Setup load-testing of our website, try to crash it (with jmeter ?)

We're developping a friend-finder style website, and need help with several things, but most urgent at the moment is with Q.A. (quality assurance) of the http daemon that serves the requests. The site is developed with the [login to view URL] framework. In theory this should be a high performance site, yet some of us are experiencing lock-ups (one thread gets frozen polling for data) even in simple dual-access scenario, whereas some others cannot put a dent in the server even with dozens of simultaneous requests to hammer it. So whatever problem there is, needs to be made reproducible, so I can work on fixing the bug. I see that "jmeter" is recommended to test websites. Tried to set it up myself. The framework engine we used is based on URL rewriting (not cookies) so in jmeter I setup a "recorder". Tried adding more "thread groups" in jmeter etc but no dice. Basically I need somebody to look into it for me so that I can focus on coding. We'll consider this project to be successful on one of two conditions: 1) you find a _reproducible_ way to "break" our httpd, so that I can get to work on the breakage, regardless of whether you use high-intensity or low-intensity hammering of the website 2) you provide a load test system that is high-intensity, sustained, thorough (all pages of the site) We are most interested in goal 1, but we also provide goal number 2 so that the candidate can complete the project even if he/she does not find a way to "break" the site. For the test system to be _reproducible_ at our end, provide us with e.g. the ".jmx" file if using jmeter, and help us setup the test. Thanks for reading ! Looking forward to read proposals. This project is a "trial baloon" of sorts. If we're really happy with your work we'll be quite open to developing a long term, well rewarded relationship with you. (we'll have positions open for sysadmin, security, SQL, testing, etc) *EDIT*: if you want to go the extra mile, give us some information about how performant our server is, i.e. how many requests per second it can handle and how quickly it responds to requests.