHN, HOLLAND

$
INotice
INext,

SRUN

that
run

account

the

OVERHEAD uses

ACTGEN

the

default class,

program.

ACTGEN<KRET>

ACTGEN>TAKE

(COMMANDS

FROM)

MAIN:<ACCOUNTS>ACCOUNTS.CMD<KRET>

!After ACTGEN finishes and returns its prompt, give
I INSTALL command to create a new version of the
1<SYSTEM>ACCOUNTS-TABLE.BIN file

the

ACTGEN>INSTALL<RET>

!After

ACTGEN

IEXIT

command

finishes

and

returns

its

prompt,

give

the

ACTGEN>EXIT<RET>

$
STYPE SYSTEM:6-CONFIG.CMD<RET>
ITerminal

Speeds

TERMINAL

SPEED

2400

TERMINAL

SPEED

9640

TERMINAL

SPEED

2400

TERMINAL

SPEED

2400

TERMINAL

SPEED

300

TERMINAL

SPEED

300

DEFINE

SYS:

DEFINE

SYSTEM:

MAIN:<SUBSYS>

DEFINE

NEW:

MAIN:<NEW>,SYS:

MAIN:<SYSTEM>

DEFINE

OLD:

MAIN:<OLD>,SYS:

DEFINE

HLP:

MAIN:<OLD>,SYS:

MAGTAPE

PRINTER

VFU

PRINTER

LOWERCASE

TIMEZONE

for

CREATE
CREATE

.45

CREATE

.25

.15

BATCH

RAM

SYS:LP96.RAM

{Commands

CREATE

SYS:NORMAL,VFU

the

class

scheduler

.05

ENABLE

CLASS-SCHEDULING

ACCOUNTS

ALLOCATED

$
To start the Class Scheduler, either reload the system,
or
ENABLE CLASS—-SCHEDULER command to OPR.
If you edited the n-CONFIG.CMD

file
the

remove

system

the
you

DISABLE
can

start

ACCOUNT VALIDATION
validating

19-11

command,

accounts.

you must

reload

SYSTEM

An Alternative

19.1.9

to Using

PERFORMANCE

Accounts

Chapter 11 describes the access control program that is used to grant
This
restrict access to various system hardware and software.
and

same program can also include the appropriate monitor calls to

class scheduler decisions.
e

classes

class memberships

the

class

changing

class

windfall

allocation per

batch

handle

Some of the requests it can define are:

class

percentages

class

in
The monitor calls that are required in this program are described
DIGITAL distributes a
Reference Manual.
the TOPS-28 Monitor Calls
sample Access

monitor

Control

Job

program

that

includes

class

scheduler

calls.

This program is called ACJ.MEM and is located on the Distribution tape
The ACJ.MEM file provides you with a guide
in the documentation area.
try
and
Do not copy it
for writing your own access control program.
to

use

is.

NOTE

You CANNOT run the access control program to implement
class scheduler decisions at the same time you use the
You must use one or the other.
accounting method.

1¢.2

SCHEDULING

LOW PRIORITY TO BATCH JOBS

tasks
background
as
The decision to favor batch jobs or to run them
your
at
have
you
environment
batch
of
type
the
on
depends
installation.
For example, if users submit batch jobs that
are
1long

that do not require completion immediately, you can give batch
and/or
jobs a low priority.
Conversely, if batch jobs are the
primary
Jjobs
on the system, you can give them a high priority.
Section 1¢.1 describes how to place batch jobs in either a high or low
class by including the BATCH n command in the n-CONFIG.CMD
percentage
batch
the
When a user submits a batch job, the scheduler uses
file.
class and

not

the

user's class.

19-12

SYSTEM

PERFORMANCE

You must use the class scheduler to give batch jobs a
high
priority.
If
you
are
not using the class scheduler, you can give batch jobs a
low priority.
To do this, enter the
BATCH-BACKGROUND

command in the n-CONFIG.CMD file.
This
command
specifies
that
all
batch
Jjobs
run
on
the
1lowest
priority
queue,
also known as the
background queue.
This means that after
processing
all
interactive
jobs,
the scheduler selects and runs batch jobs waiting in the queue.
They receive left—-over CPU time.
You can enter
the
BATCH-BACKGROUND
command into the n-CONFIG.CMD file.
The command takes effect the next
time you reload the system.
NOTE

The BATCH-BACKGROUND command is intended for those who
are
not using the class scheduler on their system but
want to give the
batch
jobs
a
1low
priority.
You
should
not use this command when you enable the class
scheduler.

19.3

FAVORING

INTERACTIVE

VERSUS

COMPUTE-BOUND

This

PROGRAMS

section describes how you can influence the
either
interactive
or
compute-bound programs.
the bias control, which is analogous to turning a
settings
(from
1
to
28).
When
you
select
scheduler
a
higher
programs.

favors users running interactive
number,
the
scheduler
favors

Figure

1@0-1

illustrates

this

scheduler
to
favor
You do this by using
knob over a range of
a
lower number, the

programs.
When you
select
users running computational

concept.

BIAS CONTROL

INTERACTIVE

COMPUTE BOUND

THE ""KNOB"
MR-S-568-80

Figure

1@¢-1:

Bias

Control

'Knob'

19-13

SYSTEM

PERFORMANCE

NOTE

You can use bias control with
or
without
the
<class
scheduler
enabled.
However,
the effect of the bias
control is less noticeable when used
with
the
class
scheduler.

After

you

control
evenly

install

number
to

TOPS-20

1ll.

This

interactive

and

New users of TOPS-20 can
favor
particular
types

want

experiment

controls

can

with

serve

software,

setting

the

system

distributes

compute-bound

use
of
a

good

setting until
Previous users

settings.

tool

the

default

scheduler's

bias

attention

programs.

the default
programs.

new control

uses

the

for

For

they
need
of TOPS-20

example,

favoring

the

different

to
may

bias

types

of
users at different times of the day.
For instance, you
can
set
the
bias
to
a
low number during the day to favor on-line users and to a
high number during the evening to favor batch users.
The response you
receive
from your user community should determine the appropriateness
of the selected settings.

Setting the bias toward interactive programs gives better response
to
the
terminal | user.
Also,
this setting may produce a higher system
overhead, because the scheduler swaps jobs and switches from different
tasks
more
often
in
1its
effort
to
favor
interactive
programs.
Generally, setting the bias toward interactive programs is
beneficial
only
if
you
have sufficient memory.
Both the swapping rate and the
scheduler overhead are likely to increase in small
systems
with
too
little
memory.
If
your
system
has adequate memory, the scheduler
overhead should be fairly constant over most of the bias settings.

Setting the bias toward computational programs
should
reduce
system
overhead
and
increase
the
total
system
throughput.
However, the
response to the terminal user may decrease slightly.
In
some
cases,
the
improvement
in
system
throughput more than compensates for the
lessened response time, and users are more satisfied.
As

you experiment with the bias settings, remember
that
setting
the
bias
control
to
the
extremes can prevent certain types of programs
from running for long time periods.

After

you

select

bias

operation,

enter

use

the

SCHEDULER

The

format

SET

BIAS

The

the

You

change

SCHEDULER

the

number

CONTROL command

BIAS-CONTROL

command

that

fits

your

the

n-CONFIG.CMD

the

OPR

the

n-CONFIG.CMD

file

is:

OPR

is:

timesharing

system's

file or

program.

command

bias
the

control

BIAS

command

SCHEDULER

where m
can

the

CONTROL

format
SET

the

BIAS-CONTROL

control
bias

BIAS-CONTROL

(TO)

number.

setting
command.

during

However,

when

you

using

reload

the

SET

system,

the bias setting in the n-CONFIG.CMD file takes effect.
If
you
want
your
change
to
be
permanent,
edit
the
n-CONFIG.CMD
file
at
a
convenient time before you reload the system.
Any time you are unsure
give
the
INFORMATION
setting.

of the current setting of
the
bias
control,
(ABOUT) SYSTEM-STATUS command to determine its

19-14

SYSTEM

19.4

IMPROVING

PROGRAM

STARTUP

PERFORMANCE

TIME

Some of the programs on your system are run frequently.
This involves
constant
searching
for
the
same file on disk, bringing the program
pages into memory, and allocating swapping space when the
program
is
swapped out of memory.
By storing these programs in an easy-to-access
area, some of the startup time is saved.
To improve the startup
time
of
the
frequently used system programs, TOPS-20 keeps a program nhame
cache.

The <SYSTEM>PROGRAM-NAME~-CACHE.TXT file is
placed
<SYSTEM>
on
the public structure automatically at
and contains a list of programs
and
files
to
be
program name cache.
These are:

in
the
directory
installation time,
copied
into
the

SYS:PAl1050,EXE
SYS :MACRO.EXE
SYS:EDIT.
EXE
SYS:TV.EXE
SYS:LINK.EXE

SYSTEM: ERRMES.BIN

Each

time

you

reload

the

system,

the

<SUBSYS>MAPPER.EXE

program

runs

under
the
SYSJOB
program
at
the
operator's
console,
<SUBSYS>MAPPER.EXE reads the <SYSTEM>PROGRAM-NAME-CACHE.TXT
file
and
loads
the
program name cache.
Now, when requests are made for these
programs, the system looks first in the program name cache to
see
|if
it

can

retrieve

further

the

required

improve

pages

system

PROGRAM-NAME~CACHE.TXT
file
frequently accessed programs.
FORTRAN language, you may want

quickly.

performance,

you

can

and
add
the
filenames
of
For example, if your
system
to add the files:

edit
your
uses

the
own
the

SYS:FORTRA,
EXE

SYS:FOROTS.
EXE
SYS:FORLIB.REL

Your list can
contain
the
names
of
up
to
16
executable
files.
Therefore,
select
the
files
to be placed in the program name cache
carefully.
You should consider only the
executable
files
that
are
started

frequently

large

number

users.

You can also add library files to the program name cache, for example,
SYS:FORLIB.REL.
However, these types of files use up swapping space.
If you have too many or very large files, you may create a detrimental
effect
on
vyour
system's
performance.
The total library file pages
that you cache should be
no
greater
than
200
to
304.
Give
the
VDIRECTORY
command
for the library files you want to cache and check
the number of pages in each file.

10-15

SYSTEM

PERFORMANCE

(MAPPER
effect
takes
file
revised
If you edit the cache file, the
creates
a
new
version
of
the
cache) the next time you reload the
system.
Alternatively, you can create a
new
version
of
the
cache
immediately
by
entering
the
following
commands
at the operator's
console.

ltalk to SYSJOB
1kill old version of cache
lread new file and create

"ESPEAK
KILL MAPPER
RUN SYS:MAPPER.EXE

!new version of

lexit

1.5

REINITIALIZING

DISK

cache

PACKS

After many files are created, they may no longer be contiguous on
the
structure
(disk(s)).
This scattering of files may increase the time
it takes to
process
them.
You
may
decrease
processing
time
by
reinitializing
the
disk
packs in your heavily used structures.
For
example,
the
public
structure
may
be
a
likely
candidate
for
reinitialization.
This
procedure
places
files
into
a contiguous
format and can be scheduled as part of a backup procedure.
How
often
you reinitialize your packs depends on the work load of the system and
whether you notice a difference in system performance after
following
this

procedure.

Reinitializing disk packs requires that you dump all
the
directories
and
files
to
tape, re-create the structure (and, in the case of the
public
structure,
reinstall
the
system),
and
restore
all
the
directories
and
files.,
Therefore,
if
you
do
not
notice
any
appreciable difference in your system performance
after
doing
this,

don't

schedule

it on

regular basis,

all.

The
TOPS-20
Operator's
Guide
describes
re-creating
the
public
structure
and other structures.
Have the operator use this procedure
for
reinitializing
disk
packs.
If
possible,
have
the
operator
re-create
the structure and restore the files to a different pack (or
set of packs) from the structure that you dumped.
This
ensures
that
you
do
not lose your files should you have problems reading the tape
back to disk.
That is, you still have the original
structure
intact
and can run DUMPER again to copy the files to another tape.

19.6

DYNAMIC

DUAL

PORTING

Dynamic dual porting refers to a disk drive that is dual-ported to one
system
only.
When one of the channels is busy transferring data for
another disk unit, input/output is automatically switched through
the
other
disk channel.
Dynamic dual porting improves the performance of
input and output operations.
It
1is
activated
automatically
after
field
service properly sets up the RP@4, RPO6, or RPP7 disk drive(s),
and the operator places the drives' port switches in the A/B position.
Dynamic dual porting is not supported for RP20 disks.
jobs could hang, files could be damaged, and so on.

The DECSYSTEM-2§ Technical Summary describes the
with input and output operatilons.

18-16

hardware

tried,

associated

CHAPTER
ACCESS

TOPS-20
e

provides
Govern

control
the

CONTROLS

mechanisms

access

many

that
of

help
your

you:
system's

resources

and

services

Reduce or

Investigate

The

following

11.1

ACCESS

prevent

occurrences

sections

CONTROL

unauthorized

discuss

access

unauthorized

these

the

system

access

topics.

PROGRAM

Previous chapters deal with
administrative
policies
for
allocating
resources.
For example, Chapter 10 describes the policy decisions you
can make regarding the scheduler, and Chapter 8 describes
the
policy
decisions
you
can
make
regarding tape drive allocation and labeled
tape

support.

In addition, you can make policy decisions that govern the
access
to
specific
system
resources.
For
instance, TOPS-20 allows a user to
change the speed of a terminal, assign a device, log in at any time of
day,
mount a magnetic tape, and mount a disk structure.
However, you
may want to restrict or disallow use of some of these facilities.
You
may
want
only
specified
wusers
at
specified times of the day and,
perhaps,
at
specified
terminals,
to
wuse
certain
facilities.
A
particular mechanism lets you control the access to such resources and
services.
With it,
you
have
an
additional
means
for
<collecting
accounting or other information.

To use this access control mechanism, you must write an access control
program
that
carries
out
your policy decisions.
An access control
program can
control
scheduling
classes,
the
bias
control,
batch
background
queue, logging in, use of physical resources (tape drives,
terminals,
structures),
and
enabling
capabilities.
When
a
user
requests
a resource (like ASSIGN TTY34:), your program identifies the
user, the user's controlling terminal, and the type of
request
being
made.
Your program can merely log this information in a file, or make
a decision and tell the monitor to either grant or deny the request.

DIGITAL provides the necessary mechanisms (monitor calls) to implement
a
program
at
your
installation.
Your
system programmer uses the
appropriate monitor calls to write an access control program according
to
the
requirements of your system.
A sample access control program
(ACJ.MEM) is distributed in the documentation area on the distribution
tape.
Your system programmer can use this program as a sample of the
structure
of
an
access
control
program
and
the
parameters
and
decisions

that

can

controlled.
11-1

ACCESS

The

following

program

Assign

can

list

defines

control

and

CONTROLS

the

explains

resources
why

you

that

may

want

access
control

control
thenm.

Devices

You can restrict users from assigning
their
jobs.
For
example,
1if
you
allocation

Chapter

which

(described in
wusers are allowed

prevent

one

user

from

terminal lines or tape drives to
have
not
enabled
tape
drive

8), you may
still
want
to
assign tape drives.
Your program

assigning

all

the

enable

their

available

tape

control
can also

drives

terminals.
Enable
You

Capabilities

can

allow

users

capabilities

only

certain

locations,
or perhaps only at certain times of the day.
For example,
in a university environment, you may not
want
users
enabling
WHEEL
capabilities
in
a terminal room used by students.
1In an environment
where security is a major concern, you may want to
allow
a
user
to
enable
WHEEL
capabilities only on a hard-copy terminal and only in a
certain location.
You can then collect the hard-copy output from
the
terminal.
The
access
control
program
can also keep a file of the
names

Create

people

who

have

enabled

capabilities.

Jobs

You can restrict
additional
Jjobs

the users
wvia
the

this

facility

certain

Allow

You

their

can

prevent

users

who are able to write
CRJOB monitor call.

applications

from

logging

to
log
in
only
at
certain
university environment, it may

programs that
create
You may want to limit

only.

times
not be

than

once,

of
the day.
desirable to

the
system
during
1large production runs.
Also,
accounting information (and not that provided with

For
have

permit

user

example, in a
students
on

if you use
TOPS-28),

your
you

ownh
can

use
the access control program with the login function to recognize a
user.
In addition, the login function can
be
wused
to
control
the
number of jobs that a user can create under PTYCON.
Create
You

can

Processes
prevent

processes,

(Forks)
a

user

Also,

from

you

creating

may

want

than

~charge

predefined

users

number

for

using many

all

terminals.

processes,

Set

Terminal

You

can

Baud

control

Rate

the

input

and

output

speed

settings

This
control
prevents
users
from changing the baud rate to a speed
that is unsupported by the terminal, and, as a result,
rendering
the
terminal
unusable
until
the operator resets the baud rate.
You can
also restrict the speed of a terminal to
no
more
than
a
specified
maximum, for example, 300 baud.

ACCESS

CONTROLS

Logout
You

can

request

the

access

information

system.

may

also

You

are

over

their

can

notify

the

control

accounting
want

permanent

operator

keep

disk

that

program

file

each

track

notify

time

the

you

user

logs

who

log

users

record

off
out

the
and

page quota.
The access control program
migration-trim-run is needed
for
this

directory to bring the directory back under its quota.
If you use the
login function with the logout function, you can give the user who
is
logging
out
information
about
time,
resources,
and perhaps mohey
spent.,

Set

ENQ

Quota

TOPS-20

the

allows

ENQ

users

other

Monitor
Create
You

WHEELS

Reference

can

prevent

request

who

police

these
a

can

the

have

the

ENQ-DEQ

describes

control
of

the

control

these

and

access

users

quota.

program,

you

quota.

(The

uses

BUILD,

wusing

can

allow

TOPS-20

ENQ-DEQ.)

check

the

SET

change

program

commands.

certain

This

ive

You

DIRECTORY,

parameters.

Or,

you

notify

the

may

want

parameter

MOUNT

facility

Operator's

structures.)

recorded

structures

the

TOPS-2f

NON-REGULATED

giving

access

used

(The

you

the

operator

changes.

used

Also,

accounting

allowing

command

conjunction
Guide

information

only

for

select

particular

with

regulated

describes

about

REGULATED

structure

and

mounts

files.

MDDT

can

disallow

example,
WHEELS

during

looking

Class

change

from

directories

structure(s).

want

control

to create directories

structures,

You

access

Structure

group

Entexr

change

Manual

users

command

people

can

your

Directory

may simply

You

WHEELS

function

than

Calls

"ECREATE

Mount

enabled

quota

keep

privileged

users

certain

of the
problem in

fixing

record

from

times
a

who

has

used

entering

day,
the

MDDT

you may
monitor.

mode.

not

For
enabled

want

You

may

also

MDDT.

Assignment

You

can
classes.

prevent

The

wusers
from
changing
to
wunauthorized
scheduling
access control program determines the classes a job can

use,

Set

Class

The

access

program

their

Access

You

can

system.

program

contain

(or

associated

can

access

set

user's

file

that

class

contains)

log

the

in.

list

Your

users

class.

Request

have
allowed

nonblank

control

can

and

For

the

access

example,

access

code

control

a
a

decide if this user can use
Manual describes the access

program decide
restricted labeled

non-TOPS-20

the

access

labeled

field,

whether

tape

a
from

tape

you

can

this tape.
(The TOPS-20
fields on labeled tapes.)

11-3

user

should
non-TOPS-20

mounted

have

Tape

your

with

program

Processing

ACCESS

CONTROLS

ACCESS/CONNECT Request

request

ACCESS

These

CONNECT

ACCESS

directory should succeed in cases where the request is

denied by the monitor.
met.

determine

The access control program can

CONNECT

For example, the

request

TOPS-20

allows

monitor

to succeed when appropriate criteria are

are:

capabilities

The requesting process has

WHEEL

OPERATOR

The target directory is

the

same

group

target
the
and
The target structure is DOMESTIC
name matches the logged-in directory or the job.

The correct password

enabled.

"accessed"

the

Jjob's

directory.

directory

is specified.

If all of the above criteria fail, the
monitor
denies
the
request,
program can be called to approve or override the
control
access
The

denial.
ATTACH

Request

his
attaching
from
user
prevent a
The access control program can
function allows the access control
This
job.
another
to
terminal
program to control which terminals can attach to specific jobs.

11.2

PASSWORD

ENCRYPTION

One way to violate system security
1is
through
unauthorized
wuse
of
Having acquired someone's password, an intruder
passwords.
directory
The
resources,
system
restricted
to
access
could log in or gain
facility in TOPS-28, however, makes it harder to
encryption
password
steal

passwords.

With
encryption
enabled,
passwords
entered
into
the
system
are
format before they are
cyphertext
indecipherable
an
to
translated
stored or otherwise used.
Nowhere
in
the
system
is
the
original
As a further security measure, no
form of a password kept.
plaintext
current TOPS-2@ utility converts the cyphertext to plaintext.
NOTE

Password

encryption

operating

system.

irreversible.

Therefore,

sure you will never
be
encryption,
enabling
before
need to revert back
to
an
earlier
version
of
the
To enable password encryption, use the CHECKD
program.
You
<can
do
(Refer to the TOPS-20 KL
after system installation.
or
during
this
Model
B
Installation
Guide
for
details,)
With
CHECKD,
password
encryption
1is
enabled
on
a structure-by-structure basis; after the
procedure, all passwords for, a particular structure are
encrypted
as
previously
described.
If
vyou enable encryption after installation,
run the KRYPTN program after
CHECKD
to
convert
existing
plaintext
passwords on a structure to cyphertext.
The KRYPTN program is located
on the tools tape, which is part of the TOPS-20 software
installation
package.

11-4

oyttt

CONTROLS

Encryption should be enabled for all structures except those
be
wused on a TOPS-20 pre-Version 6 system.
(Section 11.2.1

—

ACCESS

this

that will
discusses

——

You

can

—

topic.)

not

add

own

encryption

TOPS-2@'s

your

algorithm.

algorithm

Refer

the

system

Section

you

choose

11.2.2.

Because the encryption
the following areas:

algorithm

irreversible,

care

required

Remembering

one's

Working

Adding

new

Using

DUMPER

password

multiple-system

algorithms

the

environment
system

—

Mistakes

—

need

—

p—— T

—

———

—

——

————

use

topics

these

could

discussed

invalidate

with

the

BUILD or

following

passwords

"ECREATE.

that

These

they

may

interrelated

sections.

11.2.1

Moving

Structures

Among

Systems

If you are in a multiple-system environment,
you
may
need
to
move
structures from one system to another.
Problems could arise, however,
if some systems are running TOPS-20 pre-Version 6 software and
others
are
running
TOPS-20
Version
6.
For
example,
when
a
structure
containing encrypted passwords is taken to
a
TOPS-200
pre-Version
6
system, any access
to files on the pack that requires a password to be
supplied fails, because, in validating a password, the
older
monitor
simply
compares
the
entered
plaintext
to the cyphertext stored on
disk.
The older monitor is unfamiliar with the encryption process.

To avoid this problem, you should
postpone
structures until all systems are upgraded.

encryption

£for

relevant

Any

TOPS-20

system

can

correctly

handle

unencrypted

structures.

You could also
encounter
problems
in
moving
structures
to
other
systems
if
vyou
wuse
your
own
encryption algorithm.
This topic is
discussed below.

11.2.2

Adding

Encryption Algorithms

the

System

You can use one or more of your own encryption algorithms
exclusively
or
in
addition
to
TOPS-20's
algorithm.
For a description of the
procedures involved, refer to the monitor module, STG.MAC,

algorithms

created
since the installation of the new (current) algorithm will be
encrypted with that algorithm.
Older passwords invoke the appropriate
algorithms during password-required accesses.

—

——

—

Each time a password is encrypted
and
stored
1in
a
directory,
the
version
number
of
the
algorithm used to encrypt it is also stored.
This allows new encryption algorithms to be added to the
system
with
no
impact
on
currently
encrypted
passwords,
provided
the
o014

—

——

—

——

—

D avmam

—

——

p—

—

——

—

are

areas

respecified

have not been

removed

from

the

monitor.

Only

passwords

ACCESS

CONTROLS

If you also want existing passwords to
wuse
the
new
algorithm,
the
operator
must
individually
respecify
the
passwords
with BUILD or
"ECREATE.
The
operator
does
this
after
the
new
algorithm
is
installed.
Note
that KRYPTN cannot be used here to convert existing
cyphertext

new

cyphertext,

In using your own encryption algorithms, be aware that directories
on
structures
and
on DUMPER-created tapes include passwords that may be
unusable at other sites.
Other TOPS~20 monitors
could
consider
the
passwords'
algorithm
version
numbers
to
be invalid.
For example,
these monitors may acknowledge only the
standard
TOPS-20
algorithm.
Even
1if
a
site
accepts
the
version
numbers,
1its
corresponding
algorithms may be different from the algorithms at your
site.
Thus,
on
attempted password use, the cyphertext produced at this other site
could
To

never

address
e

match
these

Warn

the

stored

problems,

sites

about

cyphertext.

you
the

could:
nature

the

passwords

tape

structure.
A site could then avoid using certain directories
or respecify a password with BUILD or "ECREATE if necessary.
e

Refrain from saving directory
other
sites.
That
1is,
do
when

creating

Ship
only
"passwords"
problem

11.2.3

Using

such

information on tapes bound
for
not use DUMPER's CREATE command

tapes.

directories
that
have
null
are
considered unencrypted,

any

passwords.
These
and should cause no

system.

DUMPER

Section 11.2.2 addressed using
DUMPER
with
nonstandard
This section continues the discussion of DUMPER,

algorithms.

