Forum Stats

  • 3,727,187 Users
  • 2,245,333 Discussions


Long running backup job email

3433105 Member Posts: 9
edited August 2020 in Enterprise Manager

Hello Team,

I need to schedule a job to get email notification for long running backup jobs from oem for group of servers.

For example: nonprod oem group. Need to get email notification for long running backup jobs in nonprod

oem group. Kindly help me.


  • Venkata Thiruveedhi-Oracle
    Venkata Thiruveedhi-Oracle Posts: 590 Employee
    edited August 2020


    I do not see any incident that gets triggered if the Job is taking long time.

    We can have a notification sent based on the Job Status change which is explained in the following document:
    EM 12c, EM 13c: How to Configure Incident Rules to Send Notifications for Job Executions (Status Changes) in Enterprise Manager Cloud Control (Doc ID 1386816.1)

    But for the job duration, i do not think we have any.

    But you can write a custom query to capture the duration of the backup job. Create a Metric Extension (ME) using this.
    Deploy it to all your targets and make it a regular metric.

    Then you can add this metric to your rules and get it triggered based on the threshold you set.

    Best Regards,

  • Joerg.Sobottka
    Joerg.Sobottka Senior Consultant and Oracle Ace Member Posts: 589 Bronze Trophy
    edited August 2020

    Like Venkata said. That isn't something out of the box. But I think (I can't test it at the moment) you can put your own monitoring select / your own event for that. As I said, I can't test it, but in my opinion the MGMT$JOB view is the right one:

  • EdStevens
    EdStevens Member Posts: 28,024 Gold Crown
    edited August 2020

    Are you wanting an email to say "hey, backup xyz is still running but we think it should be done".  What if the backup actually completes one second after the alert is triggered? 

    Or do you want an email after the fact to say "hey, backup xyz completed, but it took longer than it should".

    How do you define "long running" and the trigger for the alert?

    BTW, email has an uncontrollable latency.

    Do you really want to get an alert that job is running 'too long', only to discover it actually completed immediately after the alert was triggered?  What real business problem is this solving?

Sign In or Register to comment.