Chapter 7 - EBR and System Maintenance
This chapter explains procedures you should follow when using EBR to recover system files, as well as how to prepare for and recover from a major disk failure.
The first section, "Using EBR from the Server Console Menus," contains the procedures to use if you do not have UNIX Access rights, or if you have access to UNIX but do not feel comfortable working with the UNIX file system. The second section, "Using EBR from a UNIX Shell," contains the recovery procedures you can use with UNIX Access.
Using EBR from the Server Console Menus
All of the operations described in this section are accomplished by using the Enterprise Backup and Restore menu options. The Enterprise Backup/Restore menu is accessible from the Backup/Restore menu.
Important: If you have configured UNIX access on your server and feel comfortable working with the UNIX file system, you may follow the alternate Disaster Recovery procedure outlined in "Using EBR from a UNIX Shell."
The EBR service includes two major directories:
![]()
The service' s home directory, /disk1/banyan/ebr, which includes all of the service executables. ![]()
The service' s data directory, which holds the configuration data, logs, and servers running both the EBR agent and services software, indexes. This defaults to /disk1/EBR, but this default may be changed when the service is created. The directory will always be created as /diskn/EBR, where /diskn is the one specified in the Optional Parameters list when the service is created.
Every EBR data directory includes a res (resource) directory, which contains configuration information. An EBR service' s data directory also includes an index directory that contains one directory of index data for each client supported by the service. These directories are named with the client' s name.
Using EBR to Recover from Server Disk Failures
All system administrators need to prepare for disaster recovery after a serious disk failure. There are three main types of disk failures:
![]()
The server's boot disk, which contains the operating system and the EBR executables is damaged or destroyed.
In this case, you will probably need to replace and reformat the disk, or both, then reinstall VINES, the EBR agent, and possibly the EBR service.![]()
A secondary disk (/disk2), which contains other services, but not the EBR service data directory, is damaged or destroyed.
You may have to replace and reformat the disk, or both. Once the disk volume is usable again, which includes initialization using VINES, you can usually use EBR to recover the service data from previous backups.![]()
The disk that contains the EBR service's data directory, which contains the on-line indexes, is damaged or destroyed.
In this case, you may have to replace and/or reformat the disk, and then you will have to recover the EBR indexes before using EBR to recover any services on the disk. If the disk is also the boot disk, the problem is further complicated by the need to reinstall VINES and EBR first.
If your EBR server is completely destroyed, along with replacing it, you must:
![]()
Have a VINES "System Configuration" backup, created using the original VINES Backup and Restore utility. This backup only has to be done once and it will only be used in the event of a failure of the boot disk. ![]()
Re-install EBR using the same data directory location for the on-line indexes as in the original installation. Important: Once you understand the procedures for disaster recovery, make sure you have devised a disaster recovery plan for your site. If possible, you should test the ability to recover from a disaster at your site.
If you have enabled EBR services to execute scheduled backups, you are well prepared for a disaster.
Every time an EBR service backs up a group of clients, it also backs up all the on-line indexes for those clients, including the indexes for the EBR server itself.
It is recommend that you take the following precautionary steps to help you recover from a disk failure on the EBR server:
1. After EBR has been installed on the server, create a VINES System Configuration backup tape, using the standard VINES backup/restore utility. This only needs to be done once, immediately after EBR is installed and configured.
2. Select Backup/Restore from the Operator Menu.
3. Select Backup from the Backup/Restore Menu.
4. Select System Configuration Backup from the backup menu, and when prompted, select Confirm Readability of the backup.
5. After the backup completes, store it in a safe, but convenient location. It will be needed if the boot disk is damaged or destroyed.
Be sure to keep a hardcopy of every savegroup completion notification mail message. (See "Using Notifications" in Chapter 4, which describes how to use and configure ebrmail). The savegroup completion notification contains a record of EBR bootstrap information, including the date, backup level, and volume names of the backup media. You will need this information to determine which EBR media volume contains the most recent full backup. This will be the starting point for the recovery operation. We recommend printing it and keeping a copy in a safe place.
