We have 4 master servers in our enviroment, sometimes a backup log scheduled with frequency option dont start automatically.
Mostly times the problem begins after a status codes different than 0.
Sometimes when we start manually works fine for a few days or hours. The problem happens with different clients and master.
This problem is less common with SQL clients. And the workaround was use the server crontab.
All masters have Linux Red Hat 7.7 and Netbackup(Master, Media and Clients) 8.1.2
We heard about a bug in Netbackup 7.7.3, maybe we have this bug in 8.1.2? There is any log to help us?
Other thing that maybe is related, when we use the command "nbpemreq -suspend_scheduling" dont suspend the schedules.
I think you'll find this behaviour is by design. There are settings you can change to reduce the restart delay after a failure
See these two posts that I think address the issue you are having:
No its different, in the posts that you mentioned the backup failed and they want to auto restart the job. That is not my case.
Just happens with frequency type schedules, like archive logs backups(Sybase, SAP, Oracle) and the backup don't start, there is no errors, we have to start manually.
If the problem is not related to a job restart delay after a failure (default period is I believe 12 hours to wait after any failure for that policy/schedule to restart irrespective of the schedule frequency), then we will need more details.
What policy types and schedules are affected.
What errors (if any) preceed the start of the problem.
What is the "bug" you are referring to - do you have a EEB related to this.
Otherwise, the nbpem logs are probably where you want to start looking.
Thanks David, I will take a look at the nbpem logs, but the problem its intermittent so I have to wait. This is one reason of archive log area become full.
My problem is more often with Oracle and SAP types, and with Sybase and SQL types its less common. The schedules is based on frequency, 1hour, 30 or 15 minutes.
I didn't find the documentation, but the suggestion was stop/start the services.