I guess you could create a bpend_notify script file on your client and
freeze/suspend the tape(s) with 'bpmedia' netbackup command - provide
you capture the tape #'s from the backup job somehow.

"Ewan" wrote:
>
>Right, I think I know the answer to this one but I'm going to try anyway.
>
>We're planning a new backup procedure in which we backup ourdata remotely
>as
>opposed to what we currently do which is back it up at one location and
move
>it
>to the other site.
>
>This has left ME with a problem to sort out. Currently A database is backed
>up to
>a specific volume pool. The tape that uses is then moved offsite thus the
>tape
>can't be used again to put the next days image on. However, with the new
>
>system the tapes will remain in the silo. So in theory we could end up
with
>multiple backups on one tape. Not ideal if a tape should become corrupt.
>
>
>Is my only solution to create different Volume Pools for each daily backup?
>Or
>does someone else have a bright idea? I hope so...
>
>Yours hopefully...

Hi Ewan,

Depends how many free slots you have in your silo. LTO tapes very rarely go
bad. If a tape does go bad while writing (we get about 1 every two to three
months out of 1300 tapes), all the previous images will still be readable,
because they exist and were written and fully saved before the bad segment
in the tape. A bad tape will usually get frozen, in which case we eject it
from the robot, replace it with a new tape, and place the old tape on a
shelf until it expires, and then we physically destroy it permanently. We
never usually take tapes out of the silo, unless they go bad and are frozen.
We use one pool, and two retention levels (daily schedules retention 30
days, weekly schedules retention 30 days, monthly schedules retention 1
year) - and we simply append all backups to any previous existing backups,
and it doesn't matter which client gets written to which tape.

The rotation handles itself nicely, with expired tapes returning to the
scratch pool on time as expected.

I think using individual pools for individual days will create a lot of
admin headaches and waste media.

Regards,
Dave.




"Frank Cavadi" wrote in message
news:45faabaa@ROSASTDMZ05....
>
> I guess you could create a bpend_notify script file on your client and
> freeze/suspend the tape(s) with 'bpmedia' netbackup command - provide
> you capture the tape #'s from the backup job somehow.
>
> "Ewan" wrote:
>>
>>Right, I think I know the answer to this one but I'm going to try anyway.
>>
>>We're planning a new backup procedure in which we backup ourdata remotely
>>as
>>opposed to what we currently do which is back it up at one location and
> move
>>it
>>to the other site.
>>
>>This has left ME with a problem to sort out. Currently A database is
>>backed
>>up to
>>a specific volume pool. The tape that uses is then moved offsite thus the
>>tape
>>can't be used again to put the next days image on. However, with the new
>>
>>system the tapes will remain in the silo. So in theory we could end up
> with
>>multiple backups on one tape. Not ideal if a tape should become corrupt.
>>
>>
>>Is my only solution to create different Volume Pools for each daily
>>backup?
>>Or
>>does someone else have a bright idea? I hope so...
>>
>>Yours hopefully...
>