[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Recommendations for Squid proxy hardware

On Wed, 5 Jul 2000, ridgey wrote:

> What would you recommend for Squid proxy hardware, for a machine thats
> running transparent and has approximately up to 80 dialups pumping requests
> through it?
> Current machine has a drive setup of:
> Filesystem            Size  Used Avail Use% Mounted on
> /dev/sda1             288M  220M   54M  80% /
> /dev/sda3             7.8G  5.4G  2.0G  73% /var/spool/squid1
> /dev/sdb              8.2G  2.1G  5.6G  28% /var/spool/squid2
> /dev/hdc              9.2G  6.1G  2.7G  69% /var/spool/squid3
> /dev/hda1             9.2G  6.1G  2.7G  69% /var/spool/squid4

OK, so approx 34GB squid spool.  That seems generous for 80 dialups.  Have
you looked at the LRU expiration age of your cache?

> hdc and hda1 being UATA, rest being Ultra Wide, have installed 256 MB Ram,
> after around 24 hours all RAM and all swap is taken up and needs restarting.

Strewth, sounds like something might be leaking memory.  Are you certain
it's squid that's doing this?  e.g. does the situation improve if you
restart squid after say 12 hours?

Not that using it on a proxy is anything but evil, but how much swap do
you have allocated?

So far as squid is concerned:

* what version?

* what's the setting for cache_mem (not too big please, e.g. <=32MB).


Reply to: