cancel
Showing results for 
Search instead for 
Did you mean: 

Incomplete Job Alert Reporting on All Policies

Mr_White
Level 2

As each of our NetBackup 7.5.03 media servers are in different offices and managed by specific IT staff, we have several incomplete job alerts configured for each of our media servers.  However, we found that all incomplete job alerts were reporting for all incomplete jobs, regardless of which server reported the incomplete job.  I found that all policies were selected for all alerts, so I fixed this by selecting the office specific servers for each of the incomplete job alerts as appropriate.

However, after saving these changes for each of the policies on Friday morning, I found that all the alerts reverted back to selected all policies for all offices this morning.  We did restart the master sever Friday afternoon, but I would not think this would cause the issue.

I did "save" each one and went back to verify that each one was correctly set (Friday morning).  But, sometime prior to Saturday @10AM these alerts changed and an incomplete job for one of our offices was reported by all alerts (my manager was flooded by these emails as she receives all alerts).  To my knowledge no one changed the alerts, so I assume something in NetBackup reverted the alerts back to their original state - possibly the master server reboot (planned reboot to turn logging back off for another unrelated issue).

Does anyone know what may cause this?

Just to add, in OpsCenter I created a view for just my office policies and changed my office incomplete job alert to use that view (instead of default viewing all policies), then checked the policies.  I wouldn't think changing the view would "fix" anything, but figure I would try it. 

Thanks,

Matt

1 REPLY 1

Zenshai
Level 4

I've seen rollbacks of settings and collected data if the DB is not given a chance to shut down cleanly. The DB can take a long time to exit so if you issue server shut down it will consider that service as not responding after a while and kill it. Best thing to do is to manually down OpsCenter before you shut down the server.