Hi.
I've been getting some slow responses from squid, to the point where
telnetting to the proxy will sit there for a minute (sometimes longer)
trying to connect.
Usually it happens in the afternoons, so I had put it down to all the extra
users on at that time... We have a backup proxy (so it hasn't been _too_
much of a problem), but the weird thing is that at other afternoons when the
system is just as busy it will be alright.
The system is a ppro200 w/192meg ram & 8gig cache on redhat4.1 and squid1.1.14,
we have roughly 160 dialup lines and two ISPs hanging off us, with a 1meg link
out.
This was happening with squid1.1.10 plus it also happens with squid1.1.14.
The cache log seems to be filled mainly with ERR_CLIENT_ABORT messages and
the like.
Any help would be much appreciated...
Also, while I'm at it, I rebooted the server today and ran a filesystem check
on the cache drive (it's a multi disk, 2 x 4gig scsi) which came back with a few
file size errors... After it came back up squid has been going on about
WARNING: Object size mismatch, expects x bytes only read so many bytes...
What does squid do in this situation? Reget the file? Unlink the object?
Thanks.
Jeremy.
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+ jeremy clough - systems administrator - vianet australia +
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Received on Tue Aug 05 1997 - 01:51:59 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:35:55 MST