Digital PDFs

AA-ME92B-TE

June 1990

73 pages

Original

3.2MB

Document:	ULTRIX Guide to Backup and Restore
Order Number:	AA-ME92B-TE
Revision:	000
Pages:	73
Original Filename:

OCR Text

ULTRIX

Guide to Backup and Restore

Order Number: AA-ME92B-TE
June 1990

Product Version:

ULTRIX Version 4.0 or higher

This manual provides information on how to develop a backup and restore
strategy. Included in this manual is information using backup and restore
commands and utilities.

digital equipment corporation
maynard, massachusetts

Restricted Rights: Use, duplication, or disclosure by the U.S. Government is subject to restrictions as set forth in
subparagraph (c) (1) (ii) of the Rights in Technical Data and Computer Software clause of DFARS 252.227-7013.

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

The following are trademarks of Digital Equipment Corporation:

IJDmaalma
CDA
DDIF
DDIS
DEC
DECnet
DEC station

DECUS
DECwindows
DTIF
MASSBUS
MicroVAX
Q-bus
ULTRIX
ULTRIX Mail Connection

ULTRIX Worksystem Software
VAX
VAXstation
VMS
VMS/ULTRIX Connection
VT
XUI

UNIX is a registered trademark of AT&T in the USA and other countries.

Contents

About This Manual
Audience
Organization

vii
. . .. . .. . . . . . . . .. . . . . . .. . . . ... . . . .. . . . . . . . . . . . . . . . . ... . . . . . .. . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . .. . . . .. .

Related Documents

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

Conventions

Backup and Restore Methods and Strategies

1.1

Selecting a Backup Method

1.2

1.3

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

1-1

Using the dump or rdump Commands ..............................................
Using the mdtar or tar Commands ....................................................
Using the opser Utility ...................................................................

1-1
1-2
1-2

Selecting a Restore Method
1.2.1
1.2.2

Using restore or rrestore Commands ................................................ .
U sing tar or mdtar Commands ....................................................... ..

Developing a Backup Strategy
1.3.1
1.3.2

viii
viii

1.1.1
1.1.2
1.1.3

vii

Selecting a Full or Incremental Backup ............................................ .
Determining How to Use the Backup Levels .................................... .

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

1-4

Backup Commands and Procedures

2.1

Backing Up File Systems

2-1

2.2

Transferring Files to Tape or Diskette

2-2

2.3

Backing Up from the Operator Account

2-3

2.3.1
2.3.2
2.3.3
2.3.4

2.3.5

Running the opser Utility ............................................................... .
Determining Who Is Logged In ...................................................... .
Shutting Down Multiuser Mode ...................................................... .
Un mounting File Systems .............................................................. .
Checking File System Consistency .................................................. .

2-3

2-4
2-4
2-5
2-5

2.3.6
2.3.7
2.3.8
2.3.9
2.3.10
2.3.11

Backing Up File Systems ............................................................... .
Escaping to the Shell ..................................................................... .
Restarting Multiuser Mode ............................................................. .
Performing a Remote Backup ......................................................... .
Halting the Processor ................................................................... .
Exiting from opser ...................................................................... .

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

2.4

The Labeled Tape Facility

Local File Restoration

3.1

Using the restore Program

3-1

3.2

Preparing for an Interactive Restore

3-1

3.2.1
3.2.2
3.2.3

3-1
3-2
3-2

3.3

Logging In ................................................................................... .
Mounting the Dump Media ............................................................ .
Changing Your Working Directory ................................................. .

Restoring Files Interactively

3-2

3.3.1
3.3.2

3-2
3-3

Invoking the Interactive restore Program .......................................... .
Using Interactive Mode Commands ................................................. .
3.3.2.1
3.3.2.2
3.3.2.3
3.3.2.4
3.3.2.5

3.3.3
3.3.4
3.3.5

Responding to the 'Specify next volume #' Prompt
Responding to the 'Set owner/mode for' ." Prompt

3.4.2
3.4.3

3-4
3-5
3-6
3-6
3-7
3-7

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

3-7

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

3-7

Preparing for a Noninteractive Restore

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

3-8

Logging in ....................................................................... ..
Mounting the Dump Media ................................................. ..
Changing Your Working Directory ...................................... ..

3-8
3-8
3-8

Invoking the Noninteractive restore Program .................................... .
Exiting the Noninteractive Restore ................................................. ..

3-8
3-9

Restoring a Complete File System

iv Contents

3-3
3-3
3-3
3-4
3-4

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

Exiting from the Interactive Restore

3.4.1.1
3.4.1.2
3.4.1.3

3.5

Creating and Modifying the Extraction List

Restoring Files Noninteractively
3.4.1

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

Restoring the Specified Files and Directories .....................................
Entering the extract Command .........................................................
3.3.5.1
3.3.5.2

3.3.6

Listing the Commands ........................................................ .
Listing the Pathname of the restore Working Directory ........... .
Getting a File and Directory Listing ..................................... ..
Comparing Inode Numbers ................................................. ..
Changing Directories .......................................................... .

Specifying What You Want Restored from the Dump Image
3.3.3.1

3.4

2-8

3-9

3.6

Local Restoration of the root and /usr File Systems

3.7

Using the tar or mdtar Commands to Restore Files

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

3-13

3.8

The Labeled Tape Facility.........................................................................

3-14

Remote Backup

4.1

The Network Environment

4-1

4.1.1
4.1.2

4-1
4-1

The Master/Slave Relationship ....................................................... .
Hardware and System Setup Considerations ..................................... .
4.1.2.1
4.1.2.2

4.1.3
4.2

Hardware Considerations ......................................................
System Setup Considerations ................................................

Staging Area or Direct Backup Determination

4-3

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

4-3

Escaping to the Slave Shell .............................................................
Escaping to the Master Shell ...........................................................
Unmounting the Slave's File Systems .......... .....................................
File System Consistency Check on the Slave .....................................
Backing Up the Slave's File Systems ...............................................

4-6
4-6
4-6

4.2.5.1
4.2.5.2
4.2.6
4.2.7
4.2.8
4.2.9

4-9
4-11

Returning to the Local opser .......................................................... .
Halting the Slave Processor ............................................................ .
Stopping Remote Execution of opser ............................................... .
Exceptional Condition Handling ..................................................... .

4-11
4-12
4-12
4-12

Remote File Restoration

5.1

Using the rrestore Program

5.2

5.3

Network Considerations

5-1
..................................................................

5-2

Logging In ....................................................................................
Mounting the Dump Media .............................................................
Transferring Files to the Staging Area ..............................................

5-2
5-2
5-2
5-2

Restoring Files Interactively
5.3.1

5-1

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

Preparing for an Interactive Restore
5.2.1
5.2.2
5.2.3

4-7
4-9

Staging Area Backup Method ...............................................
Direct Backup Method .........................................................

5.1.1

4-1
4-2

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

Performing Remote Backups
4.2.1
4.2.2
4.2.3
4.2.4
4.2.5

3-11

Using Interactive Mode Commands
5.3.1.1
5.3.1.2
5.3.1.3
5.3.1.4

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

5-3

Listing the Commands .........................................................
Listing the Pathname of the restore Working Directory............
Getting a File and Directory Listing .......................................
Comparing Inode Numbers ...................................................

5-3
5-3
5-3
5-3

Contents v

5.3.1.5
5.3.2

Changing Directories

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

5-4

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

5-4

Restoring the Specified Files and Directories .....................................
Entering the extract Command .........................................................

5-6
5-6

Specifying What You Want Restored from the Dump Image

5.3.2.1
5.3.3
5.3.4

5.3.4.1
5.3.4.2
5.3.5

5-4

Creating and Modifying the Extraction List

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

5-6
5-7

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

5-7

Responding to the 'Specify next volume #' Prompt
Responding to the' Set owner/mode for' ." Prompt

Exiting from the Interactive Restore

5.4

Restoring Files Noninteractively

Device Mnemonics

5-7

Examples
4-1: Performing a File System Consistency Check

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

4-8

4-2: Staging Area Backup Method Display..........................................................

4-10

Tables
A-1: Devices Supported by MAKEDEV

vi Contents

A-2

About This Manual

The objective of this guide is to provide you with infonnation on backing up and
restoring files and file systems using ULTRIX system commands and utilities. The
guide assists you in developing system backup strategies locally and in a network
environment. It also presents guidelines from which you can develop specific
procedures for your site.

Audience
The ULTRIX Guide to Backup and Restore is written for the person responsible for
managing and maintaining an ULTRIX system. It assumes that this individual is
familiar with ULTRIX commands, the system configuration, the system's
controller/drive unit number assignments and naming conventions, and an editor such
as vi or ed. You do not need to be a programmer to use this guide.

Organization
This manual consists of 5 chapters, one appendix, and an index. The chapters and the
appendix are:
Chapter 1:

Backup and Restore Methods and Strategies
Identifies backup and restore methods and suggests certain
strategies for backing up and restoring files and file systems.

Chapter 2:

Backup Commands and Procedures
Identifies and explains how to use the various ULTRIX backup
commands and utilities.

Chapter 3:

Local File Restoration
Describes how to use the restore command and provides
procedures for restoring file systems.

Chapter 4:

Remote Backup
Explains the network environment and how to use the opser
utility to perform remote backups.

Chapter 5:

Remote File Restoration
Explains how to restore files and file systems using the
rrestore command. This chapter also contains procedures
for restoring the root and / u s r file systems after a head crash
has occurred.

Appendix A:

Device Mnemonics
Lists the supported device mnemonics and explains how to
obtain detailed reference page information on devices.

Related Documents
You should have the hardware documentation for your system and peripherals.

Conventions
The following conventions are used in this manual:

The default user prompt is your system name followed by a right
angle bracket. In this manual, a percent sign ( % ) is used to
represent this prompt.

A number sign is the default superuser prompt.

»>

The console subsystem prompt is two right angle brackets on
RISC systems, or three right angle brackets on VAX systems.
On a system with more than one central processing unit (CPU),
the prompt displays two numbers: the number of the CPU, and
the number of the processor slot containing the board for that
CPU.

CPUnn»

user input

This bold typeface is used in interactive examples to indicate
typed user input.

system output This typeface is used in interactive examples to indicate system
output and also in code examples and other screen displays. In
text, this typeface is used to indicate the exact name of a
command, option, partition, pathname, directory, or file.

UPPERCASE
lowercase

The ULTRIX system differentiates between lowercase and
uppercase characters. Literal strings that appear in text,
examples, syntax descriptions, and function definitions must be
typed exactly as shown.

rlogin

In syntax descriptions and function definitions, this typeface is
used to indicate terms that you must type exactly as shown.

viii About This Manual

macro

In text, bold type is used to introduce new tenns.

filename

In examples, syntax descriptions, and function definitions, italics
are used to indicate variable values; and in text, to give references
to other documents.

[ ]

In syntax descriptions and function definitions, brackets indicate
items that are optional.

{ I }

In syntax descriptions and function definitions, braces enclose
lists from which one item must be chosen. Vertical bars are used
to separate items.
In syntax descriptions and function definitions, a horizontal
ellipsis indicates that the preceding item can be repeated one or
more times.
A vertical ellipsis indicates that a portion of an example that

would nonnally be present is not shown.
cat(l)

Cross-references to the ULTRIX Reference Pages include the
appropriate section number in parentheses. For example, a
reference to cat(1) indicates that you can find the material on the
cat command in Section 1 of the reference pages.

IRETURNI

This symbol is used in examples to indicate that you must press
the named key on the keyboard.

ICTRUxI

This symbol is used in examples to indicate that you must hold
down the CTRL key while pressing the key x that follows the
slash. When you use this key combination, the system sometimes
echoes the resulting character, using a circumflex (A) to represent
the CTRL key (for example, AC for CTRL/C). Sometimes the
sequence is not echoed.
This symbol is used in examples to indicate that you must press
the first named key and then press the second named key. In text,
this combination is indicated as ESC-X.

About This Manual ix

Backup and Restore Methods and
Strategies

This chapter provides an overview of the backup and restore processes. The backup
and restore processes allow you to back up and restore files on the local system or
over the network. By performing regular backups, you are ensured a retrieveable
copy of your files should the originals be destroyed.
This chapter discusses how to select a backup or restore method. Also included in
this chapter is information that can help you develop a backup strategy.

