You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@qpid.apache.org by "Martin Ritchie (JIRA)" <qp...@incubator.apache.org> on 2009/03/12 18:34:50 UTC
[jira] Created: (QPID-1730) Persistent Queue recovery doesn't work
with durable queues in virtualhost.xml
Persistent Queue recovery doesn't work with durable queues in virtualhost.xml
-----------------------------------------------------------------------------
Key: QPID-1730
URL: https://issues.apache.org/jira/browse/QPID-1730
Project: Qpid
Issue Type: Bug
Components: Java Broker
Reporter: Martin Ritchie
Assignee: Martin Ritchie
Priority: Blocker
Fix For: M5
Summary:
The change to prevent duplicate MBeans registering has highlighted that we are recovering first from the TransactionLog then trying trying to create queues defined in the virtualhost.xml file.
We need to only store the persistent queue declaration in a single location so that changes can be correctly propagated. i.e. currently I believe that if you have custom alert values on a durable queue after restart you will get the broker code default alerting values. The global alert values in the virtualhost.xml file will not apply to recovered queues nor will the values in the file for any durable queue recovered from the store.
Long term we need to use a single location.
We also have one of three options to consider.
1) Allow duplicate MBean registration and leave recovery / startup code as is.
2) Adjust the startup code to ignore entries in the virtualhost file that have been loaded from the persistent store.
3) Adjust the startup logic to open the persistent store and merge the vhost and store configuration.
For 1 and 2 we also need to create a JIRA for the broken persistent queue configuration changes as this is not a new problem.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
Apache Qpid - AMQP Messaging Implementation
Project: http://qpid.apache.org
Use/Interact: mailto:dev-subscribe@qpid.apache.org
[jira] Assigned: (QPID-1730) Persistent Queue recovery doesn't work
with durable queues in virtualhost.xml
Posted by "Aidan Skinner (JIRA)" <qp...@incubator.apache.org>.
[ https://issues.apache.org/jira/browse/QPID-1730?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Aidan Skinner reassigned QPID-1730:
-----------------------------------
Assignee: Aidan Skinner (was: Martin Ritchie)
> Persistent Queue recovery doesn't work with durable queues in virtualhost.xml
> -----------------------------------------------------------------------------
>
> Key: QPID-1730
> URL: https://issues.apache.org/jira/browse/QPID-1730
> Project: Qpid
> Issue Type: Bug
> Components: Java Broker
> Reporter: Martin Ritchie
> Assignee: Aidan Skinner
> Priority: Blocker
> Fix For: 0.5
>
>
> Summary:
> The change to prevent duplicate MBeans registering has highlighted that we are recovering first from the TransactionLog then trying trying to create queues defined in the virtualhost.xml file.
> We need to only store the persistent queue declaration in a single location so that changes can be correctly propagated. i.e. currently I believe that if you have custom alert values on a durable queue after restart you will get the broker code default alerting values. The global alert values in the virtualhost.xml file will not apply to recovered queues nor will the values in the file for any durable queue recovered from the store.
> Long term we need to use a single location.
> We also have one of three options to consider.
> 1) Allow duplicate MBean registration and leave recovery / startup code as is.
> 2) Adjust the startup code to ignore entries in the virtualhost file that have been loaded from the persistent store.
> 3) Adjust the startup logic to open the persistent store and merge the vhost and store configuration.
> For 1 and 2 we also need to create a JIRA for the broken persistent queue configuration changes as this is not a new problem.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
Apache Qpid - AMQP Messaging Implementation
Project: http://qpid.apache.org
Use/Interact: mailto:dev-subscribe@qpid.apache.org
[jira] Assigned: (QPID-1730) Persistent Queue recovery doesn't work
with durable queues in virtualhost.xml
Posted by "Aidan Skinner (JIRA)" <qp...@incubator.apache.org>.
[ https://issues.apache.org/jira/browse/QPID-1730?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Aidan Skinner reassigned QPID-1730:
-----------------------------------
Assignee: Martin Ritchie (was: Aidan Skinner)
Hi, I fixed this. Please review.
> Persistent Queue recovery doesn't work with durable queues in virtualhost.xml
> -----------------------------------------------------------------------------
>
> Key: QPID-1730
> URL: https://issues.apache.org/jira/browse/QPID-1730
> Project: Qpid
> Issue Type: Bug
> Components: Java Broker
> Reporter: Martin Ritchie
> Assignee: Martin Ritchie
> Priority: Blocker
> Fix For: 0.5
>
>
> Summary:
> The change to prevent duplicate MBeans registering has highlighted that we are recovering first from the TransactionLog then trying trying to create queues defined in the virtualhost.xml file.
> We need to only store the persistent queue declaration in a single location so that changes can be correctly propagated. i.e. currently I believe that if you have custom alert values on a durable queue after restart you will get the broker code default alerting values. The global alert values in the virtualhost.xml file will not apply to recovered queues nor will the values in the file for any durable queue recovered from the store.
> Long term we need to use a single location.
> We also have one of three options to consider.
> 1) Allow duplicate MBean registration and leave recovery / startup code as is.
> 2) Adjust the startup code to ignore entries in the virtualhost file that have been loaded from the persistent store.
> 3) Adjust the startup logic to open the persistent store and merge the vhost and store configuration.
> For 1 and 2 we also need to create a JIRA for the broken persistent queue configuration changes as this is not a new problem.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
Apache Qpid - AMQP Messaging Implementation
Project: http://qpid.apache.org
Use/Interact: mailto:dev-subscribe@qpid.apache.org
[jira] Updated: (QPID-1730) Persistent Queue recovery doesn't work
with durable queues in virtualhost.xml
Posted by "Aidan Skinner (JIRA)" <qp...@incubator.apache.org>.
[ https://issues.apache.org/jira/browse/QPID-1730?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Aidan Skinner updated QPID-1730:
--------------------------------
Status: Ready To Review (was: In Progress)
> Persistent Queue recovery doesn't work with durable queues in virtualhost.xml
> -----------------------------------------------------------------------------
>
> Key: QPID-1730
> URL: https://issues.apache.org/jira/browse/QPID-1730
> Project: Qpid
> Issue Type: Bug
> Components: Java Broker
> Reporter: Martin Ritchie
> Assignee: Aidan Skinner
> Priority: Blocker
> Fix For: 0.5
>
>
> Summary:
> The change to prevent duplicate MBeans registering has highlighted that we are recovering first from the TransactionLog then trying trying to create queues defined in the virtualhost.xml file.
> We need to only store the persistent queue declaration in a single location so that changes can be correctly propagated. i.e. currently I believe that if you have custom alert values on a durable queue after restart you will get the broker code default alerting values. The global alert values in the virtualhost.xml file will not apply to recovered queues nor will the values in the file for any durable queue recovered from the store.
> Long term we need to use a single location.
> We also have one of three options to consider.
> 1) Allow duplicate MBean registration and leave recovery / startup code as is.
> 2) Adjust the startup code to ignore entries in the virtualhost file that have been loaded from the persistent store.
> 3) Adjust the startup logic to open the persistent store and merge the vhost and store configuration.
> For 1 and 2 we also need to create a JIRA for the broken persistent queue configuration changes as this is not a new problem.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
Apache Qpid - AMQP Messaging Implementation
Project: http://qpid.apache.org
Use/Interact: mailto:dev-subscribe@qpid.apache.org