tapestry-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Kalle Korhonen <kalle.o.korho...@gmail.com>
Subject Re: Performance Testing Lab
Date Thu, 02 Sep 2010 22:02:09 GMT
For this type of testing, I've used JMeter
(http://jakarta.apache.org/jmeter/) successfully in the past. Without
a doubt, you need at least two machines, perhaps more, otherwise the
results will be skewed as you quickly exhaust the clients even if they
don't fully parse the responses. I wouldn't worry about creating a
real-world app, a synthetic one that exercises key parts of the
framework should give more meaningful results. Memory and cpu resource
consumption should be measured on the server, throughput from the
(master) client.

I have not used Performance Hudson plugin
(http://wiki.hudson-ci.org/display/HUDSON/Performance+Plugin) but it
seems ideal for this. I might be interested in participating in this
effort. There are several things that need to be done - securing
dedicated hardware is surprisingly one of the most difficult issues
(anybody can find a few boxes for temporary use, but how do you make
sure you can keep the environment stable and unchanged for years to
come). A desktop or two running under somebody's desk is not a great
solution in the long run.


On Thu, Sep 2, 2010 at 2:30 PM, Howard Lewis Ship <hlship@gmail.com> wrote:
> One of my long term and unrealized goals for Tapestry is to have a
> legitimate performance testing lab.
> I'd like the ability to run a "standard" application through its
> paces, collect statistics, and see how well each new Tapestry release
> compares to its predecessor in terms of general performance: response
> time, memory utilization, saturation and recovery.  Ideally, it would
> be nice to create equivalent JSP/Spring MVC/Struts/Wicket
> implementations of the same app.
> I'm really concerned with ensuring that 5.2 out-performs 5.1, and uses
> less memory.  The singleton pages approach should lower the memory
> utilization ... but I'm concerned that the new AOP stuff inside
> ClassTransformation (which has a tendency to create lots of little
> one-off classes) will overshadow the other improvements.
> So ... how do we go from thinking about this to doing it?  I have a
> spare desktop I could use for this purpose (I'm going to wipe out
> windows and install Ubuntu). It's dated hardware but otherwise
> probably ok.  Does any of the committers, or Tapestry community, have
> some specific skills, or better ideas?
> --
> Howard M. Lewis Ship
> Creator of Apache Tapestry
> The source for Tapestry training, mentoring and support. Contact me to
> learn how I can get you up and productive in Tapestry fast!
> (971) 678-5210
> http://howardlewisship.com
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@tapestry.apache.org
> For additional commands, e-mail: dev-help@tapestry.apache.org

To unsubscribe, e-mail: dev-unsubscribe@tapestry.apache.org
For additional commands, e-mail: dev-help@tapestry.apache.org

View raw message