Benchmarks: AMD64 vs i386 on Dual 246 Opteron

[Ken Gunderson]

On Fri, 29 Jul 2005 09:24:14 +0200
Pav Lucistnik <pav at FreeBSD.org> wrote:

> Ken Gunderson píše v čt 28. 07. 2005 v 23:16 -0600:
> 
> > Document Path:          /
> > Document Length:        2846 bytes
> 
> > Document Path:          /
> > Document Length:        7951 bytes
> 
> Is it just my impression that you served two completely different
> websites in the test? Doesn't make it much trusty.

Not completely different.  Same ROR app w/Identicle configuration on
both machines.  Same db backend.  As I pointed out initially, the size
of the requests on the i386 machine were larger (because that site
actually has a bit of content), wh/ I doubt is responsible for the
legacy hardware serving up approx. double the requests/sec. as the dual
opteron.

The ROR app is Typo.  i386 front page <http://blog.truebrewin.com>
(My 70 yr. father's 1st blog, please be merciful).  The amd64 was just
default front door w/ o any content.

This wasn't intended as a scientific study.  Like I said, it was just
something I happened to be doing in another context and posted here
since it fit the general profile of 2x that the op was reporting.

I'm sure BOTH machines could do better but I was curoius about a +/- oob
scenario.

-- 
Best regards,

Ken Gunderson

Q: Because it reverses the logical flow of conversation.
A: Why is putting a reply at the top of the message frowned upon?