Digital PDFs

EK-DECRA-UG-001

May 2000

141 pages

Original

5.6MB

view download

OCR Version

5.4MB

view download

Document:	DEC RAID Utilities User's Guide
Order Number:	EK-DECRA-UG
Revision:	001
Pages:	141
Original Filename:

OCR Text

DEC RAID Utilities

User’s Guide
Order Number: EK-DECRA-UG-001

Revision Update Information:

This is a new manual.

Operating Systems/Versions:

MS-DOS Version 5.0
NetWare Version 3.11
SCO UNIX Version 3.2.4

Software Version:

DEC RAID Utilities
Version 1.0

Digital Equipment Corporation

Maynard, Massachusetts

Firet Printing, Novembaer 1093

The information is thus document iw subject to change without notice and should not he conatrued
as m commitment by Digital Equipment Corporation. Digital Equipment Corporation assumes no
responsibility for any srrors that may appear in this document.
The software described in this document is furnished under a license and may be used or copied
only in accordance with the terms of such license.
No responaibility is assumed for the use or reliability of software on equipment that is not supplied
by Digital Equipment Corporation or its affiliated compunies.
Rastricted Rights: Use, duplication, or disclosure by the U.8. Government is subject to restrictions

as set forth in subparagraph (cX1Xii) of DFARS 252.227-7013, or in FAR 52.227-19 or FAR
52.227-14, Alt. III, us applicable.

® Digital Equipment Corporation 1992.
All Rightc Reserved.
Printed in tiae USA.

The postpaid READER'S COMMENTS form on the last page of this document requests the usera
eritical evaluation to assist in preparing future documentation.

The following are trademarks of Digital Equipment Corporation:

BA35X, DEC, DECnet, DECnet -DOS, DEC RAID, DECeerver, DECastation, DECator/me,
DECwindows, OpenVMS, VAX, VAXcluster, VAX DOCUMENT, VAXatation, VMS, VT, and the
DIGITAL logo.
The following are third-party trademarke:

DPT and SmartROM are registered trademarks of Distributed Processing Technology.

Intel is a registered trademark of Intel Corporation.

MS-DOS and Microsoft are registered trademarks of Microsoft Corporation.
Novell and NetWare are registered trademarks of Novell, Inc.
SCO i» a registared trademark of The Santa Cruz Operation, Inc.
UNIX is a registered trademark of AT&T.
ASPI is a registered trademark of Adaptec, Inc.
RAID Manager is a registered trademark of NCR Corporation.
All other trademarks and registered trademarks are the property of their respective holders.

This document was prepared using VAX DOCUMENT, Version 2.0.

Contents
Preface ........... ...
Part |
1

Introduction to the DEC RAID Utilities

Introduction
1.1

Hardware Requirements .. ... ... ....... ..

.. ... .. .. .. .........

1-1

1.2

Software Requirements . ...............
... ... ... .. ... .. . ... ...

1-1

1.21

MS-DOS .

1-1

1.2.2

NetWare . . .. ... e

1-1

1.23

SCOUNIX . ...

1-2

1.3
1.3.1

DECRAID Utility KGits ... ... ... .. . i
..
DEC RAID Stand-Alone Utility Kit (MS-DOS based) .. .........
..

1-2
1-2

1.3.2

DEC RAID NetWare Utility Kit . . ............................

1-3

133

DECRAIDSCOUNIX Utility Kit .. ..........................

1-3

Part Il
2

Description of the RAID Manager Utility

DEC RAID Utilities
....

2-1

2.1.1

2.1

Using the DEC RAID Stand-Alone Utility . ... ...........
... ... .

2-1

2.1.2

Using the DEC RAID NetWare Utility . . . ...........
... .. ... ...

2-1

Starting the DEC RAID Utilities . . ....................
... .. .....

2-2

Starting the DEC RAID Stand-Alone Utility . . .. ................

2-2

22
221

What are the DEC RAID Utilities? ... ..... .. ...

... ... .. ...

222
2.3

Moving Around in the DEC RAID Utilities . .. ..

Starting the DEC RAID NetWare Utility . . ... ...............
...
.......
... ... .. ..

2-2
2-3

24
2.4.1

DECRAID UtilityMenus ...........
... ... .. ... ... .. .. .......
Disk Array Services Window . ... ........... .................

2-4
2-4

242
243

Array Controller Configuration Window. . . ... ... ... ... .. .. ...
DEC RAID Utility Read Map . . .. ...........
.. ... ...........

2-6
2-8

DEC RAID Manager Logical Unit And Drive Status . ... .......
...
.

2-9

26
2.6.1

DEC RAID Manager Device Names . . .. ......... .................
DEC RAID Stand-Alone Utility Device Names .. .. ... ... . .. .

2-11
2-1

26.2

NetWare Device Name . . . . . ..........
... ... .. ... ... .. ..

2-11

Part Il
3

Description of the Array Configuration Editor

Array Configuration Editor (ACE)
31
32
321
322
323
33

Whatis ACE?. . . ... ... . e
Starting ACE . . ... ...
ACEScreza Window . ...... ...
. i,
Navigatiagin ACE. . ...........
... ... .. ... . .. ... ..
Quitting ACE . . ... ... ..
Drive Winfow ... ... ... .
Drive Matrix .. ....... ..

332
333

ACEDriveStatus ............... ...
Drive Window Options . . ... .............
... ...
...
Logica! Unit Window . ...
. .....

3.4.1
342
34.3

3-2
3-3

34
34

3-5
3-5
3-6

Logical UnitDisplay ....... ... .. ... ... . ... ... . ... . ..
ACElogical UnitStatus .. ..........
... ... ... ... ..........
iogical Unit Window Options . . ..................
... ........

Part IV

3-1

3-4

331

3-1

3-6
3-7
3-8

Configuring Logical Units

Configuring Logical Units
4.1

Array Configuration Summary....................
....... ... ...

4.2

Logical Unit Parameters ................
... ... ... ... ... ...
...

... ..

. i

4-1

421

RAID Level

422

DriveMap . ... ... ... ... .. .. .

42.3

Logical Block Size ... ......... ...

... ... . ... . ... ...

4-6

42.4

Logical Unit SiZe . . .. ..ottt
e e

4-6

425

Segment Size ... ... ...

... ...

4-7

426

SegmentZeo Size . . ... . ... ...

4-7

427

Delayinterval ... ... ... ...

428

Blocks Per Delay Interval

4-4

4-5

. . . .

4-7

... ... . ... ...

.. ... ... .........
..

4-8

... .......
. ..... .

4-8

4.3

Displaying and Modifying Logical Unit Parameters.

4.4

Modify/Display Procedure

........................
... 0.0,

4-9

Creating a Logical Unit from Spare Drives .. ......................

4-11

4.6

Creating a Logical Unit from an Existing DriveGroup . ..............

4-13

47
48

Changirg RAID Level/Drives on an Existing Logical Unit
DeletingaLogical Unit . ......... ... . ... ... ... ..........
...

4-15

4-16

Modifying and Displayinga Drive Status

4-17

491

AddingaDrive........

492

DeletingaDrive.

493

FailingaDrive.......

4-18

494

ReconstructingaDrive. .. ...............
... ... ... ... .....

4-19

PartV

...

........................
.
e

4-17

... ...............
... .. .. .. ..
...

4-18

Restoring a Logical Unit

... .. ... .. ...

... ... ... .. . . ...

Restoring a Logical Unit

53.1

Restoring a Logical Unit Task Summary
When to Replace a Drive .. . . . . ..
Drive Failures on RAID O . ... ...
Single Drive Failure on RAID 0

5.4

Drive Failureson RAID1 .. .....

5.4.1

Single Drive Failures on RAID 1
Multiple Drive Failures on RAID 1
Drive Failures on RAID 5 . .. .. ..
Single Drive Failure on RAID 5
Multiple Drive Failures on RAID 5

5.1
5.2
5.3

542

5.5.1
5§5.2
56
5.7
5.9
5.10

..............................

512

DEC RAID NetWare Utility

Instaliation of the DEC RAID NetWare Utility
6.1

6.1.1
8.2
8.3
6.4
6.4.1

6.4.2
6.5

...........................

Reconstruction Rate . . . ....
.. ...
Restoring
a RAID 0 Logical Unit . .
Restoring
a RAID 1 Logical Unit . .
Restoring RAID 1 Logical Units after Multiple Failures . .............
Restoring
a RAID 5 Logical Unit . .
Restoring RAID 5 Logical Units after Multiple Failures

Part VI

About Installation ... . ...
... ..
DEC RAID Manager Files . . . .
When to Install the DEC RAID Manager
Before You Install the DEC RAID Manager

6-1

Running the Installation Program
Using an MS-DOS Partition . .

6-3

6~1

6-3
6-3

6-3

Withcut Using an MS-DOS Partition

Installation Error Messages . . . . . .

Adding Logical Units to NetWare
7.1

Using the Install Command . ..
..

7-1

7.2

Adding Logical Units to Existing Volumes

7-1

7.3

Performance Notes . . .. . ... .....

7-1

Checking/Repairing Array Parity
What is Parity Check/Repair? .. ..
When to Run Parity Check/Repair
How Automatic Parity Check/Repair Works

8-1

8.2.1
83
8.3.1

Automatic Parity Check/Repair Operation
How to Run Manual Parity Check/Repair

8-2
8-3

Manual Parity Check/Repair Operation

8-3

8.3.2

Running the Check .........
Changing Parity Check Scheduling

8.1
8.1.1

8.2

8.4

8-1

8-2

8-3

8-5

9 The Array Monitor Daemon
9.1
9.1.1
9.1.2
913
9.2
9.3
9.3.1
9.3.2
9.3.3

Array Monitor Daemon
How the Array Monitor Daemon Works . . . .........
.. . ... ....
What to do When an Erroris Displayed .. ..........
.. ... .. .. ..
Error Log Example ... ...... ... ... .. ... ... ... ... .. ...
ErrorMessage Types . ... ... ... ... ... ...
AMD Ervor Message Format ... ................................
AMD Pop-Up Message Format . . . . .................
.. ........
AMD Console Message Format . .............................
AMD Error Log Message Format . . .............
... ... ... ...

9.4
9.4.1
9.4.2
9.4.3
9.5

AMDError Messages . . .. ...ttt
.
Array Component Errors . ... ................
... ... ... ... ...
ArrayDevice Errors . ... ....... ... .
Other Status ChangeErrors ... ..................
... ........
DEC RAID Manager Messages ... ..............................

Part VIl
10

.......................................

About Installation

9-2
9-4
94

9-5

9-5
9-6
9-8
9-10

9-11

10-~1
10-2

10.1.2

Installing the RAID Manager Parity Check/Repair Utility
Installing the RAID Monitor Utility

10.1.3

RAID Siatus Utility

10-2

Array Parity Check/Repair
1.1

RAID Parity Check/Repair Utility . . . ...........
... .. ... .....
..

11.1.1

Getting Started . ................... ...

11.1.11

Using the Shell Seript . . ... ... ...

11112

Invoking the Executable

11.1.2

Options

1113

Output .. ...

1114

Error Messages

11-1
111

... ... ... . ... ... .. ..

111

......... ........... .. ... ... ...,

11-2

. ....... ... .. .. ...
... e

. .. ....... ... ... . ... i

11-2
11-3

11-3

RAID Monitor Utility
12.1

RAID Monitor

12.1.1
12.1.11

121.1.2

9-1

Installation of the DEC RAID SCO UNIX Utility
10.1.1

9-1

DEC RAID SCO UNIX Utility

10.1

. @ i

121

Getting Started .. .............. ...

. .......... ...

...

121
12-1

... ....... .. ... ................
.

12-2

. ..... ... ... e

12-2

Using the Shell Script ... . ..... ..

Invoking the Executable

12.1.2

Options

121.3

ErrorMessages

........... ...

...

... ..

...

... ...

. ... .. ... ... .. ...

12-2

RAID Status Utility
131
13.1.1

RAIDStatus Utility .. . ...........
... .. ... ... ... ... ... ..
GettingStarted

. ......... .

131

... .. ... .. ...

13.1.1.1

Usingthe Shell Seript . ... . ... ... ...

13.1.1.2

Invoking the Executable

... .. ...

... ...

...
..

13-1

. ... ... ... . ..

131

... ... ...

13-2

...

13.1.2

Options

......... ... ... i

13-2

13.1.3

Qutput . ... . e

13-3

13.1.4

Error Messages ....................
... . . . ...
.

13-3

SCSI-2 Status Codes

SCSI Sense Keys and Additional Sense Codes

DPT Host Adapter Error Codes

Glossary
index

Figures
2-1

Disk Array Services Window . ... ...... .. ... ... . ... .. .. ....

2-4

2-2

Array Controller Configuration Window. ... ....................

2-6

3-1

ACEWindow .. ... ... .. .

3-2

Conventions ... .......... ... ... .0 iiieiiiiiiiinen

-1

DEC RAID Stand-Alone Utility Kit .. .........................

1-2

DECRAID NetWare Utility Kit . . ...........
... ... .. ........

1-3

DECRAIDSCOUNIX Utility Kit . . ..........................

1-3

2-1

Moving within the DEC RAID ManagerMenus . . ................

2-3

2-2

Logical Unit Status .. ...........
.. ... .. .. ... . ... ... ... ..

2-9

2-3

Drive Status.

2-10

3-1

Understandingthe ACEWindow . ............................

3-3

3-2

ACE NavigationKeys ............
... ... ... ... ... ...,

3-3

Logical Unit Status . ... ..............
... ... . ... ..c......

3-7

4-1

DEC RAID Stand-Alone Utility Logical Unit Configuration Tasks . . . .

4-2

DEC RAID NetWare Utility Logical Unit Configuration Tasks ... ...

4-3

4-3
4-4
5-1
5-2

Logical Unit Parameters ................
... ........c......
RAID LevelDriveSelection ... ..............................
Procedure for Restoring Logical Units . . . . . ...............
... ..
Logical Unit Status . ... ...............
... .. ... ... .........

44
4-5
5-1
5-3

5-3

Reconstruction Rates . .. . ... ... ... ... ... ... ...............

5-8

6-1

Files Copied to the File Serverin NetWare . .. .. ................

6—1

6-2
9-1

Installation Error Messages . . . .. ............................
Array Component Ervors .. ...........
.. ... ... ... .. ... ... ..

6-6
9-6

9-2

ArrayDevieeErrors . . .. ... ... ... ... .. ...

9-8

Tables

. . ........ ... ..

9-3

Other Status ChangeErrors .............................
...

9-10

94
10-1

DEC RAID Manager Messages ..............................
DECRAIDSCOUNIX Utility Kit .. ..........................

9-11
10-1

1-1

Parity Check/Repair Utility Options . .......................
..

11-2

12-1

RAID Monitor Options . . ... .............
...
...

12-2

131

RAID Status UtilityOptions . ...............................

13-2

A-1

SCSI-2 Status Codes for the Array Controller . . .. .......
.. ... ..

A-1

vii

vill

B-1

SCSI Sense Keys

B-2

SCSI Error Codes

c-1

SCSI Status Codes for the DPT Host Adapter . . ... ...........
...

.........................................

Preface
This guide provides a description of the DEC RAIDTM Ultilities and includes:
*

General functional description of the DEC RAID Utilities

Description of the DEC RAID Stand-Alone Ultility (hereafter referred to as the
Stand-Alone RAID Manager)

Description of the DEC RAID NetWareTM (hereafter referred to as the
NetWare RAID ManagerTM)

Intended Audience
This guide is intended for users to understand how to use the DEC RAID Utilities
to install, configure, and manage their DEC RAID Subsystem.

Document Structure
.

Thie guide contains the following parts and chapters:
Part 1: Introduction

Chapter 1: Introduction, discusses the hardware and software
requirements for the DEC RAID Utilities and describes the software
kits.

Part 2: Description of the RAID Manager Utility

Chapter 2: DEC RAID Manager Utilities, discusses when and how to run
a given utility. It also describes the RAID Manager menus.

Part 3: Array Configuration Editor (ACE)

Chapter 3: The Array Configuration Editor, discusses the configuration
utility and its menu.

Part 4: Configuring Logical Units

Chapter 4: Configuring Logical Units, describes how to initially set up
your RAID Subsystem.

Part 5: Restoring a Logical Unit

Chapter 5: Restoring a Logical Unit, discusses recovery procedures for
logical units of various RAID levels.

Part 6: DEC RAID NetWare Utility
®

Chapter 6: Installation of the DEC RAID NetWare Utility

* Chapter 7: Adding Logical Units for NetWare
*

Chapter 8: Checking/Repairing Array Parity, describes the automatic and

Chapter 9: Array Monitor Daemon, describes how the Array Monitor

manual check and repair procedure for NetWare.
Daemon scans and reports errors.

Part 7: DEC RAID 8COTM UNIXTM Utility

Chapter 10: Installation of the DEC RAID SCOTM UNIXTM Utility

Chapter 11: Parity Check/Repair Utility, describes the parity check/repair

Chapter 12: RAID Monitor, and how it reports errors.

utility for SCO UNIX.

Chapter 13: RAID Status Utility, and how it displays current status of

logical units and drives.

Appendix A: SCSI Status Codes

Appendix B: SCSI Sense Keys and Additional Sense Codes
Appendix C: DPTTM Host Adapter Error Codes
Glossary

Associated Documents
In addition to this guide, the following docuinentation may be useful to the
reader:

DEC RAID Subsystem User’s Guide

HSZ10-AX Controller Site Preparation Guide

Conventions
This guide uses the following conventions:

Table 1

Conventions

[Ese)
Fq

Press the Escape key.
Press a function key (for example, F'1).

italics

Used for emphasis.

bold text

Used for user input, new terms, and for emphasis.

Partl
Introduction to the DEC RAID Utilities

This part of the guide contains:

Chapter 1, Introduction

1
Introduction
The DEC RAID Utilities are a collection of utilities designed to provide RAID
configuration and management functionality in. the MS-DOSTM, NetWareTM
and SCO UNIXTM operating system environments. This chapter describes the
necessary hardware and software components required for each operating system.
It also describes the various DEC RAID Utility Kits and their components.

1.1

Hardware Requirements
The DEC RAID Utilities are designed to run with the following hardware
components:

DEC RAID Subsystem (SZ200) with the HSZ10 disk array controller

Distributed Processing Technology (DPT) PM2012B EISA to SCSI differential
host adapter

Supported drives: RZ25 and RZ26

Refer to the RAID Subsystem User’s Guide for further information on the SZ200
RAID Subsystem.

1.2 Software Requirements
The DEC RAID Utilities offer RAID management services in the MS—-DOS,
NetWare and SCO UNIX environments. The following sections discuss DEC
RAID Utility support for each operating system.

1.2.1

MS-DOS
The DEC RAID Stand-Alone Utility (or Stand-Alone RAID Manager) is a

MS-DOS-based utility used primarily to configure the RAID array in an M8-DOS
and/or SCO UNIX environment. The utility is provided on a MS-DOS Version 5.0
bootable tioppy.

1.2.2 NetWare

In addition to the DEC RAID Stand-Alone Utility, the DEC RAID NetWare Utility
(or NetWare RAID Manager TM) provides the same level of functionality in the
NetWare Version 3.11 operating system environment. Additional features include:
°

Array parity check/repair

Array Monitor Daemon to notify users of drive status changes

Introduction

1-1

introduction
1.2 Software Recjuirements
1.2.3 SCO UNIX
Configuration in SCO UNIX Version 3.2.4 is accomplished through the DEC RAID
Stand-Alone Utility. Additional functionality is provided through a set of SCO
UNIX Utilities including:

Array parity check/repair utility

RAID Monitor to notify users of changes in logical unit and drive status

RAID status utility to display current logical unit and drive status

1.3 DEC RAID Utility Kits
Software kits are available for the DEC RAID Utilities for each supported
operating system. The following sections describe the kits and the files contained
within each kit.

1.3.1

DEC RAID Stand-Alnne Utility Kit (MS-DOS based)
The DEC RAID Stand-Alone Utility Kit contains a bootable MS~DOS Version §.0
diskette with the files listed in Table 1-1. This document is also included as part
of that kit.

Table 1-1

1-2

DEC RAID Stand-Alone Utility Kit

Flles

Description

COMMAND.COM

MS-DOS Version 5.0

DPTDDL.SYS

DPT MS-DOS driver

CONFIG.SYS

Configuration file

RAIDMGR.CWA

RAID Manager overlay C-Worthy file

RAIDMGR.EXE

Stand-Alone RAID Manager

RAIDMGR.MSG

RAID Manager message file

RAIDMGR.HLP

RAID Manager help file

RAIDMGR.CFG

RAID Manager configuration file

ACE.EXE

Array Configuration Editor (ACE)

ACF.EXE

Array Configuration Utility (ACF)

FAD.EXE

Array Diagnostic Utility (FAD)

README.BAT

RAID Manager online introduction

READ.ME

RAID Manager online introduction text

PAGE.EXE

Pagination for online introduction

FDISK.EXE

Fdisk utility

ACEHELP <DIR>

Directory of help files

DOS <DIR>

Directory of DOS utilities

Introduction

introduction
1.3 DEC RAID Utility Kits

1.3.2 DEC RAID NetWare Utllity Kit
The DEC RAID NetWare Utility Kit contains the DEC RAID Stand-Alone Utility

Kit as well as a diskette with the files listed in Table 1-2.
Table 1-2

DEC RAID NetWare Utility Kit

Flles

Description

RMINST.NLM

Installation file

RAIDMGR.NLM

NetWare RAID Manager

RAIDMGR.CFG

Configuration file

RAIDMGR.DAT

Menu and form definitions for RAID Manager

RAIDMGR.MSG

Message file

RAIDMGR.HLP

Help file

RAIDIF1.NLM

NetWare file

ACE.NLM

Array Configuration Editor (ACE)

APC.NLM

Array Parity Check/Repair Utility (APC)

AMD.NLM

Array Monitor Daemon Utility (AMD)

RELEASE.ID

Release version number

ACEHELP <DIR>

Directory of help files

1.3.3 DEC RAID SCO UNIX Utility Kit
The DEC RAID SCO UNIX Utility Kit contains the DEC RAID Stand-Alone

Utility Kit as well as a diskette with the files listed in Table 1-3.
Table 1-3

DEC RAID SCO UNIX Utility Kit

Flles

Description

makedpth

Create device files used by utility

help.txt

Help text file

monitor

RAID Monitor executable

monitorsh

Shell script to run RAID Monitor

parity

Parity Check/Repair Utility executable

paritysh

Shell seript to run Parity Check/Repair Utility

raidstat

RAID Status Utility executable

raidstatsh

Shell script to run the RAID Status Utility

read.me

Read me file with general information about the kit

Introduction

1-3

Part i
Description of the RAID Manager Utility

This part of the guide contains:

Chapter 2, DEC RAID Utilities

2
DEC RAID Utilities
2.1 What are the DEC RAID Utilities?
The DEC RAID Utilities allow you to configure, monitor, and repair disk arrays.
There are two versions of the utilities:

DEC RAID Stand-Alone Utility—this ia a MS—-DOS-based program provided
on a bootable MS-DOS Ve rsion 5.0 floppy.

DEC RAID NetWare Utility—this is an NLM used for most array purposes in
NetWare. It is loaded to your NetWare system during the DEC RAID Utilities
installation.

2.1.1

Using the DEC RAID Stand-Alone Utility
Use the DEC RAID Stand-Alone Utility to do the following:
Configure or reconfigure a logical unit in the MS-DOS or SCO UNIX
environment.

Configure or reconfigure a logical unit in the NetWare environment if that
unit contains (or will contain) the operating system.

To restore the logical unit containing your operating system if drives fail
(one or more drives on a RAID 0 logical unit, 2 or more drives with other

RAID levels). Note that thie involves reformatting the logical unit instead of
reconstructing the data on it.

To download controlicr firmware to a controller attached to the logical unit
containing your operating system.

To access logical units after system boot, if for some reason you cannot access
any of your logical units any other way.

2.1.2

Using the DEC RAID NetWare Utility
Use the DEC RAID NetWare Utility to perform all array tasks that do not involve

a logical unit containing the operating system. Specifically, use the DEC RAID
NetWare Utility to do the following:
Configure, reconfigure, and modify logical units that do not contain the
operating system.

Check and repair array parity on logical units.

Check device statuses for the array devices on your system.
Change DEC RAID Manager configuration parameters (scheduled parity
time, parity file name, and so on).
Restore a RAID 1 or RAID 6 logical unit after a single disk failure (on any
logical unit, even one containing the operating system).

DEC RAID Utilities

2-1

DEC RAID VUtliities
2.2 Starting the DEC RAID Utilities

2.2 Starting the DEC RAID Utilities
There are two ways to start DEC RAID Utilities, depending on whether you are
running the DEC RAID Stand-Alone Utility from diskette or running the DEC
RAID NetWare Utility.

2.2.1 Starting the DEC RAID Stand-Alone Utllity
To run DEC RAID Stand-Alone Utility, perform the following steps:
1.

Bring down your operating system in an orderly fashion, if necessary.
Insert the diskette with the DEC RAID Stand-Alone Utility in Drive A:
Reboot your system.

Type raidmgr and press the Return key.

The system starts to load DEC RAID Stand-Alone Utility from the diskette.
There is a delay while DEC RAID Manager scans all the SCSI buses on your
system, looking for arrays. This delay can take 5-10 seconds per adapter.
During this scan, DEC RAID Manager may prompt you if it does not detect
any arrays on your system (for example, arrays may not be turned on or are
not yet ready).

After DEC RAID Stand-Alone Utility has located all the arrays on your
system, the opening menu is displayed. See Section 2.4 for a description of
the DEC RAID ManagerTM menus.

2.2.2 Starting the DEC RAID NetWare Utility
To run the DEC RAID NetWare Utility, perform the following steps:
1.

From the system console, type load raidmgr and press the Enter key.

There is a delay while the DEC RAID Manager scans all the SCSI buses on
your system, looking for arrays. This delay can take 5-10 seconds per adapter.

After DEC RAID NetWare Utility has located all the arrays on your system,
the opening menu is displayed. See Section 2.4 for a description of the DEC
RAID Manager menus.
Note

Do not leave DEC RAID Manager or the NetWare Install utility running
because AMD will not work.

2-2

DEC RAID Utilities

2.3 Moving Around In the DEC RAID Utilities

2.3 Moving Around in the DEC RAID Utilities

Table 2-1 lists how io move around in the DEC RAID Manager menus.
Table 2-1

Moving within the DEC RAID Manager Menus

To:

Press:

Select an option

The up and down arrow keys to highlight the option

Return to the previous menu

The Escape key.

Exit the DEC RAID Stand-Alone
Utility

you want, then press the Enter key.

The Esc key until a popbox appears asking whether
you want to exit DEC RAID Stand-Alone Utility.
To exit DEC RAID Manager, select Y.

To remain in DEC RAID Manager, select N.
To return to the Disk Array Services menu, press
the

key.

DEC RAID Utilities

2-3

DEC RAID Utilities
2.4 DEC RAID Utility Menus

2.4 DEC RAID Utility Menus
DEC RAID Utilities operations are performed by selecting functions from the
various DEC RAID Manager menus. Most of the menus are self-explaratory and

use the same basic type of display. The following sections detail the displays used
by the DEC RAID Manager and give a map showing what selections you must
make from the menus to perform a given function.

Two menus are given, the initial DEC RAID Manager window (Disk Array
Services, Figure 2~1) and the Array Controller Configuration, Figure 2-2. The
major difference between the two windows is that the Disk Array Services
Window shows all the logical units on your system, whereas the Array Controller
Configuration Window gives only the logical units attached to a given array
controller.

