Auto restart of failed jobs

[smwoodcrafts]

I seem to be having an issue with a job that's failing intermittently. The policy this server (LINUX) is in has 7 backup selections. It takes a checkpoint every 15 minutes and multiple data streams is allowed. The schedule is frequency based.

Master and Media servers are 2008 R2 running 7.5.0.6

Client is SUSE version 10 patch level 2 running 6.5 client

There is an intermittent failure, error code 41, on one of the child jobs backing up /var.

We are working on the failure, but in this post the issue I'm looking at is that the jobs seem to be restarting automatically. Our System Admin will restart the failed child job for /var, but there more is restarted than just that job. The parent job is restarted as well as the child job. But none say, in the datails of the failed job, that it was restarted and no second attempts within the detils either. This can happen several time in a night. 

So is there a setting that I am missing that has NBU restart the failed jobs? Is there something in OpsCenter that would do this?

[Nicolai · Accepted solution]

Netbackup will automic re-try failed file systems backups.

The retry behaviour is configured in GUI - Netbackup Mangement - > Host Properties -> Master Server -> Rigt click -> Properties - > Global Attirbutes

• Job Retry delay
• Schedule Backup Attempts

The Help provioded will high level explain how the settings works.

[smwoodcrafts]

Ok, I found it, thanks. A couple questions to make sure I have it right. It will only re-try failed systems backups? These are started as a new job, not another attempt within the failed job, is that consistent with this setting? With multiple data streams enabled, if one child job fails, the parent job fails. With this setting, would it restart both the parent and child jobs?

[Nicolai]

correct - only file system backup will auto retry. Database backup like DB2, Oracle and SAP do not do auto retry (expection is Exchange and Lotus Notes - but they behave sort of like file systems backup)

As far as I can re-call, retry are spawned as new job, not retry for a existing backup - so correct.

Last question - Parent jobs are "control" for it children. I would expect a restart to start the control job and the failed stream.