Checking for Incidents Related to a Failed Job

It is possible for a job to fail because of an internal code error, a severe scaling issue, or other Enterprise Manager issue for which you may be able to investigate an incident or event trail. For example, if the OMS bounced because all Job Workers were stuck, this would cause many jobs to fail. If the loader were failing, that could also cause some jobs to fail.

  1. Check for incidents or alerts in the time-frame of the job.

    • To check for incidents, select Summary from the Enterprise menu of the Enterprise Manager console, then view the Incidents section of the Enterprise Summary page.

      For more detailed information, select Monitoring from the Enterprise menu, then select Support Workbench.

    • To check for alerts,

  2. Submit a service request with the related incident(s) or event data.

    All step output, error output, logging, remote log files, and incident dump files for a given job are captured for an incident.

    • To submit a service request, select My Oracle Support from the Enterprise menu of the console, then select Service Requests.

    • To create a technical SR, click Create SR on the Service Requests Home page. To create a contact us SR, click Create "Contact Us" SR at the top of the Contact Us Service Requests region, or click Contact Us at the top of any My Oracle Support page. If you are creating a technical service request, depending on the Support IDs registered in your profile, you can create hardware or software SRs, or both.

      The Create Service Request wizard guides you through the process of specifying product information and attaching configuration information to the SR when it is filed with Oracle Support. To ensure that Oracle Support has the most accurate target information, select the Configuration tab in the What is the Problem? section of Step 1: Problem, then select a target.

  3. Apply a patch that support provides.

    • Select Provisioning and Patching from the Enterprise menu of the console, then select Patches & Updates.

    • Provide login credentials, then click Go.

    • Access the online help for assistance with this page.

  4. Try to submit the job again after applying the patch.

To package an incident or manually trigger an incident:

  1. Access Support Workbench.
  2. Gather all job-related dumps and log files, as well as other data from the same time, and package it for Support.
  3. Review the incident-related data in Support Workbench, searching for relevant errors.
  4. If you determine the root cause without support intervention, fix the job and resubmit it.