Blandt
Well-known member
This is question for the optimization experts, even though I suspect my google reading have probably gave me similar answers already...
On my dev test bed installations on different servers, I am noticing a lot of fragmentation on the APC cache even though it is not being used anywhere near its limit (shm_limit="32m", usage only around 10m). Should I worry about this, or is it safe to ignore?
My own observation (in other words not scientific

in 64bits distros APC doesn't perform as well as in a similar configuration 32bits