1.1

Selecting a Backup Method
A backup operation lets you copy or transfer file systems or individual files and
directories to tape, disk, or diskettes. The command or utility that you choose to
back up files is determined by the type and location of files that you want to back up.
You can perform either a local backup or a remote backup. A local backup allows
you to copy or transfer file systems or individual files and directories from the system
on which you reside to tape, disk, or diskette. A remote backup lets you copy file
systems over the network to tape or disk. You cannot do a remote backup of
individual files and directories.
To back up files on the local system, you can use the dump, mdtar, or tar
commands, or you can use the opser utility. To back up file systems over the
network, you can use the rdump command or the opser utility. The following
sections briefly highlight the backup commands and the opser utility. For more
information, see Chapter 2.

1.1.1

Using the dump or rdump Commands
The dump command copies all designated file systems or individual files and
directories changed after a specified date to a file, pipe, magnetic tape, disk, or
diskettes. To use the dump command, you must have superuser privileges. The
system can be in either multiuser mode or single-user mode for a backup operation;
however, if you perform a backup of a file system or files in multiuser mode, any
user or system activity might corrupt that backup.
The rdump command copies all designated file systems changed aft.er a spe~ified
You cannot back
up individual files and directories over the network. To use the rdump command,
you must have superuser privileges and ~em_~!:!,~_~~. ,~Il~~~gl_~~~~~:. .rnode.

~ver the network to magnetic tape, disk, or a dump image file.

Both commands operate by checking the inodes of the files you want to back up.
The inodes contain data such as table entries and other statistics. When you use this
command to back up file systems or files, an inode is attached to each file. If system
or user activity changes a file after the inode data is recorded, but before the file is
backed up, the backup may be corrupted. Whether you are doing a local dump or a
remote dump of a file system, it is recommended that you perform the operation in
single-user mode.

1.1.2

Using the mdtar or tar Commands
The mdtar and tar commands save multiple files on multiple archives. Both
commands have the same functionality. While these commands let you back up
individual files and directories, you cannot back up file systems, nor can you perform
a remote backup using these commands. To use the mdtar or tar commands, you
must have read access on the files you want to back up and write privileges on the
output device. To back up files from root (I) or /usr, you must be root. The
system can be in multiuser mode.
The mdtar and tar commands operate by transferring files collectively. Unlike the
dump command, these commands read the inode statistics but do not attach inodes to
the files before transfering them to the output device. This lowers the possibility of
file corruption when doing backups in multiuser mode.

------1.1.3

USing the opser Utility
The osper utility is an interactive facility that lets you back up file systems on the
local system or over the network. The local op s e r facility only supports tape as a
backup media. The remote opser utility supports backups to files, disk, or tape.
To use opser, log in to the operator account; the opser utility is automatically
invoked. This utility runs as a shell environment in place of the system shell. While
in the op s e r environment, you can escape to the system shell, execute ULTRIX
system commands, and return to the op s e r environment. The system must be in
single-user mode when performing backups using the opser utility. A built-in help
menu provides you with a list of the available commands that allow you to shut
down the system, dismount devices, perform a file consistency check, and perform a
backup. You can back up file systems only to tape.
--1'his utility uses tlJie dump and the rdump commands to perform a backup of your
file systems. For information on how these commands work, see Section 1.1.1.

1.2 Selecting a Restore Method
A restore operation lets you retrieve file systems and individual files and directories
that you copied to a designated dump device or dump image file. The method you
choose to restore files is determined by the type of backup you performed.
If you used the rdump command or remote opser utility to back up files, you must
use the rrestore command to retrieve files. If you used the tar or mdtar
command to back up files, you must restore the files with the same command. If you
used th~ dump command or local op s e r utility to back up files, you must use the
_____ -------cre-s-to-re-~O'ommand to restore files. The next two sections briefly describe these
""o':[5erations'.

1.2.1

Using restore or rrestore Commands
The restore command retrieves, from magnetic tape, disk, dump image file,or a
pipe, files that were backed up using the dump command or the opser utility. If
you have write privileges on the files you are restoring, and if you do not plan to
change file or directory attributes, you do not need superuser privileges to perform a
restore operation.
The rrestore command retrieves from magnetic tape or disk, files that were
backed up using the rdump command or the remote opser utility. To restore

1-2 Backup and Restore Methods and Strategies

individual files and directories, you do not need superuser privileges, but you must
have write access to the files. To restore file systems, you must have superuser
privileges.
Both the restore and rrestore commands operate by reading, from the dump
media, the files that you specify. You can do an interactive restore or a
noninteractive restore. The interactive restore process has a built-in help file and
command options that allow you to do the restore. This is particularly helpful when
you want to restore only parts of a dump. The noninteractive restore is most useful
when you want to restore entire file systems or when you want to restore the entire
dump. These commands do not perform consistency checks nor do they use the
inode statistics to restore files.

1.2.2

Using tar or mdtar Commands
If you backed up your files with either the tar or mdtar command, the files must
be restored using the same command. To restore files using tar or mdtar, you
need write access to the files or directories that you want to retrieve. If you do not
have write access, superuser priveleges are necessary. As long as you do not attempt
to access absolute pathnames, you can perform the restore in multiuser mode.

A restore operation, using the mdtar or tar command, transfers files from your
backup medium to the directory or file you specify.

1.3 Developing a Backup Strategy
Regardless of whether you perform a local backup or a remote backup, you should
develop a backup strategy that ensures you have a retrievable copy of the latest files
should a mishap occur. Developing a backup strategy entails determining whether
you need to perform a full backup of your file systems or determining whether an
incremental backup is sufficient.
A full backup copies all files from a device to the backup medium you specify. An
incremental backup copies a subset of files from a device to the backup medium you
specify. The next two sections discuss how to select and specify a full or
incremental backup.

1.3.1

Selecting a Full or Incremental Backup
To instruct the backup commands, dump and rdump, and the opser utility, to back
up either all the files or a subset of files, you must specify a backup level. Backup
levels range from 0 to 9. A level-O backup instructs the commands or utility to back
up all files. Level-l to level-9 backup instructs the commands or utility to perform
an incremental backup based on the dates and levels stored in the
/ etc/ dumpdates file. _
Any backup strategy should begin with a full backup of the file system. You must
then determine how often you should perform incremental backups and when your
next full backup should occur. This varies from site to site, as it depends on your
work environment.
If there is constant testing or software development on your system, mishaps can
occur that affect your file system. In this instance, setting up a monthly backup
strategy is most feasible. You could begin with a full backup at the beginning of the
month, followed by daily incremental backups during the week, and weekly
incremental backups at the end of each week. If the environment that you work in

Backup and Restore Methods and Strategies 1-3

does not have much system activity and files are not updated or created frequently,
you may feel comfortable performing a full backup followed by weekly incremental
backups. Only you can determine the importance of your files and the need for
backups.

1.3.2

Determining How to Use the Backup Levels
The backup levels range from 0 to 9. A level-O backup always copies all files to the
device you specify. For levels I through 9, the information stored in the
/ etc / d umpda t e s file determines how much of the file system is backed up. The
/ etc / d umpda t e s file contains the dates and level numbers of each backup
performed on a file system. When performing a backup, the command or utility
searches for the previous backup that is equivalent in level to that currently specified.
If backups do not exist at that level, the backup command or utility searches for the
next lowest level.
The following example demonstrates a monthly backup strategy. In this example, the
files in the root file system are backed up to the default tape device. The u option
instructs the dump command to update the / etc/ dumpdates file after each
successful backup. Level-O backups are performed at the beginning of the month,
level 9 backups are performed daily, and level 5 backups are performed weekly:
# jete/dump Ou

# / ete/ dump 9u

[2]

# jete/dump 5u
# jete/dump 9u

# jete/dump Ou

The preceding example backs up files as follows:
~

Tells the dump command to do a full backup of the file system. This occurs as
the first backup of the month.

[2]

Instructs the dump command to do an incremental backup. The dump
command searches for the previous backup that occurred at level 9. Because a
level 9 backup does not exist, it backs up all files created or changed from the
next lowest level (level 0).

lal

Initiates a weekly incremental backup. That is, the / etc/ dumpdates file is
searched for the previous backup that occurred at level 5. If, as in this example,
the backup command or utility cannot locate a level 5, it searches for the next
lowest level. In this instance, all files modified or changed since level 0 are
backed up.

1-4 Backup and Restore Methods and Strategies

Continues daily backups. The! etc! dumpdates file is seached for the
previous backup that occurred at level 9, and all files changed or" created since
!~c~t."ti~~.c~~"~,~,~~e_? u~:
----.-- -. . ,,,.,,------,,-.,-- -_.,-""",,",, .-.-

15]

Forces the dump command to back up those files changed since the last level 5
backup.

Backup and Restore Methods and Strategies 1-5

Backup Commands and Procedures

This chapter explains how to back up files and file systems using the ULTRIX
backup and file copy/transfer commands and utilities. The backup and file transfer
commands described in this chapter enable you to complete the following:
•

Back up file systems

•

Transfer files and file systems to tapes or diskettes

•

Perform backups from the operator account

The following sections describe how to use ULTRIX commands and utilities to
perform the above tasks.

2.1

Backing Up File Systems
Use the dump command to back up file systems. This command enables you to back
up a file system to a tape or disk device. To back up file systems, the system must
be in single-user mode. This section describes how to perform a file system backup
of the root (/) and the /usr file systems:
1.

Shut down the system. To shut down the system in five minutes and give users
periodic warning messages, type:
# fete/shutdown +5 'System going down to perform backups'

Determine the names of the devices containing the file systems to be backed up
by listing the contents of the / etc / f stab file as follows:
# cat /ete/fstab
/dev/rzOa:/:rw:l:l:ufs: :
/dev/rzOg:/usr:rw:l:3:ufs::

In this example, the / and /usr file systems are on the rzO device, partitions
a and 9 respectively.
If you have performed backups using the dump command with the -u option,
you can also use the dump command with the -w option to determine which
files systems need to be backed up. The -u option updates the
/ etc/ dumpdates file.

Use the umount command with the -a option to dismount the files systems
that you want to back up:
# umount -a

Note that the root (/) file system remains mounted.
4.

Use the fsck command to perform a file system consistency check on the file
system that you want to back up. The f s ck command has the following
format:
fsck device

Perfonning a file consistency check on the file system you want to back up
ensures the integrity of the file system. The f s ck command attempts to correct
any inconsistencies.
5.

Perfonn a level-9 backup of the root (I) file system to the default tape device.
Use the following fonnat:
# fete/dump 9u /

Perfonn a level-9 backup of the /usr file system to a disk device. Use the
following fonnat:
fete/dump 9uf device /usr

In these level-9 backups, the dump command provides you with several messages
and prompts you to insert your media, depending on the device type that you specify.
It also prompts you with a number of yes and no questions. Answer all of the
questions.
Change the media when the program prompts you to do so. Be sure to label all your
media with the date of the backup, the file system name, the starting inode number
for that volume, and the backup level. This makes it easier to locate files and file
systems that you want to restore. Store your backups in a safe location.
If you want to tailor the backup process so that certain options are in effect, see
dump(8) in the ULTRIX Reference Pages.

2.2 Transferring Files to Tape or Diskette
The tar and mdtar commands allow you to transfer an individual user's directories
or files to tapes or diskettes. You do not need superuser privileges, nor do you have
to be in single-user mode to use these commands. However, you must have read
pennission on the files you want to back up.
Use the tar command when magnetic tape is the primary backup medium. The
tar command is a tape archive facility that supports single-volume and multivolume
archives.
System users should invoke the tar command to back up their files, provided that
the system has a tape drive. The following example shows how to back up a
directory named /usr/staff/ajw to tape. This example assumes that tape drive
0, the default device, is being used:
# ed /usr/staff
# tar e ./ajw

To use the tar command along with the device name, type:
# tar ef /dev/rmtlh ./ajw

Use the mdtar command when diskettes are the primary backup medium. More
than one diskette may be needed. Given a directory pathname, the mdtar command
makes a list of the files and decides how to fit them on multiple diskettes.
The following example shows how to back up the directory / u s r / s t a f f / a j w to
diskettes:
# ed /usr/staff
# mdtar e ./ajw

2-2 Backup Commands and Procedures

To use the mdtar command along with the device name, type:
# mdtar cf /dev/rmtlh ./ajw

The mdtar command prompts you to insert and remove additional diskettes as
needed to back up the files.

2.3 Backing Up from the Operator Account
The opser utility, / opr / opser provides an interface for ULTRIX system
maintenance, primarily file system backups. The opser utility runs as a shell
environment in place of the system shell. To automatically invoke the opser utility,
log in to the operator account. The operator password is set by the superuser.
This utility enables you to complete the following:
•

Determine who is currently logged in to the system

•

Shut down multiuser mode, enter single-user mode, and restart multiuser mode

•

Check file system consistency

•

Perform incremental or full file system backup to tape

•

Perform incremental or full file system backup over the network

•

Halt the processor

•

Execute op s e r options on a remote system.

