I am not personally familiar with WebLoad. There are a lot of open source tools out there, with varying degrees of value. One key to remember is that sometimes with open source, you get much more than you pay for (Watij, Selenium, etc.) and sometimes you do not.
The key to selecting a good load tool is to know up front what you need to do. Are you simply generating http traffic, or will your tool need to leverage other protocols, too? That will severely limit your options. A second issue is how realistic the load tool is. A very common open source load tool which I have used (in the past, as in ‘terminated’) showed signs of serious flaws. First, it uses http 1.0 and not 1.1 (there are significant differences in the protocols, and ALL web browsers today use 1.1), making it impossible to truly mimic user interaction. Secondly, the tool reported false levels of load – while it was telling me it was simulating so many thousands of users, our servers showed a much different story. Finally, said tool also tapered its load off over time – so it might start at 500 users, but 15 minutes later it was only generating the traffic of 200 or so. The end result was a very costly, false, sense of security which bit us in the long run.
Other open source tools suffer from the ‘multi-tool bulk’ effect. I like to backpack, and in so doing I try to keep my load light. Some backpackers bring along a multi-tool – it’s a tool that will do practically anything. Usually more than you need on a hike. And a tool like this can weigh a lot! It’s the same with some open source load tools – with clients and servers and managers, the complexity of just installing and configuring the tool can be overwhelming. Make sure whatever solution you pick won’t cost you more in the long run than you’ll save in the short run.
Bottom line: when you choose tools, don’t allow cost to be the overriding factor. Sometimes saving a little money up-front can cost a lot in the long run. Some open source tools, however, can cost you a fortune in install, config, or inaccuracies. Avoid these tools! Remember: you’re an engineer, so there is no right tool for you personally, just a right tool for the job.
All the best and successful testing!
John Overbaugh is a testing professional with 13 years of experience. He blogs irregularly at <a href=”http://thoughtsonqa.blogspot.com”>http://thoughtsonqa.blogspot.com</a> and is a frequent contributor and testing expert with IT Knowledge Exchange..