NBU 7.6 - Do partial restore (have 16/17 tapes)

[JaquesC]

Hi there

Server is 2008 R2

I have a year end backup that consists of 17 tapes spread across, now I have found 16/17 and want to continue on with the restore, and want to know, how can I force the job to continue and restore what is possible without failing because of one tape being missing.

OS is Server 2008 R2 Standard

NBU - 7.6.0.2

[Michal_Mikulik1]

If I understand correctly, it is a backup performed by MS-Windows policy type. It is one backup ID consisting of 17 (or more) fragments - you can confirm this by Images On Tape report.

Then you have to just cancel the backup - and verify at least what was restored.

Verify that you dont have copy of this backup id somewhere.

Not sure if NetBackup can report which files were in a certain fragment - anybody here knows?

m.

[Marianne]

Is this a single client, single image spread across 17 tapes? Hopefully not. If you can list the image id's for this backup set, you should able to list media for each id and then do restores in smaller chunks.

[JaquesC]

Hi there

It is a single client spread across 17 tapes, however also keep in mind mind its a massive file server so the spread is neccesary.

Can you give me an example of best way to search for these images Marianne?

[Marianne]

Hopefully not a single image? I would hope that Backup Selection is broken up into multiple streams and not a single stream. How many lines show up in Client Backup report for this client and period? You will only be able to restore partial data if the backup is broken up into multiple streams/images. Have you tried Images on Media report for the missing tape?

[mph999]

"Not sure if NetBackup can report which files were in a certain fragment - anybody here knows?"

Not easily ...

I believe it can be worked out from a combination of the FRAG line in the EMM ImageFragments table (or bpimagelist -l command)  and the info in the images .f file

However, as I found in this post :

https://www-secure.symantec.com/connect/forums/data-backed-vs-data-stored-tape

NBU seems to be doing something a bit odd with the maths behind it, so it's not quite as simple as I first thought.