On Fri, 4 Jul 1997 11:57:18 +0200 (MET DST), Morten Guldager Jensen
wrote:
>I'm running Squid om a IBM PowerPC thing, 266 MHz, 256MB ram, and about
>15G disk. AIX 4.1.5.
>
>The cache area is made of 6 2 GB fast and wide SCSI disks.
>
>But it seems that squid doesn't utilize them at their full bandwidth.
>
>Could I do something to enable some kind of parallelism against the
>disk drives?
>
>tcp-banger running with a input file consisting of HIT's only URL's,
>I get a max of 50 requests done per second. (running on several machines.)
>
>Everything are running on a 100Mbit/s switch.
It's possible that you don't have enough memory for the cache
size/request rate you are seeing. Squid can eat memory like crazy,
and if you've got *lots* of requests coming in then it eats even more.
You probably want to look at the cachemgr output (or at the OS itself)
to see whether the machine is paging. If so, then it's probably not
your disks that are the problem.
Ian
-- Ian Cooper, Technical Manager e-mail: ian@mirror-image.com Mirror Image Internet (UK) Ltd. tel: +44 (0)171 712 1515 1 Canada Square, Canary Wharf fax: +44 (0)171 712 1615 London E14 5DY mobile: +44 (0)966 285 145Received on Fri Jul 04 1997 - 07:28:26 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:35:40 MST