Care must be taken when restoring directories
that
were
saved
with
DUMPER's
CREATE command.
Incompatible versions of tapes, DUMPER, and
TOPS-20, when combined,
can
produce
a
number
of
password-related
problems.
Note the system's behavior.during tape restoration for the
combinations in Table 11l-1.
1In the
table,
tape
Version
4
is
the

version
the

version

Table

11-1:

tape
of

tape

DUMPER

that
that

DUMPER

Version

DUMPER

Directory

Version

4.1
5

creates.
creates.

Tape Version
‘

Restorations

Tape Version

DUMPER

MONITOR

5
4

5
5

6
6

Result
OK
OK

4.1

5
4

5
5

E2
OK

4.1

11-6

(N1)

ACCESS

CONTROLS

Legend:

Passwords

the
El

are
correctly
encrypted for the
monitor's current encryption algorithm.

The

tape

DUMPER

version

reports

directories
are

The

with
on

files

will
on

the

this

from

incompatible

restoring

this
all

will

BUILD

1is

before

causing

passwords

unencrypted

and

fact

restored

respecified

time,

are

re—-encrypted,

fail.

Any

number

this

tape,
uses

then

"ECREATE.

tape,

then

with

the

this

tape

these

encrypted

are

restored

Pre-version
6
monitors
encryption-related data that

The

this

tape

tape
version
reports this

DUMPER

incompatibility

are

password

for

directory

passwords

for

considered

1l1.3

sure

number.

This

correctly.

results

information

the

same

situation

E2.

users and operators could
on

tapes,

they

that are to be saved.
unencrypted and should cause no

refrain

could

use

Null

passwords

access

problems,

is just one part of password management.
You
that users at your site do not choose passwords

short

or that are
initials.
Also,

passwords

11.4

restored

version

must
be
directory
descriptor

from
null
are

PASSWORD MANAGEMENT

Encryption
make

first

have
no
logic
to
handle
the tape may contain. Therefore,

directories

the

number
is incompatible with this DUMPER.
fact before restoring the tape data. This

these problems occur often,

saving

individually

correctly.

information, such as
the
encryption
descriptor data is not restored.
files

passwords

restored
encrypted
passwords
are
unusable
and
respecified with
BUILD
or
“ECREATE.
Note
that
blocks
on
the
tape
may
contain
password

Any

DUMPER.

data.

usable.
tape

using

passwords

However,
is

time

encrypted
of

have

first

LAST

otherwise easy
it
would
be

to
guess,
such
as
helpful
for
wusers

should
also
that are too

one's
name
or
to change their

regularly.

INFORMATION

When users log into
in
are
displayed

the
on

system, the dates and times they
last
logged
their terminals.
This information helps alert
them to instances of illegal system or account entry.
For example, if
users
keep
track of their actual login times, they can compare those
times to the ones displayed.
Then, if there is
a
discrepancy,
they

will know the exact time
using their password.

that

someone

else

logged

into

their

directory

ACCESS

11.5

PREVENTING

FAST

CONTROLS

LOGINS

By using the /FAST switch with the LOGIN
command,
users
can
bypass
processing
of
their
LOGIN.CMD
and
COMAND.CMD
files.
Managers
sometimes set up these files to limit users'
computing
environments,
however.
For example, sets of users may be allowed only to read mail
or run some other computer application.
You can prevent
fast
1logins
at

your

site

entering

the

following

command

in the n-CONFIG.CMD

file:
DISABLE

FAST-LOGIN-OPTION

The ENABLE FAST-LOGIN-OPTION command is in effect by default.
You can
also enable and disable fast logins with the “ESET privileged command.
Refer to the TOPS-20 User's Guide for information on the LOGIN.CMD and
COMAND.CMD files.
The TOPS-2@ Commands Reference Manual describes the
LOGIN

command.

11-8

CHAPTER
THE

12,1

The

FILE

SYSTEM

OVERVIEW

Common

File

System

from
more
than
structure in the
for

COMMON

reading

(CFS)

feature

TOPS-20

composed

The

main

all

users

the

disk

local

features

allows

users

Any
user

writing.

Each TOPS--2@ system in the CFS configuration
has
system,
main
memory, public structure, console,
and processes to be scheduled and run.
But
the
through
a
shared
file
systemnm.
This
unified
appear

that

one
system
to
simultaneously
access
files.
CFS configuration can be made available to
any

CFS

structures

their

own

all

1its
own
operating
unit-record devices,
systems
are
1linked
file
system can be

systems.

These

structures

systems.

are:

increases file accessibility.
For example, if a system is
for
maintenance, users can log onto another system and
access all files that do not depend on the down
system

down
still
for

access.

It lets you adjust loads on systems by reassigning
loads
require.
(0r,
users
themselves
may
be
switch systems as they
see
fit.)
These
changes
result in file-access limitations.

It lets you
reduce
the
time
that
maintaining duplicate sets of files.

lets

files

you

save

disk

different

space

systems.

12-1

would

minimizing

users
as
allowed to
need
not

involved

duplication

THE

The

CFS

FILE

HARDWARE

following

are

typical

CFS configurations:

DECSYSTEM-20

DECSYSTEM-20
Public
Structure

l Cl-20 I

| Cl-20 I

Star Coupler

N K

HSCS50

Figure 12-1:

SYSTEM

fehcoer

12.1.1

COMMON

MR-8-3905-85

Two Systems with Massbus Disks and HSC5@0-based Disks

12-2

COMMON

FILE

DECSYSTEM-20
Public
Structure

SYSTEM

THE

DECSYSTEM-20

| Cl-20 |

Public
Structure

| Cl-20 I

4——-——— Star Coupler
Figure

Star

12-2:

Two

Systems

with

Massbus

MR-S-3906-85

Disks

Coupler

The star coupler provides
the
cables among DECSYSTEM-2@s and
system and the star coupler is
A DECSYSTEM-2#

can

only one

part

connected
CFS

physical
interconnection
for
the
CI
HSC5@0s.
The maximum distance between a
45 meters.
to

just

one

star

coupler.

That

is,

cluster.

CcI

The Computer Interconnect (CI) bus is the communications link used
by
CFS.
It also connects systems to HSCS5@-based disks (RA60s and RA8ls).
In addition, it provides access to massbus disks for systems without a
direct
connection
to
those
disks, for example, to another system's
public

Each

structure.

system

has

four

communications

links

the

star

coupler.

Two

them
are
for
transmitting
data and the other two are for receiving
data.
The
redundant
CI
connections
are
used
for
increased
availability
and performance.
When one of the connections has failed
or is in use,
the
CI
microcode
<chooses
the
other
one
transmission.
At start-up, TOPS-2@ verifies that at least
transmit and receive connections is working.

for
data
one set of

C120

The
and

CI20 port adapter provides
the CI bus.
Only one CI20

the interface between the
is allowed per system.

DECSYSTEM-20

THE

Massbus

COMMON

FILE

SYSTEM

Disks

Multi-system

access may be granted

all massbus disks.

It is
recommended
that
massbus
disks
intended
to be
shared
be
dual-ported
between
two DECSYSTEM-20s (drive port switches placed in
the A/B position).
With a two-system CFS
cluster,
this
avoids
the
overhead
1involved in file-server activity, as described later in this
section.
However, the systems must be able to communication with each
other
over
the
CI; they must be connected to the same star coupler.
Otherwise, neither system will be allowed access to the disk.
Thus,
following

configurations

are

not

the

supported:

_
\

MR-S-3915-85

.
MR-8-3916-85

—

MR-§-3917-85

In the first two figures, systems G and H are
not
Jjoined
1in
a
CFS
configuration.
The
same
applies
to
systems
H and I in the third
figure.
TOPS-20 maintains the integrity of data on
shared
disks
by
ensuring
that
the
systems
can, over the CI, coordinate accesses to
those

disks.

12-4

THE

COMMON

FILE

SYSTEM

Massbus disks not directly connected to a system
are
called
"served
disks"
because
TOPS-20's
MSCP
(Mass
Storage
Control
Protocol)
file-served facility makes this "outside" access possible.
To
enable
an
outside path to a massbus disk, that is, to make it a served disk,
enter an ALLOW command in the n-CONFIG.CMD file, on a system to
which
the

disk

drive

ALLOW <drive

The drive type
You can obtain

connected,
type>

serial

the

number

CFS

RP@7,

RP20.

DISK-DRIVE

that TOPS-2@ creates an RP2@

disk drive unit number.
among

form:

is one of the following:
RP#4, RP@6,
the serial number with the command:

OPR>SHOW CONFIGURATION

Note

serial

number by adding 8088

the

Therefore, RP2# unit numbers should be unique

systems.
NOTE

Disks

12.1.2

that make

dual

ported

CFS

SOFTWARE

up the

another

Intersystem communication

is an

public

TOPS-20

structure
systemn.

integral

part of

must

not

CFS.

When

TOPS-20

starts
up, it makes a CFS connection with each TOPS-20 system that is
already
running.
This
establishes
the
contact
necessary
for
intersystem

file-system management.

In reality,

only one system writes to a 256K section of a

file

time.
When
a
system
needs
write
access
to
a
file section, it
broadcasts
a
request
for
that
resource
to
all
systems
it
has
established
contact with.
If another system already owns the desired

write access, that system will respond negatively.
Clearance will
be
granted
to
the
requesting
system
only
after the other system has
completed the write operation by writing the data back
to
disk
from
its
memory.
Thus,
systems
negotiate for write access to files and
keep each other informed of the state of the disks
that
they
share.
This ensures the integrity of data on those disks.

Because intersystem communication is
wvital
to
CFS
operations,
the
systems
stay
alert to CI problems and to other indications that they
may have lost contact with each other.
Section 12.11.1, Communication
systems take when there is a
that
actions
the
discusses
Problems,
breakdown

The

communications.

INFORMATION CLUSTER command displays the names of HSCS50s

systems

that

are currently accessible.

and

CFS

THE

DATE

and

When

COMMON

CFS

system

that

starts

are

up,

already

The

date

and

time

the

date

and

command,
time

other
could

CFS

This

systems.

CFS

Users

may

are

types

on the date
is prompted

changed

any

systems

time
not

message

and
for

CFS

are

synchronization

from

the

prompted

similar

for

the

time.
If
no
information.

system,

ensures

one

Otherwise,

many

programs

that

system

other

the

notified

from

are

with

that

the

are

that

such

creation

consistent
use

this

the

they can
date

with

the

information

normally

except

system

"A"

Users

users:

such

are

names,

are

12.1.4

CFS

DEChet

and

configuration

the

prevent

its

being

they

files,

differs

need
all

renamed

about

the

another

time

that

file

being

system

CFS

system

they

are,

read

"B."

configuration.

know

how massbus

disks

they

need

structure

know

are

systems.

command

currently

know

not

from

same

following.

required

CLUSTER

systems

someone

the

non-CFS

TOPS-20

that

file
as

access

cases

will
not

INFORMATION

unaware

accessing

ported.

CFS

written

Specifically,

and

USERS

transparent
e

The

check

other

files

date
operator

malfunction.

12.1.3
CFS

the
The

XXXXXXXXXX

operator

time

all

re-synchronize,

and

takes

the system
terminal:

is:

This typeout serves as
system is running, the

running.

this information.
1Instead,
following on the operator's

"ESET

SYSTEM

TIME

systems

When

FILE

lets

users

accessible

from

DECnet

know

their

what

HSCS5@8s

and

systems.

network.

Although

CFS

configuration
comprises
multiple
independent
systems,
the systems
share a unified file system and
cooperate
in
its
operation.
They
function more as a single system than as systems merely communicating.
If

the optional DECnet-20
running DECnet is a DECnet

software
1is
network node

installed,
with its own

each

node

CFS
name.

system

The files in CFS disk structures may be accessible to
remote
systems
by
way
of
such
DECnet
facilities as NFT.
However, a node name is
needed to access files in this way.
CFS users, on the other hand,
do
not

need

All

systems

DECnet

specify

network,

node

CFS

names.

configuration

however,

other

must

systems

included.

DECnet on a system allows access to other
DECnet

communication

the

HOST

SET

between

systems

command).

12-6

TOPS-20
that

CFS
a

systems.

support

clusters

cluster

(for

DECnet

can

well

example,

as
with

——a——————-—_a———_———_——u——————————-——_——-—.—-———-—————~————-——

THE

Table

12-1:

Comparison of

COMMON

CFS

and

FILE

SYSTEM

DECnet

Characteristic

CFS

DECnet

Multiple

TOPS-20

One

systems
systems

file

Node

system

name

DECnet

only

file

spec

software

12.1.5

CFS5

cluster

CF5

and

TIGHTLY-COUPLED
also

differs

SYSTEMS
from

tightly-coupled

multiprocessing

environments.
Each
CFS system has its own main memory, which is not
shared with another system.
It also has its own public structure
for
booting,
logging
1in, and swapping.
Also, CFS systems do not perform
automatic load balancing.
That is, the CPUs do not relieve each other
of
processing during high job loads.
All jobs, including batch jobs,
run only on the computer that the user logs onto.

