[squid-users] Strange High CPU usage

The problem is the following:
We have approximately 700Reqs/sec distributed on

4 x Dell 2650 (sibling with digest) after a load balancer
each one with
2 x Xeon 1.8Ghz (HT disabled)
2GB Ram
PERC3/Di RAID controller
5 x 36GB HD SCSI 10k (2 RAID1 for os and swap, 3 for cache)
3 NIC, 1 internet, 1 intranet, 1 proxy lan (digest and proxy comunication)

52Mbit link

Slackware 8.1
Squid 2.5.STABLE4

some line of fstab
/dev/sdb1 /var/cache/spool/0 reiserfs noatime,notail 1 2
/dev/sdc1 /var/cache/spool/1 reiserfs noatime,notail 1 2
/dev/sdd1 /var/cache/spool/2 reiserfs noatime,notail 1 2


configure options:
--prefix=/usr --exec_prefix=/usr --bindir=/usr/sbin
--libexecdir=/usr/lib/squid --sysconfdir=/etc/squid --localstatedir=/var
--with-aufs-threads=48 --with-pthreads --with-aio --enable-async-io
--enable-storeio=diskd,aufs --disable-wccp
--enable-default-err-language=English --enable-err-languages=English
--disable-ident-lookups --enable-underscores
--enable-removal-policies=heap,lru --enable-snmp --enable-cache-digests
--enable-gnuregex

and actually configuration:

cache_mem 64 MB
cache_swap_low 85
cache_swap_high 90
maximum_object_size 65536 KB
maximum_object_size_in_memory 24 KB
cache_replacement_policy heap LFUDA
memory_replacement_policy heap GDSF
cache_dir diskd /var/cache/spool/0 28000 96 256 Q1=72 Q2=64
cache_dir diskd /var/cache/spool/1 28000 96 256 Q1=72 Q2=64
cache_dir diskd /var/cache/spool/2 28000 96 256 Q1=72 Q2=64
memory_pools_limit 50 MB
cache_access_log /var/cache/log/access.log
cache_log /var/cache/log/cache.log
buffered_logs on

Squid sames to work fine, but cpu usage after ~60Req/Sec on each server
is always 100%
I have tried many changes to squid.conf with no results.

Last question
I have also to upgrade our os to RedHat AS3, any experience with it?

thanks
--

*Giulio Cervera*

EDS PA SpA
Via Atanasio Soldati 80
00155 Roma (Italy)
tel: +39 06 22739 270
fax: +39 06 22739 233
e-mail: giulio.cervera@edspa.it <mailto:giulio.cervera@edspa.it>