1 Reply Latest reply on Jun 7, 2013 12:50 PM by Wolf-Dieter Fink

    How do we deploy a JBoss singleton timer that needs to be persisted across EAP 6.1.0 nodes

    Sudhir Jena Newbie

      Hi,

       

      We have deployed a singleton timer service with the following frequency:

      0 */1 * * * * *

      This works fine on standalone but when I deploy it in a cluster using ha-profile with 2 nodes, the timer starts on both nodes simultaneously.

       

      Log from node 1:

       

      13:45:00,019 DEBUG [com.xxxxxxxxxxx.tp.integration.maxis.ScanTimer] (EJBdefault - 3) oldschedule=ScheduleExpression[second=0 minute=*/1 hour=* dayOfWeek=* dayOfMonth=* month=* year=* start=null end=null timezone=null], newSchedule=ScheduleExpression[second=0 minute=*/1 hour=* dayOfWeek=* dayOfMonth=* month=* year=* start=null end=null timezone=null]

      13:45:00,020 DEBUG [com.xxxxxxxxxxx.tp.integration.maxis.ScanTimer] (EJBdefault - 3) time to scan.

       

      Log from node 2:

      13:45:00,020 DEBUG [com.xxxxxxxxxxx.tp.integration.maxis.ScanTimer] (EJBdefault - 6) oldschedule=ScheduleExpression[second=0 minute=*/1 hour=* dayOfWeek=* dayOfMonth=* month=* year=* start=null end=null timezone=null], newSchedule=ScheduleExpression[second=0 minute=*/1 hour=* dayOfWeek=* dayOfMonth=* month=* year=* start=null end=null timezone=null]

      13:45:00,020 DEBUG [com.xxxxxxxxxxx.tp.integration.maxis.ScanTimer] (EJBdefault - 6) time to scan.

       

      The intent here is to see that the timer starts on node 2 only if node 1 is unavailable.

      Please help me have definitive implementatable answers as soon as possible.