For the record, I tried with the rpms released under 1.2 and I was able to get notified via the exception listener when a read timeout happens. I am suspecting something else maybe wrong here.
Regards, Rajith On Wed, May 5, 2010 at 5:14 PM, Wes Parish <[email protected]> wrote: > Answers below > > Do you have any suggestions that we should try on our side to further > troubleshoot the problem? > > Thanks, > Wes > > -----Original Message----- > From: Rajith Attapattu [mailto:[email protected]] > Sent: Wed 5/5/2010 3:31 PM > To: [email protected] > Subject: Re: JMS Question > > I ran my test against trunk, however IIRC it should work against 1.2 as well. > Could you let me know what happens in your situation? > Does the connection terminate? > > Yes, the connection terminates and we fail over to the secondary host. The > problem is that we are using dynamic queues through JMS, and after a > successful failover, the client application has no way of knowing (without > receiving this exception) to re-create the queue on the secondary host. > > If so do you get notify via the exception listener? (I assume not > based on your email) Or is it that you are not getting the right > exception from the listener? > > No, we do not get notified via the exception listener. We do see the > exception on stdout. > > Rajith > > On Wed, May 5, 2010 at 1:55 PM, Wes Parish <[email protected]> > wrote: >> Rajith, >> >> Thank you for the quick reply. We are running Red Hat MRG v1.2 official >> release. Which version did you run your tests against? >> >> Thanks, >> Wes >> >> >> -----Original Message----- >> From: Rajith Attapattu [mailto:[email protected]] >> Sent: Wed 5/5/2010 9:32 AM >> To: [email protected] >> Subject: Re: JMS Question >> >> When the socket timeout happens, you do get notified via onException() >> - I just verified this. >> However as soon as the "read timeout" happens the connection is also >> closed, and another exception is generated. >> Therefore you are most likely to see the "connection aborted" >> exception from your listener than the "read timeout". >> >> I think it's more desirable if the "read timeout' exception is thrown >> rather than the connection closed. >> (But please note that the "Read Timeout" exception is logged, so >> information is not lost) >> >> On Tue, May 4, 2010 at 6:18 PM, Wes Parish <[email protected]> >> wrote: >>> When a connection exception is thrown because of a socket timeout >>> exception, which occurs when a heartbeat times out, it is not being sent to >>> the onException() method. Should this exception be sent like the others to >>> this method? >>> >>> Wes >>> >> >> >> >> -- >> Regards, >> >> Rajith Attapattu >> Red Hat >> http://rajith.2rlabs.com/ >> >> --------------------------------------------------------------------- >> Apache Qpid - AMQP Messaging Implementation >> Project: http://qpid.apache.org >> Use/Interact: mailto:[email protected] >> >> >> >> >> >> --------------------------------------------------------------------- >> Apache Qpid - AMQP Messaging Implementation >> Project: http://qpid.apache.org >> Use/Interact: mailto:[email protected] >> > > > > -- > Regards, > > Rajith Attapattu > Red Hat > http://rajith.2rlabs.com/ > > --------------------------------------------------------------------- > Apache Qpid - AMQP Messaging Implementation > Project: http://qpid.apache.org > Use/Interact: mailto:[email protected] > > > > > > --------------------------------------------------------------------- > Apache Qpid - AMQP Messaging Implementation > Project: http://qpid.apache.org > Use/Interact: mailto:[email protected] > -- Regards, Rajith Attapattu Red Hat http://rajith.2rlabs.com/ --------------------------------------------------------------------- Apache Qpid - AMQP Messaging Implementation Project: http://qpid.apache.org Use/Interact: mailto:[email protected]