2.4.1

Disk Array Services Window
Figure 2-1 shows the Disk Array Services menu. The following ie a brief
description of each menu item. Note that some options are available only on the
DEC RAID NetWare Utility, while others are accessible by both the NetWare and
the DEC RAID Stand-Alone Utility.
Figure 2-1

Disk Array Services Window

RAID Manager

Disk Drive Informetion
Device:

Loglcal Unit;
RAID Lewei:

Channel#

SCSI#

Drive Status

]

OK - Optimal

]

OK - Optimal

Controller:

A
SHA-XR0Z4-DA

2-4

DEC RAID Utilities

DEC RAID Utilities
2.4 DEC RAID Utility Menus

highlighted items can be selected on your system.
Device Selection—Select the logical unit (LUN) you want to access or
for which you want to display information. These options will not be
highlighted if only a single LUN exists.
Device Service (NetWare only)—Restore a logical unit after one or more

disk failures (see Chapter 5).
Log Files (NetWare only)—Display the DEC RAID Manager Activity and
Error logs (see Chapter 9).
Array Parity (NetWare only)—~Check and Repair array parity (NetWare
only, see Chapter 8).
Array Controller Configuration—Configure logical units (see Chapter 4).
DEC RAID Manager Configuration (NetWare only)}—Set the DEC RAID
Manager configuration.

Array Diagnostics—Run diagnostics on the disk array.

Help—Display help. Note that you can get online help at any point by
presging the F1 key. Use [Esc] to exit help.

@ Disk Array Devices Display—Shows all the disk array devices (logical units)
attached to your system.

Device—The device number of the logical unit. For the DEC RAID

Stand-Alone Utility, this number is assigned by the DEC RAID Manager
(see Section 2.6.1). In the DEC RAID NetWare Utility, this number is
assigned by NetWare and is the same device number used for all NetWare
operations (see Section 2.6.2).

Dev ID—The device ID of the logical unit. For the DEC RAID Stand-

Alone Utility, this device ID is assigned by the DEC RAID Manager (see
Section 2.6.1). In the DEC RAID NetWare Utility, this is the same device
ID used for all NetWare operations (see Section 2.6.2).

Disk Array Status—The current status of the logical unit. Logical units
marked as anything other than OK - Optimal need immediate attention
(see Section 2.5 for a description of the statuses displayed).

@ Array Information Display—Gives further information on the logical unit
currently highlighted in the Disk Array Devices display.

Device—The device number of the logical unit (the same device number
displayed in the Disk Array Devices display).
Logical Unit—The logical unit number of the logical unit.
RAID Level—The RAID level of the logical unit.
Controller—The DEC RAID Manager-assigned number of the array
controller attached to the logical unit. DEC RAID Manager numbers
array controllers sequentially, in the order in which they are found.
Disk Drive Information Display—Gives information on the drives contained
in the logical unit currently highlighted in the Disk Array Devices display.
Channel] #~The channel number of the drive.

DEC RAID Utilites

2-5

DEC RAID Utilities
2.4 DEC RAID Utility Menus
SCSI #—The SCSI ID of the drive.
Drive Status—The current status of the drive. Drives marked as anything
other than OK - Optimal need immediate attention (see Section 2.5 for a
description of the statuses displayed).

2.4.2 Array Controller Configuration Window
Figure 2-2 shows the Axrray Controller Configuration Window. The following is a
brief description of each menu item.
Figure 2-2

Array Controller Configuration Window

RAID Manager

Controlier #
Disk Array Devices
Serlal Number 17834
Device

(DeviD)

Disk Array Status

ASPI0004D

OK - Optimal

ASPI0041

OK - Optimal

@ Array Controller Configuration Menu—~Lists the functions you can select.
Only the highlighted items can be selected.

Select Controller—Select the array controller you want to access. In a
single controller configuration, you cannot select this option.
Array Configuration Editor— Configure logical units attached to the
highlighted controller (see Chapter 3 and Chapter 4).
View Controller Information—View information on the highlighted

controlier.

Downloac Controller Firmware—~Download new array controller firmware
to the highlighted array controller.
Service Redundant Controllers—Perform redundant path management.
This feature is currently not available.

2-6

DEC RAID Utlities

DEC RAID Utilities
2.4 DEC RAID Utility Menus
® Disk Array Devices Display—Shows the disk array devices (logical units)
attached to the currently selected controller. (The controller is identified
by controller number and serial number.) To change the currently selected
controller, use the Select Controller function on the Disk Array Services
menu.

Controller #~The DEC RAID Manager-assigned number of the array
controller attached to the logical unit. DEC RAID Manager numbers
array controllers sequentially, in the order in which they are found.
Controller serial number—The serial number of the array controller
attached to the logical units displayed.
Device #—The device number of the logical unit. For the DEC RAID
Stand-Alone Utility, this number is assigned by DEC RAID Manager

(see Section 2.6.1). In the DEC RAID NetWare Utility, this number is
assigned by NetWare and is the same device number used for all NetWare

operations (see Section 2.6.2).
Dev ID—The device ID of the logical unit. For the DEC RAID StandAlone Utility, this device ID is assigned by DEC RAID Manager
(Section 2.6.1). In the DEC RAID NetWare Utility, this is the same

device ID used for all NetWare operations (see Section 2.6.2).
Disk Array Status—The current status of the logical unit. Logical units

marked as anything other than OK - Optimal need immediate attention

(see Section 2.5 for a description of the statuses displayed).

DEC RAID Utilities

2--7

DEC RAID Utilities
2.4 DEC RAID Utiiity Menus
2.4.3 DEC RAID Utility Road Map
The following table lists the DEC RAID Utilities tasks, the DEC RAID Manager
menu selections required from the opening menu to perform the task, and where
to 1ok in this guide for more information.
Tesk

Menu Seiections

Chapler(s)

Change Array Monitor Deemon (AMD)
pecamsters (NetWare)

DEC RAID Manager Configuration
General Configuration

Chapter 9

Change scheduled parity check time

DEC RAID Manager Configuration

Chapter 8

(NetWare)

General Configurstion
or

Array Parity
Change Automatic Parity Time
Check array parity (NetWare)

Array Parity

Chapter 8

Check and Repair Parity
or

Check Parity (no repair)
Configure a logical unit

Arrsy Controller Configuration

Chapter 3

Array Configuration Editor

and 4

Array Controller Configuration

Chapter 3

Array Configuration Editor

and 4

Format a logicel unit after drive failure

Device Service

Chapter &

Get help for selected funetion

Select function, then press the F1 key

Get help with procedurss

Help

Modify & logical unit

Array Controller Configuration

Chapter 3

Array Configurstion Editor

and 4

Status Messages
View DEC RAID Manager Activity Log

Chapter 9

Status Messages

Chapter 9

Delete s logical unit

(NetWare)

Read activity log (NetWare)
Read error messages (NetWare)

Format disk array device

View DEC RAID Menager Error Log

Read parity report (NetWare)

Arrsy Parity

Chapter 8

View Parity Report
Roplace a failed drive (NetWare)

Device Service

Replace the disk drive

2-8

DEC RAID Utilities

Chapter 5

DEC RAID Utilities
2.5 DEC RAID Manager Logical Unit And Drive Status

2.5 DEC RAID Manager Logical Unit And Drive Status
Table 2-2 explains the logical unit status displayed in the Disk Array Devices
window.

Table 2-2

Logical Unit Status

Status

Meaning

Degraded

A drive in the logical unit has failed and the logical unit is
now in degraded mode (RAID 1 and 5 only). The logical unit is
operational in the degraded mode, but the failed drive should

be replaced as soon as possible. Select the logical unit and
check the Disk Drive Information window to determine the
status of the drives.

Degraded—Warning

A drive in the degraded logical unit has been put in warning.
Note that this means one drive has failed and a second
drive about to fail In a RAID 1 logical unit, this status
indicates the most serious condition possible. Steps must be
taken immediately to replace the bad drives to prevent the

permanent loss of data.

Drive Failures

More than one drive in a RAID 5 logical unit has failed, and
the logical unit is no longer operating. All data on the logical
unit has been lost.

Formatting

The logical unit is not available because it is being formatted.

Ok - Optimal

The logical unit is fully operational.

Ok - Reconstructing

The logical unit is functioning in degraded mode and a drive
in the unit is being reconstructed. Select the logical unit and
check the Disk Drive Information window to determine the
status of the drives.

Ok - Warning

The logical unit is still optimal, but one or more drives are in a
warning state because of a read or write error. Note that in a
RAID 0 logical unit, the logical unit may not be usable. Select
the logical unit and check the Disk Drive Information window
to determine the status of the drives.

Wait For Format

The logical unit is not accearible because it needs to be

formatted. In NetWare, use the Format disk array device
gelection in the Device Service menu to format the logical unit.
In the DEC RAID Stand-Alone Utility, use the format option in
the Array Configuration Editor.

Wrong Drive

The wrong drive was replaced (that is, the drive that was
replaced was not a failed drive but was an optimal drive).

DEC RAID Utilities

2-8

DEC RAID Utilities

2.5 DEC RAID Manager Logical Unii And Drive Status

Table 2-3 explains the status of drives as displayed in the Disk Drive Information
window.

Table 2-3

Drive Status

Status

Meaning

Ok - Optimal

The drive is functioning correctly.

Failed

The drive has failed and is no longer functioning. The

Formatting

The drive is currently being formatted.

Newly Replaced

The drive was just replaced.

Reconstructing

The data on the drive is currently being reconstructed.

Warning

The drive has been put into a warning state as the

drive should be replaced as soon as posaible.

result of a read or write error. The severity of this status

depends to some extent on the RAID level of the logical
unit.

On all RAID levels, this status could indicate a minor
read or write error, in which case the logical unit is
still usable.

In RAID level 0, this is the most serious status the
array controller assigns to a drive. The logical unit
may no longer be accessible.

In RAID 1, this status is assigned if the second drive
in a mirrored pair fails after the first drive has failed.
The logical unit may no longer be accessible.

In all cases, the Warning drive should be replaced as soon
as possible.

2-10

DEC RAID Utiiities

DEC RAID Utilities
2.6 DEC RAID Manager Device Names

2.6 DEC RAID Manager Device Names
The DEC RAID Manager displays a device number and a device ID for each
array logical unit on your system. These names are for reference purposes only,
as you do not have to enter them. Both the device numbers and the device IDs
are different depending on whether you are running the DEC RAID Stand-Alone
Utility or the DEC RAID NetWare Utility.

2.6.1

DEC RAID Stand-Alone Utility Device Names
During the initial SCSI bus scan, DEC RAID Stand-Alone Utility assigns a device
number and a device name (Dev ID) to each array logical unit it finds on your
system.

The device number is assigned for reference purposes. You have as many
device numbers as you have array logical units attached to your system.
Other non-array storage devices are not included in this numbering.
DEC RAID Stand-Alone Utility assigns these numbers in the order it detects
the logical unit (which depends on the type of host adapter and where it is
installed).
The device name (Dev ID) identifies the hardware location of the logical

unit. The device name used by the DEC RAID Stand-Alone Utility has the
following format:
adapter type

device number

For example:

ASPI00040

adapter_type—This identifies the interface type of the host adapter

device_number—This 5-digit number provides the hardware address
location of the logical unit. The definition of these digits is shown in the

connected to the controller. The currently supported interface is ASPITM.

following table.

Dight

Meaning

Controller number

SCSI bus number

I/O bus number

Controller SCSI ID

Logical unit number

In this example, the /O bus number is 0. controller number is 0, SCSI bus
number is 0, controller SCSI ID is 4, and the logical unit number is 0.

2.6.2 NetWare Device Name
NetVare assigns each logical unit a device number and a device ID.

The device number is assigned for reference purposes. This number is the
same number used by NetWare to refer to the logical unit in other NetWare
utilities (such as Install). Logical units are numbered along with the other
storage devices on your system.

DEC RAID Utilities

2~11

DEC RAID Utilities
2.6 DEC RAID Manager Device Names
*

The device ID is the NetWare-assigned device ID of the logical unit. This
device ID is the same ID used by NetWare to refer to the logical unit in
other NetWare utilities (such as Install). The digits in the device ID have the
following meanings:

The first 2 digits indicate the type of host adapter. These digits are
assigned by NetWare. For example, 8B is the DPT PM2012B host
adapter.

The 3rd digit is the board number assigned to the host adapter by the

The 4th digit is the SCSI ID of the array controller.

The 5th, or last, digit is the logical unit number.

NetWare system.

For example, a device ID of 8B061 identifies the host adapter as a DPT PM2012B,
board 0, array controller SCSI ID 6, logical unit 1.

2-12

DEC RAID Utilities

Part il
Description of the Array Configuration Editor

This part of the guide contains:
¢

Chapter 3, Array Configuration Editor (ACE)

3
Array Configuration Editor (ACE)
3.1 What is ACE?
The Array Configuration Editor (ACE) is used for configuring logical units on your
array subsystem. Although your arvay is shipped pre-configured from the factory,
you may want to change that configuration. For example, if your subsystem came

with 3 pre-configured, 5-drive, RAID 5'logical units, you may want to change one
of those units to a RAID 0 logical unit. Use ACE to change the configuration.

Although ACE allows you to do other functions (such as reconstruct data on a
drive), some of these functions can more easily be done using other functions in
the DEC RAID Manager. The only operation you must use ACE for is configuring
or reconfiguring logical units.

For more information on how to configure or reconfigure logical units, see
Chapter 4.

3.2 Starting ACE
ACE is an option on the DEC RAID Manager menus. To start ACE:
1.

Start the DEC RAID Manager. See Chapter 2 for information on starting the
DEC RAID Manager. Note that you need to use the DEC RAID Stand-Alone
Utility in the MS-DOS or SCO UNIX environments, or in NetWare if the
logical unit you want to configure contains (or will contain) your operating
system.

From the Disk Array Services menu select the Array Controller Configuration
option.

From the Array Controller Configuration menu select the Array Configuration
Editor option.

The ACE screen displays. Refer to Section 3.2.1 for a description of that

display. See Section 3.2.2 for a description of how to move around in the
screen display.

Array Configuration Editor (ACE)

3--1

Array Configuration Editor (ACE)

3.2 Starting ACE

Figure 3-1

ACE Window

- T
Y
R
L

~eBDON
-0

Logileal View (not physical)

mm—
MODIFY

STATUR

0 Optimal

1 1 Non-gxigtet

A - Add

C - Create

P - Format

" 3 Pailea

P - Falleu

D - Delete

H - Help

2 spare

| 4 Replaosd
L & Warwg

]

T Controler

8 Pormatting

O - Delote

A - Raconstruct
H - Help

M - ModiyDisptey [ = Updsts
parameters

Q@ - Quit

S - auit

? Wrong Orive

ace

SHR-XRI026-GRA

3.2.1

ACE Screen Window
Figure 3-1 shows the ACE display. The display is made up of two parts: The

Drive window (the left half of the screen) and the Logical Unit window (the right
half of the screen).
You can tell which window is active by the label (for example, Drive Window or
Logical Unit Window) that is currently highlighted. When you start ACE, you are
in the Logical Unit Window.

The active window determines which array devices you can select and what parts
of the other window are highlighted.

3-2

Array Configuration Editor (ACE)

Array Configuration Editor (ACE)
3.2 Starting ACE
Table 3-1

Underatanding the ACE Window
Whaen the Logical Unit Window ia Active...

You Select:

Other Window Highlights:

A logical unit (LUN), Drive Group 0,

The drives contained in the selected logical
unit. When you select Drive Group 0, all

or a row with space remaining to be

configured by pressing the arrow keys or
the space bar.

drives that have a Spare status (2) are

highlighted.

When the Drive Window Is Activs...

You Select:

Other Window Highlighte:

An individual drive by pressing the
arrow keys, the space bar, or the Enter

The drive group that contains the selected
drive.

key.

3.2.2 Navigating in ACE
To navigate or select elen.ents in the ACE, use the keys shown in Table 3-2.
Table 3-2

ACE Navigation Keys

Key For

Navigsting
or Sslecting

Dascription of the Key In the
Drive Window

Description of the Key in the
Loglcal Unit Window

TAB key

Maves the cursor to the logical
unit window.

Moves the cursor to the drive
window.

Space Bar

Moves the cursor to all available
drives in the drive matrix.

Moves the cursor down in the
logical unit display.

Down Arrow

Moves the cursor down in the
drive matrix.

Moves the cursor down in the
logical unit display. Moves the
cursor to each logical unit that can
be changed.

Moves the cursor up in the drive

Moves the cursor up in the logical

Up Arrow

matrix.

unit display.

Right Arrow

Moves the cursor right in the
drive matrix.

N/A

Left Arrow

Moves the cursor left in the drive

N/A

matrix,

Enter Key

Moves the cursor to each drive

that is configured into a drive
group.

Moves the cursor to each logical
unit parameter that can be
modified.
Executes certain actions. For
example. when you change the
logical unit parameters. you type Y
and press the Enter key.

Q key

Quits the fellowing:
ACE

ACE

The window for selecting a RAID

level

level.

{continued on next page)

Array Configuration Editor (ACE)

3-3

Array Contiguration Editor (ACE)
3.2 Starting ACE
Table 3-2 (Cont.) ACE Navigation Keys
Key For

Navigating

or Selecting

Description of the Key in the

Description of the Key iIn the

The window for selecting the
Channel and ID of drives

The window for selecting the
Channel and ID of drives.

| Display Parameters
Modify

/ Display Parameters option.
Modify

Drive Window

Logical Unit Window

option

Help option

Help option.

3.23 Quitting ACE
Press the Q key to quit ACE when the cursor is in the drive window or the logical
unit window, as indicated in Table 3-2.

3.3 Drive Window
The Drive Window is the left half of the ACE utility display (Figure 3~1). You are
in the Drive Window when the Drive Window label in the middle of the window is
highlighted. You switch between this window and the Logical Unit Window with
the Tab key.

Use the Drive Window to view:

The status of all the drives connected to a particular array controller

The location of the array controller itself

To modify drive status

The upper half of the Drive Window, the Drive Matrix, displays the current drive
status. See Section 3.3.1 for more information about the Drive Matrix. Notice
that an abbreviated definition of each drive status is always displayed in the
lower left-hand corner of the Drive window, labeled Status. For example, the
drive with SCSI Channel 1, SCSI ID 0 is an optimal drive-—its status is 0, or

optimal. For descriptions of each drive status, see Section 3.3.2.
The Drive window also shows you the operations you can perform while the Drive
Window is active. These operations, or options, are displayed under the Modify
label.

3.3.1 Drive Matrix
The Drive Matrix is located in the upper half of the Drive Window. The Matrix
shows the status of each potential drive position on the array controller you are
accessing. See Section 3.3.2 for a description of the drive status dieplayed in the
Drive Matrix.
The Drive Matrix displays the drives according to a logical order, by SCSI ID, not
according to the physical location of the ranks.

The Drive Matrix also indicates the location of the array controller, which has the
status controller (status of 7), as shown in Figure 3-1.

3-4

Array Configuration Editor (ACE)

Array Configuration Editor (ACE)
3.3 Drive Window

3.3.2 ACE Drive Status

ACE displays the following possible status for drives in the Drive Matrix.
Drive Status

Description

0 - Optimal

The drive is operating at an optimal level.

1 - Non-existant

No drive is physically connected to the array at this pesition.

2 . Spare

The drive is connected to the array, but not configured into a
logical unit.

3 - Failed

The drive was failed by the array controller or by the user and
must be replaced.

4 - Replaced

The drive has just been replaced, or is being formatted, or is

5 - Warning

The drive has been put into a warning state as the result of a
read or write error. The severity of this status depends to some
extent on the RAID level of the logical unit.

being reconatructed.

On all RAID levels, this status could indicate a minor read
or write error, in which case the logical unit is still usable.

In RAID level 0, this is the most serious status the array
controller will assign to a drive. The logical unit may no

longer be accessible.

In RAID 1, this statue is aesigned if the second drive in
a mirrored pair fails after the first drive has failed. The
logical unit may no longer be accessible.
In all cases. the Warning drive should be replaced as
soon as possible.

6 - Mismatch

The array controller sensed that the drive has either a sector
size, capacity, serial number, SCSI Channel, or ID different
than what the array controller expected.

7 - Controller

Thie status is not a drive status—it indicates the location of

8 - Formatting

The drive is currently being formatted.

9 - Wrong drive

The wrong drive was replaced.

the array controller in the array.

3.3.3 Drive Window Options
The following options are given in the Drive window. For more information on
how to perform drive options, see Section 4.9.
*

Add A Drive—Use the Add drive option to add a drive that has a status of
Non-existent (1). After adding the drive, the drive status changes to Spare (2).

Delete A Drive—Use the Delete drive option to delete a drive that has a status
of Spare (2). After physically removing the drive, use this option to change
the drive status of the removed drive to Non-existent (1).
It is not necessary to delete the drive because the array controller does
not attempt to access the drive if it is not physically in the array. If you
reboot the subsystem or server, the status on the removed drive changes to
Non-existent (1).

Array Configuration Editor (ACE)

3-5

Array Configuration Editor (ACE)
3.3 Drive Window
*

Fail A Drive—~—Use the Fail drive option to fail a drive with a drive status
of Warning (5) or Optimal (0). After failing the drive, the status changes to
Failed (3).
Important Note:

Do not fail a drive in a RAID 0 logical unit because there is no parity.

Do not fail a drive in a RAID 1, 3, or 5 logical unit if the logical unit is
already degraded.

Reconstruct Drive Data~Normally, you do not need to initiate drive data
reconstruction because the array controller automatically reconstructs a
replaced drive. However, if your array does not automatically reconstruct
replaced drives, you must use this selection to start reconstruction. You can
only reconstruct drive data on a RAID 1 or 5 logical unit with a single-drive
failure.

Update Drive Status (NetWare only)~The ACE display is updated in
NetWare every 10 seconds. This option allows you to rescan the devices
and update the drive status.

Help—Help menus for drive options.

Quit (NetWare only)—Exit the ACE utility when the drive window is selected.

3.4 Logical Unit Window
The Logical Unit Window is the right half of the ACE utility display (see
Figure 3-1). You are in the Logical Unit Window when the Logical Unit Window
label in the middle of the window is highlighted. Switch between this window
and the Drive window with the Tab key.
Use the Logical Unit Window to view the logical unit status and the device name
of each logical unit. Also use this window to create, delete, or format logical
units, and to modify/display the logical unit parameters.
The upper half of the Logical Unit window, the Logical Unit Display, displays
information about the drive groups and logical units attached to the controller you
are accessing. For more information on the Logical Unit display, see Section 3.4.1.
Also shown in Figure 31 are the operations you can perform while the Logical
Unit window is active. For more information on the Logical Unit window options,
see Section 3.4.3.

3.4.1

Logical Unit Display
The Logical Unit Display is located in the upper half of the Logical Unit window.
It shows the following information:
¢

The logical units (and drive groups) configured on the array controller you are
accessging

3-6

The drive groups of the logical units (there may be more than one logical unit
per drive group)

Logical unit status (see Section 3.4.2 for the possible statuses)

RAID levels of logical units

®*

NetWare device numbers (NetWare only)

Armay Configuration Editor (ACE)

Array Configuration Editor (ACE)
3.4 Logical Unit Window
¢

Logical unit device names

Logical unit size

Disk space remaining in spare drives (this is drive group 0)

Disk space remaining in drive groups with configured logical units

For example, in Figure 3-1, logical unit (LUN) 0 has been configured as RAID
level 5. It has an Optimal status, the device name of ASPI100040 and 1620 MB of
space. LUN 0 is in drive group 1.

3.4.2 ACE Logical Unit Status
LUN Status means Logical Unit Status (LUN stands for logical unit number and

is often abbreviated to logical unit). ACE can display five possible statuses.
Table 3-3 lists each LUN status with a description.
Table 3-3

Logical Unit Status

Status

Description

Creating

You are presently creating this logical unit.

Dead

The logical unit is no ionger functioning. Either:
¢

You have changed logical unit parameters and have not yet
reformatted the unit.

Two or more drives have failed.

The wrong drive was replaced.

Some component connected to the logical unit failed or returned
an unexpected value (bad SCSI ID, bad channel number, and so
on).

Check the status of the logical unit's drives in the Drive Matrix to

help determine the cause of the error.

Degraded’

The logical unit is operating in degraded mode (that ie, it is still

functioning, but data must be reconstructed using data and parity
from the good drives). Either:

A single drive has failed.

The replaced drive is still being formatted.

Some component connected to the logical unit failed or returned
an unexpected value (bad SCSI ID, bad channel number, and so
on).

Check the status of the logical unit’s drives in the Drive Matrix to
help determine the cause of the error.

Optimal?

The array is operating at an optimal level.

Reconstructing

The array controller is currently reconstructing the logical unit.

Yf this is @ RAID 1 logical unit, it may not be accessible even if the LUN Statur is Degraded. To

gfitermmpn'e h‘&f a RAID 1 array has drive problems, check the status of the drives in the logical unit in
e

Drive

Matrix.

2If this is a RAID 0 logical unit, it may not be accessible even if the LUN Status is Optimal. To

determine if a RAID O array has drive problems, check the status of the drives in the logical unit in
the Drive Matrix.

Array Configuration Editor (ACE)

3-7

Array Contfiguration Editor (ACE)
3.4 Logical Unit Window

NoTE

@®

Not all statuses are displayed for all RAID levels. For example, a RAID 0
logical unit will never have a status of Degraded.

3.4.3 Logical Unit Window Options
The following options are given in the Logical Unit Window. For more information
on how to perform these options, see Chapter 4.
»

Create A Logical Unit—~Use the Create logical unit option to create a logical
unit in two ways: by configuring Spare drives, or by configuring available
drives in an existing drive group.

Delete A Logical Unit—~Use the Delete logical unit option to delete a logical

unit to create a new logical unit, or to change the RAID level or drives in an
existing logical unit.

CAUTION

Copy the data on a logical unit to back-up media before deleting the
logical unit. Once the logical unit is deleted, you can not access the logical
unit—~the data is lost.

* Modify/Dieplay Logical Unit Parameters—Use the Modify/Display Logical
Unit Parameters option to modify and display the current values for the
following logical unit parameters:
*

Logical Unit Status (display only)

Logical Unit Drives (display only)

RAID Level (display only)

Logical Block Size in bytes (display only)

Logical Unit Size in megabytes

Segment Size in blocks

Segment Zero Size in blocks

Reconstruction Delay Interval in tenths of a second

Reconstruction Blocks Per Delay Interval

Note that the first three items are also displayed in the Drive Matrix and

Logical Unit Display.

3-8

Array Configuration Editor (ACE)

Array Configuration Editor (ACE)
3.4 Logical Unit Window
Format A Logical Unit—Use the Format option to format a logical unit and

restore it to a working state. Also, use the Format option to restore any

logical unit with multiple drive failures. Note that the DEC RAID Manager
automatically formats any newly-created logical unit.

Update LUN Status (NetWare only)—The ACE display is updated in NetWare
every 10 seconds. This option allows you to rescan the devices and update the
LUN status.

Help—Help menus for LUN options.

Quit—Exit the ACE utility when the LUN window is selected.

Array Configuration Editor (ACE)

3-9

PartlV
Configuring Logical Units

This part of the guide contains:
*

Chapter 4, Configuring Logical Units

4
Configuring Logical Units
4.1 Array Configuration Summary
The drives in the DEC RAID Subsystem were pre-configured into logical units at

the factory. If this configuration does not meet your needs, you can change that
configuration by doing one or more of the following:
*

To modify logical unit parameters cther than RAID level and Drive Map
(segment size, segment 0 size, and so on), see Section 4 4.

To modify RAID Level/Drive Map parameters, see Section 4.7.

To create a new logical unit from spare drives, see Section 4.5.

To create a new logical unit from space on an existing drive group, see
Section 4.6.

To delete a logical unit, see Section 4.8.

In addition, at any time after the initial installation of your array, you may
reconfigure. However, any time you reconfigure, you must first back-up any data
on the logical unit file-by-file because reconfiguring a logical unit destroys data on
the unit.

If your operating system is MS~DOS or SCO UNIX, use the DEC RAID StandAlone Utility for configuring your logical units. Refer to Table 4-1 for a summary
of the steps involved in configuring your logical unit. In NetWare, use the DEC
RAID NetWare Utility (Table 4-2).

