I keep single instances to under 1k threads otherwise things start to go wrong.
In fairness, this is without any real JVM parameter tweaking. When I need more than that I go distributed, although at times that just means "please start your jmeter now, Bob" etc. Dave On Mon, Dec 15, 2014 at 8:59 AM, Mark Lybarger <[email protected]> wrote: > > i'm looking to test a system that should handle a load of 100,000 clients > ... eventually. right now, the clients are in fact time shifted and really > the system supports 100 or so concurrent clients. that said, i'm looking > to do some load / performance testing, so i naturally look to jmeter. > > i've read several articles or blogs suggesting a limit of 300 threads for > jmeter testing. i'm going to need many many more. perhaps 3,000 would be a > good starting number. so, finally, my question is this. is the thread limit > based on system resources? would using a distributed jmeter allow me to > get up to 3,000 threads? > > what types of loads have jmeter users typically been able to put on their > systems? > > my system under test is a back office system, but it provides an http > interface to the end client. > > thanks! > -mark- > -- e: [email protected] m: 908-380-8699 s: davelnewton_skype t: @dave_newton <https://twitter.com/dave_newton> b: Bucky Bits <http://buckybits.blogspot.com/> g: davelnewton <https://github.com/davelnewton> so: Dave Newton <http://stackoverflow.com/users/438992/dave-newton>
