You might just be running into the limits of what your Squid and system
will handle.
How many disks are you using for cache?
Are you using an async i/o compile (2.2.STABLE5+hno, of course)?
I think that's about the performance one can expect from a ~500MHz
machine running non-async with ~2 drives. Is your box going into Swap
memory at all? That will kill performance faster than anything I know.
With an async compile of 2.2.STABLE5+hno you can handle a T1 from a
single 7200 RPM IDE disk and 450MHz processor. But with a non-threaded
compile, you'll probably top out somewhere below a T1.
Just some thoughts.
Mike Richards wrote:
>
> Folks,
>
> At 2000-> 3000 req/min., my SQUID tanks!!!! Im running on Solaris 2.6
> with 2 x 250mz CPUs, 256mb memory, lots of swap, hme 10/100 interface,
> my T1 to the Internet is at >70% utilization...been tuning TCP
> parms....can't find the problem.......
>
> thanx for any suggestions,
> Mike
--
Joe Cooper <joe@swelltech.com>
Affordable Web Caching Proxy Appliances
http://www.swelltech.com
-- To unsubscribe, see http://www.squid-cache.org/mailing-lists.htmlReceived on Mon Sep 25 2000 - 18:46:42 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:55:29 MST