The bootstrap information looks similar to this:
May 09 20:59:03 1995 EBR bootstrap information Page 1
date time level ssid file record
5/09/95 20:29:08 full 8380 39 0
5/09/95 20:52:25 full 8382 41 0
The savegroup completion notification includes all the bootstrap save set ids for the past month. A bootstrap save set may span more than one backup volume. The file and record numbers help you to find the associated save set quickly.
In this example, the boot disk that includes the "/" and "/disk1" file systems, which on VINES servers holds the Operating System and the EBR executables, is damaged or destroyed.
Recovering Services on the Boot Disk
After replacing the damaged disk, format it and use the following procedure.
Important: It is impossible to provide step-by-step instructions on how to recover your system from a disaster, because every site is unique. The examples in this chapter are designed to give you general principles on how to recover files on /disk1 or /disk2 and to help you understand the procedures. They are meant to be examples only, not specific instructions.
1. Re-install VINES from the distribution media.
As part of the VINES installation procedure, you will be asked "Do you want to reload a complete system backup (y/n):" Type "y" and press [Enter].
2. Select Continue with System Rebuild from the System Rebuild menu.
3. When prompted, insert the VINES System Configuration backup media and choose Continue to start the restore operation.
4. When the restore is complete, you are prompted to "Load Next Incremental in Series" or "Continue with Installation." Choose "Continue with Installation."
VINES initializes the Mail service, if there is one, and loads the device drivers. Shortly thereafter, the Service Monitor initializes. All services, including EBR, show a status of inoperative, until they have been re-installed.
5. Re-install the EBR agent and, if necessary, the EBR service from the distribution media.
You will not have to create the EBR service because the service was created when the VINES System configuration backup was restored. However, you must re-configure EBR through the EBR Administrator utility if you made modifications to the original configuration - for example, changed the tape device from 8mm to 4mm.
6. After installing EBR, you need to reboot the server to start the portmapper before you can start the EBR service or agent.
7. Re-install any additional "Third-Party" services by following the appropriate procedures.
With the VINES operating system, EBR, and other third-party services installed, you are ready to start recovering VINES services from the EBR backup media.
Note: If the disk you are recovering also contains the EBR data directory, recover the EBR service's on-line indexes before you can recover the service data. Before proceeding further, follow the procedure in "Recovering an EBR Service's On-line Indexes" found later in this chapter.
1. Select Backup/Restore from the Operator Menu.
2. Select Enterprise Backup/Restore (EBR) from the Backup/Restore menu.
3. When prompted, enter the VINES administrator' s ID and password.
4. Verify that the server you are doing the recover on is looking for its files on the correct EBR server.
5. Select Change Recover Source from the menu and set the parameter that says "Use the EBR service on server:___" to the name of the EBR server that backs up this particular server.
6. Select Recover Server Data, from the Enterprise Backup/Restore menu.
7. When prompted, select Continue to continue the current operation.
A list of services will be displayed followed by the EBR recover prompt ("\\>" ).
This list can be displayed by typing the "dir" command. The entire list of commands can be displayed by typing "?"
8. Using the mark command, select all services you want recovered. In this case, it would be all services on the Boot/Primary disk:
\\> mark servicename
You may use this command as many times as necessary. You could also select all services:
\\> mark *
and then un-select those services you do not want restored, with the unmark command:
\\> unmark servicename
Important: If the servicename contains spaces, you must place quotes (" " ) around the name or EBR will not accept it.
9. To display which tape volumes will be needed for the restore operation, enter:
> volumes
10. To force an overwrite of existing files, enter:
> force
Note: This is necessary because the data restored from the VINES System configuration backup could be out of date.
