activemq-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Gary Tully (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (AMQ-3874) Unable to destroy a durable subscription
Date Thu, 14 Jun 2012 14:00:44 GMT

    [ https://issues.apache.org/jira/browse/AMQ-3874?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13295037#comment-13295037
] 

Gary Tully commented on AMQ-3874:
---------------------------------

A thread dump of the broker will give an indication if there are any connections pending or
pending close.

when the perl client goes away, it is the tcp client side socket that closes but it may not
initiate the broker side close, that may need to timeout the read or even a pending write.
I presume with lsof you are trying to eliminate that. Are u using it on the broker side?
                
> Unable to destroy a durable subscription
> ----------------------------------------
>
>                 Key: AMQ-3874
>                 URL: https://issues.apache.org/jira/browse/AMQ-3874
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: Broker
>    Affects Versions: 5.5.1
>         Environment: solaris, linux
>            Reporter: Bhanu
>
> I think theres an issue that occurs sometimes while removing durable subscriptions from
the broker. This has hit us second time in a week that when a perl durable subscriber using
stomp goes down, it is not shown as inactive in broker. Even though there were no active connections(did
a thorough lsof check) somehow broker thought that the subscriber was active. Whats more worse
is trying to destroy that subscription via jconsole gives an error :
> "Problem invoking destroy: java.rmi.UnmarshallException: Error unmarshalling return;
nested exception is java.lang.ClassNotFoundException: javax.jms.InvalidDestinationException
(no security manager: RMI class loader disabled)"
> Since we cannot destroy this subscription, the client cannot come up with the same client
Id and subscription name. This is not reproducible easily
> but is some weird corner case. Can somebody please check. Let me know if you have more
queries.
> Thanks & Regards
> Bhanu

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Mime
View raw message