1 Reply Latest reply on Aug 8, 2017 12:42 PM by MKenji

    Human Task completed but BPM instance doesn't advance or finish

    3354511

      We implemented timers on all Human Tasks for a BPM Process in Oracle BPM Suite 11.1.1.6. Development and QA certificated this change and it was deployed on our productive environment. However for certain number of instances (Rarely, not all of them) a human task is completed but the instance does not advance to the next task or service call.

       

      When we look at the log we get this message at the precise moment of task completion, we updated the logging level of the process to development in EM.

       

      <Jul 26, 2017 2:00:45 PM CLT> <Error> <oracle.soa.bpel.system> <BEA-000000> <Error while invoking bean "cube delivery": Error while cancelling perform.

      error while attempting to execute the cancelPerform method on the work performer for node "EVT18378222982992"; the exception reported is: "".

      An exception was thrown while canceling an activity.

      Set the logging level to debug mode. Check the log for the underlying exception stack trace related to the node.

       

      java.lang.NullPointerException

      at oracle.integration.platform.blocks.scheduler.QrtzSchedulerServiceImpl.listScheduledJobs(QrtzSchedulerServiceImpl.java:110)

      at oracle.bpm.bpmn.engine.model.runtime.util.SchedulingHelper.listScheduled(SchedulingHelper.java:143)

      at oracle.bpm.bpmn.engine.model.runtime.microinstructions.MITimerEventDefineExpirationDate.unschedule(MITimerEventDefineExpirationDate.java:85)

      at oracle.bpm.bpmn.engine.map.builder.visitors.impl.compilers.TimerEventExpireGenerator$DoUnschedule.doExecute(TimerEventExpireGenerator.java:121)

      at oracle.bpm.bpmn.engine.microkernel.MIBase.execute(MIBase.java:34)

       

      ...

       

      <Jul 26, 2017 2:00:45 PM CLT> <Error> <oracle.soa.bpel.engine.dispatch> <BEA-000000> <failed to handle message

      java.lang.NullPointerException

      at oracle.integration.platform.blocks.scheduler.QrtzSchedulerServiceImpl.listScheduledJobs(QrtzSchedulerServiceImpl.java:110)

      at oracle.bpm.bpmn.engine.model.runtime.util.SchedulingHelper.listScheduled(SchedulingHelper.java:143)

      at oracle.bpm.bpmn.engine.model.runtime.microinstructions.MITimerEventDefineExpirationDate.unschedule(MITimerEventDefineExpira

       

      Today it ocurred with 2 out of 250 instances, yesterday 0 times out of 200, the day before 3 times, at different human tasks. And the instances that end OK pass through the same human tasks that fail.