12.1.6
CFS

Limitations

does

not

systems:

DBMS

application

OPENF
OF$DUD
restricted to a

facilities
CFS

use

will

cannot

span

1IPCF,

the
and

multiple

following
OPENF

systems,

error

because

DBMS
applications

facility.
Therefore,
such
single system.
Attempts to cross

generate

facilities

OF%DUD.

systems

across

example,

uses

should

using

the

these

messages.

allows

magnetic
printing
system,

12.2

coordinate

ENQUEUE/DEQUEUE,

for shared disk files only.
It does not provide for shared
tapes and line printers.
Thus, for example, all of a user's
will be done on a
1line
printer
that
is
attached
to
his
even if he prints a file that resides on a server disk.

PLACEMENT

This section offers
maximum performance

FILES

guidelines
and

for arranging
efficiency.

12-7

files

CFS

systems

for

THE COMMON FILE SYSTEM

12.2.1

Update

Files

Simultaneous shared writing to a file from multiple systems incurs the
most overhead of any CFS file access operation. This is because
systems involved in shared writing spend time seeking and granting
and coordinating their moves in other ways.
permission
write
Therefore, you might want to place the

involved

users

the

same

system.

12.2.2

Files on Served Disks

For optimum performance, you should not place on served disks files
that require frequent access from multiple systems. This applies to
both reads and writes. MSCP file-server operations incur considerable

overhead, because the system with the direct connection acts as a disk
Therefore, such files should
controller for the accessing system.
m CFS configuration, on
two-syste
a
in
reside on HSC58 disks or,
massbus disks dual ported between systems.

12.2.3

Mail

Files

By default, users' mail files are created and updated in their
logged-in directories on the public structure. To access this mail,

users log in and issue appropriate mail commands. They may have to go
through this login procedure for every system that contains mail for
them. You can change this default arrangement and simplify matters
for the CFS user who has accounts on multiple systems. By redefining
the system-wide logical name POBOX:, as described in Section 3.3.9,
you can establish a central location on a sharable structure for all
mail files in the CFS configuration. Then, no matter where users log
in, the mail facility sees an accumulation of mail that could have
been addressed to them at any system in the configuration. Mail is no
longer isolated on individual public structures.

An added advantage to redefining POBOX:

not fill

up with mail

files.

1is that public structures

You must create a directory on the structure defined by POBOX:
every user in the CFS configuration who is to receive mail.

12.2.4

Sharing

do
for

System Files

Most of the files that normally reside on public structures can be
Rather than duplicate files in such
moved to a shared structure.
areas as SYS: and HLP: across systems, you can keep one set of these
files on a shared structure.. This saves disk space and eases the task
of maintaining the files. Also, time and tape are saved during DUMPER
Because system files are primarily
backup and restore operations.
read and not often updated, system performance does not suffer because
of this file sharing, provided the structure is not on a server disk.

12-8

mamen

definitions
contain the

you

consolidate

system

FILE

files,

for
the
system-wide
files.
For example, if

structure

COMBO:,

DEFINE

(AS)

SYS!

COMMON

the

definition

COMBO:<NEW-SUBSYS>,

SYSTEM

remember

include

1logical
names the
the
SYS:
files

for

8Y¥S:

COMEO:!<SURSYS:»

would

the

structures
reside
on

that
the

be:

MAIN!<NEW-SURSYS>,

MAIN!<SUKSYS:

where

S e

S, ey, S

et B

THE

the

name

public

structure

You

the

should

define
appropriate

structure

able

The
the
Most

structures

access

in this
structure

public

the

are

files

mounted

without

drawback to sharing system
shared structure, users on
of

the

sharing

SYSTEM:

these

files

not

must

way

on
name,

all

UNREGULATED

having

the

Make
give

the

that

giving

the

users

shared

will

MOUNT command.

files is that if there
all systems suffer.
remain

systems,

sure

©public

trouble

structures,

with

recommended.

START-UP

FILES

Certain files must remain on each public
involved
in
system
start-up
and
are
structure is made available to a system.

remain

each

CHECKD
. EXE
DEVICE--STATUS.BIN
DUMP. EXE
ERRMES,.BIN
EXEC.EXE

MONNAM, TXT

e St G

HOSTS. TXT
IPALOD.EXE

e G

area:

ACCOUNTS-TABLE.BIN

KNILDR.EXE

structure.,
These
files
are
required before a non-public
The following
files
should

6-1-~CONFIG,.CMD
6-1-PTYCON.ATO
6-1-SETSPD,
EXE
6-1-SYSJOB.EXE
6-1-SYSJOB.RUN

DR G

. — —— —

— —

—— a— ——

— — — —

— S

— T_—

— —

—

— S

s T

et e

s S

MAIN:

MONITR,
EXE
TGHA.EXE

In addition, all the GALAXY files should remain in each <SUBSYS)> area.
These
files
come from the GALAXY saveset on the TOPS-28 Distribution
Tape.
(Refer to the TOPS-2¢ KL Model B Installation Guide.)
Command
should

files that are used at
be
kept
on
separate

SYSTEM.CMD

and

your installation during start-up
also
public structures.
These files include

NETWORK.CMD.

12-9

COMMON

THE

FILE

SYSTEM

LOAD BALANCING

12,3

This section discusses the distribution of jobs across CFS systems.

Systems

Dedicating

12.3.1

will
that
One way to balance loads is to establish the types of jobs
For example, you might relegate batch jobs
run on particular systems.
jobs
interactive
run
to
systems
other
freeing
system,
to one
users to adopt this arrangement, you could
encourage
To
unimpeded.
the batch-designated
but
give batch jobs the lowest priority on all
Users will have to wait a relatively long time for completion
system.
Refer to Section 1#.2, SCHEDULING
of batch jobs on non-batch systems.
LOW PRIORITY TO BATCH JOBS,

for

further

information.

the
Jjobs
batch
accord
you could
Conversely, on the batch system,
Refer to Section 16.1.4, Procedures to Turn On the
priority.
highest
Iis
Dedicating a system in this manner
Class Scheduler, for details.
there are many long-running batch jobs at an
when
useful
especially
installation.

Another suggestion is to put software development jobs on
one
system
Also, you may want to keep one system
and production jobs on another,
lightly loaded

for

critical

jobs.

DBMS
applications,
and
programming
applications
requiring
ENQUEUE/DEQUEUE
or
IPCF
facilities
must be confined to one system.
certain
establish
These are other items to consider if you choose to
uses

for

particular

systems.

run
to
(without
the
on
in
1logged
user
Batch jobs must be submitted by a
DECnet).
system where they are to run.
The control and log files may reside on

are
that
systems
Keep in mind that users must log onto the
their
particular
jobs.
This
applies
to
batch jobs also
shared

12.3.2

disks.

Assigning

Users

Systems

In the CFS environment, much of the load balancing is expected
to
be
The systems, for example, do not detect that one
by users.
performed
CPU is overburdened
and
that
another
one
is
underutilized,
and,
Instead, users themselves could
Jjobs.
users'
reassign
accordingly,
determine whether or not they should log off a
system
and
1log
onto

Such user tools as the
is slow.
response
system
when
another one
can
function
CTRL/T
the
SYSTAT and INFORMATION SYSTEM commands and
These tools report on the current state of a
help users in this area.

system.
Among the items reported are the number of jobs running on
a
system, load averages, the current setting of the bias control "knob,"
and whether batch jobs are assigned

report

only

the

wuser's

a special

1logged-in

class.

However,

they

system, not on others in the

configuration.

If
you
choose
this
1load
balancing
scheme,
you
should
create
directories
for
all
wusers
on
all the public structures in the CFS
configuration.
Also, directory usernames should be unique
throughout
the
configuration,
as described below.
Then, users can log onto any
system

with

problem.

12-190

THE

COMMON

FILE

SYSTEM

USERNAMES

Directory
For

usernames

example,

should

there

should

unique

only

throughout
one

user

the

same

an installation.
This lets users access system
encountering password-related obstacles or causing

two

users

different

passwords,

their

systems,

same

these

passwords,

they

systems.

switch

domestic,

will

neither

systems

have

passwords
wusers

have

should

complete

Also,

user

will

invalid

chance

structure
to

give

each

usernames

when

they

have

other's

mounted
a

configuration.

username

<BROWN>
resources
without
security breaches.

access

have

CFS

with

password

the

same

files

when

both
when

but

switch

systems
accessing

the directory on that structure that has their
username.
(Refer
to
Section
4,5.7,
Mounting
Structures from Another Installation, for a
discussion of foreign and domestic structures.)
DIRECTORY

AND

facilitate

user

group

could

have

USER

GROUPS

user

access

numbers

change structures or
predictable outcomes.

12.4

STRUCTURE

Because

the

system,

structures

structure

operator

should

12.5

their

unique
same

nor

different

name

disk

will

avoid

Refer

same

and

could

make

directory

structures.

their

name

are
be

That

access

systems

way,

and
users

attempts

areas.

confusion
3.3,

would

unified

throughout

cannot

each

have

structures

its

throughout

with

structures

Having

the

alias,
name,

the

duplicate

using

Same

which

Name.)
the

otherwise

configuration

CFS

public

names,

alias.

unless

CFS

file

the

PS:.

the

part
unique

must

alias.

the

(Refer

The

system

same

the

specified.

refer

NAMES

implemented

1logical

Section

two

Structures

Logical

consistent

system-wide

you

all

systems

identification

LOGICAL

names are
aliases.

logical

must

mount
one

structure

the

all

the

Mounting

everyone
by

SYSTEM

Logical

mount

structure

that

structure

example,

necessary

permanent

Note

For

with

4.5.2,
recognizes
a

files,

systems

names

mounted

Section

CFS

NAMES

configuration.,

consistent

differently
names are local

throughout
two

However,
names

for

the

different

users

CFS

who

configuration.

systems

because

should

from
structure
definitions that

users

12-11

refer

are

consistent

switch

SYSTEM-LOGICAL

can

likely
across

systems.

NAMES,

for

further

names
and
need not be

Thus,

the

two

completely

be mobile,

systems.

This

information.

THE COMMON FILE SYSTEM

12.6

SHARING STRUCTURES AMONG SYSTEMS

to
By default, all structures in the CFS configuration are accessible
systems, provided outside paths have been established for massbus
all
to Section
(refer
the ALLOW command
disks where necessary, using

It is necessary to "mount" a structure on any
CFS HARDWARE).
12.1.1,
operator
That is, the
system that is to access files on it, however.
or a user on that system must issue a MOUNT command for the structure.
a
After
system.)
(There can be up to 32 structures online on one
it as on non-CFS
on a system, users can access
is mounted
structure
Users have automatic access to their public structure files,
systems.
as

non-CFS

systems.

If a structure has been restricted to a system through previous use of
it can be made
SET STRUCTURE str: EXCLUSIVE,
the operator command,
The
SET STRUCTURE str: SHARED command.
the
again with
sharable
operator issues this command from a terminal running OPR on the system
Then, MOUNT commands can be
that has exclusive use of the structure.
The default
that has been made sharable.
structure
the
for
issued
setting

for structures

sharable.

indicates the

The operator command, SHOW STATUS STRUCTURE,
for

exclusive

status

STRUCTURE

ATTRIBUTES

The operator

specifies

command,

STRUCTURE

all

structures

attributes

described

known

for

1in

to a

system.

structure

shared

the

BSET

not

with

the TOPS-2¢ Operator's Command

that
They are permanent settings
Language Reference Manual.
revert to default values after system crashes and reloads.

for
Note that all systems need not have the same attributes in effect
For example, one system can have a structure mounted as
structure.
a
structure
foreign and regulated, and another system can have the same
Except for SHARED and EXCLUSIVE,
mounted as domestic and unregulated.
attributes are on a

12.6.1

Bear

can

Sharing

in mind

create

Public

Structures

OPR>

users

public structure, with the

TO ONE SYSTEM

times when you want

restricts a
SET

any

This may or may not be desirable.

system,
particular
applications (refer
call
measures may
operator

accounts

RESTRICTING STRUCTURES

There may be

basis only.

that when public structures are shared, privileged

privileged

“ECREATE command.

12,7

single-system

restrict use of a

structure

DBMS
for
used
be
might
structure
a
Such
to
Section
12.1.6,
Limitations),
or
security
whatever reason, the
For
use.
restricted
for
structure with the

STRUCTURE

str:

EXCLUSIVE

12-12

following

command:

THE

When

the

operator

gives

COMMON

this

FILE

command,

SYSTEM

the

system

first

checks

see

thatt
the
structure
is
not in use on another system.
If it is, the
operator must dismount the structure from the other system
(with
the
NO-REMOVAL
option), using a terminal running OPR on that system.
The
operator follows the normal dismount procedure of making the structure

unavailable

new users

and

notifying

existing

users

dismount.
The structure should be kept unavailable
for
except
the
exclusive
one
so
that
the
structure
inadvertently shared when the owning system crashes.
After

structure

STRUCTURE
the

has

EXCLUSIVE

system,

been

dismounted

command

all

SET

can

take

STRUCTURE

and
reloads.
If users give the
exclusive
to
another
system,
indicating that the structure is

£from

other

effect.

