Restore question

[nasahasha]

Hi,

Netbackup 7.5.0.1,Redhat linux

one of 4 mediaservers G30 has _bk problem and its not responding(We have 2 networks primary and backup network,_bk is used for backups and restores).

As _bk was not responding Backups were failing with status 58,its going to take sometime to get _bk fixed,so I have pulled out that G30 media servers storage units from SUG to stop 58 failures.

I need to run a DB restore ,some pieces are residing on that media server G30,restoring is failing as _bk is not up.

What do I do?

[Marianne]

Is there another server with access to backup media?

If so, you can add bp.conf entry on master server to redirect restore via another media server:

FORCE_RESTORE_MEDIA_SERVER = <from_host> <to_host>

[nasahasha]

We have another 3 media servers which are working fine.Backups are on AdvancedDisk.

How do I check whether others have access to this trouble media server?

[Marianne]

The question was whether other media servers had access to backup media... (e.g. tape library).

Problem with disk is that it is only assigned to a particular media server.

Not sure what the steps would be update NBU if disk is attached/zoned to another media server - will have to search manuals, TechNotes, etc....
Probably just the mounted to same mountpoint on different media server with FORCE_RESTORE_... entry. Not sure.

Probably easier to configure public network interface as media server, but you would need to config client to use public interface as well, update SERVER entries, etc.
If client was backed up with _bk name, this would mean that you need to perform a redirected restore plus FORCE_RESTORE... entry. You would first need to perform test-backup to ensure all comms are working on public network.

The more I think about this, the messier it becomes. Problably easiest to get _bk interface on Media server fixed - use this restore request as urgent motivation to speed up the process.

[nasahasha]

nbdevquery -listdv -stype AdvancedDisk -U | grep Status shows that media server's volumes are down.Cant I get this restoring working until I bring this volumes up,I know about nbdevconfig but they will be marked dowm immediately again as _bk is down.

Just adding FORCE_RESTORE_MEDIA_SERVER = <from_host> <to_host> will not do the magic.

[Mark_Solutions]

As the disk are attached to a media server that is down it would be very complex and risky to try and provision them elsewhere.

As this is advanced disk can i assume that thse backups were duplicated to tape?

If so then promote the tape to be primary and the other servers will be able restore from the tapes as long as they are in a library attached to one of them.

To re-use the disk elsewhere would mean forcibly expiring all images on it first to clear down the catalog (while the disk is not accessible otherwise the data will be deleted) - then copy all files on it into another media servers disk  storage unit (or just attatch it to another media server) and then running an import against it specifying to import Copy 1.

All very dangerous and a copy of everything should be maintained to be on the safe side.

If you have tape then use that

Hope this helps

[nasahasha]

Its duplicated but duplicated to AdvancedDisk attached to another media server using SLP.

[Mark_Solutions]

Thats fine then - just promote the copy2 to be primary (via the catalog section) and you will be able to do the restore from the second media server

[nasahasha]

This is a oracle Db restore ,backup has run on 4 channels,wth 4 files per set.

How would I gather all backup Id's of last successful DB backup.

Is there any command by which I can change all backup id's associated to that client on that media server to primary?