You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@activemq.apache.org by "Craig Condit (JIRA)" <ji...@apache.org> on 2009/05/12 01:12:50 UTC

[jira] Created: (AMQ-2246) Failover transport causes connection / thread leak when broker is unresponsive

Failover transport causes connection / thread leak when broker is unresponsive
------------------------------------------------------------------------------

                 Key: AMQ-2246
                 URL: https://issues.apache.org/activemq/browse/AMQ-2246
             Project: ActiveMQ
          Issue Type: Bug
          Components: Connector
    Affects Versions: 5.2.0
            Reporter: Craig Condit


We have observed in a production environment cases where an overloaded ActiveMQ broker stops responding to requests in a timely manner -- connections can be created, but communication is slow enough to trigger reconnect.

Unfortunately, the old connections never get cleaned up, leading to resource exhaustion on the both the clients and the broker.


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

[jira] Assigned: (AMQ-2246) Failover transport causes connection / thread leak when broker is unresponsive

Posted by "Gary Tully (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/activemq/browse/AMQ-2246?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Gary Tully reassigned AMQ-2246:
-------------------------------

    Assignee: Gary Tully

> Failover transport causes connection / thread leak when broker is unresponsive
> ------------------------------------------------------------------------------
>
>                 Key: AMQ-2246
>                 URL: https://issues.apache.org/activemq/browse/AMQ-2246
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: Connector
>    Affects Versions: 5.2.0
>            Reporter: Craig Condit
>            Assignee: Gary Tully
>         Attachments: SlowConnectionTest.java
>
>
> We have observed in a production environment cases where an overloaded ActiveMQ broker stops responding to requests in a timely manner -- connections can be created, but communication is slow enough to trigger reconnect.
> Unfortunately, the old connections never get cleaned up, leading to resource exhaustion on the both the clients and the broker.

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

[jira] Updated: (AMQ-2246) Failover transport causes connection / thread leak when broker is unresponsive

Posted by "Craig Condit (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/activemq/browse/AMQ-2246?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Craig Condit updated AMQ-2246:
------------------------------

    Attachment: SlowConnectionTest.java

Test case illustrating the problem.

> Failover transport causes connection / thread leak when broker is unresponsive
> ------------------------------------------------------------------------------
>
>                 Key: AMQ-2246
>                 URL: https://issues.apache.org/activemq/browse/AMQ-2246
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: Connector
>    Affects Versions: 5.2.0
>            Reporter: Craig Condit
>         Attachments: SlowConnectionTest.java
>
>
> We have observed in a production environment cases where an overloaded ActiveMQ broker stops responding to requests in a timely manner -- connections can be created, but communication is slow enough to trigger reconnect.
> Unfortunately, the old connections never get cleaned up, leading to resource exhaustion on the both the clients and the broker.

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

[jira] Resolved: (AMQ-2246) Failover transport causes connection / thread leak when broker is unresponsive

Posted by "Gary Tully (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/activemq/browse/AMQ-2246?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Gary Tully resolved AMQ-2246.
-----------------------------

       Resolution: Fixed
    Fix Version/s: 5.3.0

thanks for the test case.
issue resolved in r773985  a timedout or failed connect attempt needed to close off the transport to kill the reader thread

> Failover transport causes connection / thread leak when broker is unresponsive
> ------------------------------------------------------------------------------
>
>                 Key: AMQ-2246
>                 URL: https://issues.apache.org/activemq/browse/AMQ-2246
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: Connector
>    Affects Versions: 5.2.0
>            Reporter: Craig Condit
>            Assignee: Gary Tully
>             Fix For: 5.3.0
>
>         Attachments: SlowConnectionTest.java
>
>
> We have observed in a production environment cases where an overloaded ActiveMQ broker stops responding to requests in a timely manner -- connections can be created, but communication is slow enough to trigger reconnect.
> Unfortunately, the old connections never get cleaned up, leading to resource exhaustion on the both the clients and the broker.

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