Restarting tomcat did not cancel the stucked task

Administration & Troubleshooting
, , , , , , , ,
1

We have a linux servers aggregation task and it has been stuck and we did restart tomcat.The task is still pending after restart.normally all pending task will be cancelleted after tomcat restart.any idea ? how to kill that.

2

you need to restart all the sp linux servers as well in order to clear the job tasks that’s hanging.

3

I never tried working IIQ on Linux but o my end when any tasks are stuck, I delete them in the debug page under TaskResults in debug. When I come back to the Main Task Results page and look for the task that I deleted, it is deleted. Can you try doing the delete in the debug and see if it works?

4

@ksuneeth85

To cancel a hung task, perform these steps:

  1. Attempt to terminate the task in the Task Results tab by right clicking and select Terminate. Wait a few minutes to see if it cancels.
  2. If the task does not terminate in step 1, go to the debug menu and find the hung task in TaskResult objects, select the check box next to it, then click Delete from the drop down in the upper right. Once the TaskResult is deleted, if a partitioned task, go to step 3. Otherwise you are done.
  3. If it is a partitioned task, go to Debug->Request objects and look for any related partition Request objects. These will be named the same as the partition was listed in the task result tab under the Name column. If there are objects which did not get cleaned up when the TaskResult was deleted, you will need to delete these objects too. Verify these objects are indeed orphaned before removing them.
  4. Terminating or deleting the tasks and request partitions will clear up the task, but the threads associated with these tasks may not be cleared until you perform an application server restart.

Terminating a task in the UI or deleting objects through debug or iiq console will call the Terminator class, so normally everything is cleaned up properly. Occasionally there may be situations where not everything is cleaned up and the resulting corruption to a database object will require the additional intervention mentioned above. It is always good to check and verify that all associated objects are removed.

If you have completed or verified all steps, then you should be good to resubmit the task. If a partitioned task is repetitively cancelled/deleted and the Request objects become orphaned and are not cleaned up, this can cause subsequent executions of the task to hang due to the orphaned partition (Request) objects.