The specific steps you must take depend on two factors:
¢

Whether you want to use the factory-set configuration (if any)

Whether you want to put the operating system on the logical unit (for
NetWare, this refers to whether the logical unit will contain the SYS volume).

If your array is pre-configured, and you want to change the configuration, treat

the logicai units as being NOT configured when using the tables.

Configuring Logical Units

4-1

Contiguring Logical Units
4.1 Array Configuration Summary
Table 4-1

DEC RAID Stand-Alone Utility Logical Unit Configuration Tasks

Is Logical Unit

Configured?"

Wil Loglcal Unit
Contain The Operating

System Volume?

Steps To Complete

Yes'

Yen

1. Install the operating system on
the logical unit.

Yes !

1. No action required.

Yeo

1. If any data is on the logical unit,

back-up the data.

2. Boot the DEC RAID StandAlone Utility disketts (Chapter 2).
3. Configure or reconfigure the
logical unit (See the procedures in
this chapter).
4. Install the operating system on
the logical unit.

8. If necesnary, copy the data from

the back-up media to the logical
unit.

1. If any date is on the logical unit,

back-up the data.

2. Shut down the system .

3. Boot the DEC RAID StandAlone Utility diskette (Chapter 2).
4. Configure or reconfigure the

logical unit (See the procedures in
this chapter).

5. Reboot the system.

6. If necessary, recopy the data
back onto the logical unit.

Yf the logical unit is already configured, but you want tbolchmgo that configuration, treat the logical

unit as if it is not already configured when using this table.

4-2

Configuring Logical Units

Contiguring Logical Units
4.1 Array Contiguration Summary
Table 4-2 DEC RAID NetWare Utility Logical Unit Configuration Tasks
Is Logieal Unit

Configured?
Yes!

Wil Logical Unit

Contain SYS Volume?
Yes

Steps To Complete
1. Create the SYS volume on the

logical unit (Chapter 7).

2. Install the operating system on
the logical unit (with the NetWare
Server and Install commands).

3. Install RAID Manager
{Chapter 6).

Yoo !

1. If necessary, install RAID
Manager (Chapter 6).
2. Create partitione and volumes

on the logical unit with the
NetWare Install command

{Chapter 7).

Yes

1. 1f there is any data on the
logical unit, back-up the data
(using file-by-file back-up).
2. Boot the DEC RAID StandAlone Utility diskette (Chapter 2).

3. Configure or reconfigure the
logical unit (see the procedures in
this chapter)
4. Install the operating asystem on

the logical unit (with the NetWare
Server and Install commands).
5. Install RAID Manager
(Chapter 6).
6. If necessary, copy the data from

the back-up media to the logical
unit.

1. If necessary. install RAID
Manager (see Chapter 6).

2. If there is any data on the
logical unit, back-up the data
(using file-by-file back-up).

3. If any volumes on the logical
unit are mounted, dismount them
(with the NetWare Dismount
command).

4. Configure or reconfigure the
logical unit using the DEC RAID
NetWare Utility.

5. Create partitions and volumes
on the logical unit (Chapter 7).
6. If necessary, recopy the data
back to the logical unit.

1if the li?,c‘l uni; is already configured, but you want to change that configuration, treat the logical

unit as

it is not already configured when using this table.

Configuring Logical Units

4-3

Contiguring Logical Units
4.2 Logical Unit Parameters

4.2 Logical Unit Parameters
Bach logical unit has a set of parameters that determine how data is stored on it.
Each logical unit can have different parameters, with the following exception: All
logical units in the same drive group must have the same RAID Level and Drive
Map (that is, they must contain the same drives, with no overlap).
The following table summarizes the parameters. Section 4.2.1 to 4.2.8 explain the
parameters in more detail.
Table 4-3

Loglcal Unit Parameters
Destroys
Data When

See

Yes!

4.2.1

Determines what drives make up

Yes!

4.2.2

Logical Block Size

The logical block size (in bytes)
used by this logical unit.

N/A

4.2.3

Logieal Unit Size

Sets the size of the logical unit.

Yes

4.24

Segment Size

Determines the amount of data
written to a single drive in the
logical unit before the controller
writes data on the next drive.

Yes

4.2.5

Sets the size of the first segment

Yes

4.2.6

427

4.2.8

Paramater
RAID Level

Meaning

Changed?

Determines how data is stored

on the logical unit, and if there

Sectlon:

is data redundancy.

Drive Map

Segment Zero Size

the logical unit.

in the array.

Delay Interval

The amount of time between
reconstruction operations.

Blocks Per Dalay

Interval

The number of blocks

reconstructed in one

reconstruction operation.

‘Chlgg.ng RAID Level or Drive Map forces you to change those parameters for all the logical units in

that

4.2.1

drive group.

RAID Level
The RAID level parameter determines how data is stored on the logical unit.
Data is striped or mirrored. The RAID level parameter also determines if data
redundancy has occurred. RAID levels 1 and 5 offer data redundancy; RAID level
0 does not.

Cautlon
Changing the RAID level parameter deletes any data on the logical unit.
Use this parameter only after doing a file back-up.

The RAID level also determines the number of drives that can be included in the
logical unit and the maximum size of that unit. Refer to Table 44.

4-4

Configuring Logical Units

4.2 Logical Unit Parameters

If you want to change the RAID level for an existing logical unit, you will first
have to delete the unit and then recreate it. In addition, if there are other logical
units in the same drive group, you will have to delete all of them and change
their RAID level as well. Similarly, if you want to create a new logical unit from
space in an existing drive group, that logical unit must have the same RAID level
as the other logical units in the drive group.

The RAID levels offered by RAID Manager are levels 0, 1, and 6. The levels you
select depend on your storage and performance needs.

4.2.2 Drive Map
The Drive Map parameter defines the drives included in the logical unit.
Individual drives are identified by Channel number and SCSI ID, which is

determined by the physical location of the drive in the array subsystem. Each
drive rank in an array has the same SCSI ID, and each rank contains channels
1-5.

Caution

Changing the drive map parameter deletes any data on the logical unit.
Use this parameter only after doing a file back-up.

In ACE, the drive matrix shows the drives in a logical arrangement, with SCSI
ID 0 first, then SCSI ID 1, and so on. This logical arrangement is not the same
as the physical arrangement of the drives in the DEC RAID Subsystem. Refer to

the DEC RAID Subsystem User’s Guide for further details about the DEC RAID
Subsystem physical configuration.
The RAID level of the logical unit sets some restrictions on drive selection, as
shown in Table 4-4:
Table 4-4

RAID Level/Drive Selection

RAID Level

Drive Map Restrictions

Number of drives allowed per logical unit = 1 - 10

Number of drives allowed per logical unit =2 - 10
Must specify an even number of drives

The mirrored pair is created by grouping the first and second drive you
enter, third and fourth, and so forth

Drives in a mirrored pair can not be on the same channel

Number of drives allowed per logical unit=3 -5
Each drive must be on a separate channel

Note

It is recommended that you use the maximum number of drives when
you create a logical unit. Remember that a single-rank array has only 5
drives.

Configuring Logical Units

4-5

Configuring Logical Units
4.2 Loglcal Unit Parameters

From the table, you can see that defining a RAID 6 logical unit as containing

drives (5,3), (4,3), (3,3), (2,3), and (2,0) is not legal, as the last two drives are

on the same channel (channel 2). Selecting (5,3), (4,3), (3,3), (2,3), and (1,3) is
legal, because although the drives have the same SCSI ID, they are on different
channels. Similarly, defining a RAID 1 logical unit as (5,3), (5,0), (4,3), and (4,0)
is not legal, as the first and second drives entered must be a mirrored pair and
mirrored pairs can not be on the same channel. The third and fourth drives have
the same problem. Note that entering (5,3), (4,0), (4,3), and (5,0) is legal because,
even though it contains the same drives as the previous example, the mirrored
pairs are on different channels.

If you want to change the Drive Map for an existing logical unit, you will first
have to delete the unit and then recreate it with the new drives. In addition, if
there are other logical units on the same drive group, you will have to delete all
of them and recreate them as well. Similarly, if you want to create a new logical
unit from space in an existing drive group, that logical unit must include the
same drives as the other logical unit in the drive group.
Within the restrictions above, the drives you select for a logical unit depend
on your storage and performance needs. In general, always use the maximum
number of drives in a logical unit.

4.23 Logical Block Size
The logical block size parameter defines the block size in bytes as seen by the

logical unit. Currently the logical block size is fixed at 512 bytes and cannot be
changed.

4.2.4 Logical Unit Size
This parameter sets the size of the logical unit. Size is determined by the RAID
level, number of drives, and the amount of space allocated on those drives.

Unallocated space on a set of drives (a drive group) can be used to create other
logical units. In general, however, you should allocate all available space for a
logical unit, and create one logical unit per drive group (multiple logical units per
group are useful with operating systems with size restrictions on disk storage).

Note

You may create any size logical unit—~NetWare does not have a size
limitation. MS-DOS, however, does have a limit, It does not display a
logical unit size of greater than 2 GB. This means that if you create a
logical unit with a size greater than 2 GB, and then use an MS-DOS

command to display the size, MS-DOS will only indicate that you have 2
GB space available.

Caution

Changing the logical unit size parameter deletes any data on the logical
unit. Use this parameter only after doing a file back-up.

Assign all available space to the logical unit.

4-6

Configuring Logical Units

‘

Contiguring Logical Units
4.2 Logical Unit Parameters

4.2.5 Segment Size
A segment is the amount of data written on a single drive in the logical unit
before the controller continues writing the data on the next drive in the logical
unit. For example, if the segment size of a RAID 0 logical unit is 64 blacks, the
controller will write 64 blocks of data on drive 1, the next 64 blocks of data on
drive 2, the next on drive 3, and so on.
Caution

Changing the segment size parameter deletes any data on the logical unit.
Use this parameter only after doing a file back-up.

The recommended segment size is 512 blocks. ACE accepts a segment size as
small as 3 blocks and as large as 65,635 blocks (however, do not use an odd
number of blocks if you want to enhance array performance).

4.2.6 Segment Zero Size
The first segment in a logical unit is segment 0. Segment 0 is reserved for
future development. The recommended segment 0 size is 0. Any other value may
degrade array performance.
Caution

Changing the segment size parameter deletes any data on the logical unit.
Use this parameter only after doing a file-by-file back-up.

4.2.7 Delay Interval
Note

This parameter only applies to RAID 1 and 5 logical units. The data on a
RAID 0 array cannot be reconstructed.

During data reconstruction on a RAID 1 or § logical unit (after you replace a
drive in a degraded logical unit), the array controller divides its time between
data reconstruction and regular I/O operations. In this way, you can continue
using the logical unit while it is being reconstructed. The Delay Interval is the
amount of time, in tenths of a second, between reconstruction operations. During
this time, the array controller is able to perform normal I/O operations. As the

delay interval increases, system I/O performance also increases, but the longer
reconstruction takes.
You can change this parameter during reconstruction to affect the reconstruction

rate. The Delay Interval value interacts with the Blocks per Delay Interval
value to determine the over-all rate of reconstruction. See Section 5.7 for further
information.

Configuring Logical Units

4-7

Configuring Logical Units
4.2 Logical Unit Parameters
4.2.8 Blocks Per Delay Interval
Note

This parameter only applies to RAID 1 and 5 logical units. The data on a
RAID 0 array cannot be reconstructed.

The amount of data, in blocks, that the array controller reconstructs at a time.
The more blocks, the longer the time necessary to reconstruct them. This is time
that cannot be used to perform system I/O. Therefore, the larger the Blocks Per
Interval value, the more system performance degradation increases.
You can change the Blocks Per Interval value after reconstruction begins to adjust
system performance.

The Blocks Per Delay Interval value is a decimal number between 1 and 32768.
However, the value interacts with the Delay Interval value to determine the

over-all rate of reconstruction. See Section 5.7 for further information.

4.3 Displaying and Modifying Logical Unit Parameters
This section describes how to display and modify logical unit parameters for an
existing logical unit. The logical unit parameters are described in Section 4.2.
Caution
Note that changing any parameter on an existing logical unit except the
Reconstruction Rate parameters destroys all data on that unit. Make sure
you back-up all data on the logical unit before modifying any parameters
other than Delay Interval or Blocks Per Delay Interval.

You need to use the Display/Modify Logical Unit Parameter function when:
*

You want to change logical unit parameters on an existing logical unit

(except the RAID level and Drive Map parameters, which use a separate
procedure—see Section 4.7.)
Note that changing parameters on an existing logical unit will usually destroy
all data on that unit.
*

You want to change the Reconstruction Rate parameters (Delay Interval

and Blocks Per Delay Interval). Note that you can change these parameters
without destroying data on the logical unit.

4-8

You are creating a new logical unit (in which case this function is started
automatically during the creation procedure).

You want to display the values for the Logical Block Size, Segment Size,
Segment 0 size, and Reconstruction Rate parameters.

Configuring Logical Units

Configuring Loglcal Units
4.4 Modity/Display Procedure

‘

4.4 Modify/Display Procedure
Perform the following steps to display or modify logical unit parameters:
Caution
Changing any logical unit parameter except the Delay Interval or Blocks
Per Delay Interval parameters destroys all data in the logical unit. Make
sure you back-up your data file-by-file before modifying any of these

logical unit parameters.

If necessary, back-up any data on the logical unit you want to change by
copying it file-by-file to the back-up medja.

In NetWare, unmount any volumes contained on the logical unit you want

to change if any are mounted (if you are changing the Reconstruction Rate
parameters, you do not need to dismount any volumes).

If you are going to be modifying parameters in MS-DOS or SCO UNIX, or in
NetWare if you are going to be modifying parameters on the logical unit that
contains any part of your SYS volume, you will have to use the DEC RAID
Stand-Alone Utility. Reboot the system using the DEC RAID Stand-Alone

Utility diskette and continue with this procedure.
4.

Start DEC RAID Manager from the console by entering:
In the DEC RAID Stand-Alone Utility:

raidmgr

In the DEC RAID NetWare Utility:
load raidmgr

Select the Array Controller Configuration option on the Disk Array Services
menu. Check the Disk Array Devices window on the right of the screen to
make sure it is displaying the logical unit you want to modify.

Select the Array Configuration Editor option from the menu. Note that you
are in the Logical Unit window of the Array Configuration Editor (ACE). If
not, press the Tab key to switch to the Logical Unit window. Use the Arrow
keys to highlight the logical unit you want to modify.

Press the M key to modify or display the logical unit parameters. The
following screen is displayed:
LUN 1 Status = Optimal
LUN 1 Drives = (1,2) (2,2)

(3,2)

(4,2)

PARAMETER
RAID Level
Logical Block Size (bytes)
Logical Unit Size (MB)
Segment Size (blocks)
Segment Zero Size (blocks)
Delay Interval (tenths of a second)
Blocks Per Delay Interval

(5,2)
VALUE
5
512
1600
512
0
1
256

Are the values shown OK (y/n/q}?

Configuting Logical Units

4-9

Configuring Logical Units

4.4 Modity/Display Procedure
8.
*

To save the current logical unit parameters for this logical unit, press the
Y key. The array controller automatically formats the logical unit if you
modified the logical unit size, segment size, or segment zero size.
In NetWare, if you created a very large logical unit, and you press Alt-Esc
to do an MS-DOS operation while the format is occurring (for example,
you try to load Install), it is possible that the screen will lock up until
the array controller completes the format. This situation occurs because
of the way the NetWare driver handles outstanding 1/0 commands. Wait
until the array controller completes the format before doing an MS~-DOS
operation.

Caution
Changing the values for either the logical unit size, segment size or

segment zero size forces the array controller to automatically reformat
the logical unit. This will Jelete all data on the logical unit. If you do not
want to change the values, press the @ key to quit the function.
Press the C key to continue.
¢

To modify a value, prees the N key. Use the Down-arrow key or the Enter
key to melect the value you want to change and enter the new value.
Remember, you cannot change the RAID level parameter, or the Logical
Block Size (to determine the values you may use for the logical unit
parameters, see Section 4.2.)

After you have set all the parameters you want to set, use the Downarrow key or the Enter key to move to the Are the ralues shown OK
prompt, then press:

the Y key to save the current logical unit parameter values.

the Q key to quit without changing the logical unit parameters.
Cautlon

Changing the values for either the logical unit size, segment size or
segment zero size forces the array controller to automatically reformat

the logical unit. This will delete all data on the logical unit. If you do not
want to change the values, select the Q key to quit the function.

Press the C key to continue.
*

To quit the display without changing the parameters, press the Q key.

Press the Q key to quit the Array Configuration Editor. You return to the
Array Controller Configuration menu. Press the Esc key to quit DEC RAID

Manager. Then press Y and the Enter key.

10. In DEC RAID Stand-Alone Utility, remove the diskette and reboot the system.
In DEC RAID NetWare Utility, use the NetWare Install command to create
partitions and volumes on the modified logical unit (Chapter 7).

You are done with the Display/Modify Logical Unit Parameter procedure.

4-10

Configuring Logical Units

Configuring Logical Units
4.5 Creating a Logical Unit from Spare Drives

4.5 Creating a Logical Unit from Spare Drives
Use the following procedure to create a logical unit from Spare drives.
1.

If you are going to be creating a logical unit in MS-DOS or SCO UNIX, or in
NetWare if you are going tc be creating the logical unit that will contair your
SYS volume, you will have to use the DEC RAID Stand-Alone Utility. Reboot
the system using the DEC RAID Stand-Alone Utility diskette and continue
with this procedure.
Start DEC RAID Manager from the console by entering:

In the DEC RAID Stand-Alone Utility:
rajidmgr

In the DEC RAID NetWare Utility:
load raidmgr

Select the Array Controller Configuration option on the menu. Check the Disk
Array Devices window on the right of the screen to make sure it is displaying

logical units attached to the controller you want to create the new logical unit
on.

Press the C key to create a logical unit. When prompted, specify the RAID

level choice. Enter the Channel and ID of each drive you want in the logical
unit using the guidelines given below. The drives you can configure into a
logical unit are marked with the status 2 in the Drive Matrix (see the Drive
window). The cursor moves as you enter each selection.

Refer to Table 4-4 for considerations when assigning drives.
Press the Enter key after entering the drive ID number for the last drive
in the logical unit. Is the data OK prompt will appear. Press the Y key to
respond. The following screen is displayed:
LUN 1 Status = Optimal

LUN 1 Drives = (1,2)

(2,2)

(3,2)

(4,2)

PARAMETER

(5,2)
VALUE

RAID Level
Logical Block Size (bytes)
Logical Unit Size (MB)
Segment Size (blocks)

5
512
1600
512

Segment Zero Size (blocks)

Delay Interval

(tenths of a second)

Blocks Per Delay Interval
Are the values shown OK

256

(y/n/q)?

To save the current logical unit parameters for this logical unit, press the
Y key. The array controller automatically formats the new logical unit.
The utility will return to the ACE menu when done.

Configuring Logical Units

4-11

Configuring Logical Units

4.5 Creating a Logical Unit from Spare Drives

* To modify a value, press the N key. Use the Down-arrow key or the Enter

key to select the value you want to change and then enter the new value.
Remember, you cannot change the RAID level parameter, or Logical Block
Size (to determine the values you may use for the logical unit parameters,
see Section 4.2).
After you have set all the parameters you want to set, use the Downarrow key or the Enter key to move to the Are the values shown OK
prompt, then press:

the Y key to save the current logical unit parameter values.

* the Q key to quit without changing the logical unit parameters.
Press the C key to continue. The array controller automatically formats
the new logical unit. The utility will return to the ACE menu when done.
*

To quit the display without changing the parameters, nress the Q key.

This will return you to the ACE menu.
8.

Press the Q key to quit the Array Configuration Editor. You return to the
Array Controller Configuration menu. Prees the Esc key to quit the DEC
RAID Manager. Then press Y and the Enter key.

In DEC RAID Stand-Alone Utility, remove the diskette and reboot the system.
In DEC RAID NetWare Utility, use the NetWare Install command to create
partitions and volumes on the modified logical unit (Chapter 7).
You are done with this procedure.

4-12

Configuring Logical Units

Configuring Logical Units
4.6 Creating a Logical Unit trom an Existing Drive Group

4.6 Creating a Logical Unit from an Existing Drive Group
Use the following procedure to create a logical unit from space in an existing
drive group. Note that the new logical unit must use the same drives and the
same RAID level as the other drives in the group.
1.

If you are going to be creating a logical unit in MS-DOS or SCO UNIX, or in

NetWare if you are going to be creating the logical unit that will contain your
SYS volume, you will have to use the DEC RAID Stand-Alone Utility. Reboot

the system using the DEC RAID Stand-Alone Utility diskette and continue
with this procedure.
2.

Start DEC RAID Manager from the console by entering:
In the DEC RAID Stand-Alone Utility:
raidmgr

In the DEC RAID NetWare Utility:
load raidmgr

Select the Array Controller Configuration option on the menu. Check the
Disk Array Devices window on the right of the screen to make sure it is
displaying the logical units in the drive group in which you want to create the
new logical unit.

keys to highlight the space remaining in the drive group in which you want
to create the new logical unit.

Note that when the desired drive group is highlighted. all the drives that
are in that drive group are highlighted in the Drive Matrix, and will
automatically be configured into the logical unit you create. The new logical
unit will also have the same RAID level as the other logical units in the
group.

Press the C key to create a logical unit. The following screen (or one like
it—parameter values may vary) is displayed:
LUN 1 Status = Optimal
TUN 1 Drives = (1,2} (2,2)

(3,2)

(4,2)

PARAMETER

(5,2)

VALUE

RAID Level

Logical
Logical
Segment
Segment

Block Size (bytes)
Unit Size (MB)
Size (blocks)
Zero Size (blocks)

512
1600
512
0

Delay Interval (tenths of a second)
Blocks Per Delay Interval

1
256

Are the values shown OK (y/n/q)?

To save the current logical unit parameters for this logical unit, press the
Y key. The array controller automatically formats the new logical unit.

The utility will return to the ACE menu when done.

Configuring Logical Units

4-13

Configuring Logical Units
4.6 Creating a Logical Unit from an Existing Drive Group

¢ To modify a value, press the N key. Press the Down-arrow key or the
Enter key to select the value you want to change and then enter the

new value. Remember, you cannot change the RAID level parameter, or

Logical Block Size (to determine the values you may use for the logical
unit parameters, see Section 4.2).
After you have et all the parameters you want to set, use the Downarrow key or the Enter key to move to the Are the values shown OK
prompt, then press:

the Y key to save the current logical unit parameter values.

the Q key to quit without changing the logical unit parameters.

Press the C key to continue. The array controller automatically formats

the new logical unit. The utility will return to the ACE menu when done.
¢

To quit the display without changing the parameters, press the Q key.
This will return you to the ACE menu.

Press the Q key to quit the Array Configuration Editor. You return to the
Array Controller Configuration menu. Press the Esc key to quit RAID
Manager. Then press Y and the Enter key.

In the DEC RAID Stand-Alone Utility, remove the diskette and reboot the
system. In DEC RAID NetWare Utility, use the NetWare Install command to
create partitions and volumes on the modified logical unit (Chapter 7).
You are done with this procedure.

4-14

Configuring Logica! Units

Configuring Logical Units
4.7 Changing RAID Level/Drives on an Existing Logical Unit

4.7 Changing RAID Level/Drives on an Existing Logical Unit
If you want to change the RAID level or drive grouping of an existing logical unit,
you must first delete the current logical unit and recreate it with a new RAID
level or new drives. Note that if the logical unit is part of a drive group, you will

have to change the RAID level/drives on all the logical units in that group. This

means that you will have to delete all the logical units in the group.
Use the following procedure to change the RAID level or drive group of an
existing logical unit.
1.

If necessary, back-up the data on all the logical units in the drive group by
copying it file-by-file to the back-up media.

In NetWare, unmount any volumes contained on the logical units in the drive
group you want to change if any are mounted.
If you are going to be creating a logical unit in MS-DOS or SCO UNIX, or in
NetWare if you are going to be creating the logical unit that will contain your
SYS volume, you will have to use the DEC RAID Stand-Alone Utility. Reboot
the system using the DEC RAID Stand-Alone Utility diskette and continue
with this procedure.

Start DEC RAID Manager from the console by entering:
In the DEC RAID Stand-Alone Utility:
raidmgr

In the DEC RAID NetWare Utility:
load raidmgr

Select the Array Controller Configuration option on the menu. Check the Disk
Array Devices window on the right of the screen to make sure it is displaying
the logical units you want to modify.

Use the Arrow keys to highlight the logical unit you want to modify. Press
the D key to delete the logical unit.
Press:

the C key to confirm that you want to delete the logical unit.

the Q key if you do not want to delete the unit.

Repeat this step for each logical unit in the drive group. When you have

deleted all the logical units in the group, the status of the highlighted drives
in the Drive Matrix change to Spare (2).

Refer to Section 4.6 to create the logical unit(s) you want with the new RAID
levels and/or drive groupings.

Configuring Logical Units

4-15

Contfiguring Logical Units
4.8 Deleting a Logical Unit

4.8 Deleting a Logical Unit
You delete a logical unit to complete one of the following tasks:

Changing the RAID level of the logical unit

Changing the drive grouping containing a logical unit

Deleting a drive from the array

The following procedures gives the steps to delete a logical unit. Normally,
there is no reason to use the procedure (deleting a logical unit to change RAID
level/drive group is described in the procedure in Section 4.7) as there is usually
no reason to remove a drive.

1.
2.
8.

If necessary, back-up the data on all the logical units in the drive group by
copying it file-by-file to the back-up media.
In NetWare, unmount any volumes contained on the logical units in the drive

group you want to change if any are mounted.

If you are going to be creating a logical unit in MS—DOS or SCO UNIX, or in
NetWare if you are going to be creating the logical unit that will contain your
SYS volume, you have to use the DEC RAID Stand-Alone Utility. Reboot the
system using the DEC RAID Stand-Alone Utility diskette and continue with
this procedure.

Start DEC RAID Manager from the console by entering:
In the DEC RAID Stand-Alone Utility:
raidmgr

In the DEC RAID NetWare Utility:
load raidmgr

Select the Array Controller Configuration option on the menu. Check the Disk
Array Devices window on the right of the screen to make sure it is displaying
the logical units you want to modify.

Upse the arrow keys to highlight the logical unit you want to modify. Press the
D key to delete the logical unit.
Press:

the C key to confirm that you want to delete the logical unit.

the Q key if you do not want to delete the logical unit.

Repeat this step for each logical unit in the drive group. When you have
deleted all the logical units in the group, the status of the highlighted drives
in the Drive Matrix change to Spare (2).

Press the Q key to quit the Array Configuration Editor. You return to the
Array Controller Configuration menu. Press the Esc key to quit RAID
Manager. Then press Y and the Enter key.

10. In the DEC RAID Stand-Alone Utility, remove the diskette and reboot the
system. You are done with this procedure.

4-16

Configuring Logical Units

Configuring Logical Units
4.9 Moditying and Displaying a Drive Status

4.9 Modifying and Displaying a Drive Status
You can add, delete, modify, and reconstruct a drive.

4.9.1

Adding a Drive
Adding a drive through software causes the array controller to change the drive
status from Non-existent (1) to Spare (2), thus making the drive available for you
to create a logical unit.
Note that the Add option in ACE may not be used to revive a drive in a logical
unit that has a Failed (3) status, but should not have been failed (such as in the
circumstance of a multiple-drive failure where the logical unit status is Dead replaced wrong drive).

The following are steps for adding a drive.
1.

Start RAID Manager from the console by entering:
In the DEC RAID Stand-Alone Utility:
raidmgr

In the DEC RAID NetWare Utility:
load raidmgr

Select the Array Controller Configuration option on the menu. Check the Disk
Array Devices window on the right of the screen to make sure it is displaying
logical units attached to the controller connected to the drive you want to add.

If not, use the Select Controller option to select the correct controller.

Select the Array Configuration Editor option from the menu. Note that you
are in the Logical Unit window of the Array Configuration Editor (ACE).
Press the Tab key to switch to the Drive window. Use the arrow keys to
highlight the drive you want to add.
Note

To add a drive, you must first phy=ically connect the drive to the array
while it is operating.

Press the A key to add the drive. The status of the drive changes from
Non-existent (1) to Spare (2)

You have added a drive.

You may now create a logical unit using the drive (Section 4.5).

Configuring Logical Units

4-17

Configuring Logical Units

4.9 Modifying and Displaying a Drive Status
49.2 Deleting a Drive
Delete a drive when you plan to remove the drive from the DEC RAID Subsystem,
and are not planning to replace it or reboot the DEC RAID Suksystem.
Deleting a drive through software causes the array controller to change the drive
status from Spare (2) to Non-existent (1). This means you can never delete a drive
that is part of a logical unit.

First, delete all the logical units contained on the drive (refer to Section 4.8 to
delete the logical unit).

Second, delete the drive.

Third, physically remove the drive fror: the subsystem.

Use the following procedure to delete a dr-ive.

Start DEC RAID Manager from the console by entering:
In the DEC RAID Stand-Alone Utility:
raidegr

In DEC RAID NetWare Utility:
losd raidmgr

Select the Array Controller Configuration option on the menu. Check the Disk

Array Devices window on the right of the screen to make sure it is displaying
logical units attached to the controller connected to the drive you want to
delete. If not, use the Select Controller option to select the correct controller.
Select the Array Configuration Editor option from the menu. Note that you
are in the Logical Unit window of the Array Configuration Editor (ACE).
Press the Tab key to switch to the Drive window. Use the Arrow keys to
highlight the drive you want to delete.
Press the D key to delete the drive. The status of the drive changes from
Spare (2) to Non-existent (1)

You have deleted a drive.
Remove the drive(s) you deleted from the array.

4.9.3 Failing a Drive
Fail a drive when you want to replace a Warning and restore a logical unit.

Failing a drive through software causes the array controller to change the drive
status from Warning (5) to Failed (3). Once a drive is failed, the array controller
cannot access that drives’s data until you reconstruct the drive data.
important Note

Do NOT fail a drive in a RAID 0 logical unit because there is no parity.
Do NOT fail a drive in a RAID 1 or b logical unit if the logical unit is

already degraded.

Use the following procedure to fail a drive.
1.

4-18

Start RAID Manager from the console by entering:

Configuring Logical Units

Configuring Logical Units
4.9 Modifying and Displaying a Drive Status
In the DEC RAID Stand-Alone Utility:
raidngr

In the DEC RAID NetWare Utility:
load raidmgr

Select the Array Controller Configuration option on the menu. Check the Disk
Array Devices window on the right of the screen to make sure it is displaying

the logical units containing the Warning drive you want to fail.
3.

Select the Array Configuration Editor option from the menu. Note that you
are in the Logical Unit window of the Array Configuration Editor (ACE).
Press the Tab key to switch to the Drive window. Use the Arrow keys to
highlight the Warning drive you want to fail.

Press the F key to fail the drive. The following message is displayed:
If you fail a drive, you may lose
data redundancy or data.
Press ‘¢’

to continue.

Press ‘q’

to quit.

Press the C key to confirm that you want to continue with the operation you
selected, or press the Q key to quit.
If you press the C key, the status of the failed drive changes from Warning (5) to
Failed (3).

You have failed a drive.

Reconstruct the drive data, or reformat the logical unit.

4.9.4 Reconstructing a Drive
Use this function to reconstruct a drive after you have replaced a failed drive in
degraded RAID 1 or 5 logical unit.

Reconstructing data on a drive causes the array controller to do the following, in
order:

Sense that there is a new drive in the array

Change the drive status to Replace (4)

Copy special array controller files to the new drive

Recalculate data and parity from the data and parity on the other drives

Write the recalculated data and parity to the new drive

Change the drive status to Optimal (0)

If you replace a drive while the DEC RAID Subsystem is still operating (called

hot swapping), the array controller automatically starts reconstructing the drive
data. You need to use this function only if your subsystem does NOT have hot
swapping.
important Note

Remember, a RAID 0 logical unit does not have parity so you cannot
recover the data by reconstructing it. You must restore the RAID 0

Configuring Logical Units

4-18

Configuring Logical Units
4.9 Modifying and Displaying a Drive Status
logical unit by formatting the logical unit, then copying the data from the
back-up media file-by-file (Chapter 5 for more information).

If the drive you want te reconstruct contains more than one logical unit, all the
logical units are reconstructed.
Use the following procedure to reconstruct a drive.
1.

Physically replace the failed drive.

Start DEC RAID Manager from the console by entering:
In the DEC RAID Stand-Alone Utility:
raidmgr

In the DEC RAID NetWare Utility:
load raidmgr

Select the Array Controller Configuration option on the menu. Check the Disk
Array Devices window on the right of the screen to make sure it is displaying
the logical units containing the drive you want to reconstruct.
Select the Array Configuration Editor option from the menu. Note that you

are in the Logical Unit window of the Array Configuration Editor (ACE).
Press the Tab key to switch to the Drive window. Check the status of the
drive you want to reconstruct. If the status is currently Replaced (4) or
Formatting (8) and the logical unit status is Reconstructing, your system has
hot swapping and you do not need to continue with this procedure.

Use the arrow keys to highlight the Failed drive you want to reconstruct.
Preas the R key to start reconstruction.
You have started the roconstruction of drive data.

4-20

Configuring Logical Units

PartV
Restoring a Logical Unit

This part of the guide contains:
¢

Chapter b, Restoring a Logical Unit

S
Restoring a Logical Unit
5.1 Restoring a Logical Unit Task Summary
Restoring a logical unit is the process of returning the logical unit to a working
state after one or more drive failures. The tasks involved depend on the RAID
level of the logical unit and on how many drives have failed.
Table 5-1
RAID Level
0

Procedure for Restoring Logical Units
LUN/Drive

Status

OK - Warning (1
or more drives

in warning)

Recovery Steps:

Stop using the logical unit immediately.

Attempt to back-up the logical unit.

Fail all warning drives.

Replace all failed drives.

Reformat the logical unit.

In NetWare, recreate the NetWare partitions and

Copy backed-up data to the restored logical unit.

volumes on the restored logical unit.

OK - Warning (1
or more drives

in warning)

If two or more drives are in warning. stop using the

logical unit immediately and back-up the logical
unit before continuing.

One at a time, fail each warning drive, replace the

failed drive, and reconstruct the data on it.

Degraded (1
or more failed
drives)

Replace the failed drives (all drives can be replaced
at the same time).

One at a time,, reconstruct the data on the replaced
drives.

(continued on next page)

Restoring a Logical Unit

5-1

Restoring a Logical Unit
5.1 Restoring a Logical Unit Task Summary
Table 5-1 (Cont.)
RAID Level
1

Procedure for Restoring Logical Units

LUN/Drive
Status

Recovery Stepa:

Degraded -

Warning (1 or
more failed
drives and 1 or

more drives in

)
o
,
Stop using the logical unit immediately.

Attempt to back-up the logical unit.

warning)

If the back-up is successful, then one at a time,

If the back-up is not successful, then fail
all warning drives. replace all failed drives,
and reformat the logical unit. In NetWare,

replace and reconstruct each failed drive, then
fail, replace, and reconstruct each warning
drive.

recreate the NetWare partitions and volumes.
Copy backed-up data back to the restored
logical unit.

OK - Warning (1
or more drives

in warning)

If two or more drives are in warning, stop using the

logical unit immediately and back-up the logical
unit before continuing.

One af a time, fail each warning drive, replace the

Replace the failed drive.

Reconstruct the data on the replaced drive.

failed drive and

Stop using the logical unit immediately and back-

In warning)

2. Replace the fajled drive and reconstruct the data

failed drive, and reconstruct the data on it.

Degraded (1

failed drive)

Degraded Warning (1

1 or more drives

up the logical unit before continuing.

on it

One at a time, fail each warning drive, replace the

Drive Failures
(2 or more failed
drives and

Fail all warning drives.

2. Replace the failed drives.

es® TMTM

3.
4.

failed drive, and reconstruct the data on it.

Reformat the logical unit.
In NetWare. recreate the NetWare partitions and
volumes on the restored logical unit.

5-2

Tiestoring a Logical Unit

Copy backed-up deta to the restored logical unit.

Restoring a Logical Unit
5.2 When to Replace a Drive

5.2 When to Replace a Drive
You need to replace a drive when the following occurs:
¢

For NetWare, the Array Monitor Daemon (AMD) sends a message indicating
that a drive has failed or been put in a warning state.

For SCO UNIX, the Array Monitor sends a message indicating a change in
status to Failed or Warning.

You see a logical unit status other t. a1n Ok - Optimal displayed in the DEC
RAID Manager Disk Array Devices window.

Table 52 shows the logical unit status displayed for the varying degrees of
logical unit failures.
Table 5-2

Logical Unit Status

RAID Level

LUN Status

Ok - Warning

Meaning
One or more drives in the logical unit has

been put in warning because of a read or write
error. The array controller never assigns a

Failed status to a drive in a RAID 0 logical
unit, no matter how severe the error, so this
is the most serious status possible for a RAID
0 logical unit. Depending on the severity of

the error, the logical unit may no longer be
accessible, and all data on the unit may be
lost. Stop using the logical unit immediately

(if it is still accessible), as further errors may
result in loss of all data.

Ok - Warning

One or more drives in the logical unit has been
put in warning because of read or write errors
that were not serious enough to fail the drives.

The logical unit remains accessible. The
warning drives should be replaced as soon as

poesible. If two or more drives are in warning,

stop using the logical unit immediately. as the
failure of these drives would result in the loss
of data.

Degraded

One or more drives in the logical unit have

failed, but no two drives in a mirrored pair

have failed. so data on the logical unit can

be recovered from mirrored disks. The failed

drives should be replaced as soon as possible.
If they cannot be replaced as soon as possible.
If they cannot be replaced immediately. backup the data on the logical unit to prevent
future data loss.

{continued on next page)

Restoring a Logical Unit

5-3

Restoring a Logical Unit
5.2 When to Replace a Drive
Table 5-2 (Cont.)
RAID Level
1

Logical Unit Status

LUN Status

Meaning

Degraded - Warning

Either:
*

One or more drives in the logical unit have
failed and the mirrared pair of a failed

drive has also failed. The array controller
never assigns a Failed status to the second
failed drive in a mirrored pair, no matter
how severe the ervor (it assigns a status
of Warning), so this is the most serious
status possible for RAID 1 logical unit.

The logical unit is not longer accessible,
and all data on the unit is lost.

Ok-Warning

One or more drives in the logical unit have
failed and a minor error has occurred on
another disk (the disk may or may not be
the mirrored pair of a failed disk). The
logical unit remains acceseible.

One or more of the drives in the logical unit

have been put in warning because of read or
write errors that were not serious enough to
fail the drives. The warning drives should be
replaced as soon as possible, before they fail.

The logical unit remains accessible. If two or

more drives are in warning, stop using the
logical unit immediately, as the failure of these
drives would result in the loss of data.

Degraded

A single drive in the logical unit has failed,

but the date on that drive can be reconstructed
using data and parity from the other drives.

The failed drive should be replaced as soon
as possible. If they cannot be replaced

immediately, back-up the data on the logical
unit to prevent future data loss.

Degraded - Warning

A eingle drive in the logical unit has failed,
and and one or more drives have been put in
wamning because of minor errors. The logical
unit is still accessible. Stop using the logical
unit immediately, as the failure of a warning
drive will resuit in data loss.

Drive Failures

Two or more drives in the logical unit have
been failed and the logical unit is inaccessible.
All data in the unit is lost.

Note

Refer to the following sections for a further description of drive failures
and RAID levels, and the actions to take when they occur.

5-4

Reatoring a Logical Unit

Restoring a L.ogical Unit
5.3 Drive Failures on RAID 0

5.3 Drive Failures on RAID 0
This section describes drive failures on a RAID 0 logical unit and the logical unit
and drive status that result from such failures.

5.3.1

Single Drive Failure on RAID 0
When a single drive in in a RAID 0 logical unit experiences unrecoverable read or
write errors, the array controller puts the drive in a Warning state and changes
the logical unit status to Ok - Warning. This occurs even if the drive is completely

inaccessible (for example, if power to the drive is lost). All the data on the logical
unit may be lost, depending on the severity of the error that generated the
warning.

Whenever one or more drives are put in Warning, back-up the data on the logical
unit file-by-file (if possible), fail the drives, replace the drives, and then reformat
the logical unit. For the NetWare environment, you must then add the logical
unit to your system by creating partitions and volumes (using the NetWare
Install command). Copy data back to the restored logical unit from your back-up

media.

5.4 Drive Failures on RAID 1
This section describes drive failures on a RAID 1 logical unit and the logical unit

and drive status that result from such failures.

5.4.1

Single Drive Failures on RAID 1
Whenever a single drive in an optimal RAID 1 logical unit experiences read or
write errors, the array controller changes the status of the drive to Warning if the
errors are minor (read errors on a few sectors) or Failed if the errors are serious

(loss of drive power, drive component failure, and so on). The logical unit status
changes to Ok - Warning or Degraded, depending on whether the drive status is
Warning or Failed. In either case, you can continue to access the logical unit,
because the data on the failed drive can be read from its mirrored drive.

Whenever a single drive is marked as Failed (or Warning), you should replace it
as goon as possible. The array controller will then reconstruct the data on the
drive automatically. If you cannot replace the drive immediately, back-up the
data file-by-file, then replace it when you can.

5.4.2 Muitiple Drive Failures on RAID 1
Whenever the second drive in a RAID 1 logical unit experiences read or write

errors, the array controller takes the following actions:
¢

If the drive is the mirrored pair of a drive that is already failed, the drive

status changes to Warning, even if the drive is completely inaccessible. The
logical unit status changes to Degraded - Warning, but the logical unit may
not be accessible.

If the drive is not the mirrored pair of a failed drive, the drive status changes
to Warning or Failed, depending on the severity of the error. The logical unit
status changes to Ok - Warning, Degraded, or Degraded - Warning, depending
on the previous status of the logical unit.

Restoring a Logical Unit

5-5

Restoring a Logical Unit
5.4 Drive Failures on RAID 1

A RAID 1 logical unit can have more than one failed drive and still remain

degraded as long as none of the failed drives are a mirrored pair. For example, in
a six drive RAID 1 logical unit, it is possible for as many as 3 drives to fail while
the logical unit remains accessible (in degraded mode). You should not, however,
continue to operate a RAID 1 logical unit with any number of failed or warning
drives longer than it takes to replace the drives.
The worst status in RAID 1 logical unit can have is Degraded - Warning. This
status means that either a mirrored pair of drives has failed, and all data on the
logical unit has been lost, or that a minor read error has occurred on a drive in

an already degraded logical unit. To determine the action to take to restore the
unit, you should attempt to back-up the data.
¢

If the back-up succeeds, the logical unit does not contain a failed mirrored
pair. You can probably reconstruct the data on the failed and warning drives.

If the back-up fails, a mirrored pair has failed and all data has been lost. You
will have to restore the logical unit by reformatting it.

5.5 Drive Fallures on RAID 5
This section describes drive failures on a RAID 6 logical unit and the logical unit
and drive status that result from such failures.

5.5.1

Single Drive Fallure on RAID 5
Whenever a single drive in an optimal RAID 5 logical unit experiences read or
write errors, the array controller changes the status of the drive to Warning if the

errors are minor (read errors on a few sectors) or Failed if the errors are serious

(loas of drive power, drive component failure, and so on). The logical unit status

changes to Ok - Warning, or Degraded, depending on whether the drive status
is Warning or Failed. In either case, you can continue to access the logical unit,
as the data on the failed drive can be reconstructed using parity and data on the
ather drives.

Whenever a drive is marked as Failed (or Warning), you should replace it as
soon as possible. The array controller will then reconstruct the data on the drive
automatically. If you cannot replace the drive immediately, back-up the data
file-by-file, then replace it when you can.

5.5.2 Multiple Drive Fallures on RAID 5
If a second drive in a RAID b logical unit experiences read or write errors, the
array controller marks the drive as Warning or Failed, depending on the severity
of the error. The logical unit status changes to Optimal - Warning, Degraded Warning, or Drive Failures, depending on the previous status of the logical unit.
If the status changes to Drive Failures (because of the failure of a second drive),

all data in the logical unit is lost.
If the logical unit status is Optimal - Warning, or Degraded - Warning, you will
probably be able to reconstruct the data on the failed and warning drives by
replacing the drives, one by one, beginning with the failed drive. If the logical
unit status is Drive Failures, you will have to replace the failed drives and
reformat the logical unit.

5-6

Restoring a Logical Unit

Restoring a Logical Unit
5.6 What is Reconstruction?

5.6 What is Reconstruction?
Reconstruction is a process used to restore a degraded RAID 1 or 5 logical unit to

its original state after a single drive has been replaced. During reconstruction,
the array controller recalculates the data on the drive that was replaced, using
data and parity from from the other drives in the logical unit. The controller then
writes this data to the replaced drive. Note that although RAID level 1 does not
have parity, the array controller can reconstruct data on a RAID 1 logical unit by
copying data from the mirrored disk.
Note

Reconstruction only applies to a degraded RAID 1 or 5 logical unit with a
single-drive failure.

The array controller automatically initiates the reconstruction process after you
replace a drive in a degraded RAID 1 or § logical unit. Once reconstruction is
initiated the array controller completes the following actions:
¢

Formats the new drive (if the array controller determines it is necessary)

Copies special array software files to the new drive

Recalculates the data and parity from the data and parity on the other logical
unit drives

Writes the recalculated data and parity to the new drive

Reconstruction can take place while the logical unit is in use. You do not need to
shut it down. While the utility is formatting the new drive, there is no additional
effect on the system I/O performance (the logical unit continues to operate in
the degraded mode). Once data reconstruction begins, you can adjust the rate of
reconstruction so that it does not interfere significantly with system performance.
If there are multiple-drive failures in a RAID 1 or 5 logical unit (that is, the
status is Degraded - Warning or Drive Failures), you cannot reconstruct the drive
data. To restore a RAID 1 or 5 logical unit to a working state in this case, you
need to replace the drives and reformat the logical unit. See Section 5.4.2 and
Section 5.5.2 for more information.

5.7 Reconstruction Rate
The rate of reconstruction depends on two parameters, the delay interval
parameter and the blocks per delay interval parameter.

The delay interval is the time between reconstruction operations. When the array
controller reconstructs data on a drive, it divides its time between reconstruction

and normal 1/O operations. System 1/O operations take place during this delay
interval. As the delay interval increases, system 1/0 performance also increases,
but so does reconstruction time.

The blocks per delay interval is the number of blocks the array controller
reconstructs during each reconstruction operation. The more blocks, the longer
the time necessary to reconstruct them. This is time that cannot be used to
perform system 1/0. Therefore, the larger the reconstruction amount, the more
gystem performance degradation increases.

Restoring a Logical Unit

6-7

Restoring a Logical Unit
5.7 Reconstruction Rate

Note that these parameters control only the rate of data reconstruction, not total
reconstruction time. The first step in reconstruction, formatting the replaced
drive, can take 12-15 minutes. During this time, however, you can continue
to access the array with no decrease in current system performance (the array
remains in the degraded mode).

Some suggested rates, as well as sample timings, are given in the table below.
Table 5-3

Reconstruction Rates

Reconstruction

Time for

Frequency

Reconstruction

Completion

Slow

]

2560

Moderate

266

Fast

1024

Rate (relative)

in tenths of a second

Timings are approximate and apply

Amount In blocks

only to data reconstruction; formatting

in minutes/GB*

time (typicslly 15-20

minutes) ‘i:& not included. Times w‘ifl be slower if other processes are accegs%ng the array during
reconstruction

In general, reconstruction times tend to level off after block sizes of 20K.

Although higher block sizes (up to 64K) are possible, there is no great reduction
in time at the higher levels.

5.8 Restoring a RAID 0 Logical Unit
Restore a RAID 0 logical unit when one or more of the following occurs:
¢

A drive has a status of Warning.

The logical unit has a status of Ok - Warning

In NetWare, the Array Monitor Daemon (AMD) notifies you of a change in

disk or logical unit status or NetWare notifies you that it can no longer access
the logical unit.

In SCO UNIX, the Array Monitor notifies you of a changes in logical unit or

drive status.
Use the following procedure to restore a RAID 0 logical unit.
1.

Attempt to back up the data on the logical unit, if possible.

In NetWare, unmount any volumes on the unit, if they are still mounted.

If you modify parameters in MS-DOS or SCO UNIX, or in NetWare and if
these parameters are on the logical unit that contains any part of your SYS
volume, you will have to use the DEC RAID Stand-Alone Utility. Reboot the

system using the DEC RAID Stand-Alone Utility diskette and continue with
this procedure.

Start DEC RAID Manager by entering the following command from the
consgole:

In the DEC RAID Stand-Alone Utility:
raidmgr

In the DEC RAID NetWare Utility:
load raidmgr

5-8

Restoring a Logical Unit

Restoring a Logical Unit
5.8 Restoring a RAID 0 Logical Unit
6.

From the Disk Array Service menu, select Device Selection. Scroll through
the Disk Array Devices window until you find a logical unit with a status of
Ok - Warning This should be the RAID 0 logical unit you need to restore.

Select that unit. Examine the Array Information and Disk Drive Information
windows at the bottom of the screen to make sure that you have selected the
correct logical unit. The selected unit should be a RAID 0 logical unit with
one or more drives with a Warning status.

For the DEC RAID Stand-Alone Utility:

Replace the drive with the Warning status with a new drive, restart the
system, and restart DEC RAID Manager.

Caution

You must wait 10 seconds from the time you take out a drive, to the time
that you reinsert a new one for the bus to recognize the drive change.

From the Disk Array Services menu select the Array Controller
Configuration option.

From the Array Controller Configuration menu select the Array Controller
Editor option.

From the logical unit window, select Format to format the logical unit.

After the reformat is finished, press the Esc key to quit the DEC RAID
Manager. Then, press Y and the Enter key.

Remove the diskette and reboot the system.

Copy backed-up data back to the logical unit.

For the DEC RAID NetWare Utility:

From the Disk Array Services menu select Device Service.

From the Service Disk Array Device menu, select Replace the disk drive.
Follow the instructions given to replace the Warning drive or drives.
Press the Enter key as you complete each step to check the step off.
Cautlon

You must wait 10 seconds from the time you take out a drive, to the time
that you reinsert a new one for the bus to recognize the drive change.

From the Service Disk Array Device menu, select Format disk array
device. Answer the prompts to reformat the logical unit.

After the reformat is finished, press the Esc key to quit DEC RAID
Manager. Then, press Y and the Enter key.

In the DEC RAID NetWare Utility, use the NetWare Install command to

create partitions and volumes on the restored logical unit. (See Chapter 7)
*

Copy backed-up data back to the logical unit.

You are done with this procedure.

Restoring a Logical Unit

5-9

Restoring a Logical Unit
5.9 Restoring a RAID 1 Logical Unit

5.9 Restoring a RAID 1 Logical Unit

‘

You need to restore a RAID 1 logical unit when one or more of the following
occurs:

A drive has a status of Warning or Failed.

The logical unit has a status of Ok - Warning, Degraded, or Degraded Warning.

In NetWare, the Array Monitor Daemon (AMD) notifies you of a change in

disk or logical unit status or NetWare notifies you that it can no longer access
the logical unit.
*

In SCO UNIX, the Array Monitor notifies you of a change in logical unit or
disk status.

Use the following procedure to restore a RAID 1 logical unit.
1.

If the status of the logical unit is Degraded - Warning, see Section 65.10. If the
status of the logical unit is Ok - Warning or Degraded, continue with the next
step.

If you cannot replace the disk right away, attempt to back up the data on the
logical unit, if possible.

If you are restoring a logical unit in MS=DOS or SCO UNIX you have to use
the DEC RAID Stand-Alone Utility. Reboot the system using the DEC RAID
Stand-Alone Utility diskette and continue with this procedure.

Start DEC RAID Manager by entering the following command from the
console:
In the DEC RAID Stand-Alone Utility:
raidmgr

In the DEC RAID NetWare Utility:
load raidmgr

From the Disk Array Service menu, select Device Selection. Scroll through

the Digk Array Devices window until you find a logical unit with an Ok Warning status or Degraded status. This should be the RAID 1 logical unit
you need to restore.
6.

Select that unit. Examine the Array Information and Disk Drive Information
windows at the bottom of the screen to make sure that you have selected the

correct logical unit. The selected unit should be a RAID 1 logical unit with
one or more drives with a Failed status.
7.

For the DEC RAID Stand-Alone Utility:

Replace the drive with the Failed status with a new drive.
Caution

You must wait 10 seconds from the time you take out a drive, to the time

that you reinsert a new one for the bus to recognize the drive change.

* From the Disk Array Services menu select the Array Controller
Configuration option.

5-10

Restoring a Logical Unit

’

Restoring a Logical Unit

5.9 Restoring a RAID 1 Logical Unit
*

From the Array Controller Configuration menu select the Array
Configuration Editor option.

From the logical unit window, select Reconstruct to initiate reconstruction
of the logical unit.

After the reconstruction is finished, press tne Esc key to quit DEC RAID
Manager. Press the Y key and the Enter key.

Remove the diskette and reboot the system.

For the DEC RAID NetWare Utility:

From the Disk Array Services menu select Device Service.

Check the Disk Drive Information window. The drive marked Failed
should be highlighted. If it is not, select Select Disk Drive and select the
Failed drive. If there is more than one drive in Failed drive, highlight the
first one. If there are Failed and Warning drives, select a Failed one first.

From the Service Disk Array Device menu, select Replace the disk drive.

Follow the instructions given to replace the Failed drives. Replace the
highlighted drive, not all the Failed or Warning drives. Press the Enter
key as you complete each step to check the step off.
Caution

You must wait 10 seconds from the time you take out a drive, to the time
that you reinsert a new one for the bus to recognize the drive change.

After you have checked off all the tasks, watch the status of the logical
unit you just restored. It should change to Reconstructing.
If the status does not change to Reconstructing, make sure the drive you

just replaced is highlighted, then select Reconstruct disk drive from the
Service Disk Array Device menu to start data reconstruction.

Wait for the reconstruction to finish. After it is finished, if you have any
Failed or Warning drives, repeat these steps to replace and reconstruct
them.

After the logical unit status changes to Ok - Optimal, press the Esc key to
quit DEC RAID Manager. Then, press the Y key and the Enter key.

5.10 Restoring RAID 1 Logical Units after Multiple Failures
After a multiple drive failure, RAID 1 logical unit has a status of Degraded Warning, at least one Failed drive, and a mirrored pair of a Failed drive with

a status of Warning. Depending on the severity of the error that caused the
Warning, the logical unit may not be accessible and all the data in it may be lost.
Use the following procedure to restore a RAID 1 logical unit after a multiple drive
failure.

Stop using the logical unit immediately if it is still working. Continuing to
access the logical may result in complete loss of data.

Restoring a Logical Unit

5-11

Restoring a Logical Unit

5.10 Restoring RAID 1 Logical Units after Multiple Failures
2.

Attempt to back-up the data on the logical unit.

If the backup was successful, you may be able to reconstruct the data in
the logical unit. Go to Section 5.9, and complete the procedure. You will
need to replace the failed drive first, reconstruct it, and then replace the
Warning drive. If the reconstruction operation fails, continue with this
procedure to restore the logical unit.

If the backup was not successful, continue with this procedure to restore

the logical unit.

If you are restoring a logical unit in MS-DOS or SCO, or in NetWare and if

the logical unit contains any part of your SYS volume, you will have to use

the DEC RAID Stand-Alone Utility. Reboot the system using the DEC RAID
Stand-Alone Utility diskette and continue with this procedure.
4.