11. Insert the first volume listed in the output of the volumes command into the tape drive.
12. Log in to a client workstation configured for EBR administration, using an ID which has EBR administrator rights, then click on the EBR Administrator Icon and mount the volume you just inserted into the tape drive.
13. Go back to the server console and enter:
> recover
to start the recovery procedure.
14. Messages similar to these will be displayed:
Recovering 13794 files into their original locations.
Total estimated disk space needed for recover is 246MB
Not all Volumes are on-line (some off-line)
SNI1.003 at /dev/rct4
SNI1.00415. At the prompt "Proceed with Recover command at this time? [Y]" , press [Enter] for yes. To know when to mount the next volume needed for the restore, monitor the Messages and Pending windows using the EBR Administrator.
Recovering Services on a Secondary Disk
This section provides an example of how to recover services to a secondary disk (for example, /disk2), using EBR. The example may apply to the EBR server itself or a VINES server running the EBR agent software. This example assumes that if this procedure is for the EBR server itself, the data directory, which contains the index files, is not on the secondary disk.
Important: It is impossible to provide step-by-step instructions on how to recover your system from a disaster, since every site is unique. The examples in this chapter are designed to give you general principles on how to recover files on primary or secondary disk and to help you understand the procedures. They are meant to be examples only, not specific instructions.
The procedure for recovering services on a secondary disk is exactly the same as for recovering a service on the Boot/Primary disk. See "To Restore VINES Services" in the previous section.
Recovering an EBR Service's On-Line Indexes
This section describes the procedures to follow when an EBR service's indexes are lost due to a disk failure on the EBR server. These indexes hold the information necessary to recover data for any of the EBR server's clients - for example, client workstations or VINES servers running the EBR agent software.
If the EBR service data directory is on /disk1 and the disk has had a total failure, then you will first need to replace the disk, then reinstall VINES and EBR from distribution media, before you can proceed with recovering the indexes. If the indexes are on a disk other than /disk1, you can use the procedures in the following section.
Determine which media volume contains the last full backup, using the hardcopy printout of the EBR bootstrap information described earlier in the chapter, and load this tape into the tape drive.
To Recover an EBR Service's On-Line Indexes
1. Select Backup/Restore from the Operator Menu.
2. Select Enterprise Backup/Restore (EBR) from the Backup/Restore menu.
3. When prompted, enter the VINES administrator' s ID and password.
4. Select Local Tape Operations from the Enterprise Backup/Restore menu.
5. Select Rebuild Media and On-Line Indexes from the Local Tape Operations menu.
6. At the prompt, select yes to continue. This operation could take anywhere from 45 minutes to three or four hours, depending on size.
7. When index recovery has completed, press [Enter] to continue.
8. Press the Esc key to return to the Local Tape Operations menu.
9. Follow steps 6 through 9 for each EBR media volume used after the last full backup.
10. After recovering all necessary indexes, press the Esc key until you are returned to the Enterprise Backup/Restore menu.
11. You have completed restoring the on-line Indexes and you are now ready to recover the service data itself.
All of the operations described in this section require you to log into UNIX as described in the Banyan UNIX Access Guide available with VINES 6.00 and greater. Once you are logged into UNIX and have a UNIX Shell prompt, you can use various EBR utility programs that are installed in the EBR home directory, /disk1/banyan/ebr.
The EBR service includes two major directories:
![]()
A service home directory that includes all of the service executables. ![]()
A service data directory that holds configuration data, logs, and (for EBR services) indexes.
Depending on how the service was created, the EBR data directory can be on /disk1, or on some other system disk.
Every EBR data directory includes a res (resource) subdirectory of configuration information. An EBR service's data directory also includes one directory of index data for each client supported by the service. These directories are named with the client's name.
Using the UNIX Recover Browser
Normally, you recover server data using the Recover Server Data option of the EBR Console Menu. When you recover data this way, the recover browser only presents names that are familiar to users and administrators. This includes StreetTalk names for users and services, DOS or Macintosh names for files and directories in file services, and mail folder names in mail services. When used in this way, the recover browser also constrains recover operations in ways intended to prevent inappropriate data recovery that could compromise service or server integrity.
