[
https://issues.apache.org/activemq/browse/CAMEL-901?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=46349#action_46349
]
Bela Vizy commented on CAMEL-901:
---------------------------------
#1
Well, not so fast. Take out the thread() and it will behave differently, they
way I suggested it, that is it will re-deliver to the #3 processor where the
failure occurred.
I guess I mistook a feature for a bug :) The doc is not clear about the
redelivery behavior (or I should RTFM). My $0.02 is that if we use the
pipeline() then it should re-deliver to the pipe otherwise the failing
processor.
#2 That makes sense if it re-delivers to the pipe...
> If using thread() processor followed by a pipeline Camel redelivers to the
> wrong processor
> ------------------------------------------------------------------------------------------
>
> Key: CAMEL-901
> URL: https://issues.apache.org/activemq/browse/CAMEL-901
> Project: Apache Camel
> Issue Type: Bug
> Components: camel-core
> Affects Versions: 1.0.0, 1.1.0, 1.2.0, 1.3.0, 1.4.0, 1.5.0
> Reporter: Bela Vizy
> Fix For: 1.5.0
>
> Attachments: BelasErrorHandlerLoggingTest.java, camel-core-test.log
>
>
> Setup:
> Error handler is specified before the thread processor and there are 3
> processor after the thread() call. Third processor fails.
> Bug #1:
> The exchange received by the failed processor (#3) gets redelivered to the
> first processor in the pipe.
> It should be redeliver to #3 (this works without the thread processor).
> Bug #2:
> The specified errorHandler (log:testError) receives the exchange that entered
> processor #1 (first in the pipe).
> See details in the attached unit test. Also take a look at the log4j log
> (copy attached). The 3rd line from the bottom is the log by testError.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