Start DEC RAID Manager by entering the following command from the
console:

In the DEC RAID Stand-Alone Utility:
raldngr

In the DEC RAID NetWare Utility:
load raidmgr

From the Disk Array Service menu, select Device Selection. Scroll through
the Disk Array Devices window until you find a logical unit with a status
of Degraded - Warning. This should be the RAID 1 logical unit you need to
restore. Select that unit. Examine the Array Information and Disk Drive
Information windows at the bottom of the screen to make sure that you have

selected the correct logical unit. The selected unit should be a RAID 1 logical
unit with one or more drives with a Failed or Warning status.
6.

For the DEC RAID Stand-Alone Utility:

Replace the drive with the Failed or Warning status with a new drive,
restart the system, and restart DEC RAID Manager.
Cautlon

You must wait 10 seconds from the time you take out a drive, to the time
that you reinsert a new one for the bus to recognize the drive change.

From the Disk Array Services menu select the Array Controller
Configuration option.

From the Array Controller Configuration menu select the Array
Configuration Editor option.

From the logical unit window, select Format to initiate reconstruction of
the logical unit.

After the reformat is finished, press the Esc key to quit DEC RAID
Manager. Then, press the Y key and the Enter key.

Remove the diskette and reboot the system.

Copy backed-up data back to the logical unit.

5-12

Restoring a Logical Unit

5.10 Restoring RAID 1 Logical Units after Multiple Failures
T

For the DEC RAID NetWare Utility:

From the Disk Ar-ay Services menu select Device Service.

Check the Disk Drive Information window. The drive marked Warning
should be highlighted. If it is not, select Select Disk Drive and select the
Warning drive. If there is more than one drive in Warning, highlight the
firat one.

From the Service Disk Array Device menu, select Replace the disk drive.

Follow the instructions given to replace the Warning drive or drives.
Press the Enter key as you camplete each step to check the step off.
Caution

You must wait 10 seconds from the time you take out a drive, to the time
that you reinsert a new one for the bus to recognize the drive change.

From the Service Disk Array Device menu, select Format disk array

device. Answer the prompts to reformat the logical unit.
*

After the reformat is finished, press the Esc key to quit DEC RAID
Manager. Then, press the Y key and the Enter key.

Use the NetWare Install command to create partitions and volumes on
the restored logical unit (see Chapter 7).

Copy backed-up data back to the logical unit.
You are done with this procedure.

5.11 Restoring a RAID 5 Logical Unit
Restore a RAID 5 logicai unit when one or more of the following occurs:
A drive has a status of Warning or Failed.
The logical unit has a status of Ok - Warning, Degraded, or Degraded Warning.

In NetWare, the Array Monitor Daemon (AMD) notifies you of a change in

disk or logical unit status or NetWare notifies you that it can no longer access
the logical unit.

In SCO UNIX, the Array Monitor notifies you of a change in logical unit or
drive statua.

Use the following procedure to restore a RAID 5 logical unit.
1.

If the status of the logical unit is Degraded - Warning, see Section 5.12. If the
status of the logical unit is Ok - Warning or Degraded, continue with the next
step.

If you cannot replace the disk right away, attempt to back up the data on the
logical unit, if possible.

If you are restoring a logical unit in MS-DOS or SCO UNIX you have to use
the DEC RAID Stand-Alone Utility. Reboot the system using the DEC RAID
Stand-Alone Utility diskette and continue with this procedure.

Start DEC RAID Manager by entering the following command from the
console:

Restoring a Logical Unit

5-13

W
W

Restoring a Logical Unit
5.11 Restoring a RAID 5 Logical Unit
In the DEC RAID Stand-Alone Utility:
raidmgr

In the DEC RAID NetWare Utility:
load raidmgr

From the Disk Array Service menu, select Device Selection. Scroll through
the Dick Array Devices window until you find a logical unit with an Ok Warning status, Degraded or Degraded - Warning. This should be the RAID 5
logical unit you need to restore.

Select that unit. Examine the Array Information and Disk Drive Information
windows at the bottom of the screen to make sure that you have selected the
correct logical unit. The selected unit should be a RAID 5 logical unit with
one Failed drive and/or one or more drives with a Warning status.

For the DEC RAID Stand-Alone Utility:

Replace the drive with the Warning status with a new drive.
Cautlon

You must wait 10 seconds from the time you take out a drive, to the time
that you reinsert a new one for the bus to recognize the drive change.

From the Disk Array Services menu select the Array Controller
Configuration option.

From the Array Controller Configuration menu select the Array
Configuration Editor option.

From the logical unit window, select Reconstruct to initiate reconstruction
of the logical unit.

After the reconstruction is finished, press the Esc key to quit DEC RAID
Manager. Then, press the Y key and the Enter key.

Remove the diskette and reboot the system.

For the DEC RAID NetWare Utility:

From the Disk Array Services menu select Device Service.

From the Service Disk Array Device menu, select Replace the disk drive.

Follow the instructions given to replace the Failed drives. Replace the

highlighted drive, not all the Failed or Warning drives. Press the Enter
key as you complete each step to check the step off.
Caution

You must wait 10 seconds from the time you take out a drive, to the time
that you reinsert a new one for the bus to recognize the drive change.

5-14

Restoring a Logical Unit

Restoring a Logical Unit
5.11 Restoring a RAID 5 Logical Unit
¢

After you have checked off all the tasks, watch the status of the logical
unit you just restored. It should change to Reconstructing
If the status does not change to Reconstructing, make aure the drive you
just replaced is highlighted, then select Reconstruct disk drive from the
Service Disk Array Device menu to start data reconstruction

Wait for the reconstruction to finish. After it is finished. if you have any
Failed or Warning drives, repeat these steps to replace and reconstruct
them.

After the logical unit status changes to Ok - Optimal, press the Esc key to
quit DEC RAID Manager. Then, press the Y key and the Enter key

5.12 Restoring RAID 5 Logical Units after Multiple Failures
After a multiple drive failure, a RAID 5 logical unit has a status of Drive Failures
and at least two Failed drives. The logical unit is not accessible and all the data
in it has been lost.
Use the following procedure to restore a RAID b logical unit after a multiple drive
failure.
1.

If you are restoring a logical unit on MS-DOS or SCO, or in NetWare and
the logical unit contains any part of your SYS volume, you have to use the
DEC RAID Stand-Alone Utility. Reboot the system using the DEC RAID
Stand-Alone Utility diskette and continue with this procedure.

Start DEC RAID Manager by entering the following command from the
console:

In the DEC RAID Stand-Alone Utility:
raidmgr

In the DEC RAID NetWare Utility:
load raidmgr

From the Dirk Array Service menu, select Device Selection. Scroll through
the Disk A:ray Devices window until you find a logical unit with a status of
Degraded - Warning or Drive Failures. This should be the logical unit you
need to restore.
Select that unit. Examine the Array Information and Disk Drive Information
windows at the bottom of the screen to make sure that you have selected the
correct logical unit. The selected unit should be a RAID 5 logical unit with
two or more drives with a Failed status.
For the DEC RAID Stand-Alone Utility:
*

Replace the drives with Failed status with new drives.

Caution
You must wait 10 seconds from the time you take out a drive, to the time

that you reinsert a new one for the bus to recognize the drive change.

From the Disk Array Services menu select the Array Controller
Configuration option.

Restoring a Logica! Unit

5-15

Restoring a Logical Unit

5.12 Restoring RAID 5 Logical Units after Multiple Failures
From the Array Controller Configuration menu select the Array
Configuration Editor option.

From the logical unit window, select Format to initiate reconstruction of
the logical unit.

After the reformat is finished, press the Esc key to quit DEC RAID
Manager. Then, press the Y key and the Enter key.
Remove the diskette and reboot the system.

Copy backed-up data back to the logical unit.
6.

For the DT RAID NetWare Utility:

From the Disk Array Services menu select Device Service.

Check the Disk Drive Information window. The drive marked Failed

should be highlighted. If it is not, select Select Disk Drive and select the

first Failed drive.

From the Service Disk Array Device menu, select Replace the disk drive.
Follow the instructions given to replace all the Failed drives. Press the
Enter key as you complete each step to check the step off.
Caution

You must wait 10 seconds from the time you take out a drive, to the time
that you reinsert a new one for the bus to recognize the drive change.

From the Service Disk Array Device menu, select Format disk array
device. Answer the prompts to reformat the logical unit.

After the reformat is finished, press the Esc key to quit DEC RAID
Manager. Then, press the Y key and the Enter key.

Use the NetWare Install command to create partitions and volumes on
the restored logical unit (see Chapter 7).
Copy backed-up data back to the logical unit.

You are done with this procedure.

5-16

Restoring a Logical Unit

Part Vi
DEC RAID NetWare Utility

This part of the guide contains:

Chapter 6, Installation of the DEC RAID NetWare Utility
Chapter 7, Adding Logical Units to NetWare

Chapter 8, Checking/Repairing Array Parity
Chapter 9, The Array Monitor Daemon

of the DEC RAID NetWare Utility
Installation
6.1 About Installation
The DEC RAID NetWare Utility is installed into the NetWare file server system
by the installation program called RMINST.NLM. You do not need to remove
an old version of the DEC RAID NetWare Utility since the installation program

copies the new version into a different directory than the old version.
RMINST NLM performs the following task when it installs the DEC RAID
NetWare Utility:

Creates the following directories

gys:/system/raidmgr

gys:/system/raidmgr/acehelp

Copies the program and data files from the distribution diskette containing
DEC RAID Manager into the raidmgr subdirectory

Copies the online help file (raidmgr.nfo) into the sys:/public directory
Copies the Array Configuration Editor help files into the
sys:/system/raidmgy/acehelp subdirectory
Determines which drivers are in resident memory using the list of supported
disk drivers in the raidmgr.cfg file (located in the ays:/system/raidmgr
subdirectory)

Loads the amd utility if the driver is ASPI and loaded into memory
Reports any errors it encounters while copying files, checking drivers, or
loading the amd utilities.

6.1.1

DEC RAID Manager Files
When you install the DEC RAID NetWare Utility, the installation program copies
the files listed in Table 6-1 onto the file server:
Table 6—1

Flies Copled to the File Server In NetWare

File

Dezeription

raidmgr.nlm

User shell

reidifl.nlm

NetWars file

ace.nlm

Array Configuration Editor

amd.nlm

Array Monitor Dasmon
(continued on next page)

Installation of the DEC RAID NetWare Utility

6-1

Installation of the DEC RAID NetWare Utility
6.1 About Installation

Table 6-1 (Cont.)

6-2

Files Copled to the File Server in NetWare

Flle

Description

apc.nlm

Automatic parity check/repair

raidmgr.cfg

Configuration file

raidmgr.dat

Menu and form definitions for RAID Manager

raidmgr.meg

Messages displayed by RAID Manager

raidmgr.hlp

Online help for RAID Manager

release.id

Releans version

acehelp <dir>

Online help files for the Array Configuration Editor

Installation of the DEC RAID NetWare Utility

Installation of the DEC RAID NetWare Utility
6.2 When to Install the DEC RAID Manager

6.2 When to Install the DEC RAID Manager
Install the DEC RAID Manager after you have installed your array subsystem
and host adapter, if necessary.

6.3 Before You Install the DEC RAID Manager
If necessary, install the host adapter, using the instructions accompanying the
host adapter kit.

Install all the required drivers you neec. Refer to Chapter 1 to determine
which drivers your system needs.

6.4 Running the Installation Program
There are two ways to install DEC RAID Manager.
If you have an MS-DOS partition on your file server, refer to Section 6.4.1 to run
the installation program using MS-DOS.

If you do not have an MS-DOS partition on your file server, refer to Section 6.4.2
to run the installation program without using MS~DOS.

6.4.1

Using an MS-DOS Partition
Perform the following steps to run the installation program using a MS-DOS
partition.
1.

Make sure you have read Section 6.3, Before You Install the DEC RAID
Manager.

Insert the DEC RAID Manager distribution diskette in drive A:.
Type the following at the NetWare file server console, then press the Enter
key:
load a:rminst

The following screen is displayed:

NetWare RAID Manager Installation
-------------------------------------------------------------

Enter DOS drive identifier

(default A:):

Press the Enter key twice.

The installation program reads the release.id file from the device, and
dispiays the contents on the screen. This file contains DEC RAID Managerspecific information.
Press the Y key and the Enter key to continue with the installation.
The installation program does the following:

Creates the sys:/system/raidmgr and sys:/system/raidmgr/acehelp
direciories,

Copies files into the directories from the distribution diskette,

Copies the online help files into the sys:/public directory.

installatior of the DEC RAID NetWare Utility

6-3

installation of the DEC RAID NetWare Utility
6.4 Running the Installation Program
6.

Note when the following message (indicating that installation is complete) ie
displayed.
AUTOEXEC NCF setup complete

‘Disk Array Mohitor started

NetRare RAID Manager Installation complete

If the Disk Array Monitor is not automatically started, unload the DEC RAID
Manager, load amd, then load the DEC RAID Manager.

You have successfully completed the installation.
Go to Chapter 7 to configure the logical units.

6.4.2 Without Using an MS-DOS Partition
Perform the following steps to run the installation program without using a
MS-DOS partition.
NOTE

You can not use MS=-DOS if you have made it inaccessible on the file
server.

Make sure you have read Section 6.3, Before You Install the DEC RAID
Manager.

Make sure you are using a client workstation that has MS~DOS running and
the network software loaded.

For example, if the file server name is SERVER]1, type the following on the
workstation, and then press the Enter key:

Map a virtual volume to the SYS volume.
For example, if the file server name is SERVER], type the following on the
workstation, and then press the Enter key:
map #:=SERVER1/SYS:

Insert the DEC RAID Manager distribution diskette in drive A:.

If necessary, change the default path to the SYS volume path by typing the
following, and then press the Enter key:

Change the directory to system by typing the following, and then press the
Enter key:
cd \system

6-4

Installation of the DEC RAID NetWare Utility

Installation of the DEC RAID NetWare Utility
6.4 Running the Installation Program
8.

Delete the following existing DEC RAID Manager files by typing the following
and pressing the Enter key after each line:
del acf.nlm
del apr.nlm
del apc.nlm

del arc.nlm

del amd.nlm
del raidmgr.ini

Note

If the file is not found, an error message is displayed after you try to
delete it. Ignore the error message and continue deleting files.

Change the directory to public by typing the following, and then press the
Enter key:
cd \publie

10. Delete the following old RAID Manager file by typing the following, and then

press the Enter key:
dsl nwdutil.nfo

11 Change the directory to system by typing the following, and then press the

Enter key:
ed \system

. Create the following directories, if they do not already exist, by typing the
following and pressing the Enter key after each line:
mkdir raidmqgr

mkdir raidmgr\acehelp

13. Copy the following files from the distribution diskette to the directories by

typing the following and pressing the Enter key after each line:
copy a:\*.* raidmgr

copy

a:\acehelp\*.* raidmgr\acehelp

It takes a few moments to copy the files.
14. Delete any occurrences of lines which contain the following commands in the

autoexec.ncf file (located in the system directory) using a text editor.
load amd

Inatallation of the DEC RAID NetWare Utility

6-5

Installation of the DEC RAID NetWare Utility
6.4 Running the Installation Program
15. Add the following commands to the end of the autoexec.ncf file to load amd:
load raidmgr\amd

search add sys:\system\raidmgr

16. Log off the file server.
17. Reboot the server.

You have successfully installed the DEC RAID Manager.

6.5 Installation Error Messages
Table 6-2 deacribes installation error messages and probably causes during
installation of the DEC RAID NetWare Utility.
Table 6-2

Installatlon Error Messages

Massage

Probable Cause

Action To Take

*ERROR* Unable to create
RAIDMGR and/or RAIDMGR
ACEHELP directory.

SYS volume is not mounted, or
NetWare is unable to create a
directory, or software problem

First Option:
Mount the SYS volume if it is not
mounted.

*ERROR* in copying files. Re-

Media error.

Firat Option:

run RMINST.

Second Option:
Try to install again.

Try to install again.
Second Option:

Obtain a new copy of the DEC

RAID Manager, and try to install
again.

*ERRORS”* occurred in help

copy file. Files not copied:
«filenames>

SYS volume is not mounted, or
there is a media error.

First Option:

Mount the SYS volume if it is not
mounted.
Second Option:
Try to install again.
Third Option: Obtain a new copy
of DEC RAID Manager, and try to

install again.

*ERROR* DOS is not loaded:

RMINST procedure has failed.

First Option:
Load the utilities manually by
copying them from a workstation
over the network.

Second Option:

Reboot the server without
performing REMOVE DOS, and
reinstall RAID Manager.

There is not enough disk space.

First Option:

updated. No space for temporary
file.

The temporary file requires
10-12 Kbytes.

Remove some files to allow for the

*ERROR* AUTOEXEC.NCF not

There is no autoexec.ncf file

First Option:

*ERROR* AUTOEXEC.NCF not

updated. File not found.

available.

Create an autoexec.ncf and reinstall
the DEC RAID Manager.
Second Option:

Ignore the error and continue with
the installation.

(continued on next page)

6-6

instafiation of the DEC RAID NetWare Utility

instaliation of the DEC RAID NetWare Utility
6.5 installation Error Messages

Table 6-2 (Cont.)

Installation Error Messages

Message

Probable Cause

Action To Take

*ERROR* AUTOEXEC.NCF
not updated. File could not be

There is a write-protect flag
on the autoexec.ncf file or the
autoexec.ncf file is damaged.

Firat Option:
Check the permissions on the
autoexec.ncf file and remove any
write-protects.
Second Option:
Obtain a new copy of autoexec.ncf
file.

*WARNING?®* Array monitor not
started. Cause: Disk driver not
loaded.

One of the recognized drivers
that DEC RAID Manager

First Option:
Load one of the supported drivers.

*WARNING* An old version
of Array Monitor is running.
Complete this installation, then
enter "unload AMD" and “load
RAIDMGR\ AMD".

A file called amd.nlm has
already been loaded and you
cannot load an nlm file with the

*ERROR* Array Monitor not
started. Cause: AMD.NIM not
found.

There is a media error.

*ERROR* RAID Manager

The wrong location for the

written.

Software not found.

supports needs to be loaded.

First Option:
Unload AMD and load

RAIDMGR\ AMD.

same name twice.

First Option:
Obtain a new set of diskettes, and
repeat the installation procedure.

diskette was specified, or there
is a media error.

First Option:
Specify the diskette location again.

Second Option:
Obtain a new set of diskettes, and
repeat the installation procedure.

installation of the DEC RAID NetWare Utility

6~7

4
Adding Logical Units to NetWare
7.1

Using the Install Command
After you configure a logical unit, you use the NetWare Install command to create
partitions and volumes 8o that your sysiem can access it. You use this command
in the same way you would use it on a standard hard disk, using the instructions
given in the NetWare documentation. For best results, select 32K or 64K as the
block size of any volumes on the logical unit.

7.2 Adding Logical Units to Existing Volumes
You can add an array logical unit to an existing volume just as you can add any
other disk storage device. However, you should remember that the data rehability
of a combined volume is only as great as the reliability of its weakest part.
That is, if you combine a RAID 5 logical unit and a RAID 0 logical unit into the
same volume, the resulting volume has the data redundancy protection of a RAID
0 logical unit (that is, none). Even though the RAID 5 portion of the volume
has data redundancy, if the RAID 0 portion fails, the whole volume may become

usable and all data on the volume lost. A RAID 1 ana a RAID 5 volume would,
however, offer data redundancy.
In general you should not combine redundant and non-redundant disk storage

systems in the same volume.

7.3 Performance Notes
Array performance depends on many different factors, including RAID level.
block size, I/0 block size, and so on. Here are a few suggestions that may
improve performance on your array.
¢

Make your segment size at least the same size as your average /0 size.

For example, if your average I/O size is 16 Kbytes (32 blocks), change your
segment size (using the Configuration utility) to 32 blocks as well.

Always use a segment 0 size of 0.

If data redundancy is not important, try using RAID 0 instead of RAID 1 or
5. In some circumstances this improves performance. However, all data is

lost if a single drive fails.

Adding Logical Units to NetWare

7-1

8
Checking/Repairing Array Parity
8.1 What is Parity Check/Repair?
Note

Parity check/repair applies only to RAID 1 and 5. RAID 0 does not have
array parity, and therefore cannot be checked and repaired. RAID 1

does not really have "parity” either, but parity check compares data on
the mirrored drives. In addition, you cannot run parity check/repair
on a degraded RAID 1 or 5 logical unit (a logical unit with a status of

Degraded).

Parity Check/Repair performs the following functions:

Scans the logical unit and checks the array parity for each block in the logical
unit. On a RAID 1 logical unit, parity check compares the data on each
mirrored pair, block by block.

Repairs any array parity errors found during the parity check. On a RAID 1
logical unit, the array controller changes the date on the mirror disk to make
it match the data on the data disk. On a RAID 5 logical unit, the controller
changes the parity segment so that it is consistent with the data segments.
Note that if the array parity errors resulted from corrupted data, the data is

not repaired, only the array parity. Also, note that you may still lose some
data as a result of the power failure or abnormal shutdown, especially if you

do not have an uninterruptible power supply (UPS). Data cached in buffers

are lost and cannot be reconstructed if you do not have a UPS. This is one of
the reasons you should always maintain back-up files, even with a redundant
array.

8.1.1

When to Run Parity Check/Repair
Run parity check/repair at the following times:
Automatically at a time determined during the DEC RAID Manager

installation (Section 8.2). Automatic parity check/repair helps guarantee
the data integrity of the logical unit so that you can reconstruct the data on
the array if a drive fails.

Manually after an abnormal server or array shutdown (Section 8.3). As the

result of such a shutdown, required array parity may not have been updated,
resulting in potential data corruption. You must run parity check/repair on
the logical unit after you run the NetWare VRepair command to repair any
volumes on the unit.

Chacking/Repairing Array Parity

8-1

Check!v.g/Repalring Array Parity
8.2 How Automatic Parity Check/Repair Works

8.2 How Automatic Parity Check/Repair Works
This section describes how the automatic parity check/repair works. See

Section 8.3 for information on manual operation.
During the DEC RAID Manager installation, you specify when you want the
parity check/repair utilities to run automatijcally. Automatic parity check
/repair insures the data integrity of your array, which allows you to accurate:y

reconstruct data on the array after a device fails.

8.2.1 Automatic Parity Check/Repair Operation
*

During the DEC RAID Manager installation, you specify a time to run
automatic parity check/repair. You can also specify a parity check without
repairing any errors found.

At the specified time of day, the system starts the parity check operation. A
new screen is created on the system console for the parity check operation.

Parity check writes messages to this screen. If any errors are found, they are

also written to a parity error log file.
*

The log file is named device-ID.chk, where device-ID is the device ID of the
logical unit checked (for example, 8B001.chk is the name of the parity error
log file for device 8B001). The file is in the SYS/raidmgr directory. A separate
log file is created for each logical unit.

If parity errors are found, and you specified automatic parity repair, the
system repairs the parity errors.
If you did not specify automatic parity repair, you have to check the parity
error log files to see if any errors are recorded, and then run parity check
/repair manually. There is no reason not to specify automatic parity repair.

Automatic parity check/repair is performed on each logical unit in turn.

After the automatic parity check/repair process, the parity error log file

contains one of the following:
If no errors were found, the file contains only the date and time automatic
parity check/repair was run.
If errors where found, the file contains a list of the bad blocks found and
repaired (if parity repair was run).
¢

If desired, you can change the scheduled automatic parity check/repair time

see Section 8.4.
*

You can check the current settings for automatic parity check/repair by
selecting RAID Manager Configuration from the Disk Array Services
menu, then selecting General Configuration to view DEC RAID Manager

configuration parameters.

8-2

Checking/Repairing Array Parity

8.3 How to Run Manual Parity Check/Repalir

8.3 How to Run Manual Parity Check/Repair

This section describes how to run parity check/repair utilities manually. See
Section 8.2 for information on automatic operation.

You run parity check/repair manually after an abnormal shutdown (power failure,
system crash, and so on). Note that you may still lose some data as a result
of the power failure or abnormal shutdown, especially if you do not have an
uniaterruptible power supply (UPS). Data cached in buffers are lost and cannot
be reconstructed if you do not have a UPS. This is one of the reasons you should
always maintain back-up files, even with a RAID 5 array.

Y9u need to run parity check/repair after you run VRepair on the affected
volumes.

8.3.1

Manual Parity Check/Repair Operation
Manual parity check/repair is run from the DEC RAID Manager. You should
run manual parity check after you have run VRepair on any affected volumes.
You can run parity check without repairing parity (although normally there is
no reason not to repair parity).

An entry is made in the activity log showing the parity check operation.
Parity check creates a file to siore parity errors found and repaired. This file
is named device-ID .chk, where device-1D is the device ID of the logical unit
checked (for example, 8B001.chk is the name of the parity error log file for
device 8B001). The file is in the SYS/raidmgr directory.

You can temporarily change the name of the parity error log file using the
Change Parity Report File Name option in the Disk Array Parity menu. This
change applies only to this logical unit, and only for the next run of parity
check. If you leave the DEC RAID Manager without running parity check.
the change no longer applies.

After you run manual parity check/repair, the parity error log report is
automatically displayed, without your having to enter the file name of the file.
You must run parity check/repair on each logical unit separately.

8.3.2 Running the Check
Use the following procedure to check and repair parity manually.
L

If you are running parity check/repair after an abnormal system shutdown,
first run VRepair on all affected volumes before checking parity.
Start the DEC RAID Manager by entering the following command from the
console:
load raidmgr

Use the Device Selection option in the Disk Array Services menu to select the
logical unit you want to check.

After you have selected the logical unit you want to check, select Array Parity
from the Disk Array Services menu. The following menu is displayed:

Checking/Repairing Array Parity

8-3

Checking/Repairing Array Parity
8.3 How to Run Manual Parity Check/Repair

6. From the Disk Array Parity menu, you can select the following options:
*

Check and Repair Parity—use this option to check and repair parity on
the selected logical unit. The parity error log file contains a list of the
errors found and repaired.

Check Parity (no repairl—use this aption to check parity on the selected
logical unit without repairing any errors found. The parity error log file
contains a list of the errors found.

Change Parity Report File Name—use this option to change the name of
the parity error log file for the current logical unit before you run parity
check/repair. This change applies only to this logical unit, and only for
the next run of parity check.

View Parity Report—use this option to view the parity report file for any

logical unit.

Change Automatic Parity Time—use this option to change the scheduled

time for au' ymatic parity check. You can not change the automatic parity
repair parameter here.

If you want to write the parity error report to a file other than the default file,
select the Change Parity Report File Name option and enter the new name.
Any new name applies only to this logical unit and only for this run of parity
check/repair.

Select either Check and Repair Parity or Check Parity (no repair) to start

parity check.
8.

A prompt is displayed to confirm that you want to check parity. Press Yes and

the Enter key to start the selected operation.
9.

A popbox is displayed as parity check is performed showing the percentage of
the check completed. When the operation is completed, press the Enter key.

10. The parity error report for the logical unit you just checked is displayed.
*

If you selected parity check/repair, this is a list of the parity errors found
and repaired. If no errors were found the report indicates this.

If you selected parity check without repair, this is a list of the parity
errors found. To repair these errors, you have to run parity check with
repair. If no errors were found the report indicates this.

11. After viewing the report, press the Esc key to return to the Disk Array Parity
menti.

12. If you want to check more logical units, press the Esc key to return to the

Disk Array Services menu, then use the Device Selection option to select

another logical unit. Then go back to Step 4 and repeat this procedure.
13. Press the Esc key until you get the DEC RAID Manager exit prompt, then

exit the DEC RAID Manager.

8-4

Checking/Repairing Array Parity

8.4 Changing Parity Check Scheduling

Use the following procedure to change the acheduled time for automatic parity

check/repair.

Note

You can also change the scheduled time (but not the parity repair option)

by selecting Change Automatic Parity Time from the Disk Array Parity
menu.

Start the DEC RAID Manager by entering the following command from the
console:
load raidmgr

From the Disk Array Services menu, select RAID Manager Configuration.
From the RAID Manager Configuration menu, select General Configuration.
Use the Down-arrow key to highlight the Parity Check Start Time parameter

on the General Configuration display.

