[
https://issues.apache.org/jira/browse/QPIDJMS-528?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17311446#comment-17311446
]
Robbie Gemmell commented on QPIDJMS-528:
----------------------------------------
This is where the failover provider intercepts a connection drop and starts its
work, it isnt really considered an error and the log was just there for
debugging and so thats the level used. The failover provider does log info
warning and error messages elsewhere as it tries reconnection and uses up and
potentially exhausts its retries etc, and the ultimate failure of the
connection to reconnect is then processed in JMSConnection class (though that
doesnt log currently, but does produces an exception to the connection
exception listener).
Its probably reasonable that we tweak the logging, especially the latter part
since we do now report connected/reconnected, though I'm not sure it would be
this line exactly that we look to change, need to look into it a bit.
> "Failover: the provider reports failure:" error is logged at DEBUG level
> ------------------------------------------------------------------------
>
> Key: QPIDJMS-528
> URL: https://issues.apache.org/jira/browse/QPIDJMS-528
> Project: Qpid JMS
> Issue Type: Wish
> Affects Versions: 0.57.0
> Reporter: David Martin
> Priority: Minor
>
> On line 855 of FailoverProvider.java an actual error condition is logged at
> DEBUG level:
> LOG.debug("Failover: the provider reports failure: {}",
> ex.getMessage());
> This happens when the connection is lost irrecoverably. This is particularly
> an issue using a Single/CachedConnectionFactory as it stays down. Please can
> this event be logged at INFO or better still WARN level instead?
> I am happy to raise a PR for it if is approved by the maintainers
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