The opser utility also allows you to escape to the shell, execute ULTRIX system
commands, and return to the opser utility. It also provides on-line help in the form
of an options menu. For more information, see opser(8) in the ULTRIX Reference
Pages.

To perform backups, run the opser utility from the operator's console. Running the
opser services from a terminal limits you to the help, user, and shell functions.
The opser utility supports these tape devices for backup operations: TE16, TSl1,
TSV05, TU77, TU78, TU79, TU80, TU81, TU81E, TK50, and TK70. It also
supports the network as a backup device.

2.3.1

Running the opser Utility
To run the opser utility, log in to the operator account. The following information
is displayed on your screen when op s e r is invoked:
ULTRIX-32 Operator Services
Line editing: delete - erase one character, AU - kill entire line
For help, type h followed by a return
opr>

The opr> prompt informs you that the opser utility is running in place of the shell
and is ready to accept opser options.
To use an op s e r option, type the appropriate letter or name. Then, press the
RETURN key. For example, to display on-line help about all opser options, type:
opr> h

Backup Commands and Procedures 2-3

() - may use first letter in place of full name
Valid commands for Local Opser are:
!sh
(u) sers
(s)hutdown
(d)ismount
(f)sck
(r)estart
(h)elp
(b)ackup
halt
(n)etwork [Slave]
(q)uit

- shell escape (execute ULTRIX-32 commands)
(Type control d to return from shell)
- show logged in users
- stop time-sharing
- unmount file systems
- file system checks
- restart time-sharing
- print this help message
- file system backup
- halt processor
- initiate Remote Opser
- exit from opser

To end an opser session and return to a login prompt, type q.
The next sections provide detailed discussions of the opser options.

2.3.2

Determining Who Is Logged In
To determine how many users currently are logged in to the system, enter the u
(users) option at the opr> prompt. For example:
opr> u

When you enter the u option, the opser utility displays a list of users who are
currently logged in.

2.3.3

Shutting Down Multiuser Mode
Use the s (shutdown) option to place the ULTRIX system in single-user mode. To
back up the root (I) or /usr file systems, you must be in single-user mode. Other
file systems that you may have mounted, such as /usr / src, can be unmounted
before being backed up. You need to be in single-user mode only when backing up
systems that are used in multiuser mode, such as /usr or user data areas. Even user
data modes can be backed up if no users are logged in and you have
/ etc/nologins on.
To shut down multiuser mode and leave the system in single-user mode, enter the s
(shutdown) option at the opr> prompt. For example:
opr> s

When you enter the s option, the opser utility first displays the names of those
users currently logged in to the system. Then, it prompts for the number of minutes
to delay before shutting down from multiuser to single-user mode. The number of
minutes that you enter must be between 1 and 99. This delay gives users enough
time to finish their work and to log out. During this delay period, the opser utility
broadcasts warnings of the impending shutdown. Five minutes prior to the
designated time, it disables logins. Finally, at the designated time, it broadcasts a
final message and then shuts down multiuser mode.
Prior to entering single-user mode from multiuser mode, the opser utility kills all
run11ing processes and syncrIfonizes the file systems (using the sync utility). When
the opser utility redisplays the opr> prompt, you are automatically in a single-user
environment. You can now enter other opser options and continue with system
maintenance.

2-4 Backup Commands and Procedures

2.3.4

Unmounting File Systems
To unmount the file systems when running op s e r in single-user mode, enter the d
(dismount) option at the opr> prompt. For example:
opr> d

When you enter the d option, the opser utility invokes the umount command to
unmount file systems listed in the /etc/fstab file. For more information on the
umount command, see mount(8) in the ULTRIX Reference Pages.
To ensure that all of the file systems have been unmounted, escape to the shell, using
the ! sh option as described in Section 2.3.7, and enter the mount command.
Without options, the mount command displays all of the currently mounted file
systems. At this point, you will have to unmount the file systems by issuing the
appropriate umoun t command.
The following example shows the sequence of commands that you would enter
assuming that the /usr / staff3 file system is still mounted after you entered the d
option:
opr> !sh
Password:
type AD to return to opser
# Jete/mount
/dev/rzOa on / type ufs
/dev/rzlh on /usr/staff3 type ufs
# /ete/umount /dev/rzlh
# <CTRL/O>
opr>

In this example, the shell escape option, ! sh, returns you to the system shell prompt
(#) after the root password is entered. Entering the mount command without options
displays the list of mounted file systems, in this instance, root (/) and
/usr / staff3. The root file system should remain mounted; however,
/usr / staff3 is dismounted by issuing the umount command on the
/dev/rzlh device. Type CTRLjD to return to the opr> prompt.
Note

When you unmount the / u s r file system, some ULTRIX commands
may be unavailable.

2.3.5

Checking File System Consistency
The file systems should be checked prior to performing a backup. You should not
back up a corrupted file system. To check the consistency of your file systems when
running opser in single-user mode, enter the f (file system check) option at the
opr> prompt. For example:
opr> f

When you enter the f option, the opser utility invokes the fsck command to
check the file systems indicated in the / etc / f stab file.

Backup Commands and Procedures 2-5

2.3.6

Backing Up File Systems
You should unmount and then check all file systems before backing them up. Use
the d option to unmount the file systems. Use the f option to check the file systems
for consistency. To shut down multiuser mode prior to unmounting and checking file
systems, specify the s (shutdown) option.
~
To back up file systems when running the apser utility in single-user mode, enter
the b (backup) option at the apr> prompt. Enter the option using the fonnat:
b command file name

The command file name argument specifies the command file that contains the
ULTRIX system commands required for that backup. If you do not use the command
file name argument, the ap s e r utility calls the default backup command file,
/ apr /backup. This script enables you to perfonn level-O or level-9 backups to
supported tape drives.
The / apr /backup command file prompts you to specify the tape drive number, the
logical unit number of the tape drive, and the tape density. For example:
opr> b backup
ULTRIX-32 File System Backup Procedure
start errlog daemon - elcsd
elcsd:
Please specify the tape drive type. (eg. te16):
Please specify the logical unit number for the tape drive.
Please specify tape density. (eg. 1600):

(eg. 0):

If you specify either a TK50 or a TK70 for the tape drive type, this backup script will
not prompt you for a tape density. If any of your entries are invalid, the command

file responds with the message:
Error in specifying tape drive - repeating questions.

When you have correctly specified the device, the backup script prompts:
Backup to be full or incremental? < f or i >:

If you enter an f, a level-O backup is perfonned. If you enter an i, a level-9 backup

is perfonned.
After you have correctly responded to the prompts, the file system dump process
begins.
The default backup command file also invokes the error logger daemon in single-user
mode, so that error logging is enabled during backups. The error logger daemon is
described in the ULTRIX Guide to the Error Logger.
As the superuser, you should set up backup command files in the way that you want
your backups perfonned.

2.3.7

Escaping to the Shell
To escape to the shell at any time when running the apser utility, enter! sh at the
apr> prompt. The apser utiiity prompts you to enter the root password. This is
the only password you can enter. For example:

2-6 Backup Commands and Procedures

opr> !sh
Password:
type AD to return to opser
#

After receiving a shell prompt, you can execute any ULTRIX command. The! sh
command is useful for using the tal k command to tell users that you intend to shut
down the system. After you finish entering ULTRIX commands, press CTRL/D to
return to the ap r > prompt.

2.3.8

Restarting Multiuser Mode
To restart multiuser mode when running the ap s e r utility in single-user mode, enter
the r (restart) option at the apr> prompt. For example:
opr> r

This option causes the system to restart multiuser mode. The system displays the
same startup messages and starts the same daemons as though a normal startup had
occurred. When the system is back in multiuser mode, it again displays the standard
login prompt.

2.3.9

Performing a Remote Backup
The ap s e r utility enables you to back up a file system over the network. Chapter 4
explains how to use the ap s e r utility to perform a remote backup.

2.3.10

Halting the Processor
To halt the processor when running the apser utility in single-user mode, enter the
h (halt) option at the apr> prompt. For example:
opr> halt

This command synchronizes the disks (writes out in-memory file system information
to the disks) and then halts the processor. You must then reboot your system
manually as described in the ULTRIX Guide to Shutdown and Startup.
Note

To shut down multiuser mode prior to halting the processor, specify the
s (shutdown) option.

2.3.11

Exiting from opser
You can exit the ap s e r utility and return to the login prompt by using the q option,
provided that you have not shut the system down to single-user mode. Once you
shut the system down, you must use the r (restart) option to exit. Section 2.3.9
describes how to use the r option. If you enter q at the apr> prompt after shutting
down the system, apser displays the message:
Time-sharing stopped
opr>

If you have not shut the system down and you want to exit apser, enter the q (quit)
option at the apr> prompt. For example:

Backup Commands and Procedures 2-7

apr> q

The opser utility logs you out of the operator account and the system will redisplay
the login prompt.

2.4 The Labeled Tape Facility
In addition to the backup methods already described, the system contains a labeled
tape facility command, 1 t f. The 1 t f command reads and writes single-volume,
versions three and four, ANSI-compatible tape volumes. This provides you with a
way of accurately transferring information between ULTRIX and other systems. For
information on this command, refer to Itf(l) and Itf(5) in the ULTRIX Reference
Pages.

2-8 Backup Commands and Procedures

Local File Restoration

This chapter describes how to restore files and file systems using ULTRIX
commands. Specifically, this chapter enables you to do the following:

3.1

•

Restore files, directories, or file systems on different media using the restore
command ~~---. .---.
~.~."
.,,"~: . - .j

•

Transfer files using the tar and mdtar commands

•

Restore a complete file system

•

Restore the root (/) and /usr file systems after a catastrophic event has
occurred

Using the restore Program
Use the restore program to restore files, file systems, and directories that you
backed up with the dump program.
Note
If you used the tar or mdtar command to back up your files, you must
use the tar or mdtar command again to restore the backup media. See

Section 3.7 for details.
With the restore program, you can do either an interactive or a noninteractive
restore operation. The following sections explain both methods with examples of
how to perform each.

3.2 Preparing for an Interactive Restore
Before starting the actual restoration, you should prepare in three ways:

•

•
•

Mount the dump media.

VCJ V vV\ {

Change your working directory to the dump media mount point. -

The following sections describe each of these preparatory tasks.

3.2.1

Logging In
Some restore functions require superuser privileges. For example, to change file
and directory attributes during a restore operation, you must be logged in as
superuser. See restore (8) in the ULTRIX Reference Pages for details on the
restore command options and required permissions.

3.2.2

Mounting the Dump Media
Before invoking the restore program, ensure that the backup media containing the
dump image is mounted on the appropriate device. This is usually the default tape
device, / dev / rmt Oh, but it could be any device or a file system in the form of an
on-line backup file. In a multivolume dump set, you must mount volume 1 first,
because it contains information on the directory structure and inodes for all files on
the dump image. When you have established the restore environment, you can
mount other volumes in the set.

3.2.3

Changing Your Working Directory
Change your current working directory to the mount point of the file system that was
backed up. For example, if the /usr file system was backed up from root, first
position yourself at root, then enter the command to invoke restore from this
location.
Note
If you are unsure of the mount point, enter the / ete/ restore
command with the -t option for a complete listing of the dump media

contents.
Be aware that when you invoke the restore program, you enter the
restore environment. While in this environment, you get directory
information about the dump media only, not about your working
environment.
The restore program restores files to the current working directory (the directory
from which you invoked the program) using the relative pathname of the dump
media files that are being restored.

3.3 Restoring Files Interactively
To restore individual files or directories from the dump image on the backup media,
use the jete/restore command with the -i option. The -i option specifies that
you want to do the restore in interactive mode.

