Hey Namaskara~Nalama~Guten Tag~Bonjour
Looks like, the throughput goes up to 333 as your hardware might be overloaded. You might need a bigger server. Everything should work ok then. :) Deepak -- Keigu Deepak +91-9765089593 [email protected] http://www.simtree.net Skype: thumsupdeicool Google talk: deicool Blog: http://loveandfearless.wordpress.com Facebook: http://www.facebook.com/deicool "Contribute to the world, environment and more : http://www.gridrepublic.org " On Fri, Feb 10, 2012 at 7:45 PM, Birgit Pretscheck <[email protected]>wrote: > Hello, > > I want to measure the utilization of my resources for different request > rates. > For example my measurements will be: > > 1st measurement: 100 requests/s over 30 min > 2nd measurement: 200 requests/s over 30 min > > and so on > > Until now I used constant throughput timer with the option > all active threads (together). I used a stepping thread group > from jp@gc plugin with: > > start 1000 threads > First wait for 0s > Then start 1000 threads every 20s > using ramp-up 5s > Then hold load for 1800s > Finally stop 1000 threads every 20s > > As listener I use sumary report and show results in table. > If I use the constant throughput timer with 18000 requests/minute, > the throughput in summary report is 333,33 requests/s (constant) > instead of 300 requests/s as wished. Even with 16000 requests/minute > the throughput goes up to 333,33 requests/second. > > Could someone explain, why this is the case? > > I just want to have constant 300 requests/s over 30 minutes, > the rate should not alter between e.g. 280 and 320. > For 50, 100, 200 requests/s everything worked fine, but now with 300 > requests/s > it goes up to 333 requests/s in the end. The avg. latency of my requests > is 5 ms. > > It didn't work, so I took jp@gc Throughput shaping timer. > Now my summary report displays 296,6 requests/s in the end (and during the > whole measurement). > This is ok , BUT now the load of the server goes up from > 0,03 at the beginning to 3,05 (15 min), 4,14 (5 min), 5,00 (1 min). > > This was not the case without the plugin. In this case the load value was > at 0,7 (15 min) at the end. The avg CPU utilization for both cases is 30%, > the avg. network send / rec is only at 200-400 K, free RAM only goes down > from > 700 MB free to 400 MB free in these 30 minutes. Everything is nearly the > same > in both cases, except the load value. > > Can someone please explain, why the load rises to this extend compared to > the constant > throughput timer? > > Later I want to shape the request rate, eg. from 100 to 1000 and back to > 100 requests/s. > So I need the throughput shaper, but I want to have a normal load value. > How can I achieve this? > Is there another throughput shaper for my case? > > > Thanks a lot! > > Best regards > Birgit > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > >
