Actually I asked a similar question awhile ago. The bottleneck is caused the poll/select implementation. And there is no plan/way to improve it in Squid-2.5. You can find the dev discussion about replacing it with epoll in Squid-3. It's working but not stable and I haven't tested it. George > -----Original Message----- > From: Kinkie [mailto:kinkie-squid@xxxxxxxxx] > Sent: Monday, March 21, 2005 6:43 AM > To: squid-users@xxxxxxxxxxxxxxx > Subject: Re: [squid-users] Performance-problems on reverse-proxy squid > > On Sat, 2005-03-12 at 22:50, Stefan Neufeind wrote: > > Dear all, > > > > I'm running a squid-proxy (squid 2.5.stable7-1) in reverse-proxy mode in > > front of two webservers. Squid does equal loadbalancing across the > > servers, and answers requests for static pages/images/... itself. > > Because of the site-content squid is able to service about 80%-85% of > > the requests itself. Statistics report about 500 requests/second hitting > > squid, with an output to the internet of about 20mbit/s during peak-times. > > [...] > > > Does somebody have an idea how I could "debug" the cpu-utilisation of > > "system", and how to lower it? Friends told me to watch out for possible > > buffers that could reduce the number of transfers from > > kernel-to-userspace or so - but I didn't find much. > > The figures do not really look too bad. > > As for debugging, you could try using opofile (it should be enabled in > Fedora). > > Kinkie
