On Fri, 13 Feb 2004, Scott Phalen wrote:

> Hello, I am very new to the world of squid.
> First question ~ will WCCP V2 be incorporated in Squid 3.0?

I hope so.

> Second question ~ This is pertaining to cache_mem size. Here is what I am
> using:
> Dell 650 PowerEdge Server
> 3Ghz Processor/2GB RAM
> 146GB Hard Drive x 2
> Linux Advanced Server 2.1
> I configured squid to use a "big" partition (100GB). Squid is the only
> process running on this server. There are roughly 1500 computers going to
> be using this cache. I have an identical Dell 650 configured to help load
> balance but when I start Squid on that server it tells me "Squid is already
> running".
> What would you suggest the default cache_mem should be? Currently I have it
> set at 1024MB.

That is probably too high. You should set it lower, like 128MB, then
wait until Squid's disk cache fills up. Once Squid has been running for
a while with a full cache, then consider increasing cache_mem if you
have RAM that is being unused.

> I am watching my "TOP" screen to see how much CPU and MEM squid is using,
> anyone know what RSS is? It is currently at 149M after squid running for
> 135 minutes. And it appears the memory is climbing pretty fast.

RSS is essentially the process size. It stands for Resident Set Size.

Duane W.