[
https://issues.apache.org/jira/browse/QPID-6428?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14361993#comment-14361993
]
Keith Wall edited comment on QPID-6428 at 3/14/15 7:57 PM:
-----------------------------------------------------------
One question on this, why does
org.apache.qpid.server.queue.AbstractQueue#changeAttribute retain the
responsibility to cause a store update? I guess this was an oversight.
Interestingly, it seems to be causing a (the?) spurious SSRE as it tries to
update an (queue) object in the store that has undergone a failure during
creation and so does not exist. SortedQueueTest#testAutoAckSortedQueue
demonstrates this problem. (The test explicitly creates a SortedQueue, then
later on creating a consumer creates a StandardQueue of the same name by side
effect. It is the latter queue declare that triggers the SSRE).
{noformat}
2015-03-14 19:33:49,020 ERROR [VirtualHostNode-test-Configuration-Thread]
(util.ServerScopedRuntimeException) - Record with id
14f56ae2-175d-49b6-a332-cb41604bc7bc does not exist
{noformat}
I also now see why the SSRE is not causing the Broker to go down. In this
case, the SSRE is being thrown within
org.apache.qpid.server.model.AbstractConfiguredObject.CreateExceptionHandler#handleException
but it is then lost by the handler rethrowing the original exception
(DuplicateNameException in this case) in the finally block.
was (Author: k-wall):
One question on this, why does
org.apache.qpid.server.queue.AbstractQueue#changeAttribute retain the
responsibility to cause a store update? I guess this was an oversight.
Interestingly, it seems to be causing a (the?) spurious SSRE as it tries to
update an (queue) object in the store that has undergone a failure during
creation and so does not exist. SortedQueueTest#testAutoAckSortedQueue
demonstrates this problem. (The test explicitly creates a SortedQueue, then
later on creating a consumer create a StandardQueue of the same name by side
effect. It is the latter queue declare that triggers the SSRE).
{noformat}
2015-03-14 19:33:49,020 ERROR [VirtualHostNode-test-Configuration-Thread]
(util.ServerScopedRuntimeException) - Record with id
14f56ae2-175d-49b6-a332-cb41604bc7bc does not exist
{noformat}
I also now see why the SSRE is not causing the Broker to go down. In this
case, the SSRE is being thrown within
org.apache.qpid.server.model.AbstractConfiguredObject.CreateExceptionHandler#handleException
but it is then lost by the handler rethrowing the original exception
(DuplicateNameException in this case) in the finally block.
> [Java Broker] Use StoreConfigurationChangeListener to manage storage of all
> configured objects
> ----------------------------------------------------------------------------------------------
>
> Key: QPID-6428
> URL: https://issues.apache.org/jira/browse/QPID-6428
> Project: Qpid
> Issue Type: Improvement
> Components: Java Broker
> Reporter: Rob Godfrey
> Assignee: Rob Godfrey
>
> Currently configured objects in the broker store use the change listener to
> keep the store updated, whereas objects in the virtual host store use
> explicit calls to the config store within the configured object code itself.
> We should only have one mechanism for managing the storage of configured
> objects.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