When you access the recover browser from a UNIX Shell, you are presented with a view of index data that shows only the UNIX file and directory names of backup data. For example, a mail service data directory that appeared in the Console recover browser as MS@ServerName@Servers could appear in the UNIX recover browser as /disk1/BMS.
Using EBR to Recover from Server Disk Failures
All administrators need to prepare for data recovery after a serious disk failure. There are three main types of disk failures that you may encounter:
![]()
A primary disk, /disk1, which contains the operating system and EBR binaries, is damaged or destroyed. In this case, you will probably need to replace and/or reformat the disk, then reinstall VINES and the EBR agent, and possibly the EBR service. ![]()
A secondary disk which contains other file systems is damaged or destroyed. You may have to replace and reformat the disk, or both, but once the disk volume itself is usable again, you can usually just recover the disk's contents from previous backups. ![]()
The disk that contains an EBR service's on-line indexes is damaged or destroyed. You must recover the indexes before using EBR to recover any file systems. If the disk is also /disk1, the problem is further complicated by the need to reinstall VINES and EBR executables first.
If your EBR server is completely destroyed, you must replace it with another machine. You must:
![]()
Name the replacement server with the same hostname as in the original installation ![]()
Re-install EBR using the same directory locations for the on-line indexes as in the original installation Important: Once you understand the procedure for disaster recovery, make sure you have devised a disaster recovery plan for your site. If possible, you should test the ability to recover from a disaster at your site.
If you have set up your network and enabled the EBR services to execute scheduled backups, you are already well-prepared for a disaster. Every time an EBR service backs up a group of clients, it also backs up all the on-line indexes for those clients, including the indexes for the EBR server itself. The server's index backup is named bootstrap, and is assigned an identification number called the save set id (ssid).
We recommend you take two precautionary steps to help you recover from a disk failure on an EBR server:
![]()
Keep a file containing printed copies of the bootstrap records. Place these daily sheets of paper in a three-ring binder or a file folder. ![]()
Make a hardcopy record of the disks, partition sizes, and mount points for the server and any clients that have a local hard disk. This information could make recovery smoother for you in the future.
File the Bootstrap Information
EBR appends a record of the index backup to every saveset completion notification. We suggest printing these so you have the date, the name of the backup media, and the save set id number for bootstrap.
If you need to recover the server' s on-line indexes, you will need this information. Put this information in a safe place.
The information sent to the printer looks similar to this:
May 09 20:59:03 1995 EBR bootstrap information Page 1
date time level ssid file record
5/09/95 20:29:08 full 8380 39 0
5/09/95 20:52:25 full 8382 41 0
Each saveset completion notification includes all the bootstrap save sets for the past month. The bootstrap save set may span more than one backup volume. The file and record numbers are used to find the associated save set quickly.
You can also manually back up the EBR server's indexes by using the saveindex utility in the EBR service home directory (/disk1/banyan/ebr). Using this command also sends the bootstrap information to a printer. For example:
Banyan> cd /disk1/banyan/ebr
Banyan> saveindex -c server_name
Use the UNIX df command to find out how the EBR server's disks are partitioned and mounted, and make a hard copy of this information. Do the same for any servers that host EBR agents as well. Print and file this information in case you need to recover any of these servers from a disk crash.
When you run df, the output tells you what you will need to set up a replacement disk for the server. Here is an example of df output:
Banyan> df
/ (adsys ): 452 blocks 5128 i-nodes
/disk1 (/dev/adusr ): 252108 blocks 126154 i-nodes
This section provides an example of how to recover a secondary server disk using EBR. The example may apply to the EBR server itself, or to a VINES server hosting an EBR agent.
