camel-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "James Strachan (JIRA)" <j...@apache.org>
Subject [jira] Commented: (CAMEL-1142) FailoverLoadBalancer
Date Fri, 20 Mar 2009 09:26:42 GMT

    [ https://issues.apache.org/activemq/browse/CAMEL-1142?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=50698#action_50698
] 

James Strachan commented on CAMEL-1142:
---------------------------------------

be nice to use this in the DSL something like this

{code}
failover("http://frodo", "http://samwise", "file://UnProcessedMessages").retries(3).interval(1000).exponentialBackOff(false)
{code}



> FailoverLoadBalancer
> --------------------
>
>                 Key: CAMEL-1142
>                 URL: https://issues.apache.org/activemq/browse/CAMEL-1142
>             Project: Apache Camel
>          Issue Type: New Feature
>          Components: camel-core
>            Reporter: Claus Ibsen
>            Assignee: Willem Jiang
>             Fix For: 2.0-M1
>
>
> See nabble:
> http://www.nabble.com/Alternative-routes-td20809601s22882.html
> > I think you could impl. it with the loadbalancer, extending
> > QueueLoadBalancer and override the process method where you will do
> > - try .. catch exception
> > - exchange.isFailed() (* might needed)
> > to determine if there was an error with the exchange and then find the
> > fallback processor and let it try it.
> >
> > I actually think this might be a good loadbalancer to have in
> > camel-core. We just need to work on the configuration so you can set
> > the designated target, and the failover. And maybe you have multiple
> > failovers.
> >
> > Any thoughts from others?
> Agreed - we should add some kind of FailoverLoadBalancer which could
> filter errors/exceptions to determine if it should fail over to
> another endpoint. (Not all errors/exceptions would be down to a WS
> being down - it could be application errors and so forth).
> The current LoadBalancer implementations just pick an endpoint to
> load to; what we need is another implementation that essentially adds
> a custom ErrorHandler to force a redelivery to another endpoint on
> certain failures

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message