I see alot of these messages in my squid warning log...
Specifically, in filtering off the date, and sort+uniq+counting, I see:
var/log# grp "Median response" warn|cut -c36-90 |more|sort|uniq -c
107 WARNING: Median response time is 57448 milliseconds
1 WARNING: Median response time is 6996 milliseconds
1 WARNING: Median response time is 7384 milliseconds
----
This is out of 471 lines in this 'warn' log -- over 20% are warnings from
squid that things are taking alot longer than I would think they should.
Usually, it doesn't behave this way...
Notable, I have my "dns" timeout set for 1 minute (so the above are all
slightly under my dns lookup time.
I also notice that the default timeout time waiting for a connection is
about 1 minute.
Could it really be that I have this many timeouts ... ? 1 minute seems like
a pathological case -- not something I'd consider 'normal'...
Ideas? Suggestions? Do other people get these types of warnings often?
Thanks (My DSL is 3Mb/768) that times close to that rate when I time it...)
-linda