3.3.1

Invoking the Interactive restore Program
The location of the dump image determines what command format you use to invoke
the restore program:
•

If the dump image is located on a device other than the default (ldev/rmtOh),

use the format:
jete/restore -if

~vke

The - i option specifies interactive mode. The - f option and device argument
let you specify the name of the archive. Include the - f option and device
argument when the archive is a device other than / dev / rmt 0 h. If the device
is the default tape drive / dev / rmt 0 h, exclude t4e - f option and device
argument on the command line as follows:
# jete/restore -i

3-2 Local File Restoration

•

If the dump image is an on-line backup file on a mounted disk, use the fonnat:
/etc/restore -if file

Here, file specifies the dump image backup file. Specify either the full or
relative pathname, depending on your current working directory. For example,
to restore the /usr file system from a dump image named /usr /baekups,
position yourself at root and type:
# Jete/restore -if /usr/baekups

3.3.2

Using Interactive Mode Commands
When you type the jete/restore command with the -i option, the restore >
prompt appears on your screen"The prompt signals that the restore program is
waiting for a command.
'. \~~ O!J \ ,:
I'

v~.hrrl

3.3.2.1

;; Cl

Listing the Commands - To get a listing of the available commands, enter an h
or a? at the restore > prompt. The program responds with a listing. For
example:
restore >h
Available commands are:
ls [arg] - list directory
cd arg - change directory
pwd - print current directory
add [arg] - add 'arg' to list of files to be extracted
delete [arg] - delete 'arg' from list of files to be extracted
extract - extract requested files
quit - immediately exit program
verbose - toggle verbose flag (useful with "ls")
help or '?' - print this list
If no 'arg' is supplied, the current directory is used
restore >

3.3.2.2

Listing the Path name of the restore Working Directory - To list the
pathname of the working directory within the restore environment, type pwd at
the restore > prompt. For example:
restore >pwd
/lib
restore >

3.3.2.3

Getting a FHe and Directory Listing - To get a listing of the files and directories
that can be restored at the current directory level, type Is at the restore >
prompt. For example:
restore >1s
adm/
bini
dict
restore >

hosts/
include/
kits

man
mdec/
msgs/

skel/
spool/
src/

To get a listing of the files and directories that can be restored at a different directory
level, enter the pathname of that directory as an argument to the 1 s command.

Local File Restoration 3-3

3.3.2.4

Comparing Inode Numbers - Inode numbers increase with each tape. The
largest inode numbers are on the last tape and the smallest numbers are on the first
tape. By comparing the displayed inode numbers with the starting in ode numbers
recorded on a multivolume dump set, you can find the tape or disk that contains the
file you want to restore. To see the inode numbers of the files contained within your
listing, first use the verbose command to enter verbose mode, then use the Is
command to get your listing. The restore program displays the files with the
inode numbers to the left of the dumped file names. For example:
restore >verbose
verbose mode on
restore >ls
2
2

*.1
* .. I

11971
2177

11969

adml

11427
2722

1633 bini
restore >

hosts
includel
mdecl
msgsl

14253
4145

4253
3266

spool I
srcl
staff
sysl

In this example, the h 0 s t s file has an inode number of 11971 and the spool
directory has an inode number of 14253.

Note
Sometimes the in odes for a given file span multiple tapes or disks.
When this happens, you will need more than one tape or disk to restore
the file.
After checking the inode numbers, you can turn off verbose mode. To do this, type
verbose one more time. For example:
restore >verbose
verbose mode off
restore >

3.3.2.5

Changing Directories - Change directories in the restore environment with the cd
command. To specify a particular directory, enter the cd command with the
directory pathname given as an argument to the command. For example:
restore >cd /lib
restore >

3.3.3

Specifying What You Want Restored from the Dump Image
When working interactively with the restore program, you must first decide which
files or file systems you want restored from the dump media. Use the 1 s command
to get a file or directory listing of the dump media. When you know what is on the
media, you can specify what you want restored.
You specify what you want restored by creating and modifying an extraction list.
Creating an extraction list involves working from the dump image and using the
restore program's add command. Modifying an extraction list involves working
from the dump image and using the restore program's add and delete
commands.

3-4 Local File Restoration

3.3.3.1

Creating and Modifying the Extraction List - To create and modify an
extraction list, you must be within the restore environment. When you invoke the
restore program, follow these steps:
1.

Use the pwd command to determine your position within the dump media
directory structure.

Use the 1 s command to identify which files and directories are on the dump
image at the current level.

Decide which files or directories you want to extract from the dump image.

Use the cd command to move around the dump image and confirm the location
of files and directories.

Position yourself in the dump media directory structure at a point that is one
level above the desired item. For example, to restore /usr / lib, position
yourself at / u sr. If you attempt to identify / u s r /1 ib while positioned
within /usr / lib, the program issues an error message.

Create the extraction list by typing the add command and the name of each
desired file or directory as an argument to the command.
For instance, if you are in the / us r directory and you want to extract the files
Ipf and Ipd and the subdirectory refer, enter the add command followed
by the file and directory names:
restore >add 1pd 1pf refer
restore >

Note
When you add a file to the extraction list, the program marks that
file as one to restore. When you add a directory to your extraction
list, the program marks that directory, its files, and all subordinate
directories and files as ones to restore.
7.

Use the Is command to see how the program has marked these files. For
example:
restore >ls
diffh
dnetl
how_pix
restore >

libcy.a
libcg.a
libp~y.a

*lpd
*lpf
*refer/

Notice the asterisks (*) in front of the items that you added to your extraction
list. When you view the directory listing, the asterisks point out what you put
on the list. Similarly, when the restore program reads an asterisk during the
extraction process, it transfers a copy of that file or directory from the dump
image to your current working directory.
8.

Use the delete command with the file name given as an argument to remove
a particular file from the list. For example:
restore >de1ete 1pf
restore >
Local File Restoration 3-5

Use the Is command to see what the program did in response to the delete
command. For example:
restore >18
diffh
dnet/
howyix
restore >

libcy.a
libcg.a
libpc_p.a

*lpd
lpf
*refer/

Notice that the Ipf file has no asterisk now. This change indicates that the
program took Ipf from the list and will not extract it from the dump image.
10. When you create or modify the extraction list, you request a transfer of the files
to your system. To do this, use the extract command.

3.3.4

Restoring the Specified Files and Directories
The restore program's extract command initiates the actual restore. When you
enter the extract command, the program transfers a copy of each marked item to
the current working directory using the relative pathname found on the dump media.
Note

The current working directory is the one from which you invoked the
restore program initially. Your position within the dump media
directory structure when you enter the extract command has no
influence on the final location of the restored files.
For example, assume that the file /usr / lib/ Ipf was dumped to tape with /usr
as the mount point of the backup. In this case, . / lib/ Ipf is the relative
pathname on the dump media. If you invoked the rrestore program from /usr
(in other words, with /usr as your current working directory), the program restores
/ lib/ Ipf as subordinate to /usr (/usr / lib/ Ipf). If you invoked the
rrestore program from / lib (in other words, with / lib as your current
working directory), the program restores / lib/ Ipf as subordinate to / lib
(/ lib/ lib/ Ipf).

3.3.5

Entering the extract Command
To extract the files that you placed on your extraction list, enter the extract
command at the restore > prompt. For example:
restore >extract

In response, the restore program displays several statements and prompts you for
some answers. For example:
You have not read any tapes yet.
Unless you know which volume your file(s) are on you
should start with the last volume and work forward
towards the first.
Specify next volume #:
set owner/mode for' .'? [yn]

3-6 Local File Restoration

3.3.5.1

Responding to the 'Specify next volume #' Prompt - The first prompt,
Specify next volume #:, assumes that you are working from a multivolume backup
set.
•

If you are working from a multivolume set, and you know that the files you
want extracted start on a certain volume number, then enter that volume
number. If you do not know which volume contains the files that you want
extracted, enter the number of the last volume of the set and work your way
towards the first. For example, for a three-volume backup set, enter 3 at the
initial prompt. The program returns a second prompt to which you respond by
entering 2, and so on.

•

If you are working from a single-volume backup tape, enter 1 in response to the
initial prompt.

In response to your entry, the restore program searches the dump image for the
items listed on the extraction list. When it finds the individual files and directories, it
restores them. \\
l\
I~ ':::lQ"

3.3.5.2

.:s.'.,

;

~.\

,)'

,i.\f--4:..

R~;~~1,;~~ t~(the 'Set o~~~zr/~l~d;\f~r ,~,~p~~~~1t - T~etJ;~~6~d)pro~p1, seiC/~~l :./,
owner/mode for '.', lets the superuser change the attributes of the current directory
(shown as '.' within the prompt) to the attributes found for that entry on the dump
image. The current directory's attributes include access permissions, ownership, and
the accessed and update times.
•

If you want to change the attributes, enter a y.

The restore program responds by changing the current directory's attributes
to those found on the dump image for the corresponding directory. It then
redisplays the restore > prompt.
If you are not the superuser and you enter a y in response to the prompt, the
program responds with an error message and does not change the attributes.
•

If you do not want to change the attributes or if you do not have superuser
privileges, enter an n.

The restore program leaves the attributes of the current directory unchanged.
It then redisplays the restore > prompt.

3.3.6

Exiting from the Interactive Restore
When you finish restoring files, type qu it to get back to the system prompt:
restore >quit
#

3.4 Restoring Files Noninteractively
Using a noninteractive restore method is useful when you want to restore an entire
file system or the entire dump image.
The following sections describe how to prepare for a restoration and how to invoke
the noninteractive restore program.

Local File Restoration 3-7

3.4.1

Preparing for a Noninteractive Restore
Before starting the actual restoration, you should prepare in three ways:
•

•

Mount the dump media.

•

Change your working directory to the dump media mount point.

The following sections describe each of these preparatory tasks.

3.4.1.1

Logging in - Some restore functions require superuser privileges. For example, if
you intend to change directory attributes when restoring, you must be superuser. See
restore(8) in the ULTRIX Reference Pages for details on the restore command
options and required permissions.

3.4.1.2

Mounting the Dump Media - Before invoking the restore program, ensure that
the backup media containing the dump image is mounted on the appropriate device.
This is usually the default tape device, / dev / rmt Oh, but it could be any device or
file system in the form of an on-line backup file. In a multivolume dump set, you
must mount volume 1 first, because it contains information on the directory structure
and inodes for all files on the dump image. When you have established the
restore environment, you can mount other volumes in the set.

3.4.1.3

Changing Your Working Directory - Before invoking restore, change your
current working directory to the mount point of the file system that was backed up.
For example, if the /usr file system was backed up from root, first position
yourself at root and then enter the command to invoke restore from this
location.
Note
If you are unsure of the mount point, enter the / etc / res tor e
command with the -t option for a complete listing of the dump media
contents.

Be aware that, when you invoke the restore program, you enter the
restore environment. While in this environment, you get directory
information about the dump media only, not your working environment.
The program restores files to the current working directory (the directory from which
you invoked the restore program), using the relative pathname contained on the
dump media. If you are in the wrong directory, the program could overwrite the files
in that area if they happen to have the same file name.

3.4.2

Invoking the Noninteractive restore Program
To restore files noninteractively, enter the / etc/ restore command with the
options and arguments that specify what you want restored. See restore(8) in the
ULTRIX Reference Pages for a description of the command, its options, restrictions,
and diagnostics.
The following examples demonstrate various command line entries:

3-8 Local File Restoration

•

If the dump image is located on the default device, and you want to restore the
entire image, use the format:
#/ete/restore -x

The -x option specifies that the root directory and all subordinate items are to
be extracted.
•

If the dump image is on a device other than the default, and you want to restore
the entire image, use the fonnat:
fete/restore -xf device

The - x option specifies that the entire image is to be restored. The - f option
and device argument let you specify the name of the archive. For example, if
the dump media resides on a device named /dev/nntlh, and you want all items
restored, enter:
#/ete/restore -xf /dev/rmtlh

•

If the dump image is located on the default device, and you want to restore
specific files from the image, use the fonnat:
)<

fete/restore -v pathnamel pathname2

Here, the -v option specifies that you want the program to display the name
and type of each file it extracts. The pathname arguments specify that the
named files are to be extracted from the dump media. If these files match a
directory on the dump media, the directory and subordinate files and directories
are also extracted. For example, to restore (in verbose mode) the files
/usr / lib/ a and /usr /bin/b which reside on the default device, enter:
#/ete/restore ~ /usr/lib/a /usr/bin/b
/I;

3.4.3

Exiting the Noninteractive Restore
At the end of a noninteractive restore, the restore program exits automatically and
the system redisplays its prompt.

3.5 Restoring a Complete File System
,,----7

Vi el\r I/Ciluv0e!'0 restore a completefile system, you must first unmount it.

For infonnation on
unmounting a file system, see mount(8) in the ULTRIX Reference Pages. When the
L,lhl ()
-' '1
file system is unmounted, perfonn these steps:
1.

