You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@activemq.apache.org by "Timothy Bish (JIRA)" <ji...@apache.org> on 2011/08/17 01:25:27 UTC

[jira] [Closed] (AMQ-3405) DLQ messages moved/copied into original queue disappear after failing processing for a second time

     [ https://issues.apache.org/jira/browse/AMQ-3405?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Timothy Bish closed AMQ-3405.
-----------------------------

    Resolution: Not A Problem

Working as designed.  See AMQ3405Test.java in trunk for an example.

> DLQ messages moved/copied into original queue disappear after failing processing for a second time
> --------------------------------------------------------------------------------------------------
>
>                 Key: AMQ-3405
>                 URL: https://issues.apache.org/jira/browse/AMQ-3405
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: Broker
>    Affects Versions: 5.4.2
>         Environment: Windows 7, Java 64bit 1.6 update 20, Client and server 5.4.2
>            Reporter: Carey Flichel
>
> 1. Message in queue A fails enough times to be moved to the DLQ.
> 2. We move/copy the message back to queue A to attempt reprocessing. Invoking the retry operation from JMX can also be used.
> 3. The consumer once again fails to consume the message greater than the max number of times.
> 4. The message is not put back into the DLQ. If it was moved rather than copied, it means we have a message that was never successfully consumed, but has nevertheless disappeared.
> This is all with the default ActiveMQ settings, though I tried adjusted the DLQ configuration to support both expired and non-persistent messages.
> I found this posting which seems to indicate the same issue I am seeing: http://www.mail-archive.com/users@activemq.apache.org/msg09683.html. One difference that I am seeing is that restarting ActiveMQ does not cause the message to show up in the DLQ again. Once it is lost, it is lost forever.
> To replicate this, I believe you can use a consumer that just fails on every invocation.
> One interesting thing is that if I put a message into the DLQ directly, then move it to a queue with a consumer, it will end up in the DLQ again as you would expect. If I then move the DLQ message back to the original queue, it will then show the bevahiour outlined above.

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira