Simon Greaves <Simon.Greaves@usp.ac.fj> writes:
> Squid-2.2STABLE4, Pentium 100, 128MB RAM, 6Gb IDE :-(
That machine should do the trick; our dedicated P200 serving the whole
university + several regional siblings is somewhere around load 0.15;
monitor your LRU + latency etc. through the cache manager. Add RAM and
disk as needed. We have 256 MB + 8 GB. That's just fine for our 350 K
requests per day. More info (in german only -- sorry) can be found at:
http://www.informatik.uni-bremen.de/t/info/text/www-cache/
> I tried changing OPEN_MAX in /usr/include/linux/limits.h and
> reconfiguring and recompiling, but no change. Do I need a kernel patch?
> Do I need to modify header files? Should I use FreeBSD instead?
Depends on which system you know best (as always). Anyway, for Linux
check:
http://hem.passagen.se/hno/squid/#notes
> The Uni here has a single 64K link for some 800 staff and 3000 students.
> The link is totally saturated during the day. I'm planning on setting up
> 2 or 3 small caches at first, then once we have them working [...]
Well ... sounds like some sort of real challenge. Have fun. ;)
Niels
-- Niels Pollem, np@tzi.deReceived on Thu Sep 02 1999 - 06:30:15 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:48:13 MST