Re: ..:: Linux/BSD Scalability Benchmarking ::..

[Marc Espie <espie@nerim.net>]

On Sun, Oct 19, 2003 at 02:56:37AM -0600, Paul Arola wrote:
> Comments? Aside from the clear bias againest obsd.
> 
> http://members.shaw.ca/infonick/slashdot/scalability.htm	or

The form of the article is clearly biased against OpenBSD.
I'm not sure the actual contents are.

We still don't boot from high-disk areas.
I'm not surprised our disk/memory performance is not good,
especiall mmap. This is an area UBC is supposed to fix, and
UBC was not committed.

Various failures may be connected to using a 3.4-current version
instead of an older stable release. Maybe there is bias in the
specific choice of 3.4-current, freebsd-5.1, netbsd-stable, but
if there is bias in the facts, it looks quite subtle to me.