On Fri, Aug 21, 2009 at 3:49 PM, mkeenan<kee...@p2sol.com> wrote:
>
>
>
> bsnyder wrote:
>>
>>
>> Correct, sessions should be exclusive to a specific consumer and
>> should not be shared at all.
>>
>>
>
> Bruce, thanks for the clarification.
>
> The reason I brought it up is that I am using Spring.NET with NMS and we
> regularly experience thread deadlock. I sent a bit of sample code to Jim
> Gomes and we have concluded it is most likely a problem within Spring.NET
> and not NMS.
>
> I have looked at the Spring.Net (and Java Spring) code for
> 'CachingConnectionFactory' and became thoroughly confused because there is a
> 'SessionCacheSize' property and that class appears to be caching sessions
> based on AcknowledgementMode. In the test code I created to demonstrate the
> thread problem, the issue does not occur when 'SessionCacheSize' is set to a
> number that is larger than the number of concurrent consumers in the
> application.
>
> Once I made that discovery I jumped to the conclusion that perhaps multiple
> consumers were getting created from the same session, which is why I asked
> the original question.
The reference I made to the Spring CachingConnectionFactory was for
the Spring Java APIs. Since you're working with .NET and I don't, we
should include Mark Pollack in this conversation. He's the creator and
maintainer of Spring.NET so I'll bet that he can provide some better
info about Spring.NET. I'm CC'ing him on this just to get his
attention.
Bruce
--
perl -e 'print
unpack("u30","D0G)u8...@4vyy9&5R\"F)R=6-E+G-N>61E<D\!G;6%I;\"YC;VT*"
);'
ActiveMQ in Action: http://bit.ly/2je6cQ
Blog: http://bruceblog.org/
Twitter: http://twitter.com/brucesnyder
