Re: [squid-users] Recent Polygraph results for Squid?

On Mon, Feb 15, 2010 at 11:35:47AM -0000, Oborn, Keith wrote:
>
> I'd be very interested in any numbers at all - most particularly on
> recent Sun kit, as it looks as if ZFS is a good bet for Squid. I must
> admit I alway hankered after testing a proxy/cache on a Thumper (X4540)
> because of the huge spindle density.
>

Our main squid proxies are a couple of sun fire v445's, peak http
requests are 12.5k/min, median service time < 100ms. We run our
caches on ZFS, two spindles only. We cut over from UFS because we
kept running into fragmentation problems on UFS - squid would claim
the disk was full but df would say not.

> At present, any numbers that get me within like a factor of two of
> actual performance on any modern X86 server hardware would be great (and
> perhaps any idea if using Sun T-series helps - does Squid like lots of
> threads?)

Squid itself is single threaded so it probably would not benefit from
a T-series itself but if you have a lot of
authenticators/redirectors/other ancillaries it may work well, or
maybe either setup multiple instances of squid or even resort to
slicing up the machine into a few LDOMs to present multiple smaller
machines in the same chassis - that may keep the squid config simpler.
Never tested squid on a T-series myself though.

-- 
Brett Lymn
"Warning:
The information contained in this email and any attached files is
confidential to BAE Systems Australia. If you are not the intended
recipient, any use, disclosure or copying of this email or any
attachments is expressly prohibited.  If you have received this email
in error, please notify us immediately. VIRUS: Every care has been
taken to ensure this email and its attachments are virus free,
however, any loss or damage incurred in using this email is not the
sender's responsibility.  It is your responsibility to ensure virus
checks are completed before installing any data sent in this email to
your computer."