Hi,

am I assuming correctly that the silence on this topic means it's not
currently possible to do this?

Seeing as the functionality was not ported to 2.3 in CXF-2760, does filing a
new issue on this make sense?

If I interpret the current code correctly, CXF uses the same semi-automatic
approach that the underlying Spring JMS classes implement: If the user
passes a correlationId use that for correlation, otherwise use the
messageId. That sounds like it should work in many cases. However, I'm using
WebSphereMQ which requires using a conduitSelectorPrefix. Unfortunatey, when
a conduitSelectorPrefix is given CXF constructs a correlationId itself,
thereby preempting the attempt to correlate on the messageId instead.

Thanks,
Jens


Jens wrote:
> 
> Let me try to phrase the question differently. It looks like the CXF JMS
> transport by default correlates the correlationId, ie. the client sets a
> correlationId on the request that it expects to be returned as the
> correlationId of the response.
> 
> In my case I have a server that returns the messageId of the request as
> the correlationId of the response. CXF obviously fails to correlate those
> messages and eventually times out waiting for the message even though it
> is sitting in the queue. In CXF prior to 2.3 the
> useMessageIDAsCorrelationID property could be used to fix that issue. How
> can I tell current CXF versions to change its correlation strategy?
> 
> 
> Jens wrote:
>> 
>> In version 2.1 and 2.2 CXF had a property useMessageIDAsCorrelationID on
>> the JMSConfiguration to adjust how the client correlates JMS messages.
>> The property was introduced in CXF-2760, and in that issue Willem
>> mentions that in 2.3 (and later, I presume) CXF has a new way to achieve
>> the same result. What is that new way? I haven't been able to find
>> anything similar, and the old property is no longer available.
>> 
> 


--
View this message in context: 
http://cxf.547215.n5.nabble.com/JMS-Message-Correlation-in-CXF-2-3-tp4830121p4844782.html
Sent from the cxf-user mailing list archive at Nabble.com.

Reply via email to