--- Perrin Harkins <[EMAIL PROTECTED]> wrote: > On Tue, 2004-03-23 at 10:52, Chris Gamache wrote: > > I wrote a load-testing script that hits my Apache webapp with 100 > serialized > > requests. The webapp works great that way. If I start multiple simultaneous > > instances of the load-testing script, the requests eventually fail. > > Anything will fail if you put enough load on it. Are you saying that it > failed in a situation where it shouldn't have? How did it fail? Was > the hardware heavily loaded at the time (CPU/RAM)?
True... :) Is 4 simultaneous requesting threads with 100 queued serial requests for each thread enough to cause a requests to fail? CPU/RAM load was nominal. When I started only four testing threads, I thought I was going easy on mod_perl :) Incidently, I rewrote the module to not use LWP::Useragent and XML::DOM to pull the XML document. I am using IO::Socket::INET instead. I thought that by using LWP::UserAgent and XML::DOM to pull XML documents I would be taking advantage of the advanced features like timeout, and XML Document Validation. It turns out that they were crumbling under what I thought was a light load. Perhaps I was misusing them? I don't know. But running without a safety net seems to actually be producing better results! I'd love to hear about what I could do differently with the original code, but this isn't the place to ask about XML::DOM and LWP::UserAgent. :) CG __________________________________ Do you Yahoo!? Yahoo! Finance Tax Center - File online. File on time. http://taxes.yahoo.com/filing.html -- Report problems: http://perl.apache.org/bugs/ Mail list info: http://perl.apache.org/maillist/modperl.html List etiquette: http://perl.apache.org/maillist/email-etiquette.html