12,8

DISMOUNTING

any

the

SET

effect

throughout

crashes

structure
will
be

that is
Iissued,

MOUNT command for a
an
error
message
unavailable.
of

pending
systems
not
be

systems,

remains

specifications,

Note that any system can have exclusive use
except another system's public structure.

the

all
will

sharable

structure

STRUCTURES

When issuing a DISMOUNT command for a structure,
operators
have
the
option
of
specifying that the structure be physically removed from a
disk drive.
In the CFS environment, however, the system first ensures

that
the structure is not in use on other systems.
1If a structure is
mounted on another system,
the
operator
1is
notified
and
must
go
through
the
normal
procedure of dismounting the structure (with the
NO-REMOVAL option) from that system.
Refer to the TOPS-20
Operator's
Guide

for

details.

The default setting
with the no-removal

on CFS systems
option.

for

structure

dismounted

Sometimes

the system may instruct the operator to dismount structures,
This occurs when the operator attempts to either shut down a system or
make the CI unavailable to a system.
Refer to Sections 12.9 and 12.12
for details.

12,9

MAKING THE

Ordinarily,

you

CI
need

UNAVAILABLE TO A SYSTEM
do

nothing

all

operate

the

CI.

However,

you
may
need
to
disengage
a
system
from the CI so Field Service
personnel can diagnose and/or correct problems with the
CI28
or
the
HSC54d.
Or,
you
may
wish
to
remove
a
system
from
the
CFS
configuration.
At those times, you should instruct
the
operator
to
make
the
CI
wunavailable
by
means
of
the SET PORT CI UNAVAILABLE
command.
(Refer to the TOPS~2@ Operator's Guide for details.)
When
the
CI
is
unavailable
to
a
system,
users
cannot
access
multi-access
disks
(dual-ported
disks, HSCS5@-based disks, or served
disks on other systems).
These disks rely on
the
CI
to
coordinate
accesses
and/or
to
transmit
data.
Served
disks
on
the
system
disengaging
from
the
CI
will
be
unavailable
to
other
systems.
Dual-ported
massbus disks in the A/B position will have to be powered
down and switched to one system.

12-13

When

the operator gives

THE

COMMON

the

SET

FILE

PORT

SYSTEM

UNAVAILABLE

command,

the

system
indicates
the
structures
that need to be dismounted and the
disk drives that need to be made unavailable.
The operator is advised
to
follow
the
normal
procedures
of
forewarning
users
before
dismounting
structures
and
making
disk
drives
unavailable.
The
command
option
to
forcibly disengage a system from the CI should be
reserved for emergencies.
If the operator determines that disengaging
from
the
CI
will
be
too disruptive to users, the operator has the
option

the

put

aborting

OPR>SET

the

back
PORT

procedure.

in operation,
CI

the operator gives the command:

AVAILABLE

The operator is then asked if any other TOPS-2¢ system is
running
on
the
CI.
If
vyes, the system rejoining the CFS configuration must be
rebooted.
If nho, the CI2@ will be reloaded
and
started.
A
CFRECN
BUGHLT
1is issued on the processor with the lower serial number if the
operator answers "no" and another TOPS-2§ system is
found
after
the
CI20 has started.
(See Section 12,11.1 for details.) After the system
rejoins the configuration, structures that were affected when
the
CI
was made unavailable will need to be remounted.

12.16

USING

DUMPER

CFS offers operators and users flexibility
in
saving
and
restoring
disk
files.
The only restriction is that DUMPER must be running on a
system to which tape drives are attached.
Tape drives are not
served
through

12.11

CFS.

ERRORS

This section discusses the actions
you
or
the
operator
take
when
errors
occur
in
the
CFS
environment.
It
also describes how CFS
systems react to
various
errors.
Note
that
there
1is
no
single
hardware
or
software point that can disable the whole configuration.
For example, systems can start up or crash with little impact on other
systems.

12.11.1

Communication

Problems

CFS systems are sensitive to breaks in communication, whether they are
caused
by
CI20 errors or system crashes.
Because the data integrity
of shared structures depends
on
unbroken
intersystem
contact,
the
systems
take quick action to prevent data corruption.
Therefore, you
may observe any of the following when systems lose contact
with
each
other.

These

should

For up
access
disks,

rare

occurrences.

to 15 seconds, no
system
in
the
any multiaccess disks (dual-ported
served disks on other systems).

12-14

configuration
can
disks, HSC5@-based

THE

COMMON

FILE

SYSTEM

The 15 seconds allows each system to check
and
segment
of
the
CI bus are working.
system's CI20 microcode
has
stopped
and
reloaded
during
(There
may
be

disruption.)
suspended

the
interval,
or
a
other,
unpredictable

Jobs

until

that

data

were

accessing

integrity

that its own
CI20
Most likely, some
is
automatically

system
has
reasons

crashed.
for
C1

multiaccess disks

are

assured.

If the CI20 and CI bus are working
before
the
end
of
the
interval,
the
system
can
resume accessing all multiaccess
disks except server disks on a crashed system.
e

A system crashes with a KLPNRL BUGHLT.
This happens
1if
the
CI20
microcode takes longer to reload than 18 seconds.
This
BUGHLT is expected to occur
rarely,
because
the
microcode
should be reloaded within a couple of seconds.

In a
the

two-system
15-second

crashed
number

configuration, if communication resumes after
allowance,
without one of the systems having

and restarted,
crashes
with

the
system
with
the
CFRECN BUGHLT message.

this occurs when the SET PORT CI AVAILABLE
communication to resume incorrectly due to
described in Section 12.9, MAKING THE
CI

1lower
serial
For example,

command has caused
operator error, as
UNAVAILABLE
TO
A

SYSTEM.

With such a delayed
reconnection,
a
system
1is
1likely
to
contain
old,
invalid
information
about
the
status
of
multi-access disks.
This is
because
the
other
system
Iis
allowed
to
access the disks after the 15 seconds, believing
the other system is no longer running.
Therefore, the system

12.11.2

with

the

1lower 'serial

fresh

data

base

can

system

restarts,

Massbus

Problems

with

number

is selected

established

Dual-Ported

to crash

for

the

Disk

Drives

disks

so that a

when

the

Dual-ported disk drives
are
accessed
by
each
system
through
'the
massbus
hardware
connections.
However, if for some reason a massbus
path becomes unavailable to a system, the other system,
with
working
massbus

connections,

can

the

file

The

MSCP

server.

provide access to the drives
disks become "served."

affected,

with

The operator enables this facility by
powering
down
the
disks
and
flipping:
the drive port switches from the A/B position to the position
that corresponds to the servicing
system.
Then
the
operator
must
reboot
the system with the faulty massbus link.
These procedures are
required because a running system will never invoke
the
MSCP
server
after
identifying
a
massbus path for a disk.
It is assumed that an
ALLOW command has been entered in the n-CONFIG.CMD file for
the
disk

drives,
The

as described

operator

problem
that the

returns

the

Section

12.1.1,

switches

the

CFS

A/B

Hardware.

position

when

1is
corrected.
The
PHYTPD BUGINF is then issued
massbus will now be used for data transmission.

12-15

the

massbus

confirm

THE

12,12

SHUTTING

DOWN

CFS

COMMON

FILE

SYSTEM

When an operator issues the "ECEASE command to
shut
down
a
system,
outside
Jjobs
that
may be accessing the system's served disks do not
hang, with the following procedure.
If any
served
disks
have
been
mounted
from other CFS systems, the operator is warned to check those
systems for possible structure dismounting instructions,

At the other systems, meantime, if any served disks are mounted on the
system
and

shutting

advised

down,

the operator

dismount

the

is warned

structures

12-16

listed.

the

pending

shutdown

CHAPTER

LAT TERMINAL SERVERS

13.1

OVERVIEW

A local area network is a collection of computers and their

resources

(LAT)

software

that are linked together in a small geographic area, such as a college
building.

campus or a large

enables

Local

computers

special-purpose

Transport

Area

to be used as terminal servers in

with LAT software, user terminals that would
such a network.
otherwise be individually wired to host systems (DECSYSTEM-20s, for
example) are instead connected directly to a LAT terminal server. The
server, in turn, is linked to one or more hosts by way of the Ethernet
(NI), as shown below:

Network Interconnect cable
.

|
|

LAT
HOST

frm—m——— +

|
|

LaT
HOST

tm—

tmmmfP———t
|
|
|

|
|

et
|
|
|
|
|
|
t-—e=f

Terminal
Terminal

Terminal

Figure 13-1:

--===-=- |

A LAT Network

is any NI-based terminal

|
|

A LAT host might be a TOPS-28, VMS, or RSX-1ll system.
The main

-t

LAT
-—-—-—===—- |
====——- | SERVER

features of

server.

The LAT

server

LAT servers are:

Users can connect to any NI host that supports the LAT
it appears to them that their terminals have
protocol, and
Connections between LAT
direct connections to those hosts.
There can be
users and hosts are called sessions.
terminal
up to 128 sessions on a TOPS-2§ host.

user
By requesting multiple host connections, a LAT terminal
can enable multiple sessions at one host or at several hosts
simultaneously. With one keystroke, such users can switch
between their

jobs on

these systems.
13-1

LAT

servers

rating
Note

that

This

LAT

you

set

software

runs

host

discusses

13.2

You

can

Refer

hardware

for

LAT

how

also

SERVERS

user

that

DECSYSTEM-20.

server

balance

system

chapter

server.,

can

TERMINAL

loads

systems

control

issue

commands

the documentation
details

among

those

well

LAT

host

directly

according

provided

LAT

servers.

software

to
a LAT

with

your

from

terminal

LAT terminal

commands.

SOFTWARE

Each host that supports the
LAT
protocol
for
service periodically broadcasts this fact to all
Each

hosts

up.

server

maintains

list

hosts

terminal
user
can issue a command
to see which hosts are accessible.
The logical path
circuit.
There

that

the

between a LAT host and
is one virtual circuit

have

interactive
LAT terminal
sent

server

such

terminal
servers.

messages.

display

this

list

server is called a LAT virtual
for each server/host pair that

has at least one
active
terminal
session.
When
a
terminal
|user
requests
a connection to a host, the server creates a virtual circuit
to that host if none exists.
Otherwise an already existing circuit is
used

for

data

transmission.

system

manager,

you

can

max imum number of virtual circuits that can
exist
at a
number that you decide upon can affect system performance.
is discussed in Section 13.4.
Virtual

circuit

messages

periodic

intervals

server.
virtual
interval

When

are

transmitted

determined

data

virtual
message

timer

host

and

associated

circuit
may

with

message

contain

slots

than
one
slot from a
another parameter that

has

particular

wunits

from

many

terminal

are

terminals

single terminal.
you can set.

value

grouped

A virtual

and

may

The maximum

the

LAT

into
a
single
during the past
host.
You
can

recommended

called slots.

the

The
topic

server

maintained

the timer expires, the server assembles
circuit
message
any terminal input received
for a particular host and transmits it to the

set
this
timer only at the server.
milliseconds, which is the default.
The

between

circuit

decide

host.
This

circuit

contain

size

slot

Having received a virtual circuit message, the host assembles as
much
terminal
output
data as possible into a single message and transmits
it to the server.
If no output is waiting for any
terminal
at
that
server,
a message is sent with no slots.
The host's response message
acknowledges the previously received message from
the' server.
This
message
will be acknowledged by the server message transmitted at the
next tick of the circuit timer.
To

reduce

message

when
data

utilization

only

when

reason, the
the server.

dynamic

parameter

sets

the
next
acknowledges
returns

output

host is
(Refer

all

load

the server has no data
waiting
at the host.

unacknowledged,

host

and

there

flow

the

something

host,
to

to transmit,
But because
from

the

there is
the last
host

all

server
It

would

one
the

additional

Section

13.4

for

the

host's

buffers

previously
that

13-2

happen

that

stop.

For

this

"unsolicited" message to
HOST
RETRANSMIT
TIMER

information.)

The

the server to respond at
This
"forced"
response

transmitted

they may

transmits

could

more terminal output
host message remains

permitted to transmit
to the description of

flag in the message, which forces
tick
of
the
circuit
timer.
transmit

the

send.

used

messages

again.

and

LAT

— — —— — — T

— — — — ——— T —— — — — —— —— — —— — — — —

— — it —— — — —— — S—— — — —— it — ——

— T

—

e w— S

— — —

The host maintains a circuit

TERMINAL SERVERS

timer

with

default

interval

sends

(You can set it up to 2 seconds.) The function of the host's
second.
circuit timer differs from that of the server's circuit timer:
it
1is
used

solely

retransmit

timer.

When

the

host

circuit
its
starts
it
above,
described
"unsolicited" message, as
timer.
Because the server's circuit timer is shorter than the host's,
the server should have acknowledged all outstanding host messages well

before

the

host's

Section

13.4.)

timer expires.

