As I mentioned, memory problems are easier to spot from the system
perspective (like top, vmstat, ps). Squid only knows what it has
allocated...it doesn't know how much the system has
free/buffer/cache/swapped out. Squid sees memory+swap, since in Unix
swap and real memory are the same to user processes. (There are
userspace means of knowing this stuff, but Squid doesn't go out of its
way to figure out how much memory is really free.) The cache manager is
good for checking on performance, DNS reply latency, file descriptor
usage, hit ratio, etc.
Justin Hennessy wrote:
> Ok, I have all of that.
>
> Just a quick one, where should I have been looking in the cachemgr for
> this info. I looked in the Memory utilisation section but could not see
> anything glaring.
-- Joe Cooper <joe@swelltech.com> Web caching appliances and support. http://www.swelltech.comReceived on Fri Jun 21 2002 - 19:08:36 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:08:45 MST