Mailing-List: contact dev-help@activemq.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@activemq.apache.org
Received-SPF: pass (herse.apache.org: local policy)
Subject: Re: AMQCPP Openwire much slower than Stomp?
From: Timothy Bish <tim.bish@sensis.com>
Reply-To: tim.bish@sensis.com
To: dev@activemq.apache.org
In-Reply-To: <20070418095106.GA20229@dogbert.sdsl.sun.ac.za>
References: <20070418095106.GA20229@dogbert.sdsl.sun.ac.za>
Content-Type: text/plain
Content-Transfer-Encoding: 7bit
Organization: Sensis
Date: Wed, 18 Apr 2007 07:14:52 -0400
Message-Id: <1176894892.3431.3.camel@tbish-laptop.ats.sensis.com>
Mime-Version: 1.0

Man, you complain when we don't have openwire and then complain when we
do, some people :)

There's several areas where code is not optimal but this was the first
cut and we wanted to focus on getting something that worked, not
necessarily something that worked to perfection.  I know of a few areas
that need tweaking and hope to start addressing some of them in the next
release.  

Keeps us appraised of anything you find that will help mature our
Openwire support.

Regards
Tim.

On Wed, 2007-04-18 at 11:51 +0200, Albert Strasheim wrote:
> Hello all
> 
> We started doing some tests with Openwire in our AMQCPP applications 
> today and we've noticed that Openwire seems to be much slower than Stomp 
> (anywhere from 3 to 20 times) for the simple case of sending messages of 
> a few kB in size from a single producer to a topic with no consumers.
> 
> We're using the same broker, same everything -- the only thing we change 
> is the broker URL used by the C++ application.
> 
> Has anyone else done any Stomp vs Openwire benchmarks with the latest 
> AMQCPP from trunk or with the 2.0 RC?
> 
> Thanks!
> 
> Regards,
> 
> Albert