If this does not happen, the host

either
This continues until
retransmits all unacknowledged messages.
responds with an acknowledgement, or the retransmit limit
server
the
the
If the retransmit limit is reached, the host assumes
is reached.,
1s no longer useable and detaches all LAT
server
the
to
connection
HOST
the
the description of
to
(Refer
sgserver.
that
jobs from
RETRANSMIT TIMER and the HOST RETRANSMIT LIMIT dynamic parameters in

13.3

DECNET

It does not require
LAT, for the most part, is independent of DECnet.
DECnet is required on at
However,
operations.
for general
DECnet
|is
Software from a host
least one host for start-up of LAT servers.
This
into the server's memory by DECnet's Network Management.
loaded
either
the operator
"down-line loading" of LAT servers occurs when
triggers a boot of the server by
physically boots the server or
case,
In either
issuing a DECnet command from a host running DECnet.
any host that has DECnet and the appropriate load files may respond to
its memory.
load
to
server
the boot, and then be selected by the
needed for "dumping” files of LAT memory images to a
is also
DECnet
Refer to the DEChet-2¢/PSI-20
host for problem diagnosis on the host.
System Manager's Guide for'the operator procedures involved in loading
and dumping

LAT

servers.

If a system supports DECnet, its node name and number are taken to be
You do not need to respecify them in
and number also.
LAT name
the
Refer to the discussion of

the n-CONFIG.CMD file.
in

Section

13.4

static

parameters

13.4.

CONTROLLING LAT FROM THE HOST

There are three ways for you or the operator or a system programmer to
control

LAT

from a host:

By rebuilding the monitor with new permanent parameters.

By changing static parameters in the n-CONFIG.CMD file

By changing dynamic parameters with the LAT
subsystem of

the

The lists below briefly describe these parameters.

fully

Control

Program

OPR program

Many are discussed

later.

to
Refer
Note that you can also control LAT from the server itself,
the documentation that came with your LAT server for details on server
commands.

13-3

Parameters

. ——

Normally,

you

the

TERMINAL

SERVERS

—

Permanent

—

a—

LAT

them,

not

need

system programmer

change

familiar

permanent

parameters.

with monitor

change

internals must

rebuild

ity ———— ——

—

monitor.

MAXIMUM HOST SERVICES - Sets the maximum number
of
services
that
this
host
can
offer.
You
can
specify
up
to 254
services.,
The default and recommended maximum is
8.
Refer
to Section 13.7, HOST SERVICES.

—— S

bytes.

o—

——

FRAME SIZE - The
host
or
server
guarantees
that
it
can
receive
NI datagrams of at least this size.
Usually three (
2 transmit and 1 receive) buffers of this size are
wused
for
each
LAT circuit.
The default and recommended value is 15@4

——— S

—

MAXIMUM SLOTS - Sets the maximum number of slots that may
be
entered
into a virtual circuit datagram for a given circuit.
It is agreed upon by the server and
host
when
the
virtual
circuit
between
them
1is
established.
The
default
and
value

64.

—

recommended

SLOT SIZE - Sets the maximum number of bytes of
data
(not including the slot header) that the host can accept from
the server in any slot.
The slot size can range
from
1
to
255.
The default and recommended value is 4¢.

—

—— ——

—

MAXIMUM

— p—

MAXIMUM

Sets

the

maximum

characteristics

SERVER

CACHE

are

kept

in memory.

number

servers

The

LCP command,

SHOW
SERVER,
displays
these
characteristics.
Section
13.8.3,
Displaying Server Information.)
and recommended value is 280.

(Refer
to
The default

Static

Parameters
DEFAULT

LAT ACCESS

STATE

Determines

LAT

accessibility

and from the host.
Values for the state are ON (the default)
and OFF.
The LAT access state
can
be
changed
dynamically
with
a
LAT
Control
Program
command.
When the system is
reloaded, however, the value for this static parameter
again
takes effect.

S ey

D ctinan

——

ot T

oy P

s Sty

—

a—

—

g, S

arrn

—

———

—

whose

Refer

to
additional

Section

13.5,
information.

STARTING

AND

STOPPING

LAT,

for

HOST NAME - Uniquely identifies the
host
within
the
1local
area
network,
It
may
contain
up to a combination of six
alphabetic
and
numeric
characters,
with
at
least
one
alphabetic
character.
The
default host name is the DECnet
node name, if the system supports DECnet.
If the system does
not
support
information,

discusses

host

DECnhet,
you
refer
also

service

must specify a name.
For
to
Section
13.7.
That
names.

HOST
area

related
section

NUMBER - Uniquely identifies the host within
the
1local
network.
The number can range from @ to 65535.
It is
passed to the server when the virtual
circuit
between
host
and
server is established.
The default number is the DECnet
node number, if the
system
supports
DECnet.
This
1is
an
optional parameter for use by one of your system programmers.

13-4

The

n-CONFIG.CMD

file

TERMINAL

commands

that

SERVERS

set

these

static

parameters

are:

—

o—

LAT

hostname

hostnumber

LAT-STATE

default

LAT

access

state

where

default

LAT access

state

the

OFF

Dynamic

Parameters

—

parameters:

system

manager,

you will

probably be most

concerned

with

dynamic

HOST GROUP CODES - Defines the
group
codes
for
the
host.
They
can range from # to 255,
Code @ is enabled by default.
You can define any number of codes for a host.
A LAT
server
will not connect to the host unless both server and host have
at least one group code defined in common.
Refer to
Section
13.7,

LAT

GROUPS.

HOST IDENTIFICATION - Supplies information
about
the
host.
It
appears
in
various
displays
requested by managers and
users.
The identification may contain
up
to
64
printable

characters.

The

message

SYSTEM:MONNAM,
TXT.

from

default

identification

the

TOPS-20

banner

HOST NUMBER - Uniquely identifies the host within
the
1local
area
network.
The number can range from @ to 65535,
It is
passed to the server when the virtual
circuit
between
host
and
server is established.
The default number is the DECnet
node number if
the
system
supports
DECnet.
This
1is
an
optional parameter for use by one of your system programmers.
HOST MULTICAST TIMER - Determines the interval at
which
the
host
announces
to all servers that its LAT terminal service
is available.
The default value is 68 seconds.
HOST RETRANSMIT LIMIT - Sets the maximum number of times
the
host
retransmits
a
message
at the expiration of
CIRCUIT TIMER.
The virtual circuit is closed
and
all
associated with the virtual circuit become detached when
limit is exceeded.
The default value is 3¢ times.

that
HOST
jobs
this

HOST RETRANSMIT TIMER - Determines the interval at which
the
host
retransmits
any unacknowledged messages to the server.
It is started when the host sends
an
"unsolicited"
message
and
1is
stopped
when the host receives any message from the
server
that
acknowledges
all
outstanding
messages.
The
default
wvalue
1is 1000 milliseconds (one second).
It can be
set from 1000 to 2000 milliseconds, however.
Refer
to
the
description
of
the
HOST
RETRANSMIT
LIMIT
parameter
for
related information.

——

—

——

—

— T

——

—

t—

— et

ity T

D e

—

ey TS

S—

—

— o—

—

——

—

NODE

HOST

SERVICE

NAME

Specifies

the

name

service

that

the

host
provides for users.
It may contain a combination of up
to 16 alphabetic and numeric characters as well as the dollar
sign
(%), dash (-), and underscore (_).
The default service
name is the host name.
Refer to Section 13.7, HOST SERVICES.

13-5

——

LAT

— a——
—

MAXIMUM
virtual

—

G— — o—

—

LAT ACCESS STATE - Determines LAT accessibility to
and
from
the host.
The default state allows LAT access.
This dynamic
parameter is affected by the LCP START and STOP commands.
ACTIVE CIRCUITS - Sets
the
maximum
number
of
LAT
«circuits
that can exist simultaneously at the host.
default value is 20.

—— —

The

—— T

——

HOST SERVICE NAME RATING - Assigns
a
rating
to
a
service
It
can be a fixed number in the range of # to 255 or
be DYNAMIC.
Refer to Section 13.7.1, Service Ratings.

D o— T
—
—

SERVERS

name.
it can

MAXIMUM SESSIONS
Sets
the
maximum
number
of
terminal
sessions
that may be active at the host simultaneously.
The
default value
1is
equivalent
to
the
number
of
terminals
allowed in your system configuration.
Section 13.1 discusses
sessions.

The dynamic
To use LCP,
l.

parameters are changed with the LAT Control Program (LCP).
the operator enters the following from the OPR prompt:
many

LCP

commands:

ENTER

LCP>

<LCP

command>

LCP>

<KLCP

command>

LCP>
LCP>

<LCP
<LCP

command>
command>

LCP>

RETURN

OPR>

LCP

For

single

command:

—

— T

——

— T

—

——

—

OPR>

For

S—

D Shi— ——

TERMINAL

OPR>

LCP

<KLCP command>»

OPR>

The LCP commands
also
let
you
start
and
stop
operations,
check
parameter values and other information, and access counters maintained
by the NI and servers.
These LCP functions
are
discussed
1in
1later
sections.
The
Operator's
Command
-Language
Reference Manual fully
describes LCP,
'

LAT TERMINAL SERVERS

The following

is a summary of the LCP commands:

START

STOP

SET SERVICE-NAME service-name{/RATING:{nn|DYNAMIC}|
/IDENTIFICATION:"
text"|
nothing}

SET
SET
SET
SET
SET
SET
SET

IDENTIFICATION "text"
'
NUMBER number
MAXIMUM ACTIVE~CIRCUITS number
MAXIMUM SESSIONS number
MULTICAST-TIMER seconds
RETRANSMIT LIMIT number
RETRANSMIT TIMER number

CLEAR SERVICE-NAME service-name
CLEAR

IDENTIFICATION

CLEAR

NUMBER

CLEAR MAXIMUM CIRCUITS
CLEAR MAXIMUM SESSIONS
CLEAR MULTICAST-TIMER
CLEAR RETRANSMIT LIMIT
CLEAR RETRANSMIT TIMER

SHOW CHARACTERISTICS
SHOW SESSIONS

SHOW COUNTERS{/SERVER:server-name|
nothing}
SHOW SERVER{/ALL|
server-name|
nothing}

ZERO COUNTERS{/SERVER:server-—namel
nothing}

13.5

STARTING AND STOPPING LAT

That 1is,
Support for LAT servers is enabled by default in ToPS-24.
the LAT access state is enabled unless specifically disabled in the
Disabling it is useful if you wish to establish
n-CONFIG.CMD file.,
You
guidelines and set restrictions for LAT use before you enable it.
can set groups
LAT.
enabling

the host
and
The operator

before
example,
for
identification,
enable and disable the LAT access
can

state dynamically with the LCP START and STOP commands.

The host name and number are other parameters that you specify in the
They are
n-CONFIG.CMD file (unless DECnet is supported on the host).
area
in the 1local
required items that uniquely identify the host

network.

LAT
the
and
established,
After the host name and number have been
access state has been enabled, the operator starts LAT by booting the

server or by issuing a DECnet command from the host, as

Section

discussed

13.3.

It is recommended that you disable the LAT access state if LAT is not
This will improve
intended to be used for an extended period of time.
system

performance.

13-7

13.6

—
— — —
S— S
— e
— — — — — — — —— —
- —— —— —
— — —
i, et et — — — — T

server

255

can

will

assigned

connect

host

DECSYSTEM-2#s

only

and

has

LAT

servers.

least

one

group

defined in common with the host.
(Refer to the LAT documentation
set for information on assigning codes to LAT servers.) Note that code
@, the default for hosts and servers, allows universal
access.
When
servers

and

hosts

implement

the

default,

hosts.

the

example

DECSYSTEM~2f

— —

— — —— T t— — — T

range

TS cm SR i, T

SERVERS

By using group codes, you can
divide
the
1local
area
network
into
smaller
networks.,
That
1is,
you
can
allow or prevent connections
between specified servers and hosts.
When users give a LAT command to
display
the
names
of
avallable
services,
only
those
services
corresponding to
hosts
within
a
server's
"group"
are
displayed.
(Refer
to
Section
13.7
for a discussion of services.) Codes in the
LAT
code

TR e

TERMINAL

LAT GROUPS

—— t—

—— — —

— S

——— ey ot

LAT

User

below,

and

any

connection

LCP>

SET

LCP>

CLEAR

GROUPS

terminals
to access

wired

Group

settings

stay

operator

server

allowed

with

any other

name

for

network).

services.
need

for

the

all

between

this

range

effect

any one

until

reset

these

with

groups

the

will

CLEAR

GROUPS

HOST SERVICES

Another
in

server.

command.

13.7

access

code

access

to servers
system.

the

enables

assigned

have

1:5

GROUPS

able

LAT

users

Users

refer

service

the

access

hosts

order

compile

LAT
to

host

refers

served--to

example.,

LAT
This

terminal

hosts

program,

servers

the

node

fact

have

update

deliver

services

that

some
file,

these
they

node

hosts

being
provide

type
or

services

offer,

system

computing

computing
a

report,

users.

rather

than

host
name.
When
they
request
the LAT server to connect them to a
host, they specify a service name that you have previously established
for
that host.
The server maintains a list of host and service names
and

lets

users

group.

