[ https://issues.apache.org/jira/browse/CAMEL-9069?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14935031#comment-14935031 ]
Kevin Telford commented on CAMEL-9069: -------------------------------------- I can't believe I didn't respond earlier, terribly sorry. What we'd like to have is that on a failure, a near real-time way to gather information and notify (log or call a method) about the failure. It may be as simple as just a notification of a failure and information from a header, or perhaps (if available) more information from the failed HTTP attempt (status code etc). As far as what should happen if a new exception were thrown, or the exchange altered, I think it should remain consistent with the existing .onRedelivery() and .process() methods. > Add processor to immediately handle exception while in retry loop > ----------------------------------------------------------------- > > Key: CAMEL-9069 > URL: https://issues.apache.org/jira/browse/CAMEL-9069 > Project: Camel > Issue Type: Improvement > Components: camel-core > Reporter: Kevin Telford > Assignee: Claus Ibsen > Fix For: 2.17.0 > > > We have a recipientList of http4 endpoints. Occasionally an endpoint won't > be available, and will return an exception (whether a ConnectException, or > any other ex and a status code of >=400). To handle these errors we can use > a defaultErrorHandler or leverage onException, and setup redelivery. What we > then would like to do is some custom error handling on each occurrence of an > exception. > Both the defaultErrorHandler and onException support .onRedelivery(). While > this is nice, when a retry delay grows large enough the lag between the > exception and the .onRedelivery() call really begins to grow and so our > custom error handling isn't as much “on error” as it is “on error and after a > wait”. > onException also supports a .process() call, however this only occurs once > all redeliveries have failed. > What would be ideal would be an additional .onExceptionOccur(Processor > processor) method (or some such name) that allows for immediate processing > upon an exception, even if its being handled within a retry loop. -- This message was sent by Atlassian JIRA (v6.3.4#6332)