8 Replies Latest reply: Mar 4, 2013 4:46 PM by Sunthar Tharmalingam RSS

    Reports and Jobs not running in 12C

    user10745648
      I have successfully upgraded OEM OMS 10 to OMC12c, all seems well.
      Over all schedule jobs/reports ran at different time (reports showing as generated at 01.01 hrs), none of my reports/jobs scheduled for today have been generated.

      When I tried to run a report to run in real time (i.e one time immediately), nothing happens and the report does not run/generate. Also When I submit a job and schedule it to run immediately, it shows a status of “Scheduled” but and the scheduled time is showing as the current time., but the job doesn’t run
        • 1. Re: Reports and Jobs not running in 12C
          Courtney Llamas-Oracle
          Check 2 things:

          job_queue_processes > 0
          dbms_scheduler is enabled

          The jobs are submitted using dmbs_scheduler in 12c, and it also relies on the job_queue_processes.
          • 2. Re: Reports and Jobs not running in 12C
            user10745648
            Yes both
            job_queue_processes > 0
            dbms_scheduler is enabled


            NAME TYPE VALUE
            ------------------------------------ ----------- ------------------------------
            job_queue_processes integer 10

            Thanks
            • 3. Re: Reports and Jobs not running in 12C
              Courtney Llamas-Oracle
              That doc is talking about RAC Clusters. For SG clusters, what you need is http://docs.oracle.com/cd/E11857_01/em.111/e16790/ha_multi_resource.htm#autoId68
              This will tell you how to configure an agent on each node, and relocate the database target based on package location. If you have multiple packages (1 db per package) this is the recommended method.

              The alternative, is to use ORACLE_HOSTNAME=<package vip) during the install, which means you end up with 1 agent per package and is a lot of overhead and maintenance (How to Configure Grid Control Agents to Monitor Virtual Hostname in HA environments (Doc ID 406014.1)). I highly recommend the relocate target method.

              In 11g, you can also use the emctl relocate command, instead of having to install EMCLI on each target node. You first must set the agent pairs that a target can move between with emcli set_standby_agent (once).
              For information on the EMCLI Method see, Setup and Configure Target Relocate Using EMCLI 10.2 or 11.1 (Doc ID 577443.1)

              1) Make sure you have 11g Agents (or higher)
              2) Install a single Agent on each of the Cluster nodes
              3) Discover the failover targets on each of the active cluster nodes
              4) Flag the targets, so they are allowed to fail-over to another Agent
              via EMCTL. Repeat for each target:
              $ emcli set_standby_agent -src_agent="<primary>" -dest_agent="<backup>" -target_name="<name>" -target_type="<type>"

              On a 3-node (or more) cluster, simply run the EMCLI command for each standby node (changing the dest_agent parameter)
              This setup is just a one-time thing, for each target in the cluster that needs to failover to one or more nodes.

              5) Now that the ground work has been done, you have 2 ways of making the target 'move' to another Agent:

              A> You can use EMCLI to relocate the target:
              $ emcli relocate_targets -src_agent=<old agent> -dest_agent=<new agent> -target_name=<name> -target_type=<type> -copy_from_src -force=yes

              Pro:
              - Can be done 'centrally', so an admin can use this to do a manual failover

              Con:
              - If you want to automate this, using the cluster failover scripts, you will need to deploy EMCLI everywhere (more maintenance)

              B> You can use EMCTL on the 'new agent' to force a failover to that Agent:
              $ emctl relocate_target agent '<name>' '<type>'

              Pro:
              - This is the way to automate the failover in cluster scripts

              Con:
              - You need to be in the Agent home of the surviving machine to run this command (you can ONLY pull targets to the Agent you are running from)
              - An Agent can only 'assume' responsibility if it is allowed to have this target (this is the reason for running the emcli standby_agent commands)
              • 4. Re: Reports and Jobs not running in 12C
                user10745648
                Sorry my agents are all working okay, I am not sure of the relevance of the above to my problem please
                • 5. Re: Reports and Jobs not running in 12C
                  Courtney Llamas-Oracle
                  Sorry, that post was for a different thread ;)
                  • 6. Re: Reports and Jobs not running in 12C
                    Courtney Llamas-Oracle
                    Can you check the Setup -> Manage Cloud Control -> Overview page and see if the jobs for EM are all running (green)
                    • 7. Re: Reports and Jobs not running in 12C
                      user10745648
                      The the EM Jobs Service/system shows Status Pending. How can I fix this
                      • 8. Re: Reports and Jobs not running in 12C
                        Sunthar Tharmalingam
                        did you try to recreate a job to validate the result of the run?