(The

specified.)
more

display

server

With

services

the

for

service

also

assigned

any

SET SERVICE-NAME

service

command,

hosts in
the
user's
identification text you

you

can

host:

specify one

LCP>

SET

SERVICE-NAME

LCP>

LASERPRINTER/IDENTIFICATION "OMEGA

SET

SERVICE-NAME

ARPA-GATEWAY

The

default service name for a
specify
some identifying text
host ALPHA, you could give the
LCP>

You

names

displays

can

section

SET

assign

SERVICE~-NAME

the

discusses

same

this

host is the host name.
for such services.
For
command:

ALPHA/IDENTIFICATION:"A

service

name

topic.

13-8

multiple

System"

You may want to
example, on the

DECnet

hosts.

System"

The

following

LAT TERMINAL SERVERS

Service Ratings

13.7.1

You can arrange

for

LAT

servers

distribute

host

among

wusers

This is useful

systems, according to a rating system that you set up.

systems are
Perhaps several
for hosts with service names in common.
part of a CFS cluster. You can assign the same service name, CFS, on
each host but specify a unique service rating for each one:
LCP> SET SERVICE-NAME CFS/RATING!3/IDENTIFICATION:"ALFHA/BRETA/OMEGA Cluster®
LCP> SET SERVICE-NAME CFS/RATING!9/ILNENTIFICATION!"ALFHA/BETA/OMEGA Cluster"

If a user requests connection to CFS, the server will pick the host
with the highest rating for that service. If, for some reason, that
connection fails, the server will try the host with the next highest
rating.

Ratings can be a fixed number from @ to 255.

with

When

hosts

load

averages,

common

Or they can be

DYNAMIC,

names have DYNAMIC ratings for the

service

system

service, the hosts compute ratings using an algorithm based on

Generally,

the host with the lowest load average is
the

greatest

You can use commoh service names for any collection of hosts

offering

given the

highest

rating.

available computing capacity.
the

same

That

host

probably

has

function.

The default rating is 1 for

the default

service

services created with the SET SERVICE-NAME command.

name,

and

for

Service Rating Example

Hosts SOLAR and LUNAR, in addition to providing a timesharing service
as service names SOLAR and LUNAR, each have the service name CFS.
SOLAR assigns a host rating to name CFS of 5; whereas LUNAR assigns 3.
A LAT terminal user, when displaying the 1list of available LAT
services, will see SOLAR, LUNAR, and CFS.

the

CFS, the server will first attempt to access SOLAR.

will

13.8

try

user

connects

LUNAR.

MONITORING LAT FROM THE HOST

The following sections show various LAT informational displays.

13-9

If that fails, it

13.8.1

Displaying

User

TERMINAL

SERVERS

Information

—

terminals:

SHOW SESSIONS command

displays

information

about

connected

LAT

LCP>SHOW

SESSIONS

LCP>

17:06:46

Job
221
208
209
213
207

[LCP]

Line
460
461
462
463
465

Connected

LAT

Terminals

Program

Server

EXEC
EXEC

Accounting
Accounting

Name

MCCOLLUM
GLINDELL

User

EXEC
EXEC
MS

Accounting
Finance
Accounting

BRAITHWAITE
DDANTONIO
MAYBERRY

211

467

EGPEGP

Accounting

PAETZOLD

205
217
219

4709
471
472

EXEC
OPR
EXEC

Accounting
Finance
Payroll

LOMARTIRE
TUCKER
MELOHN

The

TOPS=20

SYSTAT

user

command

shows

Displaying

Host

Parameters

this

same

information.

13.8.2
The

SHOW CHARACTERISTICS
settings:

command

displays

many

the

LAT

parameter

TOPS-2@

Monitor

LCP>SHOW

CHARACTERISTICS

17:15:13

[LCP]
State:

Name:

ALPHA

—

Access

Host

—

LAT

Host

id:

—— ——

—— —

— —

—

L c—

—

— ——

—— S

t——

. — S

Gin——y r— ————

ey S—— —

—

The

—

——

LAT

Host

number:

—
T

) G

D ———

— G— R

— S

—

Maximum

Characteristics

(KL3138)),

TOPS-20

circuits:

allocated

active

System,

circuits:

Currently active circuits: 4
Maximum sessions: 32
Current sessions: 2
Retransmit Limit: 68
Retransmit Timer: 50
Multicast Timer: 30
Groups: 2,4
Service name(rating): ALPHA(1l)
Service Id: Alpha - More Networks
Service name(rating): ALBET (D)
Service Id: Alpha/Beta Cluster

The "(D)"
rating.

Development

142

allocated

Currently

—— t—

Maximum

ALPHA

Host

that

appears

for

the

per

service

13-1¢

CPU

name

rating

indicates

dynamic

6.1

LAT

13.8.3

Displaying

The SHOW
SERVER
connections
to

Server

TERMINAL

SERVERS

Information

command
displays
this
host.
The

information
about
host
tries to keep

servers
with
information in

memory

on all servers that have connected since the last monitor load.
However,
this
could
require
a
very
1large
data base.
Therefore,
information is kept only for the number of servers
specified
1in
the
MAXIMUM

SERVER

CACHE

permanent

parameter.

information on this parameter.) When
oldest inactive entry is deleted from
new

(Refer

Section

13.4

for

this
number
is
exceeded,
the
the data base, making room for a

entry.

You can
display

specify a single-line
for a single server:

LCP>SHOW

summary

for

each

server

detailed

SERVER/ALL

LCP>

14:55: 32
Server
Server
Server
Server

[LCP]

Name (Number):
Name (Number):
Name(Number):
Name (Number):

LCP>SHOW

SERVER

Summary of

all

servers

Finance(8) Address: #8-9@-2B-gg9-17-BA
Accounting(22) Address: 98-0¢-2B-02-08-C0
Payroll [LAT3] (3) Address: AA-P@-03-091-25-38
Development(2) Address: AA-00-03-01-96-AB

FINANCE

LCP>

14:56:12

[LCP]

Information

about

Server Number: 8
Server Location: Near vending machines
Server Type: DECserver-100
Ethernet Address: #8-¢@-2B-g@-17-BA
Server Status:
Max Slots: 32

Connected

Data Link Size: 1518
Circuit Timer(ms): 80
Keep-—alive Timer(s): 20

13-11

server

Finance

LAT TERMINAL SERVERS

13.8.4

Displaying LAT Counters

The SHOW COUNTERS command displays counters kept by LAT software
These counters provide information such as the number of
modules.
messages transmitted, the number of transmission errors, and so on.

, you can obtain
You can obtain these numbers for a single serveror
cumulative figures for all servers that have connected since the last

monitor

reload:

LCP>SHOW COUNTERS
LCP>

[LCP] Counter totals for all servers

14:48:11

Messages received: 33955
Messages transmitted: 36413
Messages retransmitted: @

Sequence errors received: 21
Illegal messages received: @
Illegal slots received: @

Resource

failures:

LCP>SHOW COUNTERS/SERVER: PUBLICATIONS
LCP>

[LCP] Counters for server PUBLICATIONS

14:48: 34

Messages received: 28189
Messages transmitted: 30132
Messages retransmitted: @
Sequence errors received: 0
Illegal messages received: @
Illegal slots received: @
Resource

failures:

The single-server counts are avallable even after a server has
However, this availability, as the
disconnected from the host.
information displayed with the SHOW SERVER command, depends on the
1imit set with the MAXIMUM SERVER CACHE permanent parameter.

The cumulative counters are incremented each

counters

time

server

individual

It is possible that the sum of all server counters is

are,

not equal to the cumulative counts, because you can zero the

counters

When using the counters to monitor performance or to isolate

hardware

for any server, as discussed below, or the data base may have been
cleared according to the MAXIMUM SERVER CACHE parameter limitation.

failures,

counters.

1is

often

desirable

to be able to reset (or zero) the

The ZERO COUNTERS command lets you do this.

You can

reset

counters for one server or for all the servers. Resetting the
cumulative counts does not affect the counts of the individual
servers.

13-12

APPENDIX

THE

BUILD

COMMAND

NOTE

The information previously found
in
Appendix
A
has
been
removed.
The BUILD command is described in the
TOPS-2¢0
Commands
Reference
Manual.
Refer
to
the
TOPS-2@
Operator's
Command Language Reference Manual
for a description of the "ECREATE command.

A-]1

INDEX

—A~-

Blocking factor
magnetic tape,

ABSOLUTE-ARPANET-SOCKETS

capability,
Access

5-32

control

class

job

(ACJ),

scheduler,

7-4

3-3
BUILD.MEM, 4-15

BUGS.MAC,
1l1-1

10-12

-C-

Accounts
account

data

file

commands,

Cache

6-10
ACTGEN

program

program,

6-14

creating, 6-1
creating the data

disabling,
errors,

6-6

CFS,

6-1

6-15

OPERATOR

6-15

with

6-2
validating,

existing

<ACCOUNTS>,

jobs

12-7,

12-1¢

12-14,

12-15

12-6

and,

12-6

directory groups, 12-11
dismounting structures,

files,

12-13,

12-16
DMP:,

3-18

dual-ported

3-3

DUMPER,

“"ECEASE,

3-13

614

program,

12-5

and,

BUGHLT,

DECnet

6-16

ACCOUNTS~-TABLE.BIN,
ACTGEN

command,

batch

date/time,

selecting schemes, 6-3
setting up for shift changes,

6-2
setting

12~1

ALLOW

CFRECN

account,

18-15

5=32
3-6

CDRIVE.EXE,

base,

6-~1

enabling,

name,

Capabilities,

disks,

12-14,

12-15

12-14

12-16

ENQ-DEQ,

12-7

12«14

ACTGEN.EXE,

3-6

ACTGEN,.HLP,

errors,

3-6

exclusive structures,
File server, 12-4
hardware, 12-2

12-12

AN-MONBIG.EXE,
AN-MONDC
EXE,
N,

3-3

IPCF,

AN-MONMAX.EXE,

3-3

limitations, 12-7
load balancing, 12-7, 12-10
logical names, 12-11
mail files, 12-8
massbus disks, 12-4, 12-15

3-6

capability,

5-32

ARPANET-WIZARD

capability,
Automatic

5-32

Volume

(AVR),

8-14,

8-16

AVR,

MSCP

Recognition

8-16

common policy, 7-3
console front end files,
full dumps, 7-2
7-1

system

crash

Batch

tape,

7-5,

7-6

3-6

BATCON.EXE,

names,

low

priority

to,

systems

12-7
updating

12-8

files,

groups,

CHECKD

Beware file, 1-1
Bias controls, 10-13

and,

1l2-11

files,

program,

CHECKD.EXE,

3-3,

CHECKD,HLP,

CHKPNT, EXE,

3-22
3-6

12-11

tightly-coupled

writing

system

tailoring,

12-8,

usernames, 12-11
users, l2-6

7-4

jobs

scheduling
10-12

12-5

structure

user

requirements,

BASIC.EXE,

7-7

software,

www

tape

12-4,

12-14, 12-15, 12-16
sharing structures, 12-12

3-6

system,

server,

Privileged users, 12-12
Public structure, 12-3, 12-13
served disks, 12-5, 12-8, 12-13,

=B~

B362LB.REL,
Backup

file
12-15

CHKPNT,.HLP,

12-8
4-10

3-6

[«) Wer We))

ANAUNV,UNV,

ARPANET-ACCESS

12-7

making unavailable
9-9

Index~-1

(non-CFS),

CI (CFS), 12-3
Cc120, 12-3
Class scheduler,
CMD.REL, 3-6
CMD.UNV, 3-6

Documents

available from DIGITAL, 1-1
prepared at your site, 1-2
Domestic structures, 4-12

19-2

:
DUMP.CPY, 3-4
DUMP.EXE, 3-4
@gDUMP1l1.BIN, 3-3
DUMPER program, 7-1
CFS systems, 12-14
file archiving, 8-3
file migration, 8-10

3-6
3-6

COBDDT.HLP,
COBDDT.REL,

COBOL.EXE, 3-6
COBOL.HLP, 3-6
Common File System
see

CFS

2-1

Computer room security,
CONFIDENTIAL

password encryption,

5-32

capability,

Console front—-end files,
CREF.EXE,
CREF,HLP,

3-6
3-6

DUMPER.EXE,
DUMPER.HLP,

3-7
3-7

DXMCA,ADX,

3-7

11-6

—F-

-D—

~ECEASE,
EDDT.REL,
EDIT.EXE,
EDIT.HLP,

DECNET-ACCESS

capability, 5-32
DEFAULT-EXEC:, 3-18
Device names, 4-8
DEVICE-STATUS.BIN, 3-3
Diagnostic link
remote (KLINIK), 9-9
DIL.LIB, 3-6
DIL.REL, 3-6
DILV7.FOR, 3-6
Directories
creating, 5-1

