On Fri, 17 Dec 1999, Jan Van Ham wrote:
> - Suppose you have 256MB of physical RAM ... is it bad to make a
> statement in squid.conf like cache_mem 256 MB (def : 8 MB) Or will the
> Linux OS starts swapping like hell ?? More cache_mem is offcourse good
> for keeping lot's of data 'hot' (& ready to serve fast ;-))
Actually, Henrik is usually quite adamant that "cache_mem 8 MB" is the
optimal choice, regardless of server size. Certainly don't set it equal to
your total system RAM, as you need space for the system processes and
Squid itself, which depending on the number of objects in your cache can
keep a very large database in memory (if you have a million objects and
Squid needs a few dozen bytes to track each one, there goes a lot of
memory).
> I'm running Apache on the same machine (for the cachemgr.cgi) script ,
> but is this wise ? It will only be accessible from 1 host within our
> network and won't do any extra "handling" of requests/data Actually my
> 200MMX machine doesn't seem to east lot's of CPU ... maximum like 20%
> max ... so the Apach only eats some memory I guess ??
And very little at that - configure Apache to use a single child process,
and you have a very cheap web server. Of course, cachemgr.cgi doesn't need
to be run from the proxy machine - you can install it anywhere, such as on
your primary web server.
-- Kendall Lister, Systems Operator for Charon I.S. - kendall@charon.net.au Charon Information Services - Friendly, Cheap Melbourne ISP: 9589 7781Received on Fri Dec 17 1999 - 18:53:19 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:49:56 MST