Hi, My understanding is that in an InOnly exchange, what counts is the status of the Exchange (completed or failed) and the Unit Of Work rather than the OUT message.
When the Exchange completes and the processor stack unwinds, InOnly consumer can react differently depending on whether the outcome was successful or not (i.e. rollback a transaction - we see this in camel-jms, or move a file to a "failed" directory - File component, etc.), i.e. if an unhandled Exception is marked in the Exchange. In fact, (in my mind) it should be strictly illegal for an InOnly consumer to attempt to access the OUT message from an Exchange, even though this is not enforced by the Camel APIs right now AFAIK. Regards, *Raúl Kripalani* Principal Consultant | FuseSource Corp. r...@fusesource.com | fusesource.com <http://www.fusesource.com/> skype: raul.fuse | twitter: @raulvk <http://twitter.com/raulvk>, @fusenews<http://twitter.com/fusenews> blog: F3 - Flashes From the Field<http://blog.raulkr.net/?utm_source=fusesourceemail&utm_medium=email&utm_campaign=fusesourcemail> <http://twitter.com/fusenews> On 5 June 2012 11:11, Babak Vahdat <babak.vah...@swissonline.ch> wrote: > Thanks for that link but I've been already through that FAQ many many times > which didn't help me as well! It seems I'm too dummy to understand this > API. > > Yes I'm aware of the exchange.hasOut() semantics as well as the lazy > instantiation of OUT through exchange.getOut() > > So again my question: > > - Do we currently have a Camel Producer behaving wrong concerning this > question? > - And still I think the fact that exchange.getOut() does lazily instantiate > an OUT if OUT == null is an API flaw as this does bypass the semantics of > the given exchange pattern of the exchange object. So that you can have an > exchange with InOnly pattern having an OUT != null attached to it! > > Babak > > > -- > View this message in context: > http://camel.465427.n5.nabble.com/About-what-to-do-with-the-Response-retrieved-through-a-Producer-when-the-Exchange-is-NOT-out-capable-tp5713946p5714015.html > Sent from the Camel Development mailing list archive at Nabble.com. >
