cancel
Showing results for 
Search instead for 
Did you mean: 

Tapes reporting no data stored on them.

Bruce_Clegg
Level 5

I've got a situation I'm trying to understand a bit better.

I'm backing up a server that is suddenly using a lot more disk space than before.  One of the segments I'm backing up has grown to 1.7T - which is fine.  But on this system we're encrypting the backups and still using compressed LTO-4's.  I only had two LTO-4 tapes in the correct volume pool in the library.  Looking at the logs, I see netbackup request the first tape, write to it for hours, request the second tape, write to it for hours - The stats on the job said that it had written ~1.6T and ~5500 files.  When it requested the third tape, the job cancelled with status 96 as you'd expect because there wasn't a third tape available.

My question is - when I look in the media section of the Admin GUI - it reports those two tapes as empty.  Is this what I should expect when a job fails?

1 ACCEPTED SOLUTION

Accepted Solutions

RamNagalla
Moderator
Moderator
Partner    VIP    Certified

yes.. it is very normal, netbackup does not record tape utilization details untill it got job successfull message from Media server (bpbrm process).

it is to make sure that tapes are not occupied with failed job data.

View solution in original post

3 REPLIES 3

RamNagalla
Moderator
Moderator
Partner    VIP    Certified

yes.. it is very normal, netbackup does not record tape utilization details untill it got job successfull message from Media server (bpbrm process).

it is to make sure that tapes are not occupied with failed job data.

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified
As per Ram's excellent post, everything with status code 2 and higher is simply discarded because you can only restore from status code 0 or 1 (status 1 for filesystem only, not db).

There are ways to mitigate the effect of status 96.
You can divide the Backup Selection into smaller chunks and enable multiple data streams in policy attributes.

Each stream will be a separate job. So, you have a better chance of most jobs completing successfully. Only the failed stream will be discarded.

First prize is obviously to have a couple of extra tapes in Scratch pool.

Thanks Ram

I thought it might be something like that - but wanted to confirm.