Description of problem: If one EAR contains more than one module with a META-INF/services/org.jboss.msc.service.ServiceActivator file the services are activated multiple times (as ServiceActivator files are found). If the file is moved to the ear/META-INF as the only one there is no activation. If the ear and a jar contain it it will fail with a org.jboss.msc.service.DuplicateServiceException.
Created attachment 1103808 [details] A simple example based on the quickstarts
Hello. I think that this PR fixes the issue : https://github.com/jbossas/jboss-eap/pull/2640 Added also another simple test, which is based on the quickstart example.
Created attachment 1105573 [details] Uploading the test .ear of wfink
Created attachment 1105574 [details] Adding the stack trace for wfink's test .ear
Created attachment 1105949 [details] Test for Wildfly added.
Executing the above test for wildfly shows that Multiple Triggers of the Service Activator exist also upstream.
Resolved by : https://github.com/jbossas/jboss-eap/pull/2640
Panagiotis Sotiropoulos <psotirop> updated the status of jira JBEAP-2402 to Closed
Panagiotis Sotiropoulos <psotirop> updated the status of jira JBEAP-2402 to Reopened
Panagiotis Sotiropoulos <psotirop> updated the status of jira JBEAP-2402 to Resolved
This bug was accidentally moved from POST to MODIFIED via an error in automation, please see mmccune with any questions
https://github.com/jbossas/jboss-eap/pull/2770
This is not fixed in EAP 6.4.8.CP.CR2. I tested with the reproducer provided by wfink and the scenario with 3 service jars was still failing. The fix provided by Panagiotis in https://github.com/jbossas/jboss-eap/pull/2770 seems to fix this.
Since there is a fix moving back to POST
Rescheduling to CP9. A rouge commit was introduced when trying to fix this in CP8. It should be reverted when applying the fix in CP9.
Verified with EAP 6.4.10.CP.CR2.
Retroactively bulk-closing issues from released EAP 6.4 cummulative patches.