You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@activemq.apache.org by "Gary Tully (JIRA)" <ji...@apache.org> on 2011/06/03 14:24:47 UTC

[jira] [Created] (AMQ-3357) vmQueueCursor or fileQueueCursor with expired messages in the store can lead to startup hang with kahaDB

vmQueueCursor or fileQueueCursor with expired messages in the store can lead to startup hang with kahaDB
--------------------------------------------------------------------------------------------------------

                 Key: AMQ-3357
                 URL: https://issues.apache.org/jira/browse/AMQ-3357
             Project: ActiveMQ
          Issue Type: Bug
          Components: Message Store
    Affects Versions: 5.5.0, 5.4.2
            Reporter: Gary Tully
            Assignee: Gary Tully
             Fix For: 5.6.0


expiry during recovery, which is a read operation, can lead to the need to modify the store. The recovery processing should take out a readwrite lock on the store.

Issue arises for vm or file cursors as they replay messages from the store on restart to fill their caches. 
This issue can be avoided by using the default Store cursor, so the workaround is to configure the store cursor, restart, and revert back to the chosen cursor.

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (AMQ-3357) vmQueueCursor or fileQueueCursor with expired messages in the store can lead to startup hang with kahaDB

Posted by "Gary Tully (JIRA)" <ji...@apache.org>.
    [ https://issues.apache.org/jira/browse/AMQ-3357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13045333#comment-13045333 ] 

Gary Tully commented on AMQ-3357:
---------------------------------

you can use the latest 5.6-SNAPSHOT[1] or apply the fix from http://svn.apache.org/viewvc?rev=1131010&view=rev

[1]https://repository.apache.org/content/repositories/snapshots/org/apache/activemq/apache-activemq/5.6-SNAPSHOT/

> vmQueueCursor or fileQueueCursor with expired messages in the store can lead to startup hang with kahaDB
> --------------------------------------------------------------------------------------------------------
>
>                 Key: AMQ-3357
>                 URL: https://issues.apache.org/jira/browse/AMQ-3357
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: Message Store
>    Affects Versions: 5.4.2, 5.5.0
>            Reporter: Gary Tully
>            Assignee: Gary Tully
>              Labels: expiry, fileQueueCursor, hang, kahadb, restart, vmQueueCursor
>             Fix For: 5.6.0
>
>
> expiry during recovery, which is a read operation, can lead to the need to modify the store. The recovery processing should take out a readwrite lock on the store.
> Issue arises for vm or file cursors as they replay messages from the store on restart to fill their caches. 
> This issue can be avoided by using the default Store cursor, so the workaround is to configure the store cursor, restart, and revert back to the chosen cursor.

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Resolved] (AMQ-3357) vmQueueCursor or fileQueueCursor with expired messages in the store can lead to startup hang with kahaDB

Posted by "Gary Tully (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/jira/browse/AMQ-3357?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Gary Tully resolved AMQ-3357.
-----------------------------

    Resolution: Fixed

fix in http://svn.apache.org/viewvc?rev=1131010&view=rev

> vmQueueCursor or fileQueueCursor with expired messages in the store can lead to startup hang with kahaDB
> --------------------------------------------------------------------------------------------------------
>
>                 Key: AMQ-3357
>                 URL: https://issues.apache.org/jira/browse/AMQ-3357
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: Message Store
>    Affects Versions: 5.4.2, 5.5.0
>            Reporter: Gary Tully
>            Assignee: Gary Tully
>              Labels: expiry, fileQueueCursor, hang, kahadb, restart, vmQueueCursor
>             Fix For: 5.6.0
>
>
> expiry during recovery, which is a read operation, can lead to the need to modify the store. The recovery processing should take out a readwrite lock on the store.
> Issue arises for vm or file cursors as they replay messages from the store on restart to fill their caches. 
> This issue can be avoided by using the default Store cursor, so the workaround is to configure the store cursor, restart, and revert back to the chosen cursor.

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (AMQ-3357) vmQueueCursor or fileQueueCursor with expired messages in the store can lead to startup hang with kahaDB

Posted by "mayank tripathi (JIRA)" <ji...@apache.org>.
    [ https://issues.apache.org/jira/browse/AMQ-3357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13044710#comment-13044710 ] 

mayank tripathi commented on AMQ-3357:
--------------------------------------

Hi Garry....
Is there a revision available for this defect on 5.5.0. As this defect is reproducible in our production and has major impact on us...Will it be possible to avail a patch so that we can apply and move on with this?


> vmQueueCursor or fileQueueCursor with expired messages in the store can lead to startup hang with kahaDB
> --------------------------------------------------------------------------------------------------------
>
>                 Key: AMQ-3357
>                 URL: https://issues.apache.org/jira/browse/AMQ-3357
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: Message Store
>    Affects Versions: 5.4.2, 5.5.0
>            Reporter: Gary Tully
>            Assignee: Gary Tully
>              Labels: expiry, fileQueueCursor, hang, kahadb, restart, vmQueueCursor
>             Fix For: 5.6.0
>
>
> expiry during recovery, which is a read operation, can lead to the need to modify the store. The recovery processing should take out a readwrite lock on the store.
> Issue arises for vm or file cursors as they replay messages from the store on restart to fill their caches. 
> This issue can be avoided by using the default Store cursor, so the workaround is to configure the store cursor, restart, and revert back to the chosen cursor.

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira