I'm wondering if DDOS style attack can be used to gauge performance of servers... or is it predominantly network biased test?
Some history: At work, we are trying to test new CMS system for performance. Developers wrote a program that makes concurrent calls to CMS for various actions. As they told me it makes around 600 actions per minute.
They want to run it on one computer.
Now my hunch in the stomach is that this will never be a real concurrence! When site is accessed by allot of people from different IP's and different locations server load will be much higher than when 1 computer is calling the server from inner network.
But hunch needs proofs.
Now my question is can we use DDOS style attack to simulate a real world activity?