Enter the new time for automatic parity check. Enter the time as a 24-hour

value (that is, 1:00 PM is entered as 13:00).
important

While the parity check/repair utility is running, workstations connected
to the server may slow considerably until the check function is completed.
During the utility’s daily run, try and choose a time when this slowdown
does not affect your operations.

After you enter the new time, press the Enter key The cursor automatically
moves down one to highlight the Automatic Parity Repair option.
If you want to change the current value, press any key except the space bar
or the Enter key to toggle to the other value (yes or no). There is normally no
reason not to specify automatic parity repair.

After you have specified the parity repair parameter, press the Enter key to
save the new value. The cursor automatically highlights the next parameter.

Press the Enter key to exit the General Configuration menu.
10. You are then prompted to determine whether you want to leave the menu

with the current changes, leave without the changes, or return to the General

Configuration menu to make more changes.

11. After you select Yes, you return to the RAID Manager Configuration menu.

. Press the Esc key until you get the DEC RAID Manager exit prompt, then

exit the DEC RAID Manager.

Checking/Repairing Array Parity

8-5

9
The Array Monitor Daemon
9.1 Array Monitor Daemon
The Array Monitor Daemon scans for non-optimal status on your array devices
and notifies you when one is found. While these status changes are also written
to the system error log and on the console screen, the Array Monitor Daemon
notifies you immediately of the changes in status, and displays a message on the
console terminal, regardless of which screen you happen to be in.

9.1.1

How the Array Monitor Daemon Works
The Array Monitor Daemon operates as fqllows:

At an specified interval (determined by the delay parameter value), the
daemon polls the array devices (both logical units and individual drives) for
their status.

If a status other then Ok - Optimal is returned (for any array component),
the daemon takes over the console screen and displays the other status on the
console screen for a specified period (determined by the persistence parameter
value).

If at the end of the persistence interval you have not acknowledged the
message, the daemon releases the screen. It displays the status again after
the specified delay interval.

To acknowledge the message, press any key except the Enter key. The
message then remains on the screen until you press the Enter key. After you
press the Enter key, the delay interval begins again.

The Array Monitor Daemon continued to display the same status until you correct
the condition that caused the status.
important Note

The Array Monitor Daemon (AMD) does not poll array devices if DEC

RAID Manager, the NetWare Ingtall utility, or the NetWare VRepair
utility is currently loaded.

9.1.2 What to do When an Error is Displayed
When the Array Monitor Daemon displays an ervor status on the console, you
should take the following steps:
1.

When the message first appears, press any key except the Enter key to hold
the message on the screen (pressing the Enter key clears the screen—if that
happens, you have to wait for the message to be displayed again).

Read the message, taking note of the array device named in the message (see
the AMD message format Section 9.3).
The Array Monitor Daemon

9-1

The Array Monitor Daemon
9.1 Array Monitor Daemon

3. Change to the console screen (by pressing the Alt-Esc key) to see if any
messages are displayed there (they may have scrolled off the screen).

Start the DEC RAID Manager and check the logical unit status and drive
status of the devices named in the error message.

Determine the nature of the error. See Section 2.5 for a list of logical unit

status and drive status.

Read the system error log to see if there are any associated messages in the
log that might further explain the error and its consequences. Frequently,

there are also standard NetWare error messages associated with the error.
7.

Take whatever action is required to restore your array (Chapter 5 of this
manual).

9.1.3 Error Log Example
This section gives an example of Array Monitor Daemon operations on a RAID 5
logical unit. The example represents one set of error messages you might see if a
single drive fails on a RAID b logical unit. It is not necessarily the only sequence
(the circumstances which caused the drive to fail may result in different error
messages being displayed).
1.

You are using the logical unit when the following message appears on the
console terminal screen (which is not current set to the system console) at
1:40 PM:
Disk Array Monitor
Version xx.xx.xx

Davice #7 (8B001) Degraded: Drive Failure
Array Drive Error: Channel 1, ID 0: Drive Failure

Press the Enter key to clear screen, any key to hold screen.

You press any key except the Enter key to keep the message on the screen so
that you can read it.

From this message, you can see that a drive has failed in a RAID 1 or RAID
b logical unit, leaving the logical unit in the degraded mode. The logical unit

involved is Device #7.
4.

After you have read the message, press the Enter key to clear the error
message from the screen.

Press the Alt-Esc key to switch to the console screen and see the following
messages:

7/29/92 1:39 PM: 0.0.0 Array Monitor Dasmon: Array Device Error:
Device §#7

(8B001) Status: Degraded:

Drive Failure

7/29/92 1:39 PM: 0.0.0 Array Monitor Daemon: Array Component Disk Error:
Channel 1 ID 0,

Status: Failed:

Drive Failure

Next, start the DEC RAID Manager and use the Device Selection option to
select Device #7. You can see from the display that Device #7 is a RAID 5
logical unit and that one of the drives in the array has failed.

Check your system error log until you find the corresponding messages (from

the Disk Array Services menu, select System Messages, then System Error
Log). In this case, you see the following messages:

9-2

The Array Monitor Daemon

The Array Monitor Daemon
9.1 Array Monitor Daemon
7/29/92 1:39 PM Severity = 3
0.0.0 Array Monitor Daemon: Array Device Error:
Device #7 (8B001) Status: Degraded: Drive Failure
7/29/92 1:39 PM Severity = 3

0.0.0 Array Monitor Daemon: Array Component Disk Error:
Channel 1 ID 0, Status: Failed: Drive Failure

Check any messages preceding and following to see if they clarify the reasons
for the drive failure. In this case, there are no further messages.
Note you could see standard NetWare error messages relating to the array

error messages. For example, for an error more serious than the one in this
example you might see:
9/27/91 10:57:38 am Severity = 4.
1.1.10 Device $#1 (8B001) *** NCR ADP-92/01 0100 deactivated due to drive failure,
9/27/91 10:57:38 am Severity = 4.
1.1.86 Volume ARRAY dismounted due to drive deactivation

These messages, from NetWare, indicate that the array device has been
deactivated as the result of a drive failure, and that the volume ARRAY
(contained on the deactivated device), has been dismounted. These messages
would follow array-specific messages indicating drive or logical unit failure.
Search through the rest of the error log to make sure there are no errors
affecting other logical units or other errors affecting this unit.
10. Given the error indicated, you would then go to Chapter 5 for the steps

required to restore a RAID 5 array. (In this case, you would replace the failed
drive and reconstruct the data on it.)
11. Clear the error log (copy it to a file first so that you can refer to it later if
necessary).

The Array Monitor Daemon

-3

The Array Monitor Daemon
9.2 Error Message Types

9.2 Error Message Types

The DEC RAID Manager and the disk array software generate three basic types
of error messages:

Array-specific error messages, which are displayed only for the disk array.

These measages are displayed by AMD on the current screen, the console
screen, and in the system error log (see Section 9.4).

Standard NetWare error messages. These messages are displayed on the
console and written to the system error log. These messages are also used
for devices other than the disk array. See your NetWare System Messages
documentation for an explanation of these messages. These messages are not
displayed by the Array Monitor Daemon.

Error messages that occur while you are running the DEC RAID Manager.
These messages are displayed on the console as soon as they occur. They are
not written to the system error log (Section 9.5).

9.3 AMD Error Message Format
There are three kinds of AMD messages:

9.3.1

AMD messages that appear in the pop-up screen.

AMD messages displayed on the system console.

AMD messages written in the system error log.

AMD Pop-Up Message Format
AMD pop-up messages are displayed periodically on the console terminal,
regardless of which screen is currently displayed, until you acknowledge the
message. They have the following format:
Disk Array Monitor

Version xx.xx.xx
Device #n

(devIiD)

lun-status

Array Drive Error: Channel x, ID y: drive-status

n is the device number of the logical unit affected by the status change.

devID is the device ID of the logical unit affected by the status change.
Device IDs have the following format:

The first 2 digits indicate the type of host adapter. These digits are assigned

by NetWare. For example, 8B is the DPT PM2012B differential host adapter.
The 3rd digit is the board number assigned to the host adapter by the
NetWare system.

The 4th digit is the SCSI ID of the array controller.
The 5th, or last, digit is the logical unit number.
¢

lun_status is the current status of the affected logical unit.

x jg the channel number of the drive affected by the status change.

yis the SCSI ID of the drive affected by the status change.

drive_status is the current status of the affected drive.

9—4 The Array Monltor Deemon

The Array Monitor Daemon

9.3 AMD Error Message Format

9.3.2 AMD Console Message Format

AMD console messages are displayed on the system console once when the status
change is detected. AMD console messages have the following format (the first
message given is for a logical unit; the second message is for an array drive):
mm/dd/yy bh:mm : 0.0.0 Array Monitor Daemon: Array Device Error:
Device #n (devID) Status: lun-status

mm/dd/yy bh:mm : 0.0.0 Array Monitor Daemon: Array Component Disk Error:
Channel x ID y, Status: drive-status

mm/dd/yy hh:mm is the time the status change was detected.

n is the device number of the logical unit affected by the status change.

devID is the device ID of the logical unit afiected by the status change.
Device IDs have the following format:

The first 2 digits indicate the type of host adapter. These digits are assigned

by NetWare. For example, 8B is the DPT PM2012B differential host adapter.

The 3rd digit is board number assigned to the host adapter by the NetWare
system.

The 4th digit is the SCSI ID of the array controller.
The 5th, or last, digit is the logical unit number.

lun_status is the current status of the affected logical unit.

xis the channel number of the drive affected by the status change.

yig the SCSI ID of the drive affected by the status change.

drive_status is the current status of the affected drive.

9.3.3 AMD Error Log Message Format
AMD error log messages are written into the system error log when the status
change is first detected. They have the following format (note that this is a
standard NetWare error log format):
dd/mm/yy hh:mm Severity = z
0.0.0 Array Monitor Daemon: Array Device Error:
Device #n {(devID) Status: lun-status

dd/mm/yy hh:mm Severity = z
0.0.0 Array Monitor Daemon: Array Component Disk Error:
Channel x ID y,

Status: drive-status

mm/dd/yy hh:mm is the time the status change was detected.

z—the severity of the error:
0 = Informational
1 = Warning

3 = Cnitical
4 = Fatal

n is the device number of the logical unit affected by the status change.

devID is the device ID of the logical unit affected by the status change.
Device IDs have the following format:
The first 2 digits indicate the type of host adapter. These digits are assigned
by NetWare. For example, 8B is the DPT PM2012B differential host adapter.

The Array Monitor Daemon

9-5

The Array Monitor Daemon
9.3 AMD Error Message Format
The 3rd digit is board number assigned to the host adapter by the NetWare
system.

The 4th digit is the SCSI ID of the array controller.

The 5th, or last, digit is the logical unit number.
*

lun_status is the current status of the affected logical unit.

xis the channel number of the drive affected by the status change.

yis the SCSI ID of the drive affected by the status change.

drive_status is the current status of the affected drive.

9.4 AMD Error Messages
The following messages are AMD messages, and appear on you system console if
AMD detects a status change in the array.

9.4.1

Array Component Errors
These errors have the following format.
Array Monitor Daemon: Array Component Error
<drive name > Status:

Array Formatting

where <drive name > is the device ID of the drive.

Note

The errors in the following tables are all Array Cov.punent Errors, so the
words Array Moniter Daemon: Array Component Error are not displayed
in the following tables.

Table 9-1

Array Component Errors

Message

Probable Cause

Action To Yake

<drive name> Status:

The array controller is

No action.

«<drive name> Status:
Bad Channel

You have installed a drive that
has a different channe] than the
drive that was originally in its

Replace the drive with a
drive having the correct
channe!.

The drive failed the Test Unit
Ready (that occurs during a
logical unit format) after the

Replace the drive, and
meke sure that the data ie
restored tu the logical unit

The drive failed during & write

Replace the drive, and

Array Formatting

formatting a drive.

place.

<drive name> Status:
Drive Not Ready

drive was hot swapped, added,
or after the subsystem was
started.

<drive name> Status:

Failed

to the logical unit. or failed the
initialization process.

containing it.

make sure that the data is
restored to the logical unit
containing it.

{continued on next page)

98-8

The Array Monitor Daemon

The Array Monitor Daemon
9.4 AMD Error Messages

Table 9-1 (Cont.)

Array Component Errors

<drive name> Statue:

Probabie Cause

Action To Take

The drive was failed through

Replace the drive, and
make sure that the data is
restored to the logical unit
containing it.

Failed by User

ACE.

<drive name> Status:
Format Failed

The SCSI format issued to the
drive failed. This format was
issued through ACE as a logical
unit format or drive format

Try to format the logical
unit again. If it fails
again, replace the drive
and make sure that the
data is restored to the
logical unit containing it.

<drive name> Status:
Tlogical SCSI ID

You have installied a drive that
has a different ID than the drive
that was originally in its place.

Replace the drive with a

<drive name> Status:
Newly Replaced

The drive has just been replaced.

No action.

<drive name> Status:
Status: OK - Warning
Condition

The drive has experienced errors

Backup (if possible) and
restore the logical unit.

<drive name> Statua:
Status: OK - Warning
Condition

The drive is part ofa RAID 0
logical unit and has experienced
errors while reading or writing.

Backup (if possible) and
restore the logical unit.

<drive name> Status:
Reconstrueting

The drive data is being
reconstructed by the array
controller.

No action.

<drive name> Status:
Start of Day Failure

At system boot-up, the drive

before reconstruction.

while read operations were
occurring.

drive of the correct 1D.

Backup (if possible) and

restore the logical unit.

failed because it did not accept
writes from the array controller.

<drive name> Status:
Write Failed

The drive failed when the array
controller was initializing it for

Backup (if possible) and
restore the logical unit.

array parity.

<drive name> Status:
Wrong Capacity

<drive name> Status:
Wrong Ctrl Serial No.

The drive is incompatible with
the other drives in the logical
unit because its capacity ie
different.

Remove the drive and
replace it with a certified

The drive that was just installed
was originally connected to a
different array controller and
was formatted with special
information about that array

Fail the drive using ACE,
and restore the logical

drive.

unit.

controller, AND it replaced a
drive that was optimal, spare, or
warning.

<drive name> Status:
Wrong Parameters

The drive is incompatible with

the other drives in the logical
unit because it is a different size.

Remove one drive and
replace it with a drive that
is the same size as the
other drives.
(continued on next page)

The Array Monitor Dasemon

9-7

The Array Monitor Daemon
9.4 AMD Error Messages

Table 9-1 (Cont.)

Array Component Errors

Message

Probable Cause

Action To Take

<drive name> Status:

An optimal or warning drive
in a degraded logical unit was
replaced, or an optimal or spare
drive or a drive that is being
formatted was replaced and then
the subsystem was shutdown.

Replace the drive that
was mistakenly removed
from the subsystem, turn
on the subsystem, and
remove the drive that you
had intended to remove.
CAUTION: DO NOT send

Wrong Replacement
Disk

1/0 to the logical unit until
you have completed all the

steps above.

<drive name> Status:
Wrong Sector Size

The drive is incompatible with
the other drives in the logical
unit because its sector size ia

Remove the drive and
replace it with a certified
drive.

different.

9.4.2 Array Device Errors
These errors have the following format:
Array Monitor Daemon: Array Device Error
<LUN name > Status:

Bad Channel

where <LUN name> is the device name of the logical unit.

Note

The errors in the following tables are all Array Device Errors, so the
words Array Monitor Daemon: Array Device Error are not displayed in

the following tables.

Table 9-2

Array Device Errors

Message

Procbable Cause

Action To Take

<LUN name> Status:
Bad Channel

The logical unit is dead because
z drive that has a different

Replace the drive with the
wrong channel with a drive
of the correct channel.

channel than the drive that
was originally in its place was

installed.

<LUN name> Status:
Bad Identifier

The logical unit is dead because
a drive has a different ID than
the drive that was originally in

Replace the drive with
the wrong channel with a
drive of the correct ID.

The logical unit is dead because
a drive that has a different
capacity from the other drives
in the logical unit has been
installed.

Replace the drive with
the wrong capacity with a
certified drive.

its place was installed.

<LUN name> Statuas:
Bad Parameters

{continued on next page)

8-8

The Array Monitor Daemon

9.4 AMD Error Messages

Table 9-2 (Cont.)

Array Device Errors

Message

Probable Cause

Action To Take

<LUN name> Status:
Component Failure

Two drives were removed from
a RAID 5 logical unit, and then
the multi- rank subsystem was
shutdown.

Shutdown the subsystem,

<LUN name> Status:

A drive has failed in the logical

The drive neede to be
replaced as socn as

Degraded - Drive
Failure

unit.

replace the drives in their
original positions. and
restart the subsystem.
CAUTION: DO NOT send
1/0 to the logical unit until
you have completed the
steps above.

possible. The logical unit
remains degraded until the

drive is replaced.

<LUN name> Status:
Degraded - Drive
Formatting

<LUN pame> Status:
Drive Failures

The logical unit is degraded and

No action.

the drive that was just replaced

is being formatted by the array
controller.

Multiple drives are failed in the
logical unit.

The drives need to be
replaced. Then the logical
unit nezds to be formatted
and the data copied from
the back-up media to the

logice! unit.
<LUN name> Status:
Formatting

The logical unit is being
formatted by the array

<LUN name> Status:

No action.

OK - Reconstructing

The array controller is
reconstructing the data on the
newly-replaced drive.

<LUN name> Status:

The logical unit is dead and

Format the logical unit
using ACE.

No ection.

controller.

Waiting for Format

must be formatted.

tcontinued on next page)

The Array Manitor Dasmon

98-8

The Array Monitor Daemon
9.4 AMD Error Messages

Table 9-2 (Cont.)

Array Device Errors

Message

Probable Cause

Action To Take

<LUN name> Status:

The logical unit is dead because
a drive that was either optimal
or in warning was replaced.

First Option:
If no writes are sent to

Wrong Drive.

the logical unit, do the
following:
¢

Remove the new drive.

Replace the original
drive.

Either, install the new
drive in the correct
location if the wrong
drive was replaced, or
fail the warning drive
using ACE and replace
it.

Second Option:
If writes are sent to the
logical unit, reformat the
logical unit and copy the
data from the backup

media to the logical unit.

9.4.3 Other Status Change Errors

Table 9-3

Other Status Change Errors

Message

Probable Cause

Action To Taks

Automatic Disk Array
Parity Check Complete

Displayed once daily when the
automatic parity check utility
completes.

No action.

Automatic Disk Array
Parity Check Started

Displayed once daily when the
automatic parity check utility

No action.

starts.

Array Monitor Dasmon:
<LUN 0 name> Obsolete
Downloadable Controller

An old version of irmware is on
the array drives.

Download new firmware.

Array Monitor Daemon:

A power supply failed.

Replace the failed power

Arre v Monitor Daemon:

A subsystem component failed
(for example, a fan).

component.

Firmware [Version

xx.yy.zz)
I wer Supply #x Failure

Subsy. *em Unique

Device #.. failure

Disk Array Parity

Check: Device #n
(DevID)

2-10

The Array Monitor Dasmon

The automatic parity check
started for this device.

supply.

Replace the failed

No action.

The Array Monitor Daemon
9.5 DEC RAID Manager Messages

9.5 DEC RAID Manager Messages
Table 94 lista the error message explanations given in the DEC RAID Manager
messages. These explanations may appear whenever you are using a DEC RAID
Manager function or option.
Table 9-4

DEC RAID Manager Messages

Expianation

Probable Cause

Action To Take

ABORT: Insufficient
Memory

Software problem.

Call your Customer
Services Representative.

ABORT: Invalid

The incorrect option was typed.

Type the command again
using the correct option.

ABORT: No message
text for confirmation
header <#>, <cmessage>

Software problem.

Call your Customer
Services Representative.

ABORT: Unable to

Software problem.

Call your Customer

The file is deleted, or it was
never installed properly.

Remove the DEC RAID
Manager and reinstall it.

ABORT: Unable to

The file is deleted, or it was
never installed pro,erly.

Remove the DEC RAID
Manager and reinstall it.

ABORT: Unable to

The file is deleted, or it was
never installed properly.

Remove the DEC RAID
Manager and reinstall it.

Error encountered
during disk array parity

The array controller was unable

Call your Customer

to complete the array parity
operation due to an array
controller software error.

Services Representative.

Error encountered
during diek array parity
repair. The array device

The array controller was unable

Call your Customer
Services Repregentative.

File <file> on path
<path> cannot be
ereated or opened for

There is a problem with the

Try the operation again.
If it fails again. call

option <option> in the
command line.

Services Representative.

create confirmation
menu

ABORT: Unable to

locate the DEC RAID
Manager Data File

locate the DEC RAID
Manager Help File

locate the DEC RAID
Manager Message File

checking.

may be require other

to complete the array parity
operation due to an array
controller software error.

file, or there is a DEC RAID
Manager software problem.

your Customer Service

Representative.

File <file> on path
<path> cannot be
opened for reading.

There is a problem with the
file, or there is a DEC RAID
Manager software problem.

Try the operation again.
If it faile again, call
your Customer Service
Representative.

Firmware data file <file>

the DEC RAID Manager cannot
find the file in the path you
selected.

Try the operation again

is not present in path

<path>.

uring the correct path

name. If the problem
occurs again, call your
Customer Services
Representative.

(continued on next page)

The Array Monitor Daemon

9-11

The Array Monitor Daemon
9.5 DEC RAID Manager Messages

Table 94 (Cont.)

DEC RAID Manager Messages

Explanation

Probable Cause

Action To Take

Incomplets download
file or invalid download
record format.

There iz a problem with the
diskette or with the file.

Try te download the

firmware again. If the

problem occurs again, call
your Customer Services

Representative.
the DEC RAID Manager
Help file <file> is

Software problem.

corrupt.

Remove the DEC RAID
Manager utilities and
reinatall them. If the
problem occurs again, call
your Customer Services

Representative.
Unable to get exclusive
access to the array

device. A volume on the
device may be mounted,
or another process may

have the device locked.

A volume may be mounted, or

there is another rrocess that is
presently accessing the logical
unit (for example, the Install
utility, or another NetWare
NLM).

First Option:
Unmount the volume if it
is mounted, then try the
operation again.

Second Option:
Try to terminate the other
process accessing the
logical unit, then try the
operation again.

Unable to complete the
operation becauss an
error occurred during
a SCSI command. For
detailed information,
examine the DEC RAID
Manager Activity Log.

Unable to copy download
file from disk. File may
not be present, or SYS
volume is full.

There is a software problem or a
host adapter problem.

Read the DEC RAID
Manager Activity Log
to determine the exact
problem that occurred.

SYS volume is out of space, or

First Option:
Check the SYS volume to
determine if it is full, and

there is a problem with the DOS

copy.

make some space on the
disk if it is out of space.
Second Option:

Fix the DOS copy problem.
Unable to copy file form

flexible disk. DOS must
be present to perform
the copy.

8-12

The Array Monitor Dasmon

DOS is not available and must

be for the firmware download.

Reboot your system.

At the DOS prompt,
enter server -na, then
press the Enter key.

Download the
firmware again.

Part VI
DEC RAID SCO UNIX Utility

This part of the guide contains:

Chapter 10, Installation of the DEC RAID SCO UNIX Utility

Chapter 11, Array Parity Check/Repair

Chapter 12, RAID Monitor Utility

10
Installation of the DEC RAID SCO UNIX Utility
The DEC RAID SCO UNIX Utility provides additional functions in the SCO
UNIX environment which are not present using the DEC RAID Stand-Alone
Utility. These include a RAID Monitor for notification of changes in logical
unit status or drive status, a RAID status utility for displaying current logical
unit status and drive status, and a parity check/repair utility for checking and
repairing parity on the disk array.
Chapter 2 describes the DEC RAID Stand-Alone Utility, its installation and
functionality. This chapter discusses installation of the DEC RAID SCO UNIX
utilities. Chapter 11 discusses the parity check/repair utility. Chapter 12
describes the array monitor utility, and Chapter 13 the RAID status utility.

10.1

About Installation
The DEC RAID SCO UNIX Utility kit is a floppy containing executables, shell

scripta, and an SCO UNIX Sdsk driver. Table 101 describes the contents of the
kit.

Table 10-1
Flle Name

makedpth

DEC RAID SCO UNIX Utliity Kit
Description

Shell acript to create device files /dev/dptHO and /dev

/dptH1 uged by the utilities

monitor

RAID Monitor executable

monitorsh

Shell script to run the RAID Monitor

parity

RAID parity che-k/repair executable

paritysh

Shell script to run the parity check/repair utility

raidstat

RAID status utility executable

raidstatsh

Shell script to run the RAID status utility

read.me

informational file for the user to read describing each

help.txt

Text file for help messages

of these files

The next sections describe installation procedures for each of these utilities.

Chapter 11, Chapter 12, and Chapter 13 provide further details on each of the
utilities, their options, and their error messages.

installation of the DEC RAID SCO UNIX Utility

10-1

Installation of the DEC RAID SCO UNIX Utility
10.1 About Installation

10.1.1 Installing the RAID Manager Parity Check/Repair Utility
Installation of the RAID Manager parity check/repair utility involves the
following steps:
1

Copy the following files from floppy to any directory of choice:

parity

paritysh

makedpth

help.txt

Change to superuser mode.

Set the PATH variable to point to the directory where the files reside.
PATH=$PATH/directory_name

Type makedpth to create the device files in /dev required by the utilities.
At this point, the RAID parity check/repair utility can be run. Chapter 11
describes running the RAID parity check/repair utility and its options.

10.1.2 Installing the RAID Monitor Utility
Installation of the RAID Monitar utility involves the following steps:
1.

Copy the following files from floppy to any directory of choice:
*

monitor

monitorsh

makedpth

help.txt

Change to superuser mode.

Set the PATH variable to point to the directory where the files reside.
°*

PATH=$PATH/directory_name

Type makedpth to create the device files in /dev required by the utilities.
At this point, the RAID Menitor can be started. Chapter 12 describes running
the RAID Monitor and its options.

10.1.3 RAID Status Utility
Installation of the RAID status utility involves the following steps:
1.

10-2

Copy the following files from floppy to any directory of choice:
*

raidstat

raidstatsh

makedpth

help.txt

Change to superuser mode.

Installation of the DEC RAID SCO UNIX Utility

Installation of the DEC RAID SCO UNIX Utility
10.1 About Installation

. Set the PATH variable to point to the directory where the files reside.
¢

PATHe$PATH/directory_name

Type makedpth to create the device files in /dev required by the utilities.

At this point, the RAID status utility can be run. Chapter 13 describes
running the RAID status utility and its options.

instaliation of the DEC RAID SCO UNIX Utility

10-3

11
Array Parity Check/Renair

11.1 RAID Parity Check/Repair Utility
Parity check/repair performs the following functions:

The RAID parity check/repair utility runs a parity check and/or repair on a given
LUN. Parity check/repair applies only to RAID 1 and RAID 5. RAID 0 does not
Fave array parity. Parity check on a RAID 1 configuration simply compares the
data on the mirrored drives. Parity check/repair cannot be run on a degraded
RAID 1 or RAID 5 LUN. See the discussion at the beginning of Chapter 8 for
further definition of parity check and repair.

11.1.1

Getting Started
The RAID parity check/repair can be invoked in one of two ways: through the
shell script or by invoking the executable directly. The shell script method
prompts you for options and invokes the monitor executable. The executable
requires ewitch options to be specified, and is provided in the event that you
would prefer to include in customized shell scripts.

11.1.1.1

Using the Shell Script

The RAID parity/repair utility can be invoked by typing paritysh at the system
prompt. The following output will appear:

Parity Check/Repair Utility Shell Script
Version XX.X

(C)

Digital Equipment Corporation 1992

Enter values for the following parameters or hit <CR>
for default value:

Enter SCSI ID

[0]:

The format of the input prompt line shows the parameter to enter, followed by

square brackets containing the default value of that parameter. Press the Return
key or the Enter key to select the default value, or enter some other value for
that parameter.

