One backup job out of 383 insists that it's been cancelled because of a media alert response

[James_Avery]

I have a truly strange problem with one backup job out of 383 scheduled jobs.  It's backing up a backup folder that's been rsync'd from an internal backup job on a unix server to a linux server for the purposes of off-site storage.  The client server is running Debian, has the current linux agent, and is up to date patch-wise.  The BE job runs as scheduled, but like clockwork, after an hour and 40 minutes, plus or minus 30 seconds, it shows itself as "cancelled" because: "The job was canceled because the response to a media request alert was Cancel, or because the alert was configured to automatically respond with Cancel, or because the Backup Exec Job Engine service was stopped."  That's the message in the backup report. I've not been able to find ANY configuration setting that would cause this response in the current setup. If it were a general setting, it would affect ALL the backup jobs, but this is the only one that complains.

If the job is run manually, it runs to completion with no problems whatsoever. If the job is retried, it runs to completion with no problem whatsoever.

It's almost as though there's some data configuration in the data stream that's being interpreted as a control character or instruction that kills the job, but I would think that running the backup manually would not make any difference were that the case, and the job would still halt. But it doesn't.

This is, admittedly, the only linux server in the entire backup scheme, but that SHOULDN'T matter.

Thanks in advance,

Jim

[Simon_B_]

A media alert in general would be the notification to insert a new tape (because the old one is full or cannot be appended), or was ejected (when not having a library). These alerts can be configured to be auto-responeded to after a certain amount of time.

Go to Alerts/Alert History and look out for the alter that was responed. That should point you to the root cause.

I'd guess this might be something like the last job before the linux machine is configured to eject the tape (again with no library), or the linux job is set to append (and the append period of the media allocated is over), or the linux job is configured to only overwrite media but no overwritable media is present.

[Gregmfg]

You can also set up the alert manually.  Under the alert tab if you hit "configure alert categories" you can tell backup exec to cancel or respond differently after a set period of time.  I had an alert wait a few hours before taking action so I could see specifically what was going on.

[James_Avery]

Weeellll...that's kind of the problem.  There is no alert being responded to. The job is apparently being cancelled because of a response to a non-existant alert. Been looking for a cause with no joy.

Incidentally, I'm backing up to a Quantum DXi7500 VTL, so there's really no problem with availability of empty/overwriteable "tape."

[James_Avery]

Thanks for the replay Greg; no joy there, either.

[James_Avery]

which I've shot.

I got to thinking (a new thing) and looked at the target server in question.  I had been having a problem with the RALUS going unresponsive occasionally, so had created a cron job that would restart it on a daily basis just to make sure that it would behave when the backup time came around.

Wouldn't you know that after I had changed the backup time that the restart time fell right in the middle of the backup window.  Consequently, the backup quit. Cancelled.

Doh!

jba

[HotRob]

I have a job that randomly keep cancelling with the same error and have found in Windows event viewer that LiveUpdate is running and installing a hotfix onto the main backup server.

We are running Backupexec 2010 R2 and i have scheduled the Live Update to check for updates on a Sunday night but for the second time in the past two months the server has decided to do a check at 4am on a thursday morning and install a Hotfix that cancels the backup job that's running at that time because it restarts the Backup Exec server service.

I haven't managed to find anything so far to be able to rectify this and have turned off the LiveUpdate feature. We are looking at moving to BackupExec 2010 R3 soon and then maybe i'll turn LiveUpdate back on.

Cheers,

Rob