Create the target file system. To create the file system, use the newf s
command on an unmounted (raw) device.
Caution
When re-creating a file system, the newf s command destroys all
existing data on the partition. Therefore, you should back up
important data before issuing newf s. To check the partitions, use
the chpt command with the -q option.
For example, to use the newfs command to create a new file system on
partition g of an RZ55 disk on drive 0, enter:
# /ete/newfs /dev~rzOg rzs1
y

Local File Restoration 3-9

For more infonnation on the newfs command, see newfs(8) in the ULTRIX
Reference Pages.

Check and mount the target file system. After creating a target file system, you
must check it using the fsck command and then mount it using the mount
command. To perfonn file consistency check, use the following fonnat:
fsck device
/oI~v/YY;l (i cry

.... ;

To mount the file device onto the target directory / u s r, use the following
fonnat:
/~ev/retj
/vsr
/ etc/mount device /usr

When you invoke the fsck command, it displays certain infonnation on the
file system. Refer to f s ck(8) in the ULTRIX Reference Pages for further
details.

Change to the directory on the target file system. When the target file system is
available for use, use the cd command to position yourself at the top of the
directory, in this case, / u s r:
# ed /usr

P:-:v
YJ;1,'l

,.------)
(

Restore the dumped file system from the backup media. Having sufficiently
prepared the target file system, restore the file system from its most recent and
most complete dump tapes or disks. Use the / etc/ restore command with
the r option. For example:

'-(::--># fete/restore r

)It-~)f' )
-_./

This command restores all of the files and directories in the file system from a
dump tape mounted on the default tape device, / dev / rmt Oh, to the current
directory, / u sr.
Remove the restore table by removing the file named restoresymtable that
restore created in the current directory. Then reposition yourself at the
root (/) directory. For example:
# rm restoresymtable
# ed /

Unmount the restored file system by using the following fonnat:
/ etc/umount device

This command unmounts the file system / u s r which was mounted in step 2,
and prepares the file system for the next step. Refer to mount(8) in the
ULTRIX Reference Pages for more infonnation on the umount command.
7.

Check the file system. After unmounting the restored file system, use the
following fonnat to check for inconsistencies:
/ete/fsek device

The f s ck command checks the named file system, notifies you of all
inconsistencies, occasionally prompts you for a response to its suggested course
of action, and proceeds accordingly. When unsure of the consequences of your
response, you should answer no. By answering no, you leave the condition
uncorrected, but create a summary from which you can decide on a plan of
action. You can also use the f s ck command with either the -p or - P options.
3-10 Local File Restoration

Refer to. f s ek(8) in the ULTRIX Reference Pages fer infonnation on these as
well as other f s ek command options.
Note
If you are using f s e k on the root partition and f s e k finds and
corrects errors, you should halt the system without synchronizing
the disks. See the decumentatien for your precessor for
infonnation on how to halt the processer. Halting the processor
returns yeu to the console subsystem and allows you to rebeot the
system. It also. ensures that the system buffer cache does not
everwrite the f s ek corrections. Yeu must then reboot your system
to get the cerrected roet file system into memery.

Fer further details, see the infonnation on maintaining file system consistency
in the ULTRIX Guide to System Crash Recovery.

3.6 Local Restoration of the root and /usr File Systems
This sectien describes a precedure for restoring the root (/) and / u s r file systems of
your system disk.
The procedure in this section requires you to. have access to the most recent dump
image of your root (/) and /usr file systems. The dump infonnation sheuld be
available either in the fenn of a dump tape or as a sequence of dump diskettes.
You should use this procedure only when a catastrophic error occurs en the system
disk, such as a disk crash, or when the inadvertent deletion ef either the roet (/) er
/usr file systems renders the system inoperative.
The follewing example assumes that yeu are restoring yeur root (/) and / u s r file
systems from a level 0 TK50 dump tape.
pisk device file names are,.J2!9ce~~!~E-ecific..:.. Appendix A lists the disk device
special file name mnemonics and provides infonnation en how to find their
A-, il, , D
associated special file names. , i , ' " . , . , /

M<0V

_ _ _ _1.

[j/

7'--

' c:/VI;?

;

f/7'

al·'

e., c.,...
-

r'r:(,)

Bo.Qt the TK50 device where the standalene system volume is mounted. The
boet cemmand IS described in the Basic Installation Guide fer your specific
proces~.!, ~ 5 vu+ew) 5'5 0
-.,
<..!

After entering the beot command, the system prempts you to enter 1 for the
Basic Installation, 2 for the Advanced Installation, er 3 for System
Management.
-- Vc

Enter 3 for System Management.

Run MAKEDEV in the / dev directory to make the system disk special file and
the tape device special file. The command line fonnats for making the special
files are as fellows:
cd Idev
MAKEDEV mnemonic

r;2:.0

The mnemonic argument refers to a device mnemonic. See Appendix A fer a
listing ef the supperted device mnemonics. These commands create the
necessary special files in / dev for the system disk and the TK50 tape drive.

Local File Restoration 3-11

Change to the root (I) directory.
#

yna V) b~

v V1 vrl C. e.

cd /

Make a new root file system using the rnk f s command. This command
requires you to use the following format:
mk f s dev name 32 7 68 sectors tracks

(.s ~~.

\ e:;'

The dev name variable specifies the raw device special file. The number
327 68 specifies the size of the root file system in 512-byte blocks. This is a
constant used by all disks. The sectors and tracks variables specify the diskspecific number of sectors and number of tracks for the target disk. The
following table lists the supported disk drives, the sectors, and the tracks
entries.
Disk

1."\\

ra60
ra70
ra80
ra81
ra81
ra90
rd53
rd54
rm05
rp07
rf30
rf31
rf71
rz55
t\1 ---'rz56

Sectors

Tracks

42
33
31
51
57
69
18
17
32
50
37
50
37
36
54

4
11

14
14
15
13
8
15
19
32
6
8
16
15
15

As described in the rnkfs(8) reference page, additional entries can be made for the
block size and the fragment size. However, the accepted approach is to let rnkfs use
it defaults of 8192 and 1024 respectively.
The following example shows you the correct rnk f s entry for making the root file
system on partition a of an rd54 disk. In this example, raw device is / dev / rraOa.
If this disk drive had been connected to a VAX station 2000, the device entry would
have been / dev / rrdOa.
# /etc/mkfs /dev/rraOa 32768 17 15
r-r-~D,\

I,
1/

Reload the boot block. This step is necessary only if the disk is physically
damaged or replaced for some other reason. The boot block should normally be
operationall--I l)se the following command line format to reload the boot block:
'000'(

ltJ1k'S'

if=/~~f

of=device conv=sync

/<\.rz.v II--Y'~ 00
7.

Check the new file system for consistency using the following format:
/ etc/ fsck device
/r{ L
:/ Z. [) 0
e..Ci ,.

If you do not do a file consistency check at this p'6int, you will be unable to
\f A G E~?
1\
mount the file system later.
-...,.,1

~",~.
3-12 Local File Restoration

I ,$~,

\.)_{\ :~

Mount the root file system using the following format:
fete/mount device /mnt

/cl«..v / r:c;Change to the / mn t director¥ and restore the root file system using the
Ye vV'\O ve
rtr; X )
t';j'f e, j
following format: ~~'i /I)-I~C l /
ed /mnt
.
restore -rf deVice

JrV117 U

vir

r,,,

1St.

r~~i ~~bt d~cip\fn~ge

("feft) e. .

(!,

1-- ,.I

,I ',.-

(/0

'( _"'\

dei~~I~'1ape:~~vice~ ;o~ fa 'n~t ...
I,

If you are
the
to the
have the specify the f option or the device name.

10.

Unmount the file system and check it for consistency using the following
format:
ed /
/ete/umount device
/ ete/ fsek device

11.

~.~ v //y~Oq
//

Ir;:::t . ( V

Irr-?~
Fl,
-~/I ..{

Halt and reboot the system as follows:
# sync
# sync
# halt

»>b device

In the boot command line, you must specify the name of the newly restored
disk.
12.

Use the steps in Section 3.5 to restore the /usr file system.

!\ GVV16 VZ

hqc/( (Jf'

/rqif).('

I/o/()~f ,

3.7 Using the tar or mdtar Commands to Restore Files
If you backed up your files and directories with the tar and mdtar commands, you
must use these commands to do your restore. You do not need superuser privileges,
nor do you have to be in single-user mode to use these commands.

The following procedure assumes that a relative pathname was used when the
original tar image was created. To restore individual files:

Change the directory. Use the cd command to go to the home directory of the
file or files that you want to extract. For instance, if you want to extract files in
the /usr directory, type:
# cd /usr

When you are in the /usr directory, you can list the files stored on the archive
media by typing:
# tar t

If you want to display the file pathnames on a specific device, type:
# tar tf /dev/rmtlh

The tar command displays the pathnames of the files in the tar image.

Local File Restoration 3-13

1,'

Extract the files. To extract a single file or a series of named files from tape,
type:
# tar xp file! file2 file3 ...

The tar command extracts the listed files using the path name that was placed
on the tar archive media. The p option restores the named files to their
original protection codes. The p option only takes effect if you are logged in as
the superuser.
If you specify the tar command with the xp options and do not specify file
names, you will extract all of the files on the archive media by default. For
example:
# tar xp

There are many other options that you can use with the tar and mdtar commands.
For more information, refer to tar(l) and to mdtar(1) in the ULTRIX Reference
Pages.

3.8 The Labeled Tape Facility
In addition to the restore methods already described, the system contains a labeled
tape facility, 1 t f. The 1 t f command reads and writes single-volume, versions
three and four, ANSI-compatible tape volumes. This provides you with a way of
accurately transferring information between ULTRIX and non-ULTRIX systems. For
information on this command, refer to 1 t f (1) and 1 t f (5) in the ULTRIX Reference
Pages.

3-14 Local File Restoration

Remote Backup

This chapter introduces conceptual and setup infonnation needed to perfonn a
network backup and explains how to back up file systems over the network. The
chapter also explains how to use the remote opser utility.
The chapter contains the following topics:

4.1

•

The network environment

•

Perfonning remote backups

The Network Environment
This section introduces concepts that you need in order to perfonn a remote backup
in a master/slave environment and identifies hardware and system setup
considerations. This section also explains the concept of performing a network
backup either directly to a tape device or through the use of a staging area.

4.1.1

The Master/Slave Relationship
A master system is the local system that you use to initiate a remote backup. You
initiate a remote backup by running the remote op s e r utility as described in Section
4.2. You must be logged in to the master system to perfonn a remote backup. The
master system should have a tape drive and a disk drive that you can use to transfer
data from the remote system.
The slave system is the system from which the master system takes the backup data.
In some cases, a slave system may have only fixed disks and no tape drive device.
Therefore, the only way to back up the slave system files is from another system. It
can also be more efficient and economical to centralize the backup process, regardless
of the slave systems' hardware.

4.1.2

Hardware and System Setup Considerations
Before perfonning a network backup, there are several hardware and system setup
considerations to take into account. The hardware considerations described in this
section focus on the tape and disk drives that you use to perfonn a network backup.
The system setup considerations described in this section focus on the system files
that have to be in place to perfonn a network backup.

4.1.2.1

Hardware Considerations - The primary hardware consideration deals with the
type of tape drive connected to the master system. Tape drives can have either a
streaming mode capability or a start/stop mode. Although having a tape drive with a
streaming mode capability is not a requirement, this type of tape drive will effect a
faster backup process when moving a staging file to tape. All Tape Mass Storage
Control Protocol (TMSCP) tape drives are classified as streaming tape drives. Refer
to tms(4), tu(4), and ts(4) in the ULTRIX Reference Pages for infonnation on

specifying these tape devices.
The tape drives that have a streaming mode capability are:

•
•
•

TA81

•
•

TA90

•

TK50

•
•
•
•

TK70

•

TU81E

TA78
TA79
RV20

TS05 (When extended characteristics are set)
TU80
TU81

The other tape drives that you can use are classified as start/stop mode tape drives.
The start/stop mode tape drives are:
•
•

TE16
TS05 (When extended characteristics are not set)

•

TS11

•

TU77

•

TU78

•
TU79
If you have a start/stop mode tape drive, use the direct backup method described in
Section 4.2.5.2.
The other hardware consideration concerns the disk drives in your system
configuration. To perform a remote backup using a staging area as discussed in
Section 4.1.3, you must ensure that you have a partition on the master system disk
that is large enough to receive all of the backup data. You must know the partition
sizes of the disks connected to your system and how much free space they contain.
The free disk space on the master system disk must be at least 12 percent greater than
the amount of slave system backup data. This is because the software that performs
the network backup adds an additional 12 percent of overhead to the data being
transferred.