ENQ-DEQ

capability,

5-1,

creating

(project and central

creating

(project control),

printing

information about,

5-13

5-2,

5-19

5-2,

5-33

protection code, 5-6, 5=-24

restoring, 9-2
system, 3-1

5-26

Directory group numbers,
Disk drives, 4-9
dual ported, 4-13

dual-ported

(CFS),

12-14, 12-15

dynamic dual porting,
Disk

packs

Disk

space

reinitializing,

program,

DLUSER.EXE,
DLUSER.HLP,

3-7
3-=7

106-16

10-16

allocating, 5-21
determining, 4-16
enforcing quotas, 5-22
permanent quota, 5-21
working quota, 5-21
DITV7.FOR, 3-7
DIXV7.FOR, 3-7
DLUSER

4-19,

12-7

-F-

(central control),

control),

5-32,

ERRMES.BIN, 3-4
EXEC.EXE, 3-4

creating
5-3

12-16
3=-7
3=7
3-7

9-8

3-7
FAL.EXE,
FDB, 8-3
FE.EXE, 3-7
FE.HLP, 3-7
FEDDT,. EXE, 3-4,
FILCOM. EXE, 3-7

3-7

FILCOM.HLP, 3-7
FILDDT.EXE, 3-7
File archiving,

8«1, 8-2
archive cycle, 8-3
recycling tapes, 8-2, 8-11
retrieving files, 8-4, 8-7
sample procedure, 8-5
setting up system, 8-2
when to create tapes, 8-4
File Descriptor Block (FDB), 8-3
File migration, 8-1, 8-7
processing retrieval requests,
'

8-19

recycling tapes, 8-11
setting up system, 8-8
using

DUMPER,

using REAPER,
File

system

restoring,

8-10

8-9

9-7

Files

protection code,
FORDDT.HLP, 3-7
FORDDT.REL, 3-7

Foreign structures,
FORMAT.EXE,
FORMAT.HLP,

Index-2

3-7
3-7

5-23

4-12

7-7

-G~

GALGEN,EXE, 3-7
GLOBS.UNV, 3-8
GLXLIB.EXE, 3-8
Groups, 5-6
directory, 5-26
user, 5-26

LCPORN.REL,
LCPTAB.REL,
LIBARY.EXE,
LIBARY.HLP,

LIBOl12.EXE,

LIBOL.REL, 3-8
LINK.EXE, 3-8
LINK.HLP,

-H-

3-8

Local Area Network, 13-1
Local Area Transport
See LAT, 13-1
Logical names, 3-15
CFs, 12~11
LOGIN command, 6-16
dates/times of login, 11-7
fast login, 11-8

HELP.HLP, 3-8
<HELP>, 3-15
HLP:, 3-17
Home block, 4-3
HOSTS.TXT, 3-4
-]

IBMSPL.EXE, 3-8
INFO.EXE, 3-8
INTERNET.GATEWAYS,
IPALOD.EXE, 3-4

LP64.RAM,

3-8

LP96.RAM,

3-8

LPTSPL.EXE,

3-8

3-4
-M—-

IPCF

capability,

5-32,

12-7

ISAM.EXE,

3-8

MACREL,REL,

ISAM.HLP,

3-8

MACRO.EXE,
MACRO.HLP,
MACSYM.UNV,

-K-

KNILDR.EXE,

3-4,

3-8

-L—-

Labeled tapes,
LAT, 13-1

8-13

3-8

13-2

MAKDMP.EXE,
MAKLIB.EXE,
MAKLIB.HLP,

counters, 13-12
DECnet and, 13-3
features, 13-1

groups, 13-5, 13-8
host identification,
host name, 13-4
host number, 13-4
LCP commands, 13-6

13-5

load balancing, 13-9
multicast timer, 13-5
13-3,

7-4

MAINTENANCE

capability,

timer,

parameters,

3-8
3-8
3-8

Magnetic tape, 8-1
backup requirements,
blocking factor, 7-4
recycling, 8-11
Mail, 3-18, 12-8
MAIL.EXE, 3-8
MAIL,HLP, 3-8
MAILER.EXE, 3-9

KDDT.REL, 3-8
KLINIK, 9-9

¢circuit

4-2,

0 0

3-19,

files,

LAT (Cont.)
starting and stopping service,
13-4, 13-6, 13-7
users, 13-1, 13-9
virtual circuit, 13-2, 13-5,
13-6

13-10

retransmit limit, 13-5
retransmit timer, 13-2, 13-5
server characteristics, 13-4,
13-11
service ratings, 13-9
services, 13-4, 13-5, 13-8
sessions, 13-1, 13-6
slots, 13-2, 13-4

33~
3-

MAKRAM,.EXE,

MAKRAM,HLP,

MAKVFU.EXE,
MAKVFU.HLP,
MAPPER.EXE,

333-

MDDT.REL,

5-32
YWOWOWWYWYWWOOw

Front-end

3-7
3-7

WwwWwww

FOROTS.EXE,
FORTRA.EXE,

3-9
20690-MONBIG,
EXE, 3-3
MONITR.EXE, 3-4
2060-MONMAX,
EXE, 3-3
MONNAM.TXT, 3-4
MONSYM.REL, 3-9
MONSYM.UNV, 3-9
Mountable structure sign-up log,
1-6
Mountable structures, 4-4, 9-8
MOUNTR.EXE, 3-9

Index-3

Performance (Cont.)
interactive versus

-=N—-

n-CONFIG,CMD, 2-3,
n-PTYCON.ATO, 3-3
n-SETSPD, 2-3
n-SETSPD,EXE, 3-3
n-SYSJOB.EXE, 3-3
n-SYSJOB.RUN, 3-3

compute-bound

3-3

program name cache, 10-15
reinitializing disk packs,
19-16

Permanent storage,
PHYPAR.UNV, 3-10

Network Interconnect
See NI, 13-1
<{NEW-SUBSYS>, 3-13
<NEW-SYSTEM>, 3-13
<NEW>,

3-9

Setting unavailable,
NMLT20, 3-9
NORMAL,VFU, 3-9

5-21

PLEASE.EXE, 3-140
POBOX:, 3-18, 12-8
Power failures, 9-9
Privileges, 5-32, 12-12
Program name cache, 1#-15
, 3-4
TXT,
PROGRAM-NAME-CACHE
PROLOG.UNV, 3~-10
PTYCON.EXE, 3-14
PTYCON.HLP, 3-10
Public structure

3-1l4

NFT.HLP,
NI, 13-1

programs,

19-13

9-10

backup,

4-11

contents, 4-2
definition, 4-2
re-creating, 7-5%, 7-6
rebuilding <ROOT-DIRECTORY>,

-0

OLD:, 3-17
<OLD>, 3-14

9-3

OPERATOR

capability,
Operator

accounting,
alternative

5-32
6-15
to PLEASE

requests,

3=-15

archiving procedures, 8-4
handling user requests, 2-1
<REMARKS>, 3-15
scheduling tasks, 2-2
shared disk drive procedure,
4-13

shared

tape drive

procedure,

8-16

Operator shift change log, 1-9
Operator work request form, 1-9
<OPERATOR>,

3-13

-p-

Password

3-9

encryption,

REAPER.HLP,

11-4

RMSUTL.EXE,

algorithms, 11-5
DUMPER program, 11-6

Password management,

5-6,

18-16

3-10

<ROOT-DIRECTORY>

definition,

multiple-system environment,
11-5

3-10

Reinitializing disk packs,
<REMARKS>, 3-15
RERUN.EXE, 3-140
RERUN,HLP, 3-10
RETRFB.SPE, 3-10
RFB.EYE, 3-10
RMS.EXE, 3-10
RMSCOB.EXE, 3-10
RMSINI.REL, 3-10
RMSINT.R36, 3-10
RMSINT.UNV, 3-180

OPR.HLP, 3-9
ORION,EXE, 3-9
OVRLAY.REL, 3-9

PAl@50.EXE,
Page, 4-14

RA60G, 4-9, 7-4
RA81, 4-9, 7-4
RDMAIL.EXE, 3-10¢
RDMAIL.HLP, 3-10
REAPER program, 8-9
REAPER.CMD, 3-4
REAPER.EXE, 3-10

11-7

Performance, 10-1
batch background, 10-12
bias controls, 10-13
class scheduler, 18-2
dynamic dual porting, 10-16

3-2

rebuilding on public structure,
9-3

restoring, 9-2
RpP@4, 4-9, 7-4
RPO6, 4-9, 7-4
RPO7, 4-9, 7-4
RP29, 4-9, 7-4
RSX20F, 3-19

Index-4

RSX20F.MAP,

3-4

RSXFMT,.EXE,

3-10

RSXFMT.HLP,
RUNOFF.EXE,

3-10
3-10

SYS:, 3-16
SYSJOB.HLP,

3-4,

RUNOFF.HLP,

3-190

SYSTAP.CTL,

3-11

Supplies, 2-1
Swapping space,

4-14
3-11

System
-S-

SCSTST.EXE,
SDDT.EXE,

problems,
selecting

3-11

3~11

Security, 5-6, 1ll-1
access control job,
computer

room,

System

access

System

crash

System

log,

11-1

2-1

3-11

SERR:,

SIX12.REL,

<SPOOL>,

drive allocation, 8-2, 8-11
Tape drives
sharing between systems, 8-16
Tape labeling, 8-13

log,

1-9

3-11

TAPNAM. TXT,

TCX.EXE,
TCX.HLP,

3-5

3-1l1
3-11

Terminal server
definition, 13-1
TERMINAL.HLP,

TGHA.HLP,

3-5

TOC.HLP,

3-11

3-14
3-11

Tuning

SPRRET.EXE,

3-11

TV.EXE,

SPRSUM,.EXE,

3-11

Star coupler,
Structures

alias, 4-8
differences

UDDT.EXE,

public

and

4-5

(CFS),

12-13

12-12

ULIST

program,

5-33

ULIST.EXE,

3-11

ULIST.HLP,

3-11

requests,

1-9,

numbers,

assigning,
12-11

5-4,

CFs,

9-8
-y

VERIFY. EXE,

VOLID,

3-11

8-14
——-

WATCH.EXE,

3-11

WATCH.HLP,

3-12

WHEEL

capability,

3-12

2-1
5-26

Usernames

sharing among CF5 systems,
12-12
sharing between systems, 4-13
similarities between public and
mountable, 4-~5
size, 4-9

files, 3-5
restoring,

3-22

User-group

4-7, 4-8
(CFS), 12-11

3-11

UETP,

User

foreign, 4-12
increasing the size of, 4-19
mountable, 4-4, 4-5, 4-6
re-creating,
multiple, 4-5

10-1

3-11
.|

between

mountable,

exclusive

mechanisms,

12-3

dismounting, 4-11
dismounting (CFS),
domestic, 4-12

names

3-5

Tape

SPROUT.EXE,

names,

3-14

Software
after installation, 3-1
before installation, 2-1
checking (UETP), 3-22
updating, 2-3, 3-13, 3-15
SORT.EXE, 3-11
SPEAR.EXE,

1-6

1-3

files, 3-2
restoring,

3-17

Shift change

form,
7-5, 7-6

tape,

last login information, 11-7
password encryption, 11-4
password management, 11-7
passwords vs groups, 5-6
SERCOD,UNV,

2-3

request

<SYSTEM-ERROR>,
SYSTEM.CMD, 3-4
SYSTEM:, 3-16

fast login, 11-8
labeled tapes, 8-14

SELOTS.EXE,

9-1
features,

Windfall,

Index~-5

18-3

5-32

5-8,

5-15,

5-20

Work request form, 1-9
Working storage, 5-21

—-X-

XRMS.EXE,

Index~6

3-12

TOPS-20
System Manager’'s Guide
AA-FP67A-TM

READER’S COMMENTS

NOTE:

This form is for document comments only. DIGITAL will use comments submitted on
this form at the company’s discretion. If you require a written reply and are eligible to
receive one under Software Performance Report (SPR) service, submit your comments on an SPR form.

Did you find this manual understandable, usable, and well-organized? Please make suggestions for improvement.

Did you find errors in this manual? If so, specify the error and the page number.

Please indicate the type of reader that you most nearly represent.
[] Assembly language programmer
[ 1 Higher-level language programmer

[[] Occasional programmer (experienced)

[ User with little programming experience
(] Student programmer
] Other (please specify)

Name

Date

Organization

Telephone

Street

City

State —_ Zip Code

or Country

| || || |

No Postage
Necessary

if Mailed in the
United States

BUSINESS REPLY MAIL
FIRST CLASS PERMIT NO. 33 MAYNARD MASS.

POSTAGE WILL BE PAID BY ADDRESSEE

SOFTWARE PUBLICATIONS
MRO1-2/L12

MARLBOROUGH, MA

01752

Do Not Tear - Fold Here and Tape - - ---------—-——
- —— ———— —_—————— —_———

————

Cut Along Dotted Line

— ——~—

200 FOREST STREET