NetBackup Disaster Recovery
The procedures in this article explain how to recover your data if the system disk fails on a NetBackup master server.
Two general cases are considered:
- Recover Master Server When Root is Intact
- Recover Master Server When Root is Lost
Recover Master Server When Root is Intact
- Verify the operating system is functioning. If it isn't, take appropriate corrective actions.
- Reinstall NetBackup software
- Restore the NetBackup databases by using the
bprecovercommand on the master server. The NetBackup databases are critical to NetBackup operation and, if they are lost, they must be recovered before any other files.
- Restore other files to the server as desired.
Recover Master Server When Root is Lost
- Load the operating system on an alternate boot disk, using the same procedure you normally would for the server.
- Create, on the alternate boot disk, the partition and directory where NetBackup and its databases resided on the orginal disk. (by default they reside in /usr)
- Install NetBackup software to alternate disk.
- Restore the NetBackup databases to the alternate disk by using the
bprecovercommand on the master server. The NetBackp databases can be restored only to the same location from which they were backed up. (alternate path restores are not allowed)
- Start NetBackup and restore the latest backed up version of all files to the disk you are restoring, starting with the root filesystem.
- Copy the NetBackup databases from the alternate disk to the disk you are restoring, overwriting the databases you restored. The databases you are overwriting are old versions from the last regular backup. The ones you are copying from the alternate disk are the most recent, assuming they were done immediately after the last regular backup.
- Stop all NetBackup processes that you started from NetBackup on the alternate disk.
- Test the copy of NetBackup on the disk you have restored. Use
xbp. Also, try some backups and restores.
- When you are satisfied that all is in order, you may want to delete the NetBackup files from the alternate disk. Or, you may want to unhook that disk, if it is a spare. You will then make the restored disk the boot disk again.
Slave Server Disk Recovery
The procedure for reloading a slave server on which the system disk has failed is the same for a master server, except that you must use the following options when executing the
- slave_name: /usr/openv/volmgr/database
In the above example, you substitute the host name of the slave server for slave_name. You can execute the
bprecover from either the master or a slave by specifying the correct destination host with the
bprecover -dhost option.
Client System Disk Recovery
- Reload the operating system in the way you normally would for a client workstation of that type.
- Reinstall the NetBackup client software
- Restore files to the client machine.