4.1.2.2

System Setup Considerations - There are two system files that you must set up
on both the master and slave systems before you can perform a network backup.
These system files are:
•

The / etc/hosts file

•

The /. rhosts file in the rootdirectory

The / etc/hosts file on the master system must specify all of the slave systems on
which you intend to perform remote backups. As described in the ULTRIX Guide to
Networking, each entry in the master system / etc/hosts file must contain the

4-2 Remote Backup

slave system's Internet address, the official host name, and any aliases for that
system. Additionally, the / etc/host s file on the slave systems must contain an
entry for the master system. Refer to the ULTRIX Guide to Networking and to
hosts(5) in the ULTRIX Reference Pages for more information on the format of
this file.
The / . rhosts file in the root directory on the master system must list the slave
systems on which you intend to perform remote backups. The / . rhosts files on
the slave systems must specify the official host name of the master system. Note that
the / . rhosts file on the master system specifies access to the root directory of
all the slave systems, while the / . rhosts file on the slave system specifies access
to the root directory of the master system. This enables the master system to
perform superuser activities on the slave systems during the remote backup process.
If you try to perform a network backup without this information in the / . rhosts
files, the system responds with the message:
Permission denied

If either of these two files is set up incorrectly, you will be unable to perform a slave
system backup using the remote opser capability described in Section 4.2.

4.1.3

Staging Area or Direct Backup Determination
The remote opser utility described in Section 4.2 provides you with two methods of
performing a network backup:
•

Staging area backup

•

Direct backup

A staging area is designated space in a file system on the master system that has
enough free disk space to receive the slave system's backup data. The staging area
backup method requires you to back up the slave system's data to the master
system's disk and then to transfer the data to a tape device. Transferring the backup
data to tape from the master system's disk enables you to take advantage of the
streaming mode c~pability of the tape drive.
In the streaming mode, a tape drive continuously writes data to tape and does not
stop tape movement until all of the data has been written. To support the streaming
mode, data must be transferred at a rate at least equal to the tape's write transfer rate.
Yet, because data cannot be transmitted over a network at this rate, it must first be
transmitted to a staging area on the master system's disk drive and then copied to the
tape drive. Transferring the staging area file to a tape drive in this way takes
advantage of this type of drive's streaming mode capability. The disadvantage is that
the file restoration process has the added step of having to transfer the dump image
from the tape back to the staging area on the disk.
The direct backup method allows you to back up directly to a tape or disk drive. Use
the direct backup method if you have a start/stop mode tape drive. The direct backup
method has a simpler file restoration process, because you can restore directly from
the tape or disk without having to transfer the dump image to a staging area.

4.2 Performing Remote Backups
Remote backups are performed by using the opser utility and selecting the n
(network) option from the opser menu.

Remote Backup 4-3

To run the opser utility, log in to the operator account (operator login name).
When you log in to the operator account, the system automatically invokes the
opser utility in place of the shell. Once running, the opser utility prints a utility
sign-on message, two infonnational messages, and a command options prompt:
ULTRIX-32 Operator Services
Line editing: delete - erase one character, AU - kill entire line
For help, type h followed by a return
opr>

The opr> prompt infonns you that the opser utility is running in place of the shell
and is ready to accept opser command options.
To use an opser option, type the appropriate option letter or name and press the
RETURN key. For example, to display on-line help about all opser options, press
the RETURN key without a command option letter, or type h and press the
RETURN key:
opr> h
() - may use first letter in place of full name
Valid commands for Local Opser are:
!sh
(u) sers
(s) hutdown
(d)ismount
(f)sck
(r)estart
(h)elp
(b) ackup
halt
(n)etwork [Slave]
(q)uit

opr>

To end an opser session and return to a login prompt, type q and press the
RETURN key.
Except for the n (network) option, all of the local op s e r menu selections are
discussed in Chapter 2. Refer to that chapter for infonnation on other menu
selections.
The remainder of this section discusses the functions and capabilities of the remote
op s e r netw<?rk option.
To enable remote opser, enter n at the opr> prompt. You must also know the
name of the slave system on which you want the remote opser running. After you
enter n, the opser utility prompts you to enter the name of the slave system;
however, it only allows you to enter slave system names that are stored in the master
system's jete/hosts file.
The following example shows how opser responds when you enter an invalid slave
system name.

4-4 Remote Backup

opr> n
Enter Slave System name: porton
Slave system name not found in /etc/hosts file

The following example shows how opser responds when you enter a valid slave
system name.
opr> n
Enter Slave System name: proton
Network:
slave
proton.
alvin
ttypO
Jun 15 09:24
(nucleon)
simon
tty01
Jun 15 08:30
teddy
tty02
Jun 15 08:33
Shutdown at 09:27
(in 3 minutes)
proton_opr>

This example shows that after you enter a valid slave system name, the opser
utility displays the name of the slave system followed by a list of users currently
logged in to that system. The opr> prompt becomes proton_opr>, which tells
you that you are now running opser on the slave system named proton. You can
also enter the slave system name and the n on the opser command line and achieve
the same results. For example:
opr> n proton
Network:
slave
alvin
ttypO
simon
tty01
teddy
tty02
Shutdown at 09:27
proton_opr>

proton.
Jun 15 09:24
(nucleon)
Jun 15 08:30
Jun 15 08:33
(in 3 minutes)

Regardless of how you invoke the remote op s e r, it perfonns a shutdown of the
slave system in three minutes; all users on the slave system must be logged out.
When the shutdown completes, the slave system is in single-user mode and
connected to the network. The slave system is then able to receive opser
commands across the network from the master system.

If you press the RETURN key, or enter an h for help at the proton _ opr> prompt,
the opser utility displays the remote opser menu. For example:

() - may use first letter in place of full name
Valid commands for Remote Opser are:
!sh
lsh
(d)ismount
(f)sck
(r)estart
(h)elp
(b)ackup
halt
(q)uit

- Slave System shell escape (execute ULTRIX-32 commands)
- Master System shell escape (execute ULTRIX-32 commands)
(Type control d to return from shell)
- unmount file systems
- file system checks
- restart time-sharing
- print this help message
- file system backup
- halt processor
- exit Remote Opser (do not restart Slave System)

The following sections explain how each of these op s e r options work. Read these
sections before trying to use them.

Remote Backup 4-5

4.2.1

Escaping to the Slave Shell
To escape to the slave shell, enter! sh at the slave system opr> prompt. The
opser utility prompts you to enter a password. You must enter the root password
for the slave system. For example:
proton_opr>!sh
Password:
type ~D to return to opser
#

After you enter the slave's root password, the system places you in the slave's shell
and displays the shell prompt. You can now execute any ULTRIX command,
excluding editors. This option is useful for doing manual file system checks with the
fsck command, displaying mounted file systems with the df command, or
unmounting file systems with the umount command.
To return to the remote op s e r, type CTRL / D at the shell prompt.

4.2.2

Escaping to the Master Shell
To escape to the master shell, type ish at the slave system opr> prompt. The
opser utility prompts you to enter a password. You must type the root password
for the master system. For example:
proton_opr>lsh
Password:
type ~D to return to opser
#

After you enter the master's root password, the system places you in the master's
shell and displays the shell prompt. You can now execute any ULTRIX command.
This option is useful for using the tal k command to inform users that you intend to
shut down the system. It is also useful for ensuring that the staging area that you are
going to use is large enough to perform a staging area backup with the df command.
This option is also used for invoking script files before and after performing a staging
area backup. As described in Section 4.2.5.1, script files can be used to clear the
staging area of unnecessary files or to transfer staging area files to tape.
To return to the remote op s e r, type CTRL / D at the shell prompt.

4.2.3

Unmounting the Slave's File Systems
To unmount the file systems on the slave system, type d at the slave system opr>
prompt. For example:

When you use the d option, the opser utility invokes the umount command to
unmount file systems listed in the slave system's / etc/ fstab file. The umount
command is described in mount(8) of the ULTRIX Reference Pages.
To ensure that you have unmounted all of the file systems, use the ! s h option as
described in Section 4.2.1 and enter the mount command. Without options, the
mount command displays any file systems that are currently mounted. If any file
systems are still mounted, unmount them by issuing the appropriate umount
command.
The following example shows the sequence of options and commands that you would
enter, assuming that the /usr / staff3 file system is still mounted after you enter
4-6 Remote Backup

the d option:
proton_opr> d
proton_opr> !sh
Password:
type AD to return to opser
# tete/mount
/dev/rzOa on / type ufs
/dev/rzlh on /usr/staff3 type ufs
# /ete/umount
/dev/rzlh
# <CTRL/D>
proton_opr>

Note
When you unmount the / u s r file system, some ULTRIX commands
may be unavailable.
The sequence of commands shows the shell escape, ! s h, which causes the system to
prompt you for the slave system root password, after which it displays the # shell
prompt. Next, the sequence shows the system response to the mount command. The
file system / u s r / s t a f f 3 is unmounted by unmounting / dev / r z 1 h. This is the
device, r z 1, and partition, h, on which the / u s r / s t a f f 3 file system is mounted.
Lastly, the sequence shows the how typing CTRL/D returns to the proton_apr>
prompt.

4.2.4

File System Consistency Check on the Slave
Before you back up any file systems, you should check them for consistency. To
check file systems for consistency, enter f at the slave system opr> prompt. The
opser utility invokes the fsck command at the slave system to check the file
systems listed in the slave's /etc/fstab file. If the fsck command finds any file
system inconsistencies, it displays a message informing you that you must run f s ck
manually.
To run fsck manually, escape to the slave shell and issue the appropriate fsck
command.
If fsck fails on the slave's root file system, you must:

Quit network opser by typing q at the slave system opr> prompt. As
described in Section 4.2.8, the slave system is shutdown and the system is
placed in single-user mode.

Run fsck on the slave system. Use the following format:
/ etc/ fsck device

On the slave system, halt the processor so that you can return to the console
subsystem. This allows you to reboot the system and ensures that the system
buffer cache does not overwrite the fsck corrections. The hardware
documentation for your processor describes how to return to the console
subsystem.

You must then reboot the slave system to get the corrected root file system
into memory.

Reestablish network opser by typing n at the local opser menu at the master
system.

Remote Backup 4-7

Caution
Whenever fsck corrects root file system inconsistencies, either
through the f menu option or manual f s ck, you must halt and reboot
the slave system. See the hardware documentation for your processor for
information on how to halt your processor.
For further details, see the information on maintaining file system consistency in the
ULTRIX Guide to System Crash Recovery.
Example 4-1 shows a sequence of commands with a file system inconsistency
encountered on the I dev I r z 0 h file system.

Example 4-1: Performing a File System Consistency Check

/dev/rzOa: 489 files,
/dev/rzOd: 306 files,

6050 used, 631 free (93 frags, 160 blocks)
43542 used, 51325 free (122 frags, 327 blocks)

/dev/rzOh: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY.
/dev/rzOf: 155 files, 41487 used, 147657 free (98 frags,

126 blocks)

proton_opr> !sh
Password:
type AD to return to opser
# fsek /dev/rzOh
** Last mounted on /usr/users
** Phase 1 - Check Blocks and Sizes
** Phase 2 - Check Pathnames
** Phase 3 - Check Connectivity
** Phase 4 - Check Reference Counts
** Phase 5 - Check Cyl groups
/dev/rzOh: 3427 files, 95983 used, 14774 free (184 frags, 3362 blocks)
# tete/mount -a
# df
Filesystem
total
kbytes kbytes percent
node'
kbytes
used
free
used
Mounted on
/dev/rzOa
7423
6050
631
91%
/
/usr
/dev/rzOd
105407
43542
51325
46%
/usr/users
/dev/rzOh
123063
95983
14774
87%
/dev/rzOf
210159
41487 147657
22%
/usr/server
# /ete/umount -a
# <CTRL/D>
proton_opr>

Example 4-1 shows that the fsck command was executed on the raw device,
I dev / r z 0 h, after receiving the following message:
UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY.

The mount command was then invoked with the -a option to mount all the file
systems in the / etc/ f stab file so that the df command could be invoked. The df
command was issued to display and confirm the contents of the file systems. The
umount command with a -a option was issued to dismount the file systems so that
a network backup of the file systems could be performed without having to use the
remote opser dismount option.

4-8 Remote Backup

Refer to f s ek(8), mount(8), and df(8) in the ULTRIX Reference Pages for more
information.

