> By the way, if let say we go to production and migrate users, what are the things that we > should look into for fine tuning? > - Basically you should keep track of 'http reqs/sec' during the busiest load. (Check cachemgr -> General Runtime Information). If you stay under 100reqs/sec , there's nothing to worry about because native SQUID on normal hw can handle much more. - Watchout for error in cache.log too, during SQUID operations. M.
