Re: [squid-users] Squid 3.2.1 defunct

From: <shorewall_at_malargue.gov.ar>
Date: Fri, 24 Aug 2012 19:39:23 -0300

El 2012-08-17 12:37, Ricardo Rios - Shorewall List escribió:

> I am trying squid 3.2.1 in a opensuse 11.4 Box, with 4 workers and 2
> rock cache_dirs with 500 clients browsing, all looks perfect at
> start,
> but before 5 or 10 mins, i see the traffic going down and up, when i
> check top, and this happend, i see the squid process going defunct,
> after anothers 5 or 10 mins, happend again, and again, and again.
>
> ./configure CFLAGS=-DNUMTHREADS=128 --with-filedescriptors=16384
> --enable-removal-policies=heap,lru --enable-epoll
> --enable-stopreio=ufs,aufs,rock --enable-async-io=128 --with-pthreads
> --disable-dlmalloc --with-large-files --enable-large-cache-files
> --with-aio --enable-esi --with-dl --enable-ltdl-convenience
> --enable-linux-netfilter --disable-ident-lookups --enable-snmp
> --enable-htcp --disable-ipv6
>
> squid.conf
>
> http_port 3128
> http_port 3129 tproxy
>
> workers 4
> cache_dir rock /squid-cache 58000 max-size=31000 max-swap-rate=250
> swap-timeout=350
> cache_dir rock /var/squid-cache 19000 max-size=31000
> max-swap-rate=250
> swap-timeout=350
>
> refresh_pattern -i
> microsoft.com/.*.(cab|exe|ms[i|u|f]|asf|wm[v|a]|dat|zip) 4320 90%
> 43200
> refresh_pattern -i
> windowsupdate.com/.*.(cab|exe|ms[i|u|f]|asf|wm[v|a]|dat|zip) 4320 90%
> 43200
> refresh_pattern -i ([^.]+.|)avast.com/.*.(vpx|vpu|exe) 4320 90% 43200
> refresh_pattern -i ([^.]+.|)avg.com/.*.(bin|ctf) 4320 90% 43200
> refresh_pattern -i .(gif|png|jpg|jpeg|ico)$ 40320 75% 86400
> refresh_pattern -i .(iso|avi|wav|mp3|mpeg|swf|flv|x-flv)$ 1440 40%
> 40320
> refresh_pattern -i .mp4 1440 90% 43200
> refresh_pattern -i .(css|js)$ 300 40% 7200
> refresh_pattern -i .(html|htm)$ 300 40% 7200
> refresh_pattern . 300 40% 40320
>
> max_filedescriptors 16394
> cache_mem 1024 MB
> maximum_object_size 100 MB
> minimum_object_size 10 KB
> maximum_object_size_in_memory 500 KB
>
> CPU is a i5 2.80GHz 4 cores with 4GB DDR3
>
> Thanks for any advice and help.

I have some updates for this problem, i try just without workers, and
ausf
cache_dir, and all looks good, squid works fine, after 1 hour, i set
workers 2, restart squid, and after 5 mins, problem is there again,
squid
die, go up again, 2 mins late die again, etc. If i remember well, i
already test just no workers and rock cache_dir, and problem still
there, so i am guessing is something about SMP and Rock. I hope this
help.

http://s8.postimage.org/l09bltf45/squid_2.png
Received on Fri Aug 24 2012 - 22:39:30 MDT

This archive was generated by hypermail 2.2.0 : Sat Aug 25 2012 - 12:00:03 MDT