NBU Automated Vaulting issue : 287/288 error code

[a_la_carte]

Hello guys,

The problem I am facing here in my environment is that automated vautling does eject the media fine. But if in case, any one of the media is in use and backups are being written to that tape, then vaulting gets failed with either 287 or 288 status code.

My environment is NBU 7.0.1. We have HP MSL 8096 tape library with maximum of 15 mail-slots configured.

Following is the brief view of "detailed status" window from both jobs.

=========================

18/07/2011 16:03:22 - starting eject of 19 media
18/07/2011 16:10:34 - media ejected from robot.  Please empty MAP
18/07/2011 16:11:06 - eject operation is waiting for available MAP elements of robot
18/07/2011 16:23:13 - media ejected from robot.  Please empty MAP
18/07/2011 16:23:45 - eject operation is waiting for available MAP elements of robot
18/07/2011 16:39:30 - eject complete with status 287.  18 of 19 media ejected
18/07/2011 16:39:30 - vault eject lock released
18/07/2011 16:39:39 - vault global lock released
18/07/2011 16:39:43 - end writing
Status 287
18/07/2011 16:39:43 - end Eject/Report; elapsed time: 00:36:31
18/07/2011 16:39:43 - begin Vault, Validate Image
Status 0
18/07/2011 16:39:43 - end Vault, Validate Image; elapsed time: 00:00:00
18/07/2011 16:39:43 - begin Vault, End Notify Script
18/07/2011 16:39:43 - started process RUNCMD (640)
18/07/2011 16:39:44 - ended process 0 (640)
Status 0
18/07/2011 16:39:44 - end Vault, End Notify Script; elapsed time: 00:00:01
Status 287
18/07/2011 16:39:44 - end Vault, Execute Script; elapsed time: 00:39:43
vault eject failed(287)

============================

15/07/2011 12:03:18 - starting eject of 15 media
15/07/2011 12:09:05 - media ejected from robot.  Please empty MAP
15/07/2011 12:09:05 - failed to eject 1 media. Reason: MEDIA_IN_USE
15/07/2011 12:09:07 - eject complete with status 288.  14 of 15 media ejected
15/07/2011 12:09:11 - vault eject lock released
15/07/2011 12:09:20 - vault global lock released
15/07/2011 12:09:25 - end writing
Status 288
15/07/2011 12:09:25 - end Eject/Report; elapsed time: 00:06:14
15/07/2011 12:09:25 - begin Vault, Validate Image
Status 0
15/07/2011 12:09:26 - end Vault, Validate Image; elapsed time: 00:00:01
15/07/2011 12:09:26 - begin Vault, End Notify Script
15/07/2011 12:09:26 - started process RUNCMD (4324)
15/07/2011 12:09:26 - ended process 0 (4324)
Status 0
15/07/2011 12:09:26 - end Vault, End Notify Script; elapsed time: 00:00:00
Status 288
15/07/2011 12:09:26 - end Vault, Execute Script; elapsed time: 00:09:26
vault eject partially succeeded(288)

==========================

Now, there is strange thing that happened today. Vaulting started and finished fine and ejected all other media except the media which is in use and exited with status 0 also.It didn't even select the media which was in use.

Dont know why is this happening. Most of the times failing when media in use, and at times, getting success even when media is in use and I have confirmed that at all the times duing the vault run, the media that is in use, is also the one that needs to be ejected by NBU vaulting as we change media on daily basis.

Can someone please suggest how to stablize this? I mean, some settings that needs to be done so that vaulting does not select the media which is in use and does not exit with any error code?

Thanks.

[RonCaplinger]

Vault does not have a mechanism to know when media is still physically in use.  It is simply selecting IMAGES, not media.  The report generated is listed by media ID, though.

[a_la_carte]

Thanks Ron,

So does that mean that we will always end up having an error code with a vaulting job?

It is ejecting the rest of the media fine but ending up with an error code is something that doesn't look good in activity monitor.. !! 

[RiaanBadenhorst · Accepted solution]

You should schedule your vaults to run when the backups are done, then you wont get any "media in use" issues. Sounds simple, but I know its not that easy. If certain backups are still running when you want to run vault, maybe think of creating more vault jobs i.e. vault job that duplicates the first part of the evening's jobs, and then another that runs later that catches the rest of the backup jobs?

[a_la_carte]

Thanks Riaan... I guess.. Thatz the only solution I should try my hands on if I want to remove the error code.There is only one NDMP backup job that takes around 20+ hrs to finish and thatz the only media that is always in use when the automated vaulting runs, so I guess .. I have to make a separate vault job to eject that media to avoid getting vaulting job ending with an error code.

Thanks everyone..