On Sun, 20 Aug 2000, Daniel Cox wrote:
> I am using Polygraph 2.5.4
>
> I am looking for the easiest way to configure the test to run just
> warm,inc1,top1,dec1,cool on a machine with an already full cache - to more
> quickly find an approximate peak request rate.
>
> Just commenting out the other phases in polymix-3-guts.pg causes a very low
> document hit rate and therefore a much bigger response time.
> lx on the resulting log for phase top1 shows an offered hit rate of only 5.1&
> Does the fill have to be run?
Hi Daniel,
Note that there is some "inheritence" between phases. That is,
one phase passes on its parameters to the next phase unless you
specifically change them.
My guess is that you removed the phase that changes the hit ratio
from 5% to 55%.
A good approach is to change the duration (or goal) of the phase,
rather than removing it. So, keep the fill and "fexit"
phases, but change their goals. Maybe you can use something like
goal.duration = 5min;
Duane W.
This archive was generated by hypermail 2b29 : Tue Jul 10 2001 - 12:00:14 MDT