4.2.5

Backing Up the Slave's File Systems
There are two methods that you can use to perform a remote backup of file systems.
One is the staging area backup method and the other is the direct backup method.
You will usually use the staging area backup method if you have a streaming mode
tape drive. Use the direct backup method if you have a start/stop mode tape drive.

Note
You must unmount and check all file systems prior to backing them up.
Use the op s e r d option to unmount the file systems; use the op s e r f
option to check them for consistency.
The following two sections explain how to use the two backup methods. The
sections contain complete examples of how the opser utility responds when using
either method.

4.2.5.1

Staging Area Backup Method - The staging area backup method requires you to
back up the slave system data to a designated file system on the master system's
disk, and then to perform a tape transfer of the backed-up data.
Before performing a staging area backup, you must know:
•

The backup level, which is a number from 0 to 9. Use the number 9 for a daily
backup, the number 5 for a weekly backup, and the number 0 for a monthly or
full backup.

•

The file system that you will be using for the staging area on the master system,
for example, I staging.

•

The name of the file system that you want to dump to the staging area, for
example, I.

•

The amount of slave system data that you want to dump to the master system's
staging area and the amount of free disk space in the staging area. The amount
of free disk space must be at least 12 percent greater than the amount of slave
system data that you want to dump.

When the backup is complete, you must use the tar command to transfer the staging
area file to tape, or the md tar command to transfer the staging area file to another
disk. In either case, you should remove the staging area file from the staging area
once you have transferred it to a removable media.
To use either tar or mdtar without exiting opser, you must escape to the
master's shell using the 1 s h option as described in Section 4.2.2. If an operator is
performing the staging area backup and does not have access to the superuser
password, you must create some script files to perform the tape or disk transfer and
to remove the staging area file from the staging area. The script files must be located
in the lopr directory and have root ownership.
For example, to perform a tape transfer of the staging area file, a script file named
savei t would have a one-line entry:
tar -cv /staging/*

Remote Backup 4-9

Similarly, to remove the staging area file from the j staging directory, a script file
named rernovei t would have a one-line entry:
rm -rf /staging/*

Having script files available to operators who do not have access to the superuser
password keeps them from having to exit the opser utility to perform the tape or
disk transfers or to remove files from the staging area.
Example 4-2 shows a backup sequence of the root (j) file system from the slave
system proton to the master system named nucleon. The sequence shows that
before performing the backup, the rernovi t script file was invoked to clear the
j staging directory of any files. The example also shows that the root file system
was transferred to the master's j staging directory. When the name of the file
system is entered, the opser utility determines if there is enough disk space in the
specified staging area to back up the slave file system. If there is enough room, the
backup process begins. If there is not enough room, the system sends a beep signal
to your terminal and displays the message:
Couldn't allocate ##### bytes.
proton_opr>

The number signs (#) represent the number of bytes required to perform the backup.
When the opser utility has determined that there is enough disk space in the staging
area, it starts the dump sequence. The opser utility assigns a file name to the data
being transferred to the staging area. The naming convention that it uses for the file
created on the master system is:
slave system name..filesystem name.time stamp

The opser utility also changes any slashes (/) in the file system name to tildes (.... ).
In Example 4-2, the file name is:
proton.-.MON-Jun-24-10:32:26-1989

The dump image file is located in the master system nucleon's j staging
directory. In the example, the staging area file name breaks down as the root file
system of the slave system proton, which was backed up on Monday, June 24,
1989 at 10:32 am. The tilde sign (.... ) equals the root (I) file system.

Example 4-2: Staging Area Backup Method Display
proton_opr> lsh removeit
proton_opr> b
Enter backup level (0 - 9): 9
Use staging file or go directly to device(s = staging/d
Enter directory to place staging file: /staging
Enter name of filesystem to dump: /

direct)? s

Sizing file system to determine number of bytes to pre-allocate.
DUMP: Estimated 68608 bytes output to file named
nucleon:/staging/proton.-.Mon-Jun-24-10:32:26-1989
Allocating 68608 bytes.
DUMP: Date of this level 9 dump: Mon Jun 24 10:32:23 1989
DUMP: Date of last level 0 dump: Wed Jun 12 09:19:58 1989
DUMP: Dumping /dev/rraOa (/) to
/staging/proton.-.Mon-Jun-24-10:32:26-1989 on host nucleon
DUMP: Mapping (Pass I) [regular files]
DUMP: Mapping (Pass I I) [directories]

4-10 Remote Backup

Example 4-2:

(continued)

DUMP: Estimated 68608 bytes output to file named
/staging/proton.-.Mon-Jun-24-10:32:26-1989
DUMP: Dumping (Pass III) [directories]
DUMP: Dumping (Pass IV) [regular files]
DUMP: 58368 bytes were dumped to file
/staging/proton.-.Mon-Jun-24-10:32:26-1989
DUMP: level 9 dump on Mon Jun 24 10:32:23 1989
DUMP: Dump is done
proton_opr>lsh saveit

Example 4-2 shows that when the dump is complete, the file named
proton. '" . Mon -Jun - 24 -1 0 : 32 : 26-1989 is transferred to tape by using the
script file save it.

4.2.5.2

Direct Backup Method - The direct backup method enables you to back up a
slave's file systems directly to a master system's tape drive or disk drive if your
system configuration has the capacity to designate a disk drive for the purpose of
backing up the slave systems.
Whether you are backing up to a tape or a disk drive, you must know the following
infonnation before you can use the direct backup method:
•

The backup level number, which is a number from 0 to 9. Use the number 9
for a daily backup, the number 5 for a weekly backup, and the number 0 for a
monthly or full backup.

•

The device name where you want the slave system's backup data to go, for
example, / dev / rmt Oh.

•

The name of the slave's file system that you want to dump.

You must also ensure that you have enough of the chosen media (tapes, disks, or
diskettes) on hand when the backup process begins. Depending on the size of the
backup, you may have multiple backup volumes.

4.2.6

Returning to the Local opser
To return to the local opser main menu, enter r at the slave system opr> prompt.
For example:
This causes the remote opser running on the slave system to restart multiuser
mode, disable the network connection to the slave system, and return you to the local
op s e r that you initiated at the master system.

Caution
Do not use the q command to end a remote op s e r session. The q
command stops remote execution of the opser utility and should only
be used in the event of an emergency by someone with superuser
privileges. Issuing the r command is the only way to end a remote
opser session properly.

Remote Backup 4-11

4.2.7

Halting the Slave Processor
To halt the slave system processor, enter hal t at the slave system apr> prompt.
This command causes the remote ap s e r to synchronize the slave system disks
(which means that in-memory file system data is written to the disks), then halt the
slave system processor.
When you enter hal t, the remote apser checks to make sure that you want to halt
the slave system processor:
proton_opr>halt
This will leave the Slave System halted. You must then manually reboot
the Slave System.
Do you really want to do this ?(Enter y/n)y
Leaving network mode.
opr>

When you enter a y, the remote apser utility exits remote apser execution and
leaves the slave system halted. Consequently, someone at the slave system has to
reboot it manually. The master system then returns to local execution of the apser
utility, as indicated by the apr> prompt, and the network connection that was
established automatically disconnects. If you enter an n, the remote apser utility
returns the slave system apr> prompt.

4.2.8

Stopping Remote Execution of opser
In an emergency, it may be necessary to stop running the apser utility at the slave
system. To stop remote execution of apser, enter q at the slave system apr>
prompt. For example:
proton_opr> q
This will leave the Slave System shutdown in single user mode.
Do you really want to do this? (Enter y/n)y
Leaving network mode.
opr>

The preceding example shows that, when you enter q, the remote apser checks to
make sure that you want to quit the remote ap s e r session. When you enter a y, the
remote apser utility exits remote apser execution and leaves the slave system in
single-user mode. Consequently, someone at the slave system has to restart multiuser
mode. The master system then returns to local execution of the apser utility, as
indicated by the apr> prompt, and the network connection that was established
automatically disconnects. If you enter an n, the remote ap s e r .utility returns the
slave system apr> prompt.

4.2.9

Exceptional Condition Handling
You can abort either a remote apser process (such as rdump) on the slave or a
local apser process (such as dump) on the master. To abort any processes running
under the local apser on the master system, type'" CTRL/C. Typing'" CTRL/C
causes the apser utility to return the apr> prompt. Type CTRL/C to abort any
process running under the remote apser on the slave system.

4-12 Remote Backup

Remote File Restoration

You can restore files, directories, and file systems over the network with the
rrestore program. This chapter describes how to:

5.1

•

Restore files, directories, or file systems on different media using the
rrestore program

•

Restore the root and /usr file systems after a catastrophic event has occurred

Using the rrestore Program
You use the rrestore program to perform a remote file restoration, either
interactively or noninteractively. Subsequent sections of this chapter explain both
methods, with examples of how to perform each.

5.1.1

Network Considerations
Although the rrestore program is similar in function to the restore program,
certain rules apply when working across a network:
•

You must make a distinction between the master and slave systems, as
described in Chapter 4. Typically, the dump media is located on the host or
master system. You invoke the rrestore program from the slave system and
specify the host name where the dump media resides.

You do not have to be the superuser to restore files remotely. However, to
restore a complete file system properly, you must be logged in as the superuser.

•

It is important to remember where the file system that you want to restore was
originally backed up.

•

If the file' system was backed up to a staging area file and is still intact, then
you must specify the staging area file name to the rrestore program.

•

If the staging area file was moved to tape using the tar command, then you
must first use tar to move the staging area file from the tape to the staging
area and then remote restore the file system from the staging area file. See
Chapter 4 for a description of the staging area.

•

Whatever procedure that you used to back up a file system must be repeated in
reverse order.

•

You can only specify a device to the rrestore command when the files that
you want to restore were originally backed up using the direct backup method.
See Chapter 4 for a description of backup methods.

5.2 Preparing for an Interactive Restore
To prepare for a remote restore operation, you must do the following:
•

•

Mount the first dump tape on the appropriate device.

•

Transfer the backed up file from tape to a staging area using the tar command.

The following sections describe each of these tasks.

5.2.1

Logging In
Some rrestore functions require superuser privileges. For example, if you intend
to perform an interactive file restoration of files that you do not own, you must be
superuser. Similarly, if you intend to change file and directory attributes when
restoring, you must be superuser. See rrestore(8) in the ULTRIX Reference
Pages for details on the command options and required permissions.

5.2.2

Mounting the Dump Media
Before invoking the rrestore program, ensure that the backup media containing
the first dump tape is mounted on the appropriate device. This is normally the
default tape device, / dev / rmt Oh, but it could be any device or file system in the
form of an on-line backup file. The first tape is critical, because it contains directory
and inode information about the contents of the dump image.

5.2.3 Transferring Files to the Staging Area
If the remote backup was to a staging area file and then was transferred to tape using
the tar command, you must first ensure that the file is transferred back to the
staging area from the tape by using the tar command with the -x option. For
example, if the file name is proton. '" . Mon -Jun -2'4 -1 0: 32 : 2 6-198 9, you
would type the following:
# cd /staging
# tar -x proton.-.Mon-Jun-24-10:32:26-1989

The proton. '" .Mon-Jun-24-10: 32: 26-1989 file name is the staging area file
name assigned by the rdump command, as described in Chapter 4. When the
staging area file has been transferred from tape to the staging area on the master
system, you can invoke the rrestore command.

5.3 Restoring Files Interactively
Once you are logged in and the dump media is mounted on the appropriate device,
use the following command format:
/ etc/ rrestore options host.·device

The - i option specifies interactive mode. The - f option followed by the host:device
string specifies the host and host device from which to obtain the data. For example,
to restore files that were backed up to the master system nucleon using the direct
backup method, you would enter:
# /etc/rrestore -if nucleon:/dev/rmtOh

5-2 Remote File Restoration

To restore a file that you transferred back to the staging area from tape with the tar
command, enter a command line similar to the following:
# /etc/rrestore -if nucleon:/staging/proton.~.Mon-Jun-24-10:32:26-1989

5.3.1

Using Interactive Mode Commands
When you enter the /etc/rrestore command with the -i option, the
rrestore > prompt appears on your screen. The prompt signals that the program
is waiting for a command.

