activemq-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jereme Monteau (JIRA)" <j...@apache.org>
Subject [jira] Commented: (AMQ-1720) Default Broker config does not shutdown when an internal persistence error occurs.
Date Sun, 26 Oct 2008 23:56:52 GMT

    [ https://issues.apache.org/activemq/browse/AMQ-1720?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=46844#action_46844
] 

Jereme Monteau commented on AMQ-1720:
-------------------------------------

Hi,

We are experiencing this problem when our MySQL Cluster performs some weekly maintenance tasks.
It seems to force any clients to disconnect momentarily.
Is there anyway to configure ActiveMQ to handle these disconnects in a more robust fashion?
I'm thinking along the lines of autoReconnect at the driver level or 
something similar. Has anyone else worked around this? We are running 5.2 and the behavior
seems improved from 5.1, since the process does completely shutdown
instead of ending up in a zombie-like state of running but not responding to requests. We
are running in a Master/Slave failover configuration and unfortunately the
MySQL Cluster seems to do a wholesale disconnect of all clients at certain times, so the secondary
activemq process also goes down, leaving us with no available queues.
We are considering moving to a network of brokers high availability configuration but documentation
on this setup seems sparse. Any advice is greatly appreciatted.

Thanks,
Jereme Monteau

> Default Broker config does not shutdown when an internal persistence error occurs.
> ----------------------------------------------------------------------------------
>
>                 Key: AMQ-1720
>                 URL: https://issues.apache.org/activemq/browse/AMQ-1720
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: Broker
>    Affects Versions: 5.1.0
>            Reporter: Hiram Chirino
>            Assignee: Hiram Chirino
>             Fix For: 5.2.0
>
>
> This is due to all the other non-deamon threads running that are not part of the broker
but in the default config.  For example jetty threads and camel stuff.
> We should allow the broker server to callback to a listener to that the Main that starts
it all can System.exit properly when the broker is shutdown.. that should cause the non-daemon
stuff to exit too.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message