I originally tried to post in the OEM forum but didn't get a response after a couple of days so thought I would try here.
curious if anyone that is using OEM to schedule RMAN jobs (RMANScript job types) might have some insight on where this is going wrong.
We are going through a migration of most of our databases to a new platform and decided we would try using OEM 13.4 for all our database jobs including RMAN backups. Previously we had used shell scripts scheduled via cron on each server.
creating a cloud job (Enteprise->job activity) works perfectly fine but during our testing we have noticed that if there is anything that causes the job to fail it does not get rescheduled. As an example today we were doing some testing of our applications and how they respond to services being relocated, instances being brought down etc. This meant our archive log job was failing for a while and exceeded the retries defined in the job. Because the job failed there is no longer a scheduled job. We have to do a create like with the job and give it a new name for it to continue functioning after the problem is resolved.
Editted: After yesterdays testing I increased the job failures to 10 and retry in between to 30 minutes.. I see only one execution today @ 10:11 next one which failed due to DB being down next one was @ 10:41 and never appears to have been executed again. I now have a job that has no further scheduled executions.
Is there something we are doing wrong, faulty thinking, lack of process (should we disable the jobs ahead of time) etc.
Seems wrong to me.. I would rather have the job continue to fail as opposed to just not rescheduling itself after the retries are done. With some of the testing we are doing this is a painful process to have to recreate the jobs.. once things are stable maybe its not such an issue or maybe I am going about this all wrong.
Scott