John Oliver wrote: > I'm looking for an http load test that will load XML files and then GET > any and all objects specified in that file. I also want it to do this > up to dozens and then hundreds of times per second. >
There are a bunch of promising looking leads by googling on: http load test tools If you want to roll your own, the biggest problem would probably be with simulating arrival patterns. If you aren't worried about that part of it, then there's just 2 pieces that I see, and solving them separately seems like a decent divide and conquer approach. 1. finding the "objects" in the XML files -- I presume this means that the URLs to retrieve are in those XML files. 2. launching the individual http requests For part 1, I have recently found the PyXML tools pretty decent, and reasonably fun to play around with. If you prefer perl, there's Twig and probably lots of other packages, too. If java is to your liking, there's zillions of java XML tools (but, of them I know knothing). For part 2, again your preferred toolset would be the determining factor. Python has urllib, or httplib. Perl has HTTP::Request and, I'm sure some other things since I last poked around there. Or there's wget and related progs. If you want to DIY, "cookbook" or "recipe" might be good terms to include in your google search line. ..jim -- [email protected] http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
