Steve Judge wrote:
> A couple of questions:
> 1) I didn't think that squid cached cookies ?
It doesn't cache cookies. It caches the page if allowed, but not the
cookie (cookies are uncacheable).
> 2) Dynamic PHP pages, are these the same as cgi's and should we
> include them under the stoplist part not to cache or fetch from
> neighbors etc ?
PHP is a alternative to ASP.
Yes, you could say that PHP kind of like CGI.
Yes, you should probably include .asp and .php to your
hierarchy_stoplist.
Dynamic pages are NOT cached by Squid unless you have a refresh pattern
with a min age other that 0. You SHOULD NOT have a min age other than 0
on unknown or HTML files as this is how dynamic content usually is
identified.
If I am not mistaken some versions of IE gets totally confused when
using a proxy and caches everything locally vithout revalidation (not
even reload works). This could also be the cause here.
What is said in the Squid access log? *MISS, *HIT or nothing?
> I've also emailed the company concerned suggesting that their
> product should be more proxy aware as they are a fact of life on
> the internet now instead of adding isp's to a ban list that are
> running " aggressive cacheing practices " as they put it.
If their pages are unshareable then they should have a
Cache-control: private
header
If their pages are one shot only (not even the end user should be able
to redisplay the same page without refetching it) then
Cache-control: no-cache
and/or
Pragma: no-cache
should be used.
Pragma: no-cache is defined in HTTP/1.0 and works in all caches and
browsers. It is replaced by the more flexible Cache-Control directive in
HTTP/1.1.
--- Henrik Nordstrom Spare time Squid hackerReceived on Thu Dec 10 1998 - 18:43:20 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:43:36 MST