[jira] [Commented] (AMQNET-401) NullReferenceException from Apache.NMS.ActiveMQ.Transport.InactivityMonitor

["Markus Karolus (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/AMQNET-401?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13547759#comment-13547759 ] 

Markus Karolus commented on AMQNET-401:
---------------------------------------

In the last weeks we have monitored the unhandled Nullpointer exception numerous times. It’s always on the same test/development machine - we are starting the whole environment with multiple services that are connected via NMS. The ActiveMQ server instance also runs on that box. During the start-up phase the CPU runs at 100% and the physical memory runs out so the virtual memory produces a lot of IO calls on the slow hard disk. In most of our start-ups one of our services with the highest memory footprint dies because of the unhandled Nullpointer exception.

The Nullpointer  exception occurs because in the  „StopMonitorThreads“ the variable „asyncTasks“  will be set to null – With the warning that the connectionCheckTimer can’t be stopped within a 3 seconds timeout. So the timer is currently running – and throws some time later the exception with the call this.asyncTasks.Wakeup().

Duplicate Ticket: AMQNET-402
                
> NullReferenceException from Apache.NMS.ActiveMQ.Transport.InactivityMonitor 
> ----------------------------------------------------------------------------
>
>                 Key: AMQNET-401
>                 URL: https://issues.apache.org/jira/browse/AMQNET-401
>             Project: ActiveMQ .Net
>          Issue Type: Bug
>          Components: NMS
>    Affects Versions: 1.5.6
>         Environment: Windows Server 2008 R2, .Net 4.0
>            Reporter: Markus Karolus
>            Assignee: Jim Gomes
>            Priority: Critical
>
> It's the same problem which is descibed in,
> https://issues.apache.org/jira/browse/AMQNET-339
> in the current version (1.5.6) the real problem is not fixed, so I created this new issue.
> Today we had an crash on one of our enviromental boxes:
> Directly after a failover we got the following message from the InactivityMonitor "InactivityMonitor[x]: Timer Task didn't shutdown properly."
> Afterwards the whole process shuts down after an UnhandledException handling. In the WriteCheck method a NullReferenceException was thrown.
> It seems that during the StopMonitorThreads the timer connectionCheckTimer couldn't stop correctly and ansyncTasks exists. Directly after that the asyncTasks will be set to null -> In the currenr running WriteCheck a NullReferenceException occures. 
> (This problem was already commented in AMQNET-339 as last comment from Matthew Good)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira