On Tue, 28 Nov 2000, mohan kanumuru wrote:
> Using attached workload file, I am able to simulate
> without problem upto 250 requests/sec using 100 robots
> per machine. I tried 400 request/sec using 100 robots,
> but application is stopping after printing out a
> message "virtual memory exceeded in new".
That is because you have limited each robot to have up to 4 open
connections at a time (open_conn_lmt = 4). With that limit, you cannot
get 400 req/sec with 100 robots. Moreover, even doing 250 req/sec with
100 robots is very risky -- make sure your transaction queues (plotted
on the autogenerated report) do not grow long, or you will end up with
a best-effort workload!
BTW, mailing list archive contains a few messages explaining the
"virtual memory exceeded in new" error.
Alex.
This archive was generated by hypermail 2b29 : Tue Jul 10 2001 - 12:00:16 MDT