Hi Bill,
I have had this problem recently. Squid ran fine ... till I added an
extensive blockinglist based on regex. When I started Squid with that
filtering list the CPU usage went up && up && up.
Extansive ACLs can also be the cause. It is recommended that you compile
squid with the btree option when you are going to have such, although nobody
ever could tell me how to make use of the betree / how to set this option.
--Joerg
am 13.08.2001 17:50 Uhr schrieb Schwartz Jr., William H. unter
SchwaWH@nsc-msg01.network.com:
> My squid process recently started using all available cpu cycles and I cna't
> figure out why. Looking at the FAQ I found:
>
> -----------------------------
> 11.46 Squid uses 100% CPU
> There may be many causes for this.
> Andrew Doroshenko reports that removing /dev/null, or mounting a filesystem
> with the nodev option, can cause Squid to use 100% of CPU. His suggested
> solution is to ``touch /dev/null.''
> ----------------------------------
>
> I checked and I do have /dev/null and I didn't see anything in /etc/vfstab
> setting the nodev option for any of the file systems.
>
> Here's some more information about the system.
>
>
> last pid: 1175; load averages: 1.00, 1.00, 1.00
> 10:47:40
> 45 processes: 43 sleeping, 1 running, 1 on cpu
> CPU states: 0.0% idle, 96.4% user, 3.4% kernel, 0.2% iowait, 0.0% swap
> Memory: 512M real, 184M swap in use, 702M swap free
>
> PID USERNAME THR PRI NICE SIZE RES STATE TIME CPU COMMAND
> 238 root 1 0 0 156M 155M run 294.5H 98.54% squid
>
>
> $ uname -a
> SunOS proxy 5.8 Generic_108528-06 sun4u sparc SUNW,UltraSPARC-IIi-Engine
>
>
> $ /usr/local/squid/bin/squid -v
> Squid Cache: Version 2.3.STABLE3
>
> I've used "tunefs" to set the cache file system to "space" optimization
> instead of "time". Before that I was running into disk issues.
>
>
>
> thanks,
> Bill
>
Received on Mon Aug 13 2001 - 10:54:33 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:01:36 MST