activemq-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Gary Tully (JIRA)" <>
Subject [jira] Resolved: (AMQ-2831) Set ClientID - InvalidClientIDException with http transport
Date Wed, 01 Sep 2010 10:25:42 GMT


Gary Tully resolved AMQ-2831.

    Resolution: Fixed

The http transport now has an inactivity monitor, configure the http transport url with a
low inactivity timeout value and the server end of the http transport will timeout when the
client jvm crashes.


> Set ClientID - InvalidClientIDException with http transport
> -----------------------------------------------------------
>                 Key: AMQ-2831
>                 URL:
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: Transport
>    Affects Versions: 5.3.2
>            Reporter: Colombo Marco
>            Assignee: Gary Tully
>            Priority: Blocker
>             Fix For: 5.4.1
>         Attachments: broker-config.xml,
> I have a client that do a durable subscriptio to a topic.
> It register with a static client id.
> After a connectivity loss (an so a new re-connection), this exception is thrown and connection
cannot be
> re-established:
> 2010-07-15 13:49:04,700 ERROR - javax.jms.InvalidClientIDException: Broker: localhost
- Client: (SchedulerId = '1') already connected from blockingQueue_8461294
>         at
>         at
>         at
>         at org.apache.activemq.advisory.AdvisoryBroker.addConnection(
>         at
>         at
> ....
> This is not possible, because is the only client with that name and previos connection
was closed . .close().
> If I kill all the clients an I leave active only the broker,
> If I try to reconnect using http (or tcp) I get the same error.
> So client is not de-registered.
> This appens also if I use failover protocol. After client kill/restart. Same exception
is raised (javax.jms.InvalidClientIDException - with the clientID of the killed client).
> I have to restart broker in order to register again the client. 
> A test case is attached. 
> If you connect to http the first time, connection is ok.
> If you kill the JVM and the try to restart the test case, 'myclid already connected from
blockingQueue_' is always raised. Also after some hours. I have to restart the broker in order
to have a new conection with that client id.
> This test simulate a JVM crash with no resource cleanup.

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

View raw message