[squid-users] Clustered Reverse proxy

Hi,

I'm looking to setup a reverse proxy with Squid. It will be sitting at
the "back end" of my application stack, i.e. it won't be accessed
directly from the outside world. I need to be able to do the following
things:

1. Force an item in the cache to expire now (i.e. the next time
something accesses it, it will get it from the origin)
2. Force an item in the cache to be refreshed immediately from the origin
3. Ensure requests to the origins are minimised, i.e. if a whole lot
of requests come in for the same item, it won't duplicate those
requests to the origin unneccesarily.
4. Allow querying of different items in the cache for operational purposes
5. Run two separate instances (on different machines) that are
consistent, and things will continue to work more or less should one
of them fail

I was just wondering if Squid is suitable for this purpose? #5 could
be achieved through OS clustering (will be running on Solaris), but
would be interested to hear from anyone who's run this sort of setup
before, or any other advice anyone has to offer?

Thanks in advance,
Sam.
Received on Thu Feb 07 2008 - 02:10:33 MST