We are running several Squid proxies (2.4.STABLE6) for one of our
services, however we are seeing oddness in the monitoring systems logs.
To test the servers are up and running our monitoring software tries to
connect to various sites via the proxies to ensure they are operating
correctly. At very irregular intervals (don't match up to running cron's
or anything), we get a timeout responce from one of these tests, the next
time it runs, there is no issue. I cannot see anything in the logs which
look suspicious. We are testing with different sites (google, bbc, demon
etc) as the destination and the fault does not affect the same site,
different sites at different times on different cache boxes.
From what we can tell the service is running happily, just these errors in
our monitoring are starting to become irritating (we want perfection :) ).
Does anyone have any ideas what may be causing this? or possible