Date: Tue, 24 Oct 2006 10:22:14 -0400 From: Mike Tancsa <mike@sentex.net> To: freebsd-performance@freebsd.org Subject: Spam-Assassin Benchmarks (was Re: LINUX vs FreeBSD mysql performance using a large RT database Message-ID: <7.0.1.0.0.20061024100908.0688fd68@sentex.net> In-Reply-To: <7.0.1.0.0.20061020161447.16083e08@sentex.net> References: <7.0.1.0.0.20061020104944.144d9068@sentex.net> <20061020200638.GA18727@sandvine.com> <7.0.1.0.0.20061020161447.16083e08@sentex.net>
next in thread | previous in thread | raw e-mail | index | archive | help
At 04:21 PM 10/20/2006, Mike Tancsa wrote:
>The next set of comparisons I want to run is in our spam
>scanners. The boxes which operate in round robin make heavy use of mysql, DNS
OK, we are just getting ready to run some tests for this
setup. SpamAssassin has some built in benchmarking that we will
use. e.g. scan a set 500 spam messages, scan 5 sets using 5
different processes at the same time, scan 10 etc and measure time taken.
The applications make heavy use of perl and mysql as well as netio
(numerous DNS lookups)
For FreeBSD tweaking, does anyone have any suggestions or requests
what we should tweak or test to see how things perform ?
Anecdotal evidence shows us that AMD64, consistently runs at a higher
load avg than the others, despite having an equal mix of spam thrown
at it. One thing I am not sure of however is if this is truly slower
(I suspect yes) or if there are differences in how load average is
calculated on i386 vs AMD64 ?
---Mike
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?7.0.1.0.0.20061024100908.0688fd68>