Re: troubles with Squid

Hello,

Quoting Tomasz Papszun (papszun@gryzmak.lodz.pdi.net):
> On Fri, 24 Apr 1998, Tural KAPTAN wrote:
> > On Fri, 24 Apr 1998, Ingrid Smud wrote:
> > > Connection information for squid:
> > > Number of TCP connections: 247207
> > ^^^^^^
> > > Page faults with physical i/o: 3850667
> > ^^^^^^^
> > Your cache is paging too much .. Try to decrease this ratio..
> > Try : http://squid.nlanr.net/FAQ/FAQ-10.html#ss10.18
>
> I know this FAQ item and was quite surprised because my squid
> shows similar "strange" big ratio. After some reading I found
> that Solaris' behaviour is quite different than other OSs and
> such situation is quite normal. I stress that my squid's machine
> is quite workless.

Solaris will do _all_ the disk IO via pagein/pageout. So these number
is by no means a sign that your squid binary get paged to disk very
frequently.

>
> Can anybody confirm such ratio is normal for Solaris?

Yup.

>
> -- Tomasz Papszun, Lodz, Poland tomek@lodz.tpsa.pl

Some hints for Solaris & Squid:

(very worth reading!!)
http://www.rvs.uni-hannover.de/people/voeckler/tune/EN/tune.html

Check out the ndd values for /dev/tcp (ndd /dev/tcp \?).

I tuned the tcp_conn_req_* queue values and the tcp_close_wait_interval
(to get rid of the huge number of sockets in FIN_WAIT2 state).

Also check /etc/system. ncsize & set ufs_inode could help alot on
non stripe disks.

Solaris 2.5.1 with all suggested patches works like a charme.

Cheers,
Dirk

-- 
Dirk.Vleugels@de.uu.net              http://www.de.uu.net
Computing & Proxy-Services           UUnet Deutschland GmbH
Tel. +49 231 972 00                  Emil-Figge-Strasse 80
Fax. +49 231 972 1180                44227 Dortmund, Germany