Hi all, I played with different setups and came across something interesting. Maybe somebody has had this experience before...
Setup: 2 identical 3.0Ghz 1Gb RAM PCs running Fedora with Java 1.5.0_06 and Jmeter 2.1.1. Both run a script with 66 threads against a webserver. All tests are run manually from the command line calling jmeter -n. Test is Ok and results as expected. Now I double the threads on one machine (132 threads, same script, same theoretical throughput as the test from two machines before) and run the test. The machine does not reach the same throughput and is more unstable. That sort of figures. Now I run jmeter 2 instances with 66 threads each on one machine. The results are the same (or nearly identical) as in test one, where I test from two machines. I'd love it if somebody can run a similar trial on one of their boxes and confirm this. Regards Oliver --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]