----- Original Message -----
From: "Elfredy V. Cadapan" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Saturday, June 24, 2000 06:18 PM
Subject: Re: [plug] squid advantage.....
> On Fri, 23 Jun 2000, fooler wrote:
>
> > > My own university cache (~100 concurrent users) peaked out last week
at
> > > 100 requests/minute. It's normally well under 20 req/min though.
> >
> > we're having N kilo request per minute each of N squid proxy servers.
>
> How many concurrent users do you have? (estimate?)
sorry i cant reveal to you the figures but its more than what you have (~100
concurrent users). aside from the users, there are lots of sibling proxy
servers connected to it.
> > > According to the Second IRCache Web Cache Bake-off,
> > > (http://polygraph.ircache.net/Results/bakeoff-2/) Squid finished
second in
> > > overall response time, and had the highest hit ratio.
> >
> > according to that URL, which do you prefer, one iMimic 5000 cache engine
or
> > N squid boxes to achieve the same request per second of that iMimic
cache
> > engine in terms of price over performance for large scale proxy service?
>
> Dedicated cache appliances will always outperform non-dedicated
> general-purpose boxes, due to the performance of the underlying hardware,
> TCP/IP stack, etc.
if you take a second look again to that URL at section 8, iMimic and Squid
runs under FreeBSD version 4.0 and 3.3 respectively with no other services
run except with their own proxy software. how come iMimic is a dedicated
cache appliance? :-> one good example of dedicated cache appliance is
novell's internet caching system. eventhough novell is a dedicated cache
appliance, but iMimic still performs well.
> Too bad Microsoft Proxy and Netscape Proxy weren't in that bake-off,
> though.
well, i dont want to give some bad assumptions here but with the first
bake-off results, they know where they stand for. that is why alex rousskov
pity them for not joining the second bake-off contest.
fooler.
-
Philippine Linux Users Group. Web site and archives at http://plug.linux.org.ph
To leave: send "unsubscribe" in the body to [EMAIL PROTECTED]
