AW: Squid stops responding to requests, Solaris 8 Intel & Stable 2.3

Hi,
sorry to answer with a question - but i have the same problem.
I'm running stable 2.3 on 4 linux boxes (compaq 8500 , kernel 2.2.13)
with async enabled. After 3 days ALL the squids stop working, but the port
is still
open and the process alive - more or less. After a kill -9 and a restart
erverything
is fine for another 3 days. The logs are without any informations, also with
debug level 2.
More is not posible over 3 days. A ptrace on the hanging proces shows an
error while
calling accept() on a socket - but this looks more like a symptom than the
real reason
We use a f5 load balancer in front, but this should not be the problem.
Any ideas???
egards
rolf

-----Urspr�ngliche Nachricht-----
Von: Brendan Blewett [mailto:bb@git.com.au]
Gesendet: Montag, 29. Mai 2000 10:43
An: squid-users@ircache.net
Betreff: Squid stops responding to requests, Solaris 8 Intel & Stable
2.3

Hi All,

I have a couple of new Solaris 8 Intel servers which are starting to be a
bit of a headache. After about 12 hours - 5 days the squid cache stops
reponding to requests and requires a kill -9 and a manual restart. Kill -HUP
and -TERM both dont work at this point. The only thing that apears in the
access.log while this is happening is ICP_QUERY's. Telneting to the cache
port and feeding misc data does not return the error message as per normal
but instead it just hangs there with a open tcp connection.

They are all running different hardware combinations.
I have installed Squid 2.2.STABLE and Squid.2.3.STABLE2/3 plus the latest
patches and cant seem to get rid of the problem.

Also tried checking availablity of inodes and free space and changed the
filesystem to space optimisation as per the FAQ.

Has anyone else seen/fixed this problem or has pointers to help track down
the cause of this?

Regards,
Brendan
Received on Tue May 30 2000 - 14:34:59 MDT