Backup Exec 2010r3 with Partitions Configured

[JonScott]

Hi,

I've inherited a backup solution that has been previously configured as follows

We have a Full backup configured to use tapes on slots [0001..0004]

A Daily Differential job configured to use tapes [0005..0008]

And a full monthly job configured to use slots [0009..0010]

The problem i have,  is that even though the jobs are configured to use the specific slots, 

i'm finding that when the jobs over-run,  instead of cancelling the job,  it's grabbing whatever media is available,  even if it is in another partition.

The way i understood the partions to work,  is that they will only ever use the tapes in that partition,  and will fail if there is no media available.

Can anyone help me understand why this is happening?

Thanks

Jon

[pkh]

By any chance, is the second tape that was used in the mail/IO slot?

[JonScott]

Nope, no mail or i/o slots. Monthly job used a weekly tape, weekly job then had to use a differential tape

[Larry_Fine · Accepted solution]

You have apparently encountered a known issue.

http://www.symantec.com/docs/TECH205938

Backup Exec does not take tape from specified library partition

I would suggest using hardware partitions (created in the library UI) as a workaround.

[pkh]

The user is using BE 2010 and not BE 2012, so this is not a known problem.

[pkh]

In this case, you should either log a formal support case with Symantec for them to look at the problem, or try upgrading to BE 2014 and see whether this solves the problem.  The upgrade is free if you are on a current maintenance contract.

[VJware]

There was an issue with BE 2010, however it affected only NDMP backups. I would recommend updating BE 2010 R3 completely with all updates and check if this issue reoccurs or not. If not, pls log a formal support case as suggested earlier.

[JonScott]

It is an NDMP Backup we're doing,  we're currently running on SP3,  i'll look at deploying SP4 and see if that fixes the issue.

It's strange though,  if i create a new job and specify to use the partition [0013-0024] this one fails instantly as there are no tapes in there,  but the other jobs seem to start in the correct partition,  and just grab a tape randomly if it over-runs.

Do you have a link to the issue that affected NDMP backups?

Thanks

Jon

[VJware · Accepted solution]

Here is the KB - http://www.symantec.com/business/support/index?page=content&id=TECH212850

Upon a closer look, looks like it was resolved in BE 2014 and not BE 2010.

[JonScott]

Thanks for that.

Looking into the logs i've got available,  this all seems to have started back in may when we upgraded to SP3 because of an Exchange 2010 SP3 update.

Unfortunately we can't roll this back so going to have to look at the partitioning at the library level whilst we decide if we're going to upgrade to 2014 or not.

Thanks

Jon