I might add that you should be able to get some idea of the *minimum* resource requirements by creating a dummy test plan using the JavaTest sampler instead of HTTP samplers, and running that with suitable delays in the samplers, and with all the Assertions and Listeners etc you want to use.
Such tests can be run without the need for a server, but of course won't show if the network connectivity is adequate. The JavaTest sampler also uses fewer resources than the HTTP Sampler. But if you can't run the dummy test on the hardware, it's very unlikely that you'll be able to run the real test... Remember that console-mode is cheaper than GUI, and most Listeners are expensive (disable or delete them unless they are essential). S. On Fri, 26 Nov 2004 07:23:37 -0500, Peter Lin <[EMAIL PROTECTED]> wrote: > that depends on how you run JMeter. If you're running it in GUI mode, > I would recommend > > 512mb of ram > 1ghz > 100mb ethernet > > If you're running in console mode, I would go with > > 384-512mb of ram > 1ghz > 100mb ethernet > > If you plan to test insane load with high concurrency, I would > recommend going up to 768Mb of ram and gigabit ethernet. Aside from > that, a couple cheap machines should be better at producing heavy > concurrent load than one expensive machine. > > peter > > > > > On Fri, 26 Nov 2004 13:16:15 +0100, Coret Bob <[EMAIL PROTECTED]> wrote: > > Hi all, > > > > I was googling on the hardware requirements for Jmeter, but only found > > questions (see below). Does anybody know (from experience) what the > > hardware requirements are (for a reliable, realistic performance test)? > > > > Ragards, > > Bob > > > > * From: rainer_bohnert > > * Subject: jmeter hardware requirements > > * Date: Thu, 13 Jun 2002 06:23:04 -0700 > > > > hi there, > > > > we are considering jmeter as a load testing tool for a web application. is > > there any formula to calculate the hardware requirements for running > > jmeter? > > > > something like: no of concurrent threads X no of http requests/s per thread > > = such a hardware configuration to produce reliable results ? > > > > is there any experience of the hardware configuration of jmeter running on > > windows/nt simulating up to 100 parallel threads testing the performance of > > a web-site ? > > > > any help welcome, thank you > > > > rainer > > > > --------------------------------------------------------------------- > > To unsubscribe, e-mail: [EMAIL PROTECTED] > > For additional commands, e-mail: [EMAIL PROTECTED] > > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
