-
Bug
-
Resolution: Done
-
Major
-
7.11.1.GA
-
2 nodes with job failover configuration
-
False
-
False
-
-
-
-
-
-
CR1
-
?
-
?
-
-
-
2022 Week 05-07 (from Jan 31), 2022 Week 08-10 (from Feb 21), 2022 Week 11-13 (from Mar 14), 2022 Week 14-16 (from Apr 4), 2022 Week 17-19 (from Apr 25), 2022 Week 20-22 (from May 16), 2022 Week 23-25 (from Jun 6), 2022 Week 26-28 (from Jun 27), 2022 Week 29-31 (from Jul 18)
Using a 2 node cluster we can see that jobs are added to the cache as expected:
DEBUG [org.kie.server.services.jbpm.cluster.ClusteredJobFailOverListener] (default task-19) Adding job scheduled AsynchronousJobEventImpl [job=RequestInfo{id=8278, ...] for failover
However, the corresponding "Removing" message is not executed, and when the node is shut down, all jobs are being retried on the other running node:
WARN [org.jbpm.executor.impl.jpa.ExecutorRequestAdminServiceImpl$RequeueRunningJobCommand] (thread-34,ejb,mweiler) Error while trying to requeue jobs that runs for too long Retrying completed or cancelled job is not allowed (job id 8278)
It seems hasListeners in ExecutorEventSupportImpl does not return true.
- is cloned by
-
-
- Resolved
-
