Re: ActiveMQ-5.6.0 and temp-queue performance

[Gary Tully <ga...@gmail.com>]

Would it be possible to create a little junit test case that
reproduces this, if you use the jms api on the client it will be easy
to compare both stomp and openwire in this scenario.
When u have the test case, please open a jira so this issue is tracked
and visible and we will have a peek.

The test case will ensure we are investigating exactly what you are seeing.

On 28 June 2012 03:01, Steffen Plotner <sw...@amherst.edu> wrote:
> Hi,
>
> I have been testing and debugging using ActiveMQ-5.6.0 and tcp+stomp the performance of the following application (written in perl using Net::STOMP::Client):
>
> Message headers include persistence=true.
>
> client sends message to a queue with a reply-to header of /temp-queue/return_queue
> server consumes from queue and responds via reply-to
> this takes about 0.08 seconds and works perfectly.
>
> client sends message to a queue with a reply-to header of /queue/return_queue
> server consumes from queue and responds via reply-to
> this takes about 0.003 seconds and works perfectly also - why is the /temp-queue so slow?
>
> I found a post from 2007 and I think the problem exists in the current version.  I would like this to be addressed, as I would believe there are many applications using the temp-queue feature:
>
> http://grokbase.com/t/activemq/users/072hszfp50/sending-async-messages-to-temp-queue-really-slow
>
> Steffen
>
>
> _______________________________________________________________________________________________
> Steffen Plotner                            Amherst College            Tel (413) 542-2348
> Systems/Network Administrator/Programmer   PO BOX 5000                Fax (413) 542-2626
> Systems & Networking                       Amherst, MA 01002-5000     swplotner@amherst.edu



-- 
http://fusesource.com
http://blog.garytully.com