Grokbase Groups Camel users May 2009
FAQ
Hello,

I am running a camel route which consumes messages
with a durable subscriber from an ActiveMQ topic
in an OSGi server. The route is like I described it in my last post
http://www.nabble.com/Problems-with-error-handling-in-transactional-client-td23706098.html.
When the OSGi server is restarted it happens once in
a while that the route does not receive anymore messages despite the fact
that the topic is receiving new messages and the Pending Queue Size is
around
10.000 messages. The topic has a second durable subscriber which still
works.
I checked the trace.log of the route but could not see any errors/warnings.
If the blocking occurs once it does not go away when the OSGi server is
restarted .
The only thing which seems to help is to delete the durable subscriber
in the ActiveMQ web interface
and recreate it.

I also noticed that an undeploy of a route with a durable subscriber
does not unregisters
the durable subscriber, which means to update a route I have to delete
it, restart the OSGi server
and deploy the updated route.

Thanks,
Jörn

Search Discussions

  • Claus Ibsen at May 26, 2009 at 1:25 pm
    Hi
    On Tue, May 26, 2009 at 1:29 PM, Jörn Kottmann wrote:
    Hello,

    I am running a camel route which consumes messages
    with a durable subscriber from an ActiveMQ topic
    in an OSGi server. The route is like I described it in my last post
    http://www.nabble.com/Problems-with-error-handling-in-transactional-client-td23706098.html.
    When the OSGi server is restarted it happens once in
    a while that the route does not receive anymore messages despite the fact
    that the topic is receiving new messages and the Pending Queue Size is
    around
    10.000 messages.  The topic has a second durable subscriber which still
    works.
    I checked the trace.log of the route but could not see any errors/warnings.
    If the blocking occurs once it does not go away when the OSGi server is
    restarted .
    The only thing which seems to help is to delete the durable subscriber in
    the ActiveMQ web interface
    and recreate it.
    I am not a AMQ expert but it has a failover protocol that you should
    generally use. I do think tis default in one of the latest releases.
    Basically add: failover: as prefix to the AMQ uri for the clients.

    More info here:
    http://activemq.apache.org/failover-transport-reference.html
    I also noticed that an undeploy of a route with a durable subscriber does
    not unregisters
    the durable subscriber, which means to update a route I have to delete it,
    restart the OSGi server
    and deploy the updated route.

    Thanks,
    Jörn


    --
    Claus Ibsen
    Apache Camel Committer

    Open Source Integration: http://fusesource.com
    Blog: http://davsclaus.blogspot.com/
    Twitter: http://twitter.com/davsclaus
  • Jörn Kottmann at May 28, 2009 at 2:24 pm

    Claus Ibsen wrote:
    Hi
    On Tue, May 26, 2009 at 1:29 PM, Jörn Kottmann wrote:

    Hello,

    I am running a camel route which consumes messages
    with a durable subscriber from an ActiveMQ topic
    in an OSGi server. The route is like I described it in my last post
    http://www.nabble.com/Problems-with-error-handling-in-transactional-client-td23706098.html.
    When the OSGi server is restarted it happens once in
    a while that the route does not receive anymore messages despite the fact
    that the topic is receiving new messages and the Pending Queue Size is
    around
    10.000 messages. The topic has a second durable subscriber which still
    works.
    I checked the trace.log of the route but could not see any errors/warnings.
    If the blocking occurs once it does not go away when the OSGi server is
    restarted .
    The only thing which seems to help is to delete the durable subscriber in
    the ActiveMQ web interface
    and recreate it.
    After further investigation I found out that ActiveMQ (version 5.2.0)
    just hangs and
    writes the following error messages to the log file:

    2009-05-28 15:59:31,174 [xxx:52790] ERROR RecoveryListenerAdapter
    - Message id xxx could not be recovered from the data store - already
    dispatched

    So its an ActiveMQ bug and not a problem with camel. It seems that there
    is already a jira issues AMQ-1797 and AMQ-2021.

    Jörn

Related Discussions

Discussion Navigation
viewthread | post
Discussion Overview
groupusers @
categoriescamel
postedMay 26, '09 at 11:29a
activeMay 28, '09 at 2:24p
posts3
users2
websitecamel.apache.org

2 users in discussion

Jörn Kottmann: 2 posts Claus Ibsen: 1 post

People

Translate

site design / logo © 2022 Grokbase