5.3.1.1

Listing the Commands - To get a listing of the available commands, enter an h
or a? at the rrestore > prompt. For example:
rrestore >h
Available commands are:
ls [arg] - list directory
cd arg - change directory
pwd - print current directory
add [arg] - add 'arg' to list of files to be extracted
delete [arg] - delete 'arg' from list of files to be extracted
extract - extract requested files
quit - immediately exit program
verbose - toggle verbose flag (useful with "ls")
help or '?' - print this list
If no 'arg' is supplied, the current directory is used
rrestore >

5.3.1.2

Listing the Pathname of the restore Working Directory - To list the
pathname of the working directory within the rrestore environment, type pwd at
the rrestore > prompt. For example:
rrestore >pwd
/lib
rrestore >

5.3.1.3

Getting a File and Directory Listing - To get a listing of the files and directories
that can be restored at the current directory level, type Is at the rrestore >
prompt. For example:
rrestore >ls
adm/
bini
dict
rrestore >

hosts/
include/
kits

man
mdec/
msgs/

skel/
spool/
src/

To get a listing of the files and directories that can be restored at a different directory
level, enter the pathname of that directory as an argument to the 1 s command.

5.3.1.4

Comparing Inode Numbers - Inode numbers increase with each tape. The
largest inode numbers are on the last tape and the smallest numbers are on the first
tape. By comparing the displayed inode numbers with the starting inode numbers
recorded on a multivolume dump set, you can find the tape or disk that contains the
file you want to restore. To see the inode numbers of the files contained within your
listing, first use the verbose command to enter verbose mode, then use the Is

Remote File Restoration 5-3

command to get your listing. The rrestore program displays the files with the
inode numbers to the left of the dumped file names. For example:
rrestore >verbose
verbose mode on
rrestore >1s
11971
2 *.1
2177
2
* .. 1
11969 adml
11427
1633 bini
2722
rrestore >

hosts
includel
mdecl
msgsl

14253
4145
4253
3266

spool I
srcl
staff
sysl

In this example, the hosts file has an inode number of 11971 and the spool
directory has an inode number of 14253.

Note
Sometimes the in odes for a given file span multiple tapes or disks.
When this happens, you will need more than one tape or disk to restore
the file.
After checking the inode numbers, you can tum off verbose mode. To do this, type
verbose one more time. For example:
rrestore >verbose
verbose mode off
rrestore >

5.3.1.5

Changing Directories - Change directories in the restore environment with the cd
command. To specify a particular directory, enter the cd command, with the
directory pathname given as an argument, to the command. For example:
rrestore >cd /1ib
rrestore >

5.3.2

Specifying What You Want Restored from the Dump Image
When working interactively with the rrestore program, you must first decide
which files or file systems you want restored from the dump media. Use the Is
command to get a file or directory listing of the dump media. Once you know what
is on the media, you can specify what you want restored.
You specify what you want restored by creating and modifying an extraction list.
Creating an extraction list involves working from the dump image and using the
rrestore program's add command. Modifying an extraction list involves
working from the dump image and using the rrestore program's add and
delete commands.

5.3.2.1

Creating and Modifying the Extraction List - To create and modify an
extraction list, you must be within the rrestore environment. When you have
invoked the rrestore program, follow these steps:
1.

Use the pwd command to determine your position within the dump media
directory structure.

Use the 1 s command to identify which files and directories are on the dump
image at the current level.

5-4 Remote File Restoration

Decide which files or directories you want to extract from the dump image.

Use the cd command to move around the dump image and confirm the location
of files and directories.

Position yourself in the dump media directory structure at a point that is one
level above the desired item. For example, to restore / u s r /1 ib, position
yourself at /usr. If you attempt to identify /usr / lib while positioned
within /usr / lib, the program issues an error message.

Create the extraction list by typing the add command and the name of each
desired file or directory as an argument to the command.
For instance, if you are in the / u s r directory and you want to extract the files
lpf and lpd and the subdirectory refer, enter the add command followed
by the file and directory names:
rrestore >add 1pd 1pf refer
rrestore >

Use the ls command to see how the program has marked these files. For
example:
rrestore >ls
diffh
dnet/
howyix
rrestore >

libcy.a
libcg.a
libpcy.a

*lpd
*lpf
*refer/

Notice the asterisks (*) in front of the items that you added to your extraction
list. When you view the directory listing, the asterisks point out what you put
on the list. Similarly, when the rrestore program reads an asterisk during
the extraction process, it transfers a copy of that file or directory from the dump
image to your current working directory.
8.

Use the delete command with the file name given as an argument to remove
a particular file from the list. For example:
rrestore >de1ete 1pf
rrestore >

Use the ls command to see what the program did in response to the delete
command. For example:
rrestore >ls
diffh
dnet/
howyix
rrestore >

libc_p.a
libcg.a
libpcy.a

*lpd
lpf
*refer/

Remote File Restoration 5-5

The asterisk has been removed from the lpf file. This indicates that the
program removed the 1 p f file from the list and will not extract it from the
dump image.
10.

5.3.3

Use the extract command to request a transfer of files from the extraction
list to your system.

Restoring the Specified Files and Directories
The rrestore program's extract command initiates the actual restore. When
you enter the extract command, the program transfers a copy of each marked item
to the current working directory using the relative pathname found on the dump
media.

Note
The current working directory is the one from which you invoked the
rrestore program initially. Your position within the dump media
directory structure when you enter the extract command has no
influence on the final location of the restored files.
For example, assume that the file /usr / lib/ lpf was dumped to tape with /usr
as the mount point of the backup. In this case, . / lib/ lpf is the relative
pathname on the dump media. If you invoked the rrestore program from /usr
(in other words, with /usr as your current working directory), the program restores
/ lib/ lpf as subordinate to /usr (/usr / lib/ lpf). If you invoked the
rrestore program from / lib (in other words, with / lib as your current
working directory), the program restores / lib/ lpf as subordinate to / lib
(/ lib/ lib/ lp£).

5.3.4

Entering the extract Command
To extract the files that you placed on your extraction list, enter the extract
command at the rrestore > prompt. For example:
rrestore >extract

In response, the rrestore program displays several statements and prompts you
for some answers. For example:
You have not read any tapes yet.
Unless you know which volume your file(s) are on you
should start with the last volume and work forward
towards the first.
Specify next volume #:
set owner/mode for' .'? [yn]

5.3.4.1

Responding to the 'Specify next volume #' Prompt - The first prompt,
Specify next volume #:, assumes that you are working from a multivolume backup
set.

•

5-6 Remote File Restoration

initial prompt. The program returns a second prompt to which you respond by
entering 2, and so on.
•

If you are working from a single-volume backup tape, enter 1 in response to the

initial prompt.
In response to your entry, the rrestore program searches the dump image for the
items listed on the extraction list. When it finds the individual files and directories, it
restores them.

5.3.4.2

Responding to the 'Set owner/mode for'." Prompt - The second prompt, Set
owner/mode for'.', lets the superuser change the attributes of the current directory
(shown as '.' within the prompt) to the attributes found for that entry on the dump
image. The current directory's attributes include access permissions, ownership, and
the accessed and update times.
•

If you want to change the attributes, type y.

The rrestore program responds by changing the current directory's attributes
to those found on the dump image for the corresponding directory. It then
redisplays the rrestore > prompt.
If you are not the superuser and you type y in response to the prompt, the

program responds with an error message and does not change the attributes.
•

If you do not want to change the attributes or if you do not have superuser

privileges, type n.
The rrestore program leaves the attributes of the current directory
unchanged. It then redisplays the rrestore > prompt.

5.3.5

Exiting from the Interactive Restore
When you finish restoring files, type qu i t to get back to the system prompt:
rrestore >quit
#

5.4 Restoring Files Noninteractively
Noninteractive restoration of data is accomplished with a command line entry. The
files or directories to be restored are entered as part of the command line. Before
issuing the rrestore command in this way, ensure that you are in the correct
directory and that you have the correct media mounted.
The following example shows you how to restore three files from the master system
named nucleon. The example shows that the files named in the command line are
restored from the dump tape mounted on device / dev / rmt Oh on the master system
nucleon.
#/etc/rrestore -xf nucleon:/dev/rmtOh filel file2 file3
#

Remote File Restoration 5-7

When the restoration process is complete, the system redisplays the system prompt
and the specified files are in the current working directory. Like the interactive
restoration method, you can specify directories as well as files. The rrestore
command will restore all of the files and subdirectories beneath the specified
directory.
In addition to file names, you can specify file systems. For example:
#/etc/rrestore -rf nucleon:/dev/rmtOh
#

In this example, the rrestore command restores all files on the dump tape to the
current working directory. This shows a file system restore, which is normally
performed after a new file system is created and mounted, and a change directory is
executed.

5-8 Remote File Restoration

Device Mnemonics

This appendix identifies and defines the mnemonics that are used to attach any
hardware or software device to your system. The mnemonics are used by the
/ dev /MAKEDEV shell script to create the character or block special files that
represent each of the devices. The mnemonics also appear in the system
configuration file as described in ULTRIX Guide to Configuration File Maintenance.
Table A-I lists the mnemonics in nine categories: generic, systems, consoles, disks,
tapes, terminals, modems, printers, and others. The generic category lists the
mnemonics of a general nature and includes memory, null, trace, and tty devices.
The systems category lists the mnemonic for the DEC station 3100 system setup. The
consoles category lists the system console devices that the ULTRIX operating system
uses. The disks, tapes, terminals, modems, and printers categories identify the
appropriate mnemonics for those devices. The others category lists the mnemonic for
DEC station 3100 devices.
The description heading in Table A-I identifies the corresponding device name. It
does not define the mnemonic's use. For detailed information on the use of each
mnemonic in relation to both the MAKEDEV script and the system configuration file,
refer to the reference pages in Section 4 of the ULTRIX Reference Pages. If on-line
reference pages are available, you can also use the man command. For instance,
enter the following command at the system prompt to display the reference page for
the Mass Storage Control Protocol (MSCP) disk controller driver:
% man ra

Where appropriate, the SYNTAX section of the reference page defines the device's
syntax as it should appear in the config file. Refer to / dev /MAKEDEV for
additional software device mnemonics that MAKEDEV uses. Refer to MAKEDEV(8) in
the ULTRIX Reference Pages for a description of the MAKEDEV utility.
Table A-I uses the convention of an asterisk (*) beside a mnemonic and a question
mark (?) beside a device name to mean a variable number. The value of the variable
number is dependent on the particular device.

Table A-1: Devices Supported by MAKEDEV
Category

Mnemonic

Description

Generic

boot*
mvax*
vaxstation*
std
drum
errlog
audit
kUmem
kmem
mem
null
trace
tty
local

Boot and std devices by cpu number; for example, boot750
All MicroV AX setups; for example, mvax2000
A VAXstation 2000 setup; for example, vaxstation2000
Standard devices with all console subsystems:
Kernel drum device
Error log device
Audit log device
Kernel Unibus/Q-bus virtual memory
Virtual main memory
Physical memory
A null device
A trace device
A character terminal device
Customer-specific devices

Systems

DECstation

A DEC station 3100 setup

Consoles

console
crl
cs*
ctu*
cty*
cfl
ttycp

System console interface
Console RL02 disk interface for V AX 86?0
Console RX50 floppy interface for VAX 8??0
Console TU58 cassette interface for VAX 11/725/730/750
Console extra serial line units for VAX 8??0
Console RXOI floppy interface for 11/78?
Console line used as auxiliary terminal port

Disks

hp*
ra*
rb*

MASSBUS disk interface for RM?? drives and RP?? devices
UNIBUS/Q-bus/BI/HSC/DSSI MSCP disk controller interface
UNIBUS IDC RL02 disk controller interface
for RB?? drives
V AXstation 2000 and MicroVAX 2000 RD type drives
SCSI disks (RZ22/RZ23/RZ55/RRD40)
UNIBUS RK?? disk controller interface
UNIBUS/Q-bus RL?? disk controller interface
VAX station 2000 and MicroVAX 2000 RX type drives

rd*
rz
rk*
rl*
rx*
Tapes

Terminals

A-2 Device Mnemonics

tz*

TU78 MASSBUS magtape interface
UNIBUS/Q-bus/BI/HSC/DSSI TMSCP tape controller interface
UNIBUS/Q-bus/BI TMSCP optical disk
UNIBUS/Q-bus TS II/TS05/TU80 magtape interface
TEI6/TU45/TU77 MASSBUS magtape interface
V AXstation 2000 and Micro V AX 2000 TZK50
cartridge tape
SCSI tapes (TZ30/TZK50)

cxa*
cxb*
cxy*
dfa*
dhq*
dhu*

Q-bus cxa16
Q-bus cxb16
Q-bus cxt08
Q-bus DFAOI comm multiplexer
Q-bus DHQll comm multiplexer
UNIBUS DHUll comm multiplexer

mu*
tms*
rv*
ts*
tu*
st*

Table A-1:

(continued)