You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@cxf.apache.org by "Richard McLaren (JIRA)" <ji...@apache.org> on 2018/07/02 14:50:00 UTC

[jira] [Updated] (CXF-7774) JMS Reply Queue handles leaking on IBM MQ.

     [ https://issues.apache.org/jira/browse/CXF-7774?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Richard McLaren updated CXF-7774:
---------------------------------
    Description: 
Using 3.1.11 and JMS we are seeing handles for our Reply Queue are being left open. 

This is using a SpringBoot application and IBM MQ client jars.

The most likely suspect is in JMSConduit the field 

staticReplyDestination is set to null on a JMSException but the existing MQQueue is never closed.

I have tried simulating the problem in IntelliJ by forcing a JMSException during the SendExchange.

Should there be a "ResourceCloser.close(

staticReplyDestination);" immediately prior to setting 

staticReplyDestination to null?

  was:
Using 3.1.11 and JMS we are seeing handles for our Reply Queue are being left open. 

This is using a SpringBoot application and IBM MQ client jars.

The most likely suspect is in JMSConduit the field 

staticReplyDestination is set to null on a JMSException but the existing IBMQueue is never closed.

I have tried simulating the problem in IntelliJ by forcing a JMSException during the SendExchange.

Should there be a "ResourceCloser.close(

staticReplyDestination);" immediately prior to setting 

staticReplyDestination to null?


> JMS Reply Queue handles leaking on IBM MQ.
> ------------------------------------------
>
>                 Key: CXF-7774
>                 URL: https://issues.apache.org/jira/browse/CXF-7774
>             Project: CXF
>          Issue Type: Bug
>          Components: JMS
>    Affects Versions: 3.1.11
>            Reporter: Richard McLaren
>            Priority: Major
>
> Using 3.1.11 and JMS we are seeing handles for our Reply Queue are being left open. 
> This is using a SpringBoot application and IBM MQ client jars.
> The most likely suspect is in JMSConduit the field 
> staticReplyDestination is set to null on a JMSException but the existing MQQueue is never closed.
> I have tried simulating the problem in IntelliJ by forcing a JMSException during the SendExchange.
> Should there be a "ResourceCloser.close(
> staticReplyDestination);" immediately prior to setting 
> staticReplyDestination to null?



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)