Why wouldn't you use two squid processes ?
The first one is a cache without redirection (in accelerator mode). It sends
the requests to a second one, which does load balancing with its redirector.
This way, the load balancing is transparent to the first squid process. The
second process (load balancing) can be configured not to cache anything.
François Lauprêtre
Dresdner Kleinwort Benson (Marchés)
IT Department - Head of Distributed Infrastructure
Tel : +33 1 44 70 80 38 - Fax : +33 1 44 70 81 59
mailto:Francois.Laupretre@Dresdner-Bank.de
-----Original Message-----
From: Toby Dickenson [mailto:mbel44@dial.pipex.net]
Sent: Monday, July 17, 2000 3:20 PM
To: squid-users@ircache.net
Subject: Redirect before checking cache
I am looking to use a redirector with an http accelerator to load
balance across a cluster of identical application servers.
It seems that squid checks its cache after the redirector has been
called, using the redirected url. This is less than ideal for me,
since it will never used cached content if retrieved from a different
machine in the cluster.
I cant see a way to get squid to cache content using the url from
before redirection. Am I missing something?
Thanks for your help,
Toby Dickenson
tdickenson@geminidataloggers.com
Received on Tue Aug 01 2000 - 09:04:33 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:54:41 MST