[ https://issues.jboss.org/browse/WFLY-5590?page=com.atlassian.jira.plugin.... ] Hayk Hovsepyan updated WFLY-5590: --------------------------------- Steps to Reproduce: Reproduce using two server. On both servers: 1. Configure server1 and server2 with different node identifiers to use standard object store the same directory. Restart the servers. 2. Enlist test XA resource. 3. Enlist JMS XA resource. 4. Prepare both resources. 5. Crash both servers on the entry of test XA resource commit. 6. Restart server1. 7. Restart server2. 8. Process recovery on server1. 9. Process recovery on server2. 10. Test XA resource on server2 is rollback. For the server which recovery is started the second, has test XA resource rolled-back. However JMS XA resource is committed on both servers. Trace logs of both servers are attached. It happens for shared JDBC object store as well. It can be reproduced by crashrec testsuite: 1. Clone the repo: "git clone git://git.app.eng.bos.redhat.com/jbossqe-eap-tests-transactions.git tests-transactions" 2. Change directory: "cd tests-transactions/jbossts/" 3. Run test: "mvn clean verify -Djboss.dist=$JBOSS_HOME -Dtest=SharedStoreLoadTestCase#commitHaltClientServer" Attachment: Server1.log Server2.log -- This message was sent by Atlassian JIRA (v6.4.11#64026)