[
https://issues.apache.org/jira/browse/QPID-4153?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13455873#comment-13455873
]
Rob Godfrey commented on QPID-4153:
-----------------------------------
Why is this (causing a RuntimeException) different from simply not matching the
selector (when we wouldn't DLQ)? When would a selector cause a runtime error
other than type conversion - which should be handled as a non match?
> Messages causing a runtime selector error should be dead-lettered (or
> something similar)
> ----------------------------------------------------------------------------------------
>
> Key: QPID-4153
> URL: https://issues.apache.org/jira/browse/QPID-4153
> Project: Qpid
> Issue Type: Bug
> Components: Java Broker
> Environment:
> Reporter: Keith Wall
> Assignee: Keith Wall
> Fix For: Future
>
>
> Test SelectorTest#testRuntimeSelectorError tries to test the case where the
> contents message header causes the evaluation a consumer's selector to fail
> with a RuntimeException (most likely due to a type issue). The test goes on
> to assert that the producer's session (or connection) is closed to single the
> error.
> There is a problem with the whole approach. It is not certain that the
> producer will still be present when this determination is made. The producer
> could have already gone away or the message might be being moved to a queue
> by a management operation etc.
> The broker should be changed to handle the scenario in a different manner,
> such as routing the message to an alternate exchange or dead-letter-queue.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
