activemq-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Gary Tully <gary.tu...@gmail.com>
Subject Re: Same message redelivered to multiple consumers after failover
Date Wed, 12 Mar 2014 19:40:04 GMT
Can you share your code?
On 12 Mar 2014 17:40, "gregg" <garlakg@hotmail.com> wrote:

> I can reproduce this consistently. I even created a bare bones stand-alone
> java application listener that only logs the message contents to eliminate
> any transactional variables.
>
> I start each java listener application in a new debug process and set a
> breakpoint on the log statement (no JMS session is being shared). I use the
> default KahaDB store and as mentioned before the problem happens even with
> a
> single broker (there is no slave broker waiting for a free lock).
>
> To reproduce:
>
> 1) I push one message on a queue
> 2) Start 4 listeners. One listener picks it up and a breakpoint is hit on
> the log output line
> 3) I start a shutdown sequence of the broker (Ctrl-C or issue shutdown
> command from Hawtio)
> 4) All 3 other listener breakpoints get triggered simultaneously, each one
> having deliveryCount incremented
>
> These tests are being run in a 64Bit Windows 7 - Java SE 1.6.0_20. Haven't
> tested in other environments yet.
>
>
> Raul Kripalani-2 wrote
> > Can you reproduce it consistently time after time? I wonder if this was a
> > one-off case and you witnessed a race condition if all 3 competing
> > consumers connected/handshook exactly at the same time.
> >
> > Are they creating a new Session each, or are they sharing a single one?
> >
> > Any JMS / XA transactions involved?
> >
> > What store are you using? Shared KahaDB?
> >
> > Regards,
> >
> > *Raúl Kripalani*
> > Apache Camel PMC Member & Committer | Enterprise Architect, Open Source
> > Integration specialist
> > http://about.me/raulkripalani | http://www.linkedin.com/in/raulkripalani
> > http://blog.raulkr.net | twitter: @raulvk
> >
> > On Mon, Mar 10, 2014 at 2:41 AM, gregg <
>
> > garlakg@
>
> > > wrote:
> >
> >> Don't think that is possible because I have a breakpoint in each
> >> consumer.
> >> The delivery counter information was inspected when all consumers were
> >> paused concurrently in a debug session during consumption. If I don't
> set
> >> a
> >> breakpoint I still get the same problem.
> >>
> >> Would have to do some more testing to change the way the message is
> >> produced
> >> to potentially isolate the cause.
> >>
> >>
> >> artnaseef wrote
> >> > Is it possible the listeners are rejecting the message?  That could
> >> > explain the results you are seeing.
> >>
> >>
> >>
> >>
> >>
> >> --
> >> View this message in context:
> >>
> http://activemq.2283324.n4.nabble.com/Same-message-redelivered-to-multiple-consumers-after-failover-tp4678640p4678674.html
> >> Sent from the ActiveMQ - User mailing list archive at Nabble.com.
> >>
>
>
>
>
>
> --
> View this message in context:
> http://activemq.2283324.n4.nabble.com/Same-message-redelivered-to-multiple-consumers-after-failover-tp4678640p4678896.html
> Sent from the ActiveMQ - User mailing list archive at Nabble.com.
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message