winreg wrote:Capturing Server Performance Data?
The essential (important) difference between eValid and the
www.webperformanceinc.com offering is that eValid is a browser --
records and plays in/from a browser -- vs. capturing HTTP traffic.
eValid does not have the concept of "virtual user". The eValid
Server Loading Bundle (SLB) license authorizes an unlimited number
of playbacks (i.e. simultaneous eValid's ) on one machine...about 200
is what you'll get with a moderate sized, 3 GB machine.
The advantage of the browser is that the script is simple,
it can adapt to changes in the web application, and 100%
realism is assured. If you have an AJAX application, or
one where eValid's Adaptive Playback feature is used to
adapt to minor changes, then the advantage of having the driver
be "active" rather than "passive" (as a URL sequence will be)
should be apparent.
Hence, eValid's LoadTest system imposes very realistic server load, and
we recommend use of existing performance monitoring appliances
to "close the loop" during server loading experiments. You
can see how to do for IIS and Linux servers at:
http://www.e-Valid.com/Products/Documen ... .data.html http://www.e-Valid.com/Products/Documen ... .data.html(We didn't bother to "reinvent the data collection" and populate
eValid with fancy speedometers, etc. What's already available
is far better and besides, our forte is realism in the load
that is imposed.)
Our prices are comporable. I hope you agree that the realism
eValid provides is priceless!
by technology » Mon Sep 15, 2008 10:23 am