A full help menu which describes the options is available by typing paritysh -h

at the system prompt. If help information for a specific parameter, enter i when
that parameter is prompted for, and a description will be displayed.

Array Parity Check/Repair

11-1

Array Parity Check/Repair
11.1 RAID Parity Check/Repair Utility

Once all the parameters have been specified, the shell script invokes the parity

executable, redirecting the output to a specified log fi.=, and piping the execution
statement to the SCO UNIX at utility. This schedules the parity check/repair to
occur at the specified time.

The parity check/repair runs as a background process. The firet phase reads each
block on the LUN and verifies that the parity is good. It creates a unique file in
/tmp which lista the blocks which had bad parity. If a repair is to be done, this
file is opened, and the parity of the blocks in that file is repaired.

11.1.1.2

Invoking the Executabla

The RAID Monitor can also be started by invoking the executable directly. Enter
parity [options] to start execution. The -h option prints out a help menu with a
description of the various switches.

11.1.2 Options
The options available to both the shell script and executable are outlined the
Table 11-1.

Table 11-1

Parity Check/Repair Utility Options

Options

Symbol

Description

SCSI ID

41D

The SCSI ID of the DEC RAID Subystem. The

SCSI LUN

JLUN

The SCSI LUN. This value is not used at this time

Device File

-dFILENAME

value must be from 0 to 7, with a default value of 0.

and should be set to the default value of 0.

The device file name for access to the DPT Host

Adapter.
*

For the shell script, the default is 0. for /dev
/dptHO which is the device file for the DPT
controller as a primary controller. Enter 1 for
/dev/dptH1 which is the secondary controller,

Invoking the executable, this option does not
need to be specified for the default value of 0.
If the secondary controller is required, either
-d/dev/dptH] or -1 can be specified.

Check only

-C

Select parity check option only (no repair).

Time

n/a

Time of day to run the parity/check utility. Used

by the paritysh only. If no time ia specified, then
the utility will be run within the next few minutes.
Invoking the executable will caure the Parity check

/repair to rup in the foreground immediately, unless
oiped to at or some other scheduling utility.

Log File Name

11-2

Array Parity Check/Repair

n/a

Name of the log file to write errors. Used hy the
paritysh only. The user must manually redirect the
output to a file if the executable is invoked.

Array Parity Check/Repalr
11.1 RAID Parity Check/Repair Utility

11.1.3 Output

The output goes to the screen or to a log file depending on how the utility is run.

The output is shown below.

Parity Check/Repair Utility
Version X.XX

Parity Check begun timestamp
Comparing block 0 H to FFFF H

Parity Check completed timestamp
Number of blocks with bad parity: 0
Parity Repair begun timestamp
Parity Repair completed timestamp
Number of repaired blocks:

The output lists the range of blocks whose parity is checked. If at any time a
block with bad parity is found, the following message appears:
*** Bad parity found at block a *¥#*

The total number of blocks with bad parity will be displayed after the check is
complete. The repair portion of the display will show any blocks repaired with
.

the following message:
Repairing parity at block a

A final count of all blocks repaired is given at the end of the display.

11.1.4 Error Messages
The error messages returned by the parity check/repair utility use the following
format:

For general error messages:
PARITY:

message

Generically, if an error occurs during a command to the controller which results
in either a DPT Host Adapter error, or a SCSI status other than good, that
information will be presented to the user in the following format:
*

SCSI Check Condition Status (2 H)
SCSI Check Condition
SCSI Sense Rey: 6 H ASC:

29 H

ASCQ:

O H

In this example, the SCSI status is a Check Condition (2 H), the Sense Key
is 6 H, the Additional Sense Code (ASC) is 29 H, and the Additional Sense
Code Qualifier (ASCQ) is 0 H indicating a power on, reset or bus device reset

occurred. Appendix A lists all the ASC’s, ASCQ’s and Sense Keys valid for the
array controller. Note that these values are given in hexadecimal notation.

Array Parity Check/Repair

11-3

Array Parity Check/Repair
11.1 RAID Parity Check/Repair Utility
¢

Host Adapter error or SCSI Status (other than Check Condition)
SCSI Status: 0 H
Host Adapter Status: 7 H

In this example, the SCSI status is 0 H and the Host Adapter Status is 7 H

indicating a bus parity error. Appendix A lists all the SCSI status codes for
the array controller as well as the host adapter error codes.
The following list describes the general error messages returned from the utility.
Error Opening Controller Device File x—Results from problems opening /dev
/dptHO or /dev/dptH1. Verify that those device file exist.

Error issuing read capacity to devz'cé—-Array controller was unable to respond
to the request sense command.

Error issuing verify command to device—~Airay controller was unable to
respond to the verify command.
Error repairing parity on device~An error occurred while repairing the parity
blocks.
Illegal switch - x—An illegal option was specified.

Error Opening Bad Parity Blocks File—An error occurred while attempting to
open the bad parity blocks file in /tmp.

Error Reading Bad Parity Blocks File~~An error occurred during the repair
phase trying to open the bad parity blocks file in /tmp.

11-4 Array Parity ChecivRepair

12
RAID Monitor Utility

12.1 RAID Monitor
The RAID Monitor performs the following functions: The RAID Monitor notifies
the user of any change of status in the Logical Units (LUN) or devices in the
array. Messages can appear on /dev/console and/or in a file depending on option
selection.

12.1.1

Getting Started
The RAID Monitor can be invoked in one of two ways: through the shell script
or by invoking the executable directly. The shell script method prompts the user
for options and invokes the monitor executable. The executable requires switch

options to be specified, and is provided in the event that the user would prefer to

include it in customized shell scripts.

12.1.1.1

Using the Shell Script

The RAID Monitor can be invoked by *vping monitorsh at the system prompt.
The following output appears:

RAID Monitor Utility Shell Script

{C) Digital Equipment Corporation 1992

Version XX.X

BEnter values for the following parameters or hit <CR>
for default value:
Bnter SCSI ID [0]:

The format of the input prompt line shows the parameter to enter, followed by
square brackets containing the default value of that parameter. Press the Return
key or the Enter key to select the default value, or enter some other value for
that parameter.

A full help menu which describes the options is available by typing monitorsh
-h at the system prompt. If help information is needed for a specific parameter,

enter A when that parameter is prompted for, and a description is displayed.
Once all the parameters have been specified, the shell script invokes the monitor
executable, redirecting information output to /dev/console, and error information
to a logfile.
The monitor runs as a background process in the system. When the monitor is
started,.a status of the logical units and drives is displayed. This information is
similar to the information display when the RAID status utility the screen.

RAID Monitor Utility

121

RAID Monitor Utility
12.1 RAID Monitor

12.1.1.2

Invoking the Executable

The RAID Monitor can also be started by invoking the executable directly. Enter
monitor [options] to start execution. The -A option prints out a help menu with
a deacription of the various switches. It is the user's responsibility to redirect
standard output and standard error if necessary.

12.1.2 Options
The options available to both the shell script and executable are outlined the
Table 12~1.
Table 12~1

RAID Monitor Options

Options

Symbol

Dascription

SCs8I 1D

41D

The SCSI ID of the DEC RAID Subystem. The

SCSI LUN
:

JLUN

The SCSI LUN. Thie value is not used at this time
and should be set to the default value of Q.

Device File

-dFILENAME

The device file name for access to the DPT Host

value must be trom O to 7, with a default value of 0.

Adapter.

For the shell script, the default is 0, fer /dsv
/dptHO which is the device file for the DPT

controller as a primary controller. Enter ! for
/dev/dptH1 which is the secondary controller.

Time

+TIME

Log File Name

n/a

If the sxecutable is invoked, this option does

not need to be specified for the default value of
0. If the secondary controller is required. either
-d/dev/dptH1 or -1 can be specified.

Interval of time to update status. This time ia
specified in minutes.

Nams of the log flle where errors are written. Used

by the monitorsh only. The user must manually

redirect the output to a file if the executable is

invoked.

12.1.3 Error Messages
The error messages returned by the RAID Monitor use the following format:
*

For general error messages:
MONITOR:

measage

For a change in LUN status:
MONITOR: Lun ) Status Changed Thu Oct 15 13:11:19 EDT 1992
MONITOR: Drives: (1,2) (3,4)
MONITOR: Previous status: Optimal
MONITOR: New status
: Degraded

12-2

RAID Monitor Utility

RAID Monitor Utliity
12.1 RAID Monitor
The LUN is identified in the first message with a timestamp of when the
change was displayed. The next line shows the drives associated with that
LUN referenced by channel and SCSI ID of the drive. In the above example,
2 drives are associated with LUN 0, a drive in channel 1 and at SCSI 1D 2,

and at channel 3, and SCSI ID 4. Finally a description of the previous status
and new status is displayed.

For a change in drive status:
MONITOR: Drive Channel 1 ID 2 Status Changed Thu Oct 15 13:11:19 EDT 1992
MONITOR: Previous status: Optimal

MONITOR: New status

: Degraded: Drive failure

The drive is identified by channel and SCSI ID with a timestamp of when the
change was dieplayed. This is followed by a description of the previous status,
and now the new status.

Generically, if an error occurs during a command to the controller which results
in either a DPT Host Adapter error, or a SCSI status other than good, that
information is presented in the following format:
¢

SCSI Check Condition Status (2 H)
SCSI Check Condition
SCSI Sense Key: 6 R ASC:

29 H

ASCQ: O H

In this example, the SCSI status is a Check Condition (2 H), the Sense Key
is 6 H, the Additional Sense Code (ASC) is 29 H, and the Additional Sense

Code Qualifier (ASCQ) is 0 H indicating a power on, reset, or bus device reset
occurred. Appendix B lists all the ASC's, ASCQ's and Sense Keys valid for the
array controller. Note that these values are given in hexadecimal notation.

Host Adapter error or SCSI Status (other than Check Condition)
SCSI Status: 0 H
Host Adapter Status:

7 B

In this example, the SCSI status is 0 H and the host adapter status is 7T H
indicating a bus parity error. Appendix A lists all the SCSI status codes for
the array controller. Appendix C lists all the DPT hcst adapter error codes.

The following list describ.s the general error messages retvrned from the utility.
Error Opening Controller Device File x Monitor process terminating—Results
from problems opening /dev/dptHO or /dev/dptH1. Verify that those device file
exist.

Error issuing inquiry command to device Monttor provess terminating—Array

controller was unable to respond to the inquiry command. Verify that the
array controller is booted and the drives are spun up.
Lun x Status Changed Not Configured—LUN originally was configured. but
no longer responds as configured.

Lun x Status Changed Configured—LUN originally was not configured, but
now responds as if configured.

RAID Monitor tlility

12-3

RAID Monitor Utility
12.1 RAID Monitor

Error issuing mode sense command to logical array page Monitor process
terminating—Unable to succesafully issue a mode sense command to the
array controller, and are terminating the monitor process.

Error issuing mode sense command to physical array page Monitor process
terminating-—Unable to successfully issue a mode sense command to the

array controller, and are terminating the monitor process.
lllegal switch - x Monitor process terminating—An illegal option was specified.

12-4

RAID Monitor Utility

13
RAID Status Utility

13.1 RAID Status Utility
The RAID status utility displays the current Logicel Unit (LUN) status and
device status.

13.1.1

Getting Started
The RAID status utility can be invoked in one of two ways: through the shell
script or by invoking the executable directly. The shell script metl.rd prompts
the user for options and invokes the monitor executabie. The executable requires
switch options to be specified, and is provided in the event that the user would
prefer to include in customized shell scripts.

13.1.1.1

Using the Shell Script

The RAID status utility can be invoked by typing raidstatsh at the system
prompt. The following banner appears:

RAID Status Utility Shell Script

Version XX.X

Enter values for the following paramsters or hit <CR> for
default value:

Enter SCSI ID [0}:

The format of the messages include the parameter to enter, followed by square

brackets containing the default value of that parameter. Press the Return key
or the Enter key to select the default value, or enter some other value for that
parameter.

A full help menu which describes the options is available by typing raidstatsh -h
at the system prompt. If help information for a specific parameter, enter & when
that parameter is prompted for, and a description is displayed.

Once all the parameters have been specified, the shell script invokes the raidstat
executable, displaying all status information to the screen.

RAID Status Utility

13-1

RAID Status Utility
13.1 RAID Status Utility

13.1.1.2 Invoking the Executabie

The RAID status utility can also be started by invoking the executable directly.
Enter raidstat [options] to start execution. The -h option prints out a help menu
with a description of the various switches.

13.1.2 Options
The options available to both the shell script and executable are outlined the

Table 13-1.
Table 13-1

RAID Status Utility Options

Options

Symbol

Description

SCSIID

<ID

The SCSI ID of the DEC RAID

SCSI1 LUN

-ILUN

The SCSI LUN. This value is not

Device File

-dFILENAME

The device file name for access to

Subystem. The value must be from
0 to 7, with a default value of 0.

used at this time and should be set
to the default value of 0.

the DPT Host Adapter.

For the shell acript, the
default is 0, for /dev/dptHO

which is the device file for the

DPT controller as a primary
controller. Enter 1 for /dev

/dptH1 which is the secondary

controller.

Invoking the exerutable,

this option does not need to
be specified for the default

value of 0. If the secondary

controller is required, either
-d/dev/{dptH] or -1 can be
specified.

13-2

RAID Status Utiiity

‘

RAID Status Utllity
13.1 RAID Status Utllity

13.1.3 Output

The status informaticn displayed by the RAID status utility is shown below.
RAID Status Utility

Version X.XX

LUN STATUS

Lun

RAID

Status

Drives

0
1

RAID §
RAID 1

Optimal
Optimal

(2,0)

(Chan,ID)

(3,0}
(1,0)

(4,0)
(5,0

Size

Type

3318041
1658880

Reg
Sub

DRIVE STATUS
ID

Channel

1
2
3

1
1
1

5
6
7

1
1
7

1
b
7

0-opt 1-no-exist 2-spare 3-fail 4-repl S-warn 6-mismtch 7-ctrl 8-fmt 9-wrong

The LUN status is displayed first. Each LUN is listed, followed by its RAID
number, and status. The drives associated with the LUN are listed next as (x,y)
where x is the channel number, and y is the SCSI ID of the drive. The size is
display as the number of blocks in tl.is LUN. The LUN type is either regular or
sublun.
The drive status shows the status of each drive referenced in the matrix by

channel and SCSI ID. The definition of the statuses is listed in the last line of the
display (See Section 3.3.2 for an explanation of the drive statuses).

13.1.4 Error Messages
The error messages returned by the RAID status utility use the following format:
*

For general errv- messages:
RAIDSTAT:

message

Generically, if an error occurs during a command to the controller which results
in either a DPT Host Adapter error, or a SCSI status other than good, that
information are presented in the following format:

SCSI Check Condition Status (2 H)
SCSI Check Condition

SCSI Sense Key:

6 H

ASC:

29 H

aSCQ:

0 H

In this example, the SCSI status is a Check Condition (2 H), the Sense Key
is 6 H, the Additional Sense Code (ASC) is 29 H, and the Additional Sense

Code Qualifier (ASCQ) is 0 H indicating a power on, reset or bus device reset
occurred. Appendix A lists all the ASC's, ASCQ’s and Sense Keys valid for the
array controller. Note that these values are given in hexadecimal notation.

RAID Status Utility

13-3

RAID Status Utility
13.1 RAID Status Utility
*

Host Adapter error or SCSI Status (other than Check Condition)
SC8I Statua:

0 H

Host Adapter Status:

7 H

[n this example, the SCSI status is 0 H and the Host Adapter Status ie 7 H
indicating a bus parity error. Appendix A lists all the SCSI status codes for
the array controller as well as the host adapter error codes.

The following list deacribes the general error messages returned from the utility.
Error Opening Controller Device File x—Results from problems opening /dev
/dptHO or /dev/dptH1. Verify that those device file exist.

Error issuing inquiry command to device~—Array controller was unable to
respond to the inquiry command. Verify that the array controller is booted
and the drives are spun up.
Lun x Status Changed Not Configured—LUN originally was configured, but
no longer responds as configured.

Lun x Status Changed Configured—LUN originally was not configured, but
now responds as if configured.
Error issuing mode sense command to logical array page—Unable to

successfully issue a mode sense command to the array controller.

Error issuing mode sense command to physical array page—Unable to
successfully issue a mode sense command to the array controller.

Ilegal switch - x—~An illegal option was specified.

13-4

RAID Status Utility

A
SCSI-2 Status Codes
This appendix lists the supported SCSI-2 status codes for the array controller.
Refer to Table A-1 below.
Table A-1

SCSI-2 Status Codes for the Array Controller

Status Description

Code

Good

00 H

Check Condition

02H

Busy

08 H

Reservation Conflict

18H

Queue Full

28 H

SCSI-2 Status Codes

A-1

B
SCSI Sense Keys and Additional Sense Codes
This appendix lists the SCSI sense keys, additional sense code (ASC), and
additional sense code qualifier (ASCQ) values that the controller returns in the
sense data. SCSI-2-defined codes are used when possible. Array-specific error
codes are used when necessary and are assigned SCSI-2 vendor-unique codes 80FFH.

Some errors that could be mapped into SCSI-2-defined ASC bytes are assigned
vendor-unique codes to provide more error information than the SCSI-2 defined
ASCQ bytes. In these cases, the SCSI-2-defined byte is returned in the ASC field
and the more detailed error information byte 1:: returned in the ASCQ field as
a vendor-unique qualifier of the standard ASC. These ASCQ bytes have values
80-FFH.

Table B~-1 below defines SCSI sense keys.
Table B-1

SCSI Sense Keys

Sense Key

Description

0x00

No Sense

0x01

Recovered Error

0x02

Not Ready

0x03

*“adium Error

0x04

Hardware Error

0x05

Illegal Request

0x06

Unit Attention

0x07

Data Protect (returned only if the drive returns this

sense key)

0x08

Blank Check (not used)

0x09

Vendor-Specific (not used)

0x0A

Copy Aborted (not used)

0x0B

Aborted Command

0x0C

Equal (not used)

0x0D

Volume Overflow (not used)

OxOE

Miscompare (not used)

0xOF

Rererved (not used)

SCSI Senge Keys and Additional Sense Codes

B-1

SCSI Sense Keys and Additional Sense Codes

Table B-2 defines the additional sense codes (ASC) and additional sense code
qualifiers (ASCQ) affiliated with a given sense key.

Table B-2

SCSI Error Codes

ASC

ASCQ

Sense Key

Description

No Additional Sense Information

I/O Proceas Terminated

No Index/Sector Signal

No Seek Complete

Peripheral Device Write Fault

Logical Unit Not Ready, Cause Not Reportable

1.2

Logical Unit Is In Process Of Becoming Ready

1,2

Logical Unit Not Ready, Initializing Command
hequired

Start Unit command required

1,2

1ogical Unit Not Ready, Manual Intervention

1,2

Logical Unit Not Ready, Format In Progress

Logical Unit Not Ready, ROM Installed Does Not

1,2

Logical Unit Does Not Respond To Selection

134

No Reference Position Found

Multiple Peripheral Devices Selected

Logicel Unit Communication Failure

Logical Unit Communication Timeout

Logical Unit Communication Parity Error

Track Following Error

1,6

Error Log Overilow

Write Error Recovered With Auto Reallocation

134

Write Error—Aute Feallocation Failed

1,34

ID CRC or ECC Error

1,34

Unrecovered Read Error

Required

Support Redundant Controller Configuration

The description of this error as being unrecovered
may be misleading because a drive-reported

Unrecovered Read Error can be recovered by the
array using parity. The sense key for this case ia
Recovered and the Recovery Actions field in the sense
indicates that parity was used.

1)1

134

Read Retries Exhausted

1,34

Error Too Long To Correct

1,34

Unrecovered Read Error—Auto Reallocate Failed

Miscorrected Error

{continued on next page)

B-2

SCSI Sense Keys and Additional Sense Codes

SCSI| Sense Keys and Additional Senae Codes

Table B-2 (Cont.) 8CSI Error Codes

ASC

ASCQ

Sense Key

Description

Unrecovered Read Error—Recommend Reassignment

gntr:covorod Read Error—Recommend Rewrite The
a

1,3

Address Mark Not Found For ID Field
Address Mark Not Found For Data Field

1,3

Recorded Entity Not Found

Record Not Found

134

Random Positioning Error

Mechanical Positioning Error

134

Positioning Error Detacted By Read Of Medium

134

Data Synchronization Mark Error

Recovered Data With No Error Correction Applied

Recovered Data With Ret ies

Recovered Data With Positive Head Offset

Recovered Data With Negative Head Offset

Recovered Data Using Previous Sector ID

Recovered Data Without ECC—Data Auto-Reallocated

Recovered Data Without ECC—Recommend

Recovered Data With Error Correction Applied

Reapsignment

Recovered Data With Error Correction And Retries
Applied

Recovered Data—Data Auto-Reallocated

Recovered Data—Recommend Reasgignment

Defect List Error

Defect List Not Available

1,3

Defect List Error In Primary List

Defect List Error In Grown List

Parameter List Length Error

00?

Synchronous Data Transfer Error

0,13

Defect List Not Found

0,13

Primary Defect List Not Found

0,1,3

Grown Defect List Not Found

Miscompare During Verify Operation

Recovered ID With ECC Correction

Invalid Command Operation Code

Logical Block Address Out Of Range

1,6

Illegal Function

2The FRU code in byte 14 of the serse data indicates where the error was detected.
(continued on next page)

SCSt Sense Keys and Additional Sense Codes

B-3

SUSI Sense Keys and Additional Sense Codes

Table B-2 (Cont.)

SCSI Error Codes

ASC

ASCQ

Sense Key

Description

Invalid Field In CDB

1,6

Logical Unit Not Supported

Invalid Field In Parameter List

Parameter Not Supported

Parameter Value Invalid

Threshold Parameters Not Supported

1,7

Write Protected

00?

Power On, Reset, Or Bus Device Reset Occurred

Parameters Changed

Mode Parameters Changed

Log Parameters Changed

003

Command Sequence Error

Commands Cleared By Another Initiator

Incompatible Medium Installed

1,3

Cannot Read Medium - Unknown Format

02-

Cannot Read Medium - Incompatible Format

Medium Format Corrupted

Format Command Failed

Format Command Cannot Execute Because Of

No Defect Spare Location Available

Defect List Update Failure

Rounded Parameter

Not Ready To Ready Transition (Medium May Have

Changed)

Multiple LUN Assignments Per Physical Drive

Saving Parameters Not Supported

1,2

Medium Not Present

003

1,6

Invalid Bits In Identify Message

1,2

Logical Unit Has Not Self-Configured Yet

Target Operating Conditions Have Changed

Microcode Has Been Changed

Returned when new drive microcode has been
downloaded.
3F

Changed Operating Definition

]

Inquiry Data Has Changed

The FRU code in byte 14 of the sense data indicates where the error was detected.
(continued on next page)

B-4

SCSI Sense Keys and Additiona! Sense Codes

SCSI| Sense Keys and Additional Sense Codes

Table B-2 (Cont.) SCSI Error Codes
ASC

ASCQ

Sense Key

©)"

Description

Drive Failed Because Of
A Failed Write Operation
This is the result of an error that prevents access
to user data on this drive or that may result in

erroneous data being read/written (for example, an
Unrecovered Write Error).

81®

Drive Failed- Automatic Reallocation Failed

Either reassign block or recovery of data on the
reassigned sector failed.
3F

82?

Drive Failed—Reconstruction Failed

{Error On Drive Being Reconstructed)
The reconstruction operation can not complete
because of an unrecoverable Write error or a failed
drive format (before reconstruction starts) on the
drive being reconatructed. It must be replaced (again)
before the reconstruction can be retried.

838

Drive Failed—Reconstruction Failed

(Error On Drive Required For Reconstruction)
The reconstruction operation can not complete
because of an unrecoverable Read error on one of

the drives needed for reconstruction. You can still
run the array in degraded mode, but if user data is
stored at the address in error, you can not read it.
You should attempt a back up of the array to recover
the remaining data. To bring the LUN to an optimal
state, replace the failed drive, format the array. and
restore the data from a back-up copy (the one just
made or a previous successful back-up copy).
3F

848

(6¢

Drive Failed Due To A Hardware Component
Diagnostics Failure

85°

Drive Failed Because It Failed A Test Unit Ready
Command (during start-of-day) Or Read Capacity
Command (during start-of-day or during a format or
reconstruction operation)

(6»

Drive Failed Because It Failed A Format Unit

(6)!

Drive Failed By A Host Mode Select Command

6)!

Drive Failed Because Of Deferred Error Reported By
The Drive

Drive Failed By Start-of-Day Application Code

Unrecovered Read/Write Error

3F
3F

3F
3F
3F

89®

Command

Because Of A Drive Replacement Ervor

Unrecovered Read errors always generate this
condition.

“The Unit Attention sense key in parentheses (6) indicates that this code depends on the Sense Key

for Drive Failures field in the EE

OM data.

a'l'h.elt! conditions (3F 8x) require that the drive be replaced as soon ar possible. Urer data is still

available by using parity, but a second failure at this time (when a drive has been downed) resuite in
the loas of user data for the entire LUN (RAID 5) or results in an unrecoverable error (other RAID
levels).

(continued on next page)

SCSI Sense Keys and Additional Sense Codes

B-5

SCSI Sense Keys and Additional Sense Codes

Table B-2 (Cont.)

SCSI Error Codes

ASC

Sense Key

ASCQ

Description
Unrecovered Write errors cause this conditic. n if the
state of the LUN or the RAID level guarantees that
data at the other addresses will not be affected by the
failed operation and continued access to this drive.

A second drive failure or warning conditien on

another drive can prevent reconstruction of this
drive.

Drive Reported Deferred Error Caused Drive To Be

Placed In Warning

Single Drive Array Assurance Error
* X = 0—Drive ECC Test Failed
* X = Other—Not Currently Implemented

BO*

Excessive Media Error Rate

B1*

Excessive Seek Error Rate

B2¢

Excessive Grown Defects

co®

No Response From One Or More Drives

c1®

()Y

Communication Errors

c2*

Firmware Indicates No Drive Is Present Although

Cc7

(CIN

Subsystem Component Failure

AC Power Is Lost, DC Power Is Being Supplied By A

AC Power Is Lost, DC Power Supplied For A

Information Stored On Disk Indicates Drive Should
Be Present

FRU code and qualifiers identify the failed component.
Hold-Over Battery

Maximum Of Two Minutes

The host should start its shutdown procedure.

AC Power Is Lost, DC Power Is Exhausted
The controller will complete currently executing drive
commands to maintain data integrity.

6)!

AC Power Was Lost, But Is Now Restored

Do®

>756% Of Transfer Delays Caused By One Drive

D1®

(6)*

D2¢®

Synchronous Transfer Value Differences Between

D3*®

Software Measured Performance Degradation

>756% Of Non-Aligned Reselections Caused By One

Drive

Drives

!The Unit Attention sense key in parentheses (8) indicates that this code depends on the Sense Key

for Drive Failures field in the EEPROM data.

These errors are warnings to allow early drive replacement (drive still being

of user data—3F 80).

used until a failed Write

;}hsl;o ervors sre warnings to check installation (drive still being used until a failed Write operation—
).

%These errors are indications of potential performance degradation {drive still being used).

(continued on next page)

B-6

SCSI Sense Keys and Additional Sense Codes

Taeble B-2 (Cont.)

SCSI Error Codes

ASC

ASCa

Sense Key

D4®

EO’

6»

LUN Downed

El’

Multiple Drives Have Been Downed

E2’

Mode Parameters For Drives In LUN Don't Match

E3’

Drive Channel Verification Failed

E4’

SCSI ID Verification Failed

Es’

Wrong Drive Was Replaced

3F
k)

E¢’
Fo®

4
(6

Component Failure Affecting Multiple Channels
EEPROM Error

F1®

EEPROM Hard Checksum Error

F2'

Maximum EEPROM Write Count Exceeded

()N

Application Software Copy To More Than One Drive

Description

Mode Parameter Differences Between Drives May

Cause Performance Degradation

Failed.

