[jira] [Commented] (ARTEMIS-1691) JMS Bridge: After the FailureHandler exits, a bridge cannot be started full functional anymore

["Justin Bertram (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/ARTEMIS-1691?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16369468#comment-16369468 ] 

Justin Bertram commented on ARTEMIS-1691:
-----------------------------------------

You can send a PR to https://github.com/apache/activemq-artemis.  Please include a test with your fix so it can be verified and also to avoid regressions.  You could probably add your test to {{org.apache.activemq.artemis.tests.extras.jms.bridge.JMSBridgeTest}}.

> JMS Bridge: After the FailureHandler exits, a bridge cannot be started full functional anymore
> ----------------------------------------------------------------------------------------------
>
>                 Key: ARTEMIS-1691
>                 URL: https://issues.apache.org/jira/browse/ARTEMIS-1691
>             Project: ActiveMQ Artemis
>          Issue Type: Bug
>    Affects Versions: 2.4.0
>         Environment: WildFly Servers.
>            Reporter: Markus F
>            Priority: Minor
>
> h4. Setup:
> JMSBridge on server a pulls messages from a remote server b.
> failureRetryInterval = 1 sec
> maxRetries = 1
>  
> h4. To reproduce:
> Restart the remote server (need to take more than one second.)
> JMS Bridge prints out:
> {noformat}
> JMS Bridge unable to set up connections, bridge will be stopped{noformat}
> After the remote server is up again, start the bridge again.
> JMS Bridge prints out:
> {noformat}
> Started JMS Bridge testbridge{noformat}
>  
> h4. Result:
> The bridge ist started, but does still not consume messages
>  
> h4. Reason:
> In the Class JMSBridgeImpl line 1624, the failed boolean parameter is checked. It is not set to true anywhere outside the failure handlers (FailureHandler and StartupFailureHandler) or on initial adding of the bridge.
> But the FailureHandler has already exited. So the bridge will wait all time long for 500 ms in line 1626.
>  
> h4. Solution:
> I would add a
> {code:java}
> failed = false;{code}
> after line 403.
> I don't know how the release process is handled here, if I can make a pull request or something like that, I will do, if it is needed.
>  
> Cheers,
> Markus



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