[squid-users] squid performance of 2.4 much worse than 2.2 and 2.3?

From: Klavs Klavsen <ktk@dont-contact.us>
Date: Mon, 12 Nov 2001 19:03:12 +0100

Hi Gary and everyone else on the squid-users list..

Is Gary right in his understanding that squid 2.2 and 2.3 are considerably
faster than the current 2.4 stable?
and whatabout the 2.5?

best regards,

Hello Klavs,

Have you tried one of the older versions of squid, like 2.2 or 2.3? i
understand that the performance for 2.4 is considerably less than the
older versions. of course, if you need features that are only available
in one of the newer squid versions, i suppose this wouldn't help much.


On Monday, November 12, 2001, at 09:38 AM, Klavs Klavsen wrote:

> Hi Gary,
> My goal, is to prove a good price/performance ratio.. compared to a box
> such as the CacheFlow box..
> and also a great deal of stability.. as I - and my bosses - would rather
> have a box, that answers a few - and throws off the rest (at high loads)
> than a box that answers none.. - which was what happened here.. (it was
> serving pages for cnn.dk - while the news of the plane crash in USA came
> online)..
> /Klavs
> i think that there are quite a number of caches that perform better than
> squid, but very few of them can match the price/performance ratio. is
> your overall goal absolute performance, or best cost, or somewhere in
> between? i looked at websense, microsoft's ISA, and squid (running
> squid-head right now), and of those three, ISA was the best performing.
> but it was determined that the performance difference between ISA and
> squid wasn't enough to justify the cost of the OS, cost of the ISA, and
> user license costs. of course, ymmv, especially if absolute performance
> is your goal. if that's the case, and you're caching for windows
> clients, perhaps ISA would be worth looking into. and all things
> considered, it's not that expensive (not compared with websense,
> anyhow ;)
> gary shelton
> On Monday, November 12, 2001, at 09:07 AM, Klavs Klavsen wrote:
>> I heard from someone from squid-cache.org that there was gonna be a 4.
>> cacheoff in november, which was suppose to test squid as a reverse
>> proxy..
>> And that's what I'm running.. - squid-head - with the rproxy.patch..
>> I just "tested" my squid, in a live setup, where it achived a little
>> over
>> 900 req/sec - for a short while - before it died.. (it didn't answer
>> requests - had to stop/start).
>> That was untuned.. I would like to know, If I can configure it to
>> perform
>> better.. and to perhaps - instead of not serving costumers - it could
>> just
>> not accept any more connections - when it reaches response times (or a
>> max
>> conn) of a certain amount..
>> I thought of trying to tune max-window size and so on.. I would like to
>> know if any of you have any experience with tuning a squid box in this
>> way
>> - to see how much I can squeeze out of it :-)
>> I have to compare squid with a CacheFlow box.. and the Cacheflow people
>> tell us they can serve many more req/sec..
>> but I ofcourse hope, that I can prove that the CacheFlow box, can't
>> serve
>> more than twice what squid can.. cause then it pays to run squid
>> instead..
>> -------------| This mail has been sent to you by: |------------
>> Klavs Klavsen, IT-coordinator and Systems Administrator at
>> Metropol Online - http://www.metropol.dk
>> Tlf. 33752700, Fax 33752720, Email ktk@metropol.dk
>> Private- Email klavs@klavsen.net - http://www.vsen.dk
>> --------------------[ I believe that... ]-----------------------
>> It is a myth that people resist change. People resist what other
>> people make them do, not what they themselves choose to do...
>> That's why companies that innovate successfully year after year
>> seek their peopl's ideas, let them initiate new projects and
>> encourage more experiments. -- Rosabeth Moss Kanter
Received on Mon Nov 12 2001 - 11:03:15 MST

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:04:03 MST