[
https://issues.apache.org/jira/browse/APLO-172?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13278930#comment-13278930
]
Richard Fox commented on APLO-172:
----------------------------------
I've noticed this occurring in a 1.3 snapshot too. (20120512.032517-40) running
ubuntu 11.04
Test case is as simple as, Connect Loads of non web socket connectors to one
queue. send some messages over it. Connect one web socket connector and then
disconnect the web socket connector.
Apollo goes to 99% CPU everyone disconnects/hangs
I've rolled back my server to 1.1 and do not see the issue on 1.1 strangely.
> Broker does not respond to any publisher/subscriber after disconnecting and
> connecting one client/browser
> ---------------------------------------------------------------------------------------------------------
>
> Key: APLO-172
> URL: https://issues.apache.org/jira/browse/APLO-172
> Project: ActiveMQ Apollo
> Issue Type: Question
> Components: apollo-broker, apollo-distro
> Affects Versions: 1.1
> Environment: Windows
> Reporter: Shahinur Alam
> Assignee: Hiram Chirino
> Priority: Critical
> Labels: Apollo, stomp, websocket
> Original Estimate: 24h
> Remaining Estimate: 24h
>
> Hi, I am using apollo broker (Apache ActiveMQ Apollo 1.1-SNAPSHOT) and i have
> a lot of subscriber. i am using stomp and websocket to feed the message to
> client. But i am facing two problems:
> 1. when one client/subscriber is disconnected(browser is closed) once and if
> he/she try to connect(opening browser and requesting to the broker ) again
> then broker does not respond to any clients/ publisher/subscriber those were
> connected to broker.
> 2. if one browser is flooded with data and can not receive further message
> then broker stops responding to all clients/ publisher/subscriber.
> After restarting broker, it works again . So it is difficult to restart
> broker repeatedly. I think broker is pointing to the stale connection which
> has been closed or it is facing problem to maintain connection when one
> client is trying to reconnect.
> can anyone help me? So that all the live publisher/subscriber will not be
> affected by any one.
> Thanks in advance
> Shajon
> FinEngine Inc
--
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
