[jboss-jira] [JBoss JIRA] (WFLY-5536) Recovery manager is not able to recover MDB and shows warnings on each run.

[Jeff Mesnil (JIRA)]

    [ https://issues.jboss.org/browse/WFLY-5536?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13123928#comment-13123928 ] 

Jeff Mesnil commented on WFLY-5536:
-----------------------------------

There was an issue in the WildFly module definitions and the Artemis RA was using the default Artemis XAResourceWrapper instead of the one specific to WildFly.

Before, the trace was:

{noformat}
16:58:57,500 TRACE [org.apache.activemq.artemis.ra] (EJB default - 2) XAResource=org.apache.activemq.artemis.service.extensions.xa.ActiveMQXAResourceWrapperImpl at 3696bea
{noformat}

After the change (to import the services defined in the messaging-activemq extension) in https://github.com/wildfly/wildfly/compare/master...jmesnil:WFLY-5536_recovery_manager, I now see:

{noformat}
17:17:26,576 TRACE [org.apache.activemq.artemis.ra] (EJB default - 4) XAResource=org.jboss.activemq.artemis.wildfly.integration.WildFlyActiveMQXAResourceWrapper at 7cd8de4e
{noformat}

This org.jboss.activemq.artemis.wildfly.integration.WildFlyActiveMQXAResourceWrapper class implants org.jboss.tm.XAResourceWrapper while its superclass is not implementing it.

However even with that change, the test is still failing.

For one transaction I can see that the Artemis RA start, ends and commits its xares:

{noformat}
17:16:03,598 TRACE [org.apache.activemq.artemis.ra] (EJB default - 5) start(< formatId=131077, gtrid_length=29, bqual_length=36, tx_uid=0:ffffc0a80104:2ff694c:563397b2:1d, node_name=1, branch_uid=0:ffffc0a80104:2ff694c:563397b2:22, subordinatenodename=null, eis_name=java:/JmsXA NodeId:7a934c2e-7f21-11e5-83e2-a7afda8ee76d >, 0)
...
17:16:03,616 TRACE [org.apache.activemq.artemis.ra] (EJB default - 5) end(< formatId=131077, gtrid_length=29, bqual_length=36, tx_uid=0:ffffc0a80104:2ff694c:563397b2:1d, node_name=1, branch_uid=0:ffffc0a80104:2ff694c:563397b2:22, subordinatenodename=null, eis_name=java:/JmsXA NodeId:7a934c2e-7f21-11e5-83e2-a7afda8ee76d >, 67108864)
...
17:16:03,618 TRACE [org.apache.activemq.artemis.ra] (EJB default - 5) commit(< formatId=131077, gtrid_length=29, bqual_length=36, tx_uid=0:ffffc0a80104:2ff694c:563397b2:1d, node_name=1, branch_uid=0:ffffc0a80104:2ff694c:563397b2:22, subordinatenodename=null, eis_name=java:/JmsXA NodeId:7a934c2e-7f21-11e5-83e2-a7afda8ee76d >, true)
{noformat}

But for the transaction that's making the test fail (0:ffffc0a80104:2ff694c:563397b2:25), I see not such start/end/commit logs.
I only see:

{noformat}
17:16:36,390 WARN  [com.arjuna.ats.jta] (Periodic Recovery) ARJUNA016037: Could not find new XAResource to use for recovering non-serializable XAResource XAResourceRecord < resource:null, txid:< formatId=131077, gtrid_length=29, bqual_length=36, tx_uid=0:ffffc0a80104:2ff694c:563397b2:25, node_name=1, branch_uid=0:ffffc0a80104:2ff694c:563397b2:28, subordinatenodename=null, eis_name=java:/JmsXA NodeId:null >, heuristic: TwoPhaseOutcome.FINISH_OK, product: ActiveMQ Artemis/1.1.0.wildfly.007, jndiName: java:/JmsXA NodeId:null com.arjuna.ats.internal.jta.resources.arjunacore.XAResourceRecord at eabf2ff >
{noformat} 


> Recovery manager is not able to recover MDB and shows warnings on each run.
> ---------------------------------------------------------------------------
>
>                 Key: WFLY-5536
>                 URL: https://issues.jboss.org/browse/WFLY-5536
>             Project: WildFly
>          Issue Type: Bug
>          Components: JMS
>    Affects Versions: 10.0.0.CR2
>            Reporter: Hayk Hovsepyan
>            Assignee: Jeff Mesnil
>            Priority: Blocker
>         Attachments: MDB_JMS_XA.log
>
>
> Recovery manager is not able to recover MDB and shows warnings.
> This causes to see the same warning log in server each time recovery is run.
> This happens when server is crashed before transaction state is saved.



--
This message was sent by Atlassian JIRA
(v6.4.11#64026)