I don't deny that. I even know that it is unwise. But your claim that you can't run JMeter with -Xmx1024M on a system with only 1 GB of memory is technically false: it will run, badly probably, but it will run!
So from that aspect, you won't be able to tell that it is a bad decision to assign that much memory on such a low spec machine. You will only (hopefully!) notice it from erratic results when running a loadtest. Only knowledge of how virtual memory, Java etc works will allow you to make an educated decision on sizing the memory parameters. Mark > from a server side java perspective, using virtual memory will kill the > vm's > performance > regards > deepak > > On Fri, Oct 2, 2009 at 12:48 AM, Mark Rotteveel <m...@pluton.nl> wrote: > >> >> > I assume you mean >> > -Xms512M and -Xmx1024M . if you system only has 1GB i doubt you can >> run >> > jmeter with -Xmx1024M >> > see >> > >> http://jakarta.apache.org/jmeter/usermanual/best-practices.html#lean_mean >> >> It will work due to the magic of virtual memory, but usually it is wise >> to >> leave at least 128 MB of physical memory available for the OS and other >> processes (and more memory if there are a lot of different processes). -- Mark Rotteveel Pluton IT --------------------------------------------------------------------- To unsubscribe, e-mail: jmeter-user-unsubscr...@jakarta.apache.org For additional commands, e-mail: jmeter-user-h...@jakarta.apache.org
