Dear all,
I've now got Squid successfully running on my Redhat 6.2. My final
It's just not my weekend with emails - this time a bit got left off the end
of this one so I've replied to myself :-( and added the bits left off.
----- Original Message -----
From: npauli
all boiled down to fine tuning the squid.conf script. While getting Cache Manager up and running, I had somehow persuaded myself that I needed always_direct allow school-network.
Don't ask me how or why as it had the effect of causing all those machines to try and bypass the default parent cache. This meant that they could only see files that had already been cached by localhost.
It will be interesting to see what happens when people get back from half term. I won't be surprised if I have to rapidly retract the first sentence of this email!
BTW, I'm running Squid for a network of 35-ish machines with a max of 24 concurrent web browsers on a machine with 64MB of RAM, a 450 Mhz PII, > and a 4 GB HDD that is dedicated entirely to Squid i.e. it's /usr/local/squid.
I didn't design this setup to run squid; the machine (and especially its spare 2nd hd) was hanging around not doing much so I gave it this job. What do people think of this? Would it be worthwhile adding more memory? I've increased the cache directory size from the default 100Mb to 3000Mb? Would it be worth cutting it back, do you think? Nigel Nigel Pauli, St. John's School, Northwood