Important: It is impossible to provide step-by-step instructions on how to recover your system from a disaster, since every site is unique. The examples in this chapter are designed to give you general principles on how to recover a primary or secondary disk, and to help you understand the procedure. They are meant to be examples only, not instructions.
The example that follows assumes that /disk1 is still operational so the server has an operating system and can run EBR. However, a secondary disk is lost due to a head crash.
Example
If the disk is damaged, replace it with a new disk of the same type. Try to get a disk of at least the same capacity as the old one. You will need a disk large enough to hold all the file systems to be recovered. To replace the damaged disk, do the following:
1. Install the replacement disk.
2. Make sure the operating system and kernel recognize the new disk.
3. Label and partition the new disk so you can recover the file systems. Use the hard copy of the disk information to see the size of each partition. (See "File the Disk Information" in this chapter.)
Tip: If you do not already have this information, look at the directory /etc/fstab. Use the cat command to find out how the disk was partitioned into file systems. You will have to guess how much space to give each partition. However, you still have /disk1, so the partition information is available.
4. Make file systems for each raw partition that you are going to recover, and mount the block partition, consulting the hard copy of the df (or cat /etc/fstab) output. (EBR will not initialize file systems; it recovers data into existing ones.)
After creating and mounting all the file systems on the replacement disk, use EBR to recover the files. You should recover one file system at a time, for two reasons:
![]()
EBR will recover all files in one file system but will stop at a mount point. ![]()
You may run out of swap space if you recover too many file systems at once to your list of data to recover.
To recover backup data using the UNIX recover browser, follow these steps:
1. Run the recover command:
Banyan> cd /disk1/banyan/ebr
Banyan> recover -s server_name -Zusername -zpasswordHere servername is the name of the EBR server that backs up this server. If you have already logged in to VINES as part of obtaining root access, you will not need to supply your username and password on the recover command line. You must be a member of the server' s AdminList to recover data to any of the server' s disks.
2. In the recover browser, mark the disk you want to recover. For example, if you want to recover /disk3, enter:
> mark /disk3
3. Recover the file system:
> recover
/disk3 is being recovered into its original location
Requesting 5023 files, this may take a while...4. Exit the recover program.
> quit
To watch for tape mount requests, you can monitor the progress of the recovery using EBR Administrator on a VINES workstation.
In this example, /disk1, which on VINES always holds the operating system and the EBR binaries, is damaged. First, you must reinstall VINES and reboot the server. Next, reinstall EBR from the original software distribution or a network directory so you can recover the data.
After replacing the damaged disk, format it, and reinstall VINES, using the original software distribution. Then reinstall the EBR agent and, if necessary, service. Be sure to re-create the service using the same data directory as it had before the crash.
Next, reinstall the EBR agent and, if necessary, the EBR service. You may do this using the original distribution media, or you may reinstall from a network installation area that was set up as described in the EBR Release Notice.
With the operating system and EBR back in place, you are ready to start recovering the remainder of the data lost from the backup data. Be sure to temporarily disable any groups scheduled to run during the period when you are recovering the index. Using the UNIX recover browser, follow these steps:
1. Run the recover command:
Banyan> cd /disk1/banyan/ebr
Banyan> recover -s server_name -Zusername -zpasswordHere servername is the name of the EBR server that backs up this server. If you have already logged in to VINES as part of obtaining root access, you will not need to supply your username and password on the recover command line. You must be a member of the server' s AdminList to recover data to any of the server's disks.
2. In the recover browser, mark the root directory and /disk1/banyan.
> mark /disk1/banyan
> mark /3. Enter the following commands to forcibly overwrite files on disk with data from the tape:
> force
> recover4. Now that server configuration data has been recovered, shut down, then reboot the server.
5. After the reboot, access UNIX again and run the recover command as described in step 1.
6. Mark /disk1, then unmark /disk1/banyan.
> mark /disk1
> unmark /disk1/banyan7. Enter the following commands to forcibly overwrite files on disk with data from the tape:
> force
> recoverThis will recover any system software updates (patches) that may have been applied after VINES was installed from release media.
