What is the effect of the cache_mem parameter in NOVM versions of
squid ? What about memory_pools ?
According to cachemgr, it seems that only "Pool for disk I/O" grows
significantly. What's its use with regard to the regular filesystem
cache ? Doing async. IOs ?
If cache_mem configures all these memory pools, what could be a "good"
value or rule of thumb to set it to if my goal is to minimize memory
use without impairing performance ?
Something like "n seconds of bandwidth" maybe ?
Received on Thu Jan 29 1998 - 09:34:59 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:38:34 MST