Eric Stern wrote:
> I was running the cache from empty for about 10 minutes, just to see if it
> would hold up under a load of x requests/second.
Not too interesting then for being a Squid measurement ;-)
To make a Squid test interesting you need to at least
* Have a fairly large cache already built up. How things behave while
the cache is filling up is not that interesting as it is not the normal
mode of operation.
* Have it running for an extended period of time. At least 30 minutes of
a steady request rate before your measurements begin.
Otherwise it is quite easy to fool oneself with a unrealistic peak
measurement which does not hold out in real life use.
/Henrik
Received on Mon Feb 07 2000 - 15:43:02 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:12:21 MST