8. You may want to reboot the server again if there are patches that need to be in effect before you recover additional disks.
9. Recover data to any additional disks on your server, one disk at a time, using the following procedure:
> mark /disk1<n>
> force
> recoverMarking one disk at a time conserves server resources (memory and swap space).
10. Exit the recover program.
> quit
To watch for tape mount requests, you can monitor the progress of the recovery - for example, using EBR Administrator on a VINES workstation.
Important: Always reboot a system after recovering a primary disk.
Recovering an EBR Service's On-Line Indexes
This section describes the procedures to follow when an EBR service's on-line indexes are lost due to a disk failure on the EBR server. These indexes hold the information necessary to recover data for any of the EBR server's clients - for example, client workstations or VINES servers running the EBR agent software.
If the server loses its operating system and EBR programs, these will have to be re-installed first. (See "Recovering /disk1" in this chapter.)
The next important step is to recover the indexes from tape using the recoverindex command. The recoverindex command will ask you for the bootstrap save set identification number (ssid). If you followed the procedure recommended to prepare for a disk crash, you will have this information.
For example, ssid 1148869870 below is the most recent bootstrap backup:
August 20 03:30 1995 EBR bootstrap information Page 1
date time level ssid file record volume
8/08/95 7:44:38 full 1148869706 55 0 Full .008
8/09/95 6:12:09 9 1148869754 48 0 Nonfull .001
8/10/95 6:14:23 9 1148869808 63 0 Nonfull .001
8/11/95 6:29:58 9 1148869870 88 0 Nonfull .001
If you do not have this information, you can still recover the indexes by finding the ssid using the scanner command. (See "Finding the Bootstrap Save Set ID" in this chapter.)
You may need more than one backup volume to recover the server's indexes. During the recovery, you can use EBR Administrator to watch for pending messages requesting backup media.
With the operating system and EBR in place, recover the indexes from the backup media using the following steps:
Note: You should temporarily disable any groups scheduled to run during the period that you are recovering the index.
1. Find the bootstrap save set id information. You will need it for the next two steps.
2. Retrieve the backup tape that contains the most recent backup named bootstrap, and load it into the server' s tape device.
3. Use the recoverindex command to extract the contents of the bootstrap backup:
Banyan> cd /disk1/banyan/ebr
Banyan> recoverindex -s servernamerecoverindex: Using servername as server
NOTICE: recoverindex is used to recover the EBR
server' s on-line file and media indexes from media
(backup tapes or disks) when either of the server' s
on-line file or media index has been lost or damaged.
Note that this command will OVERWRITE the server' s
existing on-line file and media indexes. recoverindex
is not used to recover EBR clients' on-line
indexes; normal recover procedures may be used for
this purpose.What is the name of the tape drive you plan on using [dev/rct4]?
Enter the latest bootstrap save set id []: 1148869870
Enter starting file number (if known) [0]: 88
Enter starting record number (if known) [0]: 0
Please insert the volume on which save set id 1148869870 started into dev/rct4.
When you have done this, press <RETURN>:Scanning dev/rct4 for save set 1148869870; this may take a while...
scanner: scanning 8mm 5GB tape space.006 on dev/rct4
uasm -r ebr/res/ebr.res
uasm -r ebr/res/ebrjb.res
uasm -r ebr/res/
ebrmmdbasm -r ebr/mm/mmvolume
ebr/mm/mmvolume: file exists, overwriting
uasm -r ebr/index/space/
ebrindexasm -r ebr/index/space/db
scanner: ssid 449955156: scan complete
scanner: ssid 449955156: 31 KB, 10 files
ebr/index/space/db: file exists, overwriting
uasm -r ebr/index/
uasm -r ebr/mm/
uasm -r ebr/
uasm -r
space: 31 records recovered, 0 discarded.
ebrindexasm: Building indexes for servername...
ebrindexasm: Caching save times for servername...
