|Summary:||Query for process history via REST in jbpm cluster using PER_PROCESS_INSTANCE session strategy fails|
|Product:||[Retired] JBoss BPMS Platform 6||Reporter:||Radovan Synek <rsynek>|
|Component:||Business Central||Assignee:||Marco Rietveld <mrietvel>|
|Status:||CLOSED CURRENTRELEASE||QA Contact:||Radovan Synek <rsynek>|
|Version:||6.0.0||CC:||kverlaen, rsynek, smcgowan|
|Fixed In Version:||Doc Type:||Bug Fix|
|Doc Text:||Story Points:||---|
|Last Closed:||2014-08-06 20:03:27 UTC||Type:||Bug|
|oVirt Team:||---||RHEL 7.3 requirements from Atomic Host:|
|Cloudforms Team:||---||Target Upstream Version:|
Description Radovan Synek 2014-01-23 15:35:24 UTC
Created attachment 854448 [details] REST response for query for process instance history Description of problem: Having a jbpm cluster with two nodes (EAP 6.1.1 domain), following query for process history via REST API fails with response "No session found for context 6" (see an attachment): /business-central/rest/runtime/org.jboss.qa.bpms:clustering:1.0.0.Final/history/instance/6 Corresponding deployment unit has session strategy PER_PROCESS_INSTANCE, I don't get this error with other session strategies Steps to Reproduce: 1. Configure 2 node jbpm cluster on EAP 6.1.1 with a production DB 2. Clone repository git://git.app.eng.bos.redhat.com/bpms-assets.git to your BPMS installation. 3. Build and deploy project clustering (org.jboss.qa.bpms:clustering:1.0.0.Final) with session strategy PER_PROCESS_INSTANCE 4. Try to start any process from the project via REST API on the first node 5. Shut the first down and complete the process on the second node 6. After the process instance has been completed, query for its history (again, on the second node). Additional info: when the cluster is shutting down, error message appears in server log of the second node saying that "java.lang.IllegalStateException: Invalid session was used for this context org.kie.internal.runtime.manager.context.ProcessInstanceIdContext@2deb02a5". Full stacktrace is attached.
Comment 1 Radovan Synek 2014-01-23 15:36:34 UTC
Created attachment 854449 [details] stacktrace from server node two - server shutdown
Comment 2 Marco Rietveld 2014-01-30 18:38:15 UTC
Comment 3 Radovan Synek 2014-02-27 16:34:47 UTC
Although querying for process history via REST works now, I suspect some part of the issue persists, because I can see the same error message in the node two server log (attaching).
Comment 4 Radovan Synek 2014-02-27 16:36:01 UTC
Created attachment 868634 [details] stacktrace node two, 6.0.1.ER1
Comment 5 Marco Rietveld 2014-02-28 02:25:33 UTC
Radovan, I agree that there's still an issue -- but not that it is related to this bugzilla. Could you please create a new issue? The exception shown in the logs is *not* related to the REST performance. If you think that it is a REST issue, please let me know why!