Re: Benchmark (Phoronix): FreeBSD 9.0-RC2 vs. Oracle Linux 6.1 Server

From: Alex Kuster <vertexSymphony_at_zoho.com>
Date: Thu, 05 Jan 2012 11:26:10 -0300
On 01/04/2012 20:52, Matthew Tippett wrote:
>
> As a service to the community or vendor that publishes the tuning
> guide, Michael is more than willing to redo a tuned vs untuned
> comparison.  To date, the communities have never taken us up on that
> offer.  In part, this affects Phoronix.com <http://Phoronix.com>'s
> perception in the public, but that is more of a result of a one sided
> discussion by a party external to a particular community (with a
> healthy touch of journalisticly pumped compare & contrast).  For the
> FreeBSD community, who else outside of the FreeBSD community actually
> runs public comparisons of FreeBSD against anything?

If you just want to benchmark defaults, please, use proper defaults and
don't do a *custom* ZFS setup (which was IMHO a pretty big and gross
mistake).
I would be interested in a benchmark that does significant tests with
same hardware and REAL default setup.

It would be awesome if those benchmarks were re-done that way, so we can
compare an out-of-the-box experience even