8mm 5GB tape space.006 mounted on dev/rct4, write protectedThe bootstrap entry in the on-line index for mars has been recovered. The complete index is now being reconstructed from the various partial indexes which were saved during the normal saves for this server.
# ebrindexasm: Pursuing index pieces of ebr/index/space/db from servername.
Recovering 2 files into their original locations
Total estimated disk space needed for recover is 11 MB
Requesting 2 files, this may take a while...
ebrindexasm -r .db
.db: file exists, overwriting
: 25711 records recovered, 0 discarded.
ebrindexasm -r .db
.db: file exists, overwriting
ebrindexasm: waiting for lock on ../db.SCAVENGE
ebrindexasm: lock on ../db.SCAVENGE acquired
Received 2 files from ebr server ` mars'
: 733 records recovered, 0 discarded.
ebrindexasm: Building indexes for mars...
ebrindexasm: Caching save times for mars...
ebrindexasm: Suppressing duplicate entries in mars - 50 duplicates discarded.The on-line index for space is now fully recovered.
Notice how the shell prompt appears once bootstrap is recovered. You can use the EBR Administrator to monitor the progress of the recovery.
Replacing EBR Configuration Data
The recoverindex command also recovers the directory used by EBR to store configuration information such as the list of EBR's clients and registration information. However, this directory, unlike the indexes, can not be overwritten or relocated; instead, the recovered directory is renamed.
To complete the recovery of the directory, you must shut down the EBR service, move the recovered, renamed directory into its original location, and then restart the EBR service.
Complete these steps after recoverindex has finished and this final message appears:
nsrindexasm: The on-line index is now fully recovered.
1. Stop the EBR service.
2. Save the original configuration database directory and move the recovered version into the correct location. For example, if the EBR service' s data directory is located in /disk2/EBR, then the configuration databases are located in /disk2/EBR/res. After recovering the data from backup media, there will be two copies of the res directory. Use the one named res.R. To complete the recovery, enter the following commands:
Banyan> cd /disk2/EBR
Banyan> mv res res.orig
Banyan> mv res.R resThis preserves to old res directory for potential re-use if the new res directory causes problems. Once you have verified that the EBR configuration is correct, you can remove res.orig directory.
Banyan> rm -r /disk2/EBR/res.orig
3. Restart the EBR service. When it restarts, it will use the recovered configuration data.
To recover other file systems, see "Recovering a Secondary Disk" in this chapter.
Finding the Bootstrap Save Set ID
If you did not file a hard copy of the bootstrap information, you can still find the save set id of the most recent bootstrap by using the scanner command. For example:
1. Place the most recent media used for backups in the server device.
2. Read the contents of the backup media with the scanner command:
Banyan> scanner /dev/rct4
The scanner command displays the contents of the backup media, as follows:
scanner: scanning mars.2 on /dev/nrst8
client symbolic name save time size files ssid S
venus / 04/10/93 03:12:40 4326980 599 17475596 E
mars / 04/10/93 03:06:11 4478460 621 17475595 E
venus /usr 04/10/93 03:13:58 6918900 965 17475597 E
venus /home/venus 04/10/93 03:14:14 9661320 825 17475598 E
mars bootstrap 04/10/93 03:33:51 26607144 2 17850937 E
In this example, the bootstrap save set id is 26607144. Once you find the most recent bootstrap save set id, you can use the recoverindex command to recover the server's index.
The following procedure summarizes what you need to do if a primary or secondary disk is damaged on a VINES server that is being backed up using EBR:
1. If the operating system is lost, reload and boot the system using the same hostname and disk partitioning.
2. If necessary, replace the damaged disk, format it, partition it, make new file systems, and then mount them in the same locations.
3. Re-install any EBR binaries that were lost.
4. If the EBR service's on-line indexes are destroyed, use recoverindex to recover them.
5. Recover the lost file systems, one at a time.