Re: Info on Big Squids wanted

From: Dancer <>
Date: Sun, 24 Nov 1996 03:15:08 +1100

I've heard exactly the same response from cache-administrators whose
systems deal with less than 500K requests/day. Not one of them has even
tried squid, and most have never directly talked to anyone who has.

I note: These people have probably wrangled for budgets, hardware, and
software (say Roxen, or Netscape), and during the whole process have had to
defend that this was 'the best decision' to make. Squid may not have been
solid and reliable enough at the time. In some cases load-handling in the
earlier versions may well have been an issue...but when you've backed a
product like these guys have, it's _not_ easy to turn around and say,
'well, I guess the money we spent on X was a waste...Look, here's Y and
it's free and works better.'.

        When you do that, people start questioning your competency. I mean,
really. It's a fluid world we work in, and (despite what people might
think) information in some areas can be pretty scarce. I stumbled onto
squid by accident. Proxy caches are like that. Silent little buggers, that
you only really notice when they _don't_ do the job right. These guys made
the best decision they could at the time (based on what information they
could dig up, what they heard on the grapevine, and whatever other criteria
they chose to - and/or were instructed to - examine).

Do you still have to fight an accountant as to wether a 386 PC is still a
contributing member of the laboratory, and getting upgrade-request after
upgrade-request knocked back? One day, I guess they'll get the idea that in
the world of technology we _can_ completely change things overnight. Until
then, we just have to defend our positions as best we can to get anything
at all out of them, and not look like a pack of bloody idiots :)


> From: Jon Peatfield <>
> To:
> Cc:
> Subject: Info on Big Squids wanted
> Date: Saturday, November 23, 1996 5:07 PM
> I'm looking for information about Giant Squids ;-)
> By this I mean any Squid (or Harvest) server (or collection of servers
> are peered) which handles over 2 million requests per day.
> Anyone in the UK will probably guess why I want this information, but for
> rest fo you this is to try to add some fuel to the debate about one of
> centrally funded UK Academic cache networks, which currently uses
> proxy, but the same hardware might do better with Squid (at least we
could do
> ICPs to it). They think that Squid couldn't cope, I disagree.
> -- Jon Peatfield (with hat on)
> Jon Peatfield, Computer Officer, the DAMTP, University of Cambridge
> Telephone: +44 1223 3 37852 Mail:
Received on Sat Nov 23 1996 - 09:16:15 MST

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:33:36 MST