Another download is required if the drive with the
current application code failed.

RAM Fajlure

Diagnostic Failure On Component NN (80H-FFH)
In this release of the software, this ASC and ASCQ
are only returned if reported by a drive. FRU codes
will distinguish between failures on components of
the array controller and components of the drive
controllers.

00?

Data Path Failure, Carryover From CCS (same as 40

NN)

Power-On Or Self-Test Failure, Carryover From CCS
(same as 40 NN)

00?

Message Error

Internal Target Failure

00?

Select/Reselect Failure

00?

Unsuccessful Soft Reset

Initiator Detected Error Message Received

00?

Invalid Message Error

00?

1,B

SCSI Parity Error

"The Unit Attention sense key in parentheses (6) indicates that this code depends on the Sense Key

for Drive Failures field in the EE}fi'\’.OM data.

3The FRU code in hyte 14 of the sense data indicates where the error was detected.

SThese errors are indications of potential performance degradation (drive still being used).
7

Slnnot

ent media access commands to this LUN do not execute and return an ASC of Command

xecute Because The LUN Has Reen Downed with an ASCQ that indicates why the LUN is

own.

The configuration change (drive being marked failed/warning) could not he saved to EFPROM, hut

the update of the RAM information on the hoard and the DACSTORE information on the disk were

msucceasful.

new configuration information is used until a board

power-up/reset. At that time, s

diagnostic failure is detected that prevents the board from being uaeg.

(continued on next page)

SCSI Sense Keys and Additional Sense Codes

B-7

SCSI Sense Keys and Additional Sense Codes

Table B-2 (Cont.)

SCSI Error Codes

ASC

AScQ

Sense Key

Description

00?

Command Phase Error

00?

Data Phase Error

Data Overrun/Underrun

Logical Unit Failed Self-Configuration

1,B

Overlapped Commands Attempted

Operator Request Or State Change Input
(Unspecified)

1,7

Operator Selected Write Protect

Operator Selected Write Permit

Log Exception

1,6

Threshold Condition Met

Log Counter At Maximum

Log List Codes Exhausted

186

RPL Status Change

B8C

Spindles Synchronized

1,6

Spindles Not Synchronized

Error Manager Detected Error

The Error Manager Was Invoked Without Any EM_

02?

gh: Error Manager Was Passed An Out-Of-Range

Code(s) Loaded

ode

03?

The Error Manager Was Passed A Code By The SCSI

Fatal Null Pointer

No AEN Code Or An Invalid AEN Code Was Loaded

Driver, But No Error Handler Exists

By The Application Software

This error is detected when the AEN condition is
requested by the command handler (either when the
next command for this host/LUN is received or when
a polled Request Sense is received), not when the
AEN condition was detected. The illegal AEN code is

loaded into the Error-Specific Information field in the
sense data.

Mazximum # Of Errors For This I/O Exceeded

Reconstruction Setup Failed

Out Of Heap

No Command Control Structures Available

No DAC Application Control Blocks Available

Drive Reported Recovered Error Without Transferring

All Of Data

The Unit Attention sense key in parentheses (6) indicates that this code depends on the Sense Key

for Drive Failures field in the EEPROM data.

3The FRU code in byte 14 of the sense data indicates where the error was detected.
{continued on next page)

B-8

SCSI| Senas Keys and Additional Sense Codes

SCSI Sense Keys and Additional Sense Codes

Table B-2 (Cont.) SCSI Error Codes
ASC

ASCQ

Sense Key

Description

Reservation Conflict

Command Cannot Execute Because The LUN Has

01%*

Multiple Drives Have Been Downed

84
84
84

023*
03°
o4’

4
4
4

Mode Parameters For Drives in LUN Do Not Match
Drive Channel Verification Failed
SCSI ID Verification Failed

os*

Format In Progress

84
84

0s6®
og’

4
4

Awaiting Format Command
Wrong Drive Was Replaced

09°

Component Failure Affecting Multiple Channels

General Application Code Command Handler Error

Drive Error

021°

Host Error

03!

Drive Type Mismatch Within LUN

86
85

04'°
o5

4
4

Operation Not Allowed During Reconstruction
Data Returned By Drive Is Invalid

os'

Non-Failed Drive Unavailable For Operations

Insufficient Rank Structures Available

Full Format Required, But Not Allowed (Sub LUNs)

09'

856

0AY

Data Recovery After Re-Assign Block Command

Drive Not Returning Required Mode Sense Page(s)

Been Downed

Drive Cannot Be Mode Selected To Meet LUN

Parameters

Failed

Drives must support Mode Sense pages 3 and 4.

Command Cannot Execute Because The LUN Is In

Code Download/Upload Error

Partial Download (Missing Application Code Segment)

(6)

Downloaded Code Cannot Be Saved To Disk

Degraded Mode

Old application code will be uploaded on next powerup/reset.

Code CRC Failure
Check is performed on the disk read after the new
code is saved to disk.

IThe Unit Atteation sense key in parentheses (6) indicates that this code depends on the Sense Key

for Drive Failures field in the EEPROM data.

IThe FRU code in byte 14 of the sense data indicates where the error was detected.
*These errors (84 XX) are returned on every media access command to the downed LUN until the
error is corrected.

“The original CDB field in the array sense data shows the failing command. Debug information is

shown in the Error Detection Point sense field.

(continued on next page)

SCS! Sense Keys and Additional Sense Codes

B-9

SCSI Sense Keys and Additional Sense Codes

Tabie B-2 (Cont.)

SCSI Error Codes

ASC

ASCQ

Sense Key

Description

Upload Of Latest Version Of Code Failed
An older version of code has been uploaded
successfully.

No Package Verification Partition Downloaded

Incomplete RAM Partitions

Incompatible Board Type For The Code Downloaded

Incompatible ROM Version For Support Of The
Downloaded Code

Download Of Microcode To A Failed Disk Completed
Successfully

ROM Partitions Required For Download Of Code

Missing

The drive is still unusable until it is marked as
replaced and has been reconstructed.

EEPROM Command Error

EEPROM Not Responding
(There is an EEPROM on the board.)

EEPROM Not "Formatted” (certain key fields have

Invalid EEPROM Offuet (2 Write to the write-

EEPROM Hard Checksum Error (indicates a

Maximum Write Count Exceeded

EEPROM Not Initialized

Error On Request Sense Command To A Drive

Illegal Command For Pass-Through Mode

not been set.)

protected maintenance area causes this error.)

EEPROM Soft Checksum Error (indicates that an

update was interrupted)

component may be going bad)

The original CDB field in the array sense data shows
the failing command.

Illega] Command For Current RAID Level

Write Buffer Command (For Drive Microcode
Download) Attempted But The Enable Bit In The
EEPROM Was Not On

Destination Transfer State Machine Error

Invalid Transfer Release Requester

Invelid Transfer Requester

Data Stripe/Parity Generation ASIC Configuration

Write Buffer Command (For Code Download) Was

Attempted While Another Command Was Active

Error

!The Unit Attention sense keyin arentheses (6) indicates that this code depende on the Sense Key

for Drive Failures fieldin the EE

(continued on next page!

B-10

SCSI Sense Keys and Additional Sense Codes

Table B-2 (Cont.) SCSI Error Codes
ASC

ASCQ

Sense Key

Description

Data Transfer Request Error

Invalid Transfer Pad Requester

Destination Driver Data Transfer Did Not Complete

Data Stripe/Parity Generation ASIC Error
Parity/Data Mismatch
Data Underrun

Premature Completion Of A Drive Command

(expected Data Transfer and received Good Status

instead)

DACSTORE Errors
DACSTORE Error Bit Definitions

Bits 0—3 = DACSTORE Identifier
* Value of 0 = DACSTORE directory
* Value of 1 = Disk store

* Value of 2 = LUN store
* Value of 3 = Controller store

* Value of 4 = Log store
* Value of 6 = High ID controller serial # store

* Value of 6 = Low ID controller serial # store
* Value of 7 = Redundant controller common store

* Value of 8 = Boot block EEPROM store

Bits 4—6 = Error Type
* Value of 0 = Setup error
* Value of 1 = Invalid directory data
* Value of 2 = Drive error

* Value of 3 = Invalid store data

Bit 7 = Operation Type
* Value of 0 = Read

* Value of 1 = Write

Mode Select Errore

LUN Already Exists; Cannot Do "Add LUN" Function

LUN Does Not Exist; Cannot Do "Replace LUN"

Drive Already Exists; Cannot Do "Add Drive"

Drive Does Not Exist; Cannot Do Requested Action

Drive Can't Be Deleted; It Is Part Of A LUN

Drive Can't Be Failed; It Is Formatting

llm"

Function "83" Or Any Logical Function
Function "80"

For It

(continued on next page)

SCSI Sense Keys and Additional Sense Codes

B-11

SCSI Sense Keys and Additional Sense Codes

Table B-2 (Cont.)

SCSI Error Codes

ASC

ASCa

Sense Key

Description

Drive Can't Be Replaced; It Is Not Marked As Failed
Or Replaced

Invalid Action To Take

Invalid Action With Multiple Sub LUNs Defined
'

~1bably an attempt to change page 3—Format

Lrevice page)

Invalid Reconstruction Amount

4):]

Invalid Reconstruction Frequency

0oC

Invalid LUN Block Size

Invalid LUN Type

Invalid Segment Size

Invalid Segment 0 Size

]

Invalid Number Of Drives In LUN

Invalid Number Of LUN Blocks

Invalid RAID Level

Invalid Drive Sector Size

121

]

Invalid LUN Block Size/Drive Sector Size Modulo

No Disks Defined For LUN

Insufficient Rank Structures Available To Define LUN

Disk Defined Multiple Times For LUN

Sub LUN Drives Not The Same As Those Used By
Other Sub LUNs On These Drives

Sub LUN RAID Level Mismatch

First Sub LUN Defined For These Drives Has Not Yet

Non-Sub LUN Drive Already Owned By Another LUN

Sub LUN Drive Already Owned By a Non-Sub LUN

Drive Type Does Not Match The Drive Type Of The

]

I?F‘x;ilro Cannot Be Included In Rank Because Rank Is

Ranks Have Different Number Of Disks Defined

Muitiple Disks On Same Channel Within Same Rank

]

Mirrored Disks On The Same Channel

No Parity Disk Defined

]

No Data Disks Defined

Too Many Disks Defined

No Space Available For LUN—Sub LUN Cannot Be

Drive Status Can Not Be Changed To Good (drive can
not be revived through Mode Select)

Been Formatted; Second Sub LUN Is Illegal

Other Drives In The LUN

Defined

(continued on next page)

B-12

SCSI Sense Keys and Additional Sense Codes

Table B-2 (Cont.) SCSIi Error Codes
ASC

ASCQ

Senss Key

Description

]

Error In Processing A Subsystem Mode Page

Drive Inquiry Data Mismatch Between Drives In The

Drive Capacity Mismatch Between Drives In The

Drive Block Size Mismatch Between Drives In The

LUN
LUN

LUN

Support Of TTD/CIOP Messages Is Not The Same For

All Drives In The LUN

Firmware Does Not Support Redundant Controller
Options Selected

BUSY Status From Drives Could Not Be Cleared By

Array Controller

Drive Vendor Unique Sense Data Returned
See the drive sense area of the array sense for the
drive ASC/ASCQ and sense key. Also refer to the
drive documentation for further description.

Invalid Request Of A Controller In Redundant

1,B

A Drive Channel Was Reset

Controller Mode.

Probable cause is the removal or replacement of a
drive during a hot swap operation

An Extended Drive Channel Reset Has Been Detected
Probable cause is a drive left partially removed or
inserted

Redundant Controller Not Supported By Current

Alternate Controller Not Supported By Current

Firmware

Firmware (however, the alternate controller has been

detacted)

Command Timeout

Watchdog Timer Timeout

Software Loop Timeout

o0?

SCSI Driver Timeout

o1?

Disconnect Timecut

02?

Chip Command Timeout

03?

Byte Transfer Timeout

00?

Bus Ervors

o1?

CDB Transfer Incomplete

02?

Unexpected Bus Phase

03?

Disconnect Expected

04?2

ID Message Not Sent

iThe FRU code in byte 14 of the sense data indicates where the error was detected.
(continued on next page!

SCSI Sense Keys and Additional Sense Codes

B-13

SCSI Sense Keys and Additional Sense Codes

Table B-2 (Cont.)

SCSI Error Codes

ASC

ASCQ

Sense Key

Description

052

Synchronous Negotiation Error

0e*

Target Tranefer Disable (TTD) Negotiation Conflict

07?

Unexpected Disconnect

08?

Unexpected Message

09?

Unexpected Tag Message

0A?

Channel Busy

00*

Miscellaneous SCSI Driver Error

012

Ilega! C96 Chip Command

022

Uncoded Execution Path

00?

Drive SCSI Chip Reported Gross Error

00?

Non-SCSI Bus Parity Error

00?

Miscellaneous Host-Related Errors

012

Maximum Messages Received

022

Message Reject Received on a Valid Message

Deé

00?

Source Driver Chip-Related Error

00?

Source Driver Programming Error

An Error Was Encountered That Required The Data
Pointers To Be Restored But The Host Is NonDisconnecting And Does Not Support The Restore
Pointers Meseage (indicated by an EEPROM option
centrol bit or by host selection without sending the
identify message)

2The FRU code in byts 14 of the sense data indicates where the error was detected.

B-14 SCSI Sense Keys and Additional Sense Codes

C
DPT Host Adapter Error Codes
This appendix lists the error codes for the DPT Host Adapter. Controller errors
have priority over SCSI errors. Refer to Table C—-1 below.
Table C-1

SCSI Status Codes for the DPT Host Adapter

Error Description

Code

No error

00 H

Selection timeout

01H

Command timeout

02 H

SCSI bus reset

03 H

Initial controller power-up

04 H

Unezpected bus phase

05 H

Unexpected bus free

06 H

Bus parity error

07TH

SCSI bus hung

08 H

Unexpected message reject

09 H

Resst stuck

OAH

Auto request sense failed

0BH

Controller RAM parity error

OCH

Abort message processed

ODH

Reset message processes

OEH

DPT Host Adapter Error Codes

C-1

Glossary
ac distribution

The method of controlling ac power in a cabinet.
adapter

(1.)

A connecting device that permits the attachment of accessories or provides

(2.)

The device that connects an 8-bit differential SCSI bus to an 8-bit

the capability to mount or link units.
single-ended SCSI bus.

array

An array is a set of multiple disk drives and a specialized controller, an array
controller, which keeps track of how the data is distributed across the drives.
array channels

Array channels are the SCSI-2 compliant buses on which the disk drives are
located. Each array channel is independent SCSI bus.
array controiler

A device that exercises control over the SCSI bus (for example, an HSZ10-AX
disk array controller).
BA35X~-VA

A collective reference to all versions of the vertical mounting kits—single and
double.
Ci
A Digital trademark for the Digital Computer Interconnect bus.
cold-swapping

A method of device replacement that requires that power be removed from all
shelves in a cabinet. This method is used when conditions preclude the use of a
warm-swapping or hot-swapping method.

See also warm-swapping and hot-swapping.
controlier

A hardware line device that manages communications over a line. Controllers

can be point-to-point, multipoint, or multiple line controllers.
dc power system

The method for providing dc power in a cabinet.

Glossary-1

DECstor/me
The mnemonic for the Digital Storage/Modular Enclosure, a modular set of
enclosure products that allows customers to design their own storage array.
Components include power, packaging, and interconnections in a modular

storage shelf into which SBBs and array controller modules are integrated to
form modular storage arrays. System-level enclosures to house the arrays and
standard mounting devices for SBBs are also included.
double stand

A BA35X-VA vertical mounting kit composed of two single stands clipped
together. This configuration can support one BA360-EA shelf.
See also single stand.
drive group
A drive group is a set of from 1 to 10 drives that have been configured into one or
more logical units. A lcgical unit can be contained in only one drive group, and
all the logical units in a drive group must have the same RAID level and be of
the same drive type.

drive rank

Drive ranks represent a numbering scheme providing information on the
maximum number of drives on every array channel. A one rank system
indicates that there is a maximum of une drive per disk channel. A two rank

array indicates that there is a maximum of two drives per disk array channel.
However, any channel can have zero for its maximum number.
DSS!
Digital Storage System Interconnect.
FD SCsl

The fast, differential SCSI bus with an 8-bit data tranefer rate of 10 Mb/s,
See also FWD SCSi and SCSI.
FWD SCSI

The fast, wide, differential SCSI bus with a 16-bit data transfer rate of 20 Mb/s.
See also FD S£CSI and SCSI.
H981x

A collective reference to the H9810 (short), H9811 (medium), and H9812 (tall)
towers.

Heartbeat LED

The bottom most LED on the HSZ10-AX controller. Beats once per second.
host

The primary or controlling computer in a multiple computer network.

Gloasary-2

hot-awapping
A method of device replacement whereby the complete system remaine online and
active during device removal or insertion. The device being removed or inserted
is the only device that cannot perform operations during this process

See also cold-swapping and warm-swapping.
LUNSs (logical units)

A logical unit is a grouping of drives which has its own device SCSI ID and
number. Each logical unit has its own array parameters (RAID level, segment
size, and so on). For most purposes, a logical unit is equivalent to an array.
mirrored

A copy of data on a disk or a set of disks. Refer to the description of RAID 1.
parity check/repair

The process of verifying and repairing parity information so that data can bhe
maintained and reconstructed in the event of a drive failure. Parity Check/Repair
functionality is provided by the DEC RAID utilities.
RAID

A redundant array of inexpensive disks.
rank
A set of drives (1 per channel). See also drive rank.

redundancy
Also data redundancy. Data stored on another physical disk that can be used to
recover data if the physical disk containing the data cannot be accessed.
SBB
System building block. A modular carrier plus the individual mechanical and
electromechanical interface required to mount it into a standard shelf. Any
device conforming to shelf mechanical and electrical standards is considered an
SBB.
SCs!

Small Computer System Interface. This interface defines the physical
and electrical parameters of a parallel 'O bus used to connect computers
and a maximum of seven SBBs. The DECstor/me modular storage system
implementation uses SCSI-2, which permits the synchronous transfer of 8-bit
data at rates of up to 10 MB/s.
segment

A group of blocks that is continuous data which can be stored on a disk drive.
shelf array

A modular storage shelf that provides power, cooling, interconnects, and
mounting for SBBs. Specific shelves are denoted by the prefix BA350 (that is,
BA350-RA, BA350-SA, and 80 on). Shelves may be mounted in kits, towers, or

cabinets.

Glossary-3

single stand

A reference to the basic BA36X-VA vertical mounting kit with a capacity of one
BA350-8A shelf.

See also double stand.
Small Computer System intertace

See SCSI.
stands

A collective reference to all versions of the vertical mounting kits-~both single
and double.
statlc storagie device (SSD)
An electronic storage device such as the EZ51R-VA.
striped

See the description of RAID 0.

system bullding block

See SBB.
towers

A collective reference to the H9810 (short), HS81. (medium), and H9812 (tall)
towers.

warm-swapping

A method of device replacement whereby the complete system remains online
during device removal or insertion. The system bus may be halted for a brief
period of time, during device insertion or removal. No booting or loading of code
ic permitted except on the device being inserted.

See also cold-swapping and hot-swapping.

Glossary-4

Index

ACE logical units
table of status, 3-7

DEC RAID Manager

Additional se:se code (ASC),

device names,

B-1, B-2

Additional sense code qualifier (ASCQ), B-1, B-2
AMD error messages, 9-4, 9-6
console, 9-5
log message,

9-5

3-1

using.

3-3 to 34

window, 3-2

Array Controller Configuration Window,

features,

2-6
figure, 2-6
Array Monitor Daemon, 9-1
9-1

2-1

installation,
2-6

10-2

10-1

installing the RAID Monitor utility,
installing the RAID status utility,

invoking RAID Monitor,

10-2
10-2

11-2, 12-1

invoking RAJID status with the executable,
13-2
invoking the RAID Monitor executable,

9-1

invoking the RAID status utility,

how it works, 9-1

checking/repairing,

4-2 to 4-3

2-2

DEC RAID SCO UNIX Utility
install the parity check/repair utility,

Array controller
SCSI-2 status codes, A-1

Array parity

6-6 to 6-7

logical unit configuration tasks,
starting,

Array performance,

7-1

2-1

6-1
installation error messages,

3-3

error messages,

10-2

2-8

installation,

quitting, 3-4

description,

parity check/repair,
table of tasks,

description,

description, 3-1
drive status, 3-6

starting,

6-1

9-11 to 9-12

adding logical units,

Array Configuration Editor (ACE)

table of keys,

messages,

DEC RAID NetWare Utility

pop-up, 94

navigating in,

2-11

files for NetWare,

parity check/repair,

8-1, 11-1

7-1

12-2

13-1

11-1

RAID Monitor, 12-1
RAID Monitor error messages,
RAID Monitor options, 12-2

12-2

RAID status functions,

B
Blocks
per delay interval, 48

13-1
RAID status utility, 13-1
RAID status utility error messages.
RAID status utility options, 13-2
shel! seript,

11-1

starting parity checl/repair.

using the shell script,

Install for NetWare,
Configuring

logical units, 4-1

7-1

description,

11-1

13-1

DEC RAID Stand-Alone Utility,

Command

1-1

2-1

device names,

2-11

logical unit configuration tasks,
starting,

using,

13-3

4-1 to 4-2

2-2

2-1

DEC RAID Utilities
description, 1-1, 2-1
device names, 2-11

index-1

DEC RAID Utilities (cont'd)

kita,

1-2

menu functions,
moving around,

MS-DOS kit,
NetWare,

2-4
2-3

Error messages
AMD format, 9-4
array component errors,

1-2
1-3

6-~6 to 6-7
for the parity check/repair utility,

1-1

NetWare kit,
road map,

2-8

1-2
SCO UNIX kit, 1-3
Stand-Alone, 1-1
Stand-Alone kit,

1-2

types, 9-4

2-2

table of tasks, 2-8
using menus,

2-3

Delay interval, 4-7
Device names, 2-11
for NetWare, 2-11
for the DEC RAID Stand-Alone Utility, 2-11
Device numbers
table of, 2-11
Disk Array Services Window, 2-4
features, 2-4
figure, 24

H
Hardwars
requirements,

Installation
before you install,

status,

6-3

DEC RAID NetWare Utility, 6-1
DEC RAID SCO UNIX Utility, 10-1
error measages for NetWare, 5-6 to 6-7
running installation program, 6-3
using an MS-DOS partition, 6-3
without an MS-DOS partition, 64

deleting, 4-18
4-18

reconstructing,

1-1

Install command
for NetWare, 7-1

DPT host adapter error codes, C-1
Drive
adding, 4-17
failing,

4-19

2-9

when to replace, 5-3
Drive failures

multiple drive failures - RAID 1, 5-5
multiple drive failures - RAID 5, 65-6
on RAID 0,

5-6

on RAID 1, 5-5
on RAID 5, 6-6

single drive failure, 5-5
single drive failure - RAID 0, 5-5
single drive failure - RAID 5, 5-6
Drive Map, 4-5
changing, 4-8
Drive Matrix, 3-4
Drive status
modifying, 4-17
table of, 2-10
Drive Window, 3-4
options,

3-5

E
Error codes, B-2
DPT host adapter, C-1
Error log
example,

K
Kits, 1-2
Ms-DOS, 1-2
NetWare, 1-3
SCO UNLX, 1-3
Stand-Alone, 1-2

L
Logical block size,

adding to NetWare, 7-1
configuration summary, 4-1
configuring, 4-1
creating from an existing drive,
creating from spare drives,

4-13

4-11

deleting, 4-16
display, 3-6
displaying parameters,

4--8

displaying parameters procedure, 4-9 to 4--10
modifying parameters, 4-8

modifying parameters procedure, 4-9 to 4-10
parameters,

9-2

4-8

Logical unit
adding to existing volumes, 7-1

parametere, table of, 4-4

procedure for restoring.

index-2

11-3

for the RAID Monitor, 12-2
for the RAID etatus utility. 13-3
from the Array Monitor Daemon, 9-1
status change errors, 9-10

SCO UNIX,

starting,

9-6

array device errors, 9-8
for DEC RAID NetWare Utility installation,

5-1 to 5-2

Parity check/repair (cont'd)
running the check. 8-3 to -4
SCO UNIX installation, 10-2

Logical unit (cont'd)
vestoring,

6-1

restoring RAID 0, 5-8
restoring RAID 1, 5-10, 6-11
restoring RAID 6, 6-13, 6-16

starting, 11-1
when to run, 8-1

Performance notes, 7-1

size,

4-8
status, 2-9, 3-7

atatus, table of, 2-9, 5-3 to 6-4
Logical Unit Window, 3-8
options,

RAID level

3-8

options list,

changing, 4-15
description, 4=4

3-8 to 3-9

RAID Monitor
Menu
Array Controller Configuration,

error messages, 12-2
functions, 12-1
invoking, 12-1

2.8

invoking the executable,

Menus
Disk Array Services, 2-4

options, 12-2
using the shell script,

Messages

for the DEC RAID Manager, 9-11 to 8-12
MS-DOS, 1-1
inatallation with a partition,

6-3

installation without a partition,

kit,

6-4

RAID Monitor utility,

RAID status utility, 13-1
error messages, 13-3
functions, 13-1

installation, 10-2
invoking, 13-1
invoking with the executable, 13-2
invoking with the shell script, 13-1

N
NetWare, 1-1
adding logical units, 7-1
deacription, 2-1
device names, 2-11
files for DEC RAID Manager,
installation, 6--1
installation error messages,

options,

6-1
6-6 to 6-7

kit, 1-3
logical unit configuration tasks, 4-2 to 4-3
starting,

12-1

installation, 10-2
invoking, 11-2

1-2

using,

12-2

2-2

2-1

using the Install command,

7-1

logical units, 5-1
RAID
0 logical unit,
RAID 1 logical unit,

Parameters
for logical units, 4—4

RMINST.NLM,

8-1, 11-1

Parity check/repair
automatic operation,

8-2
automatic schedule changes,

changing the schedule,
8-1

error messages,

11-3

functions, 8-1, 11-1

manual, 8-3
manual operation, 8-3
options,

11-2

output,

11-3

6-1

automatic, 8-2

description,

5-8
5-10, 5-11

RAID
5 logical unit, 5-13, 5-15

Parity
array parity,

13-2

Reconstruction
description, 5-7
rate, 5-7
rate, table of, 5-8
Replacing
drives, 5-3
Requirements
hardware, 1-1
software, 1-1
Restoring

8-5

SCO UNIX,

1-2

install the parity check/repair utility,
installation, 10-1
installing the RAID Monitor, 10-2

10-2

invoking RAID Monitor,

12-1
invoking RAID status with the executable,
13-2

invoking the RAID Monitor executable,
invoking the RAID status utility,
kit, 1-3

12-2

13-1

Index-3

SCO UNIX (cont’d)
parity check/repair,

description,

11-1

RAID Monitor, 12-1
RAID Monitor error messages,
RAID Monitor options, 12-2
RAID status functions,

kit,

12-2

11-1

starting parity check/repair,
using the shell script, 13-1
SCSI error codes, B-2

11-1

SCSI sense keys, B-1
SCS] status codes
for the DPT host adapter, C~1
3CSI-2 status codes
for the array controller, A-1

starting,

13-3

segment zero, 4-7

11-1, 12-1, 13-1

2-1

Startup,

2-2

for the NetWare Utility,

2-2
for the Stand-Alone Utility, 2-2

Status

of ACE Logical Units, 3-7
of drives, 2-9
of logical units, 2-9
T

Tasks
table of, 2-8

index-4

\'4
Volumes

requirements,

Stand-Alone,

2-2

using,

Segment size, 4-7

Shell script,
Software

2-11

1-2

logical unit configuration tasks, 4-1 to 4-2

13-1

RAID status atility, 10-2, 13-1
RAID status utility error messages,
RAID status utility options, 13-2
shell script,

2-1

device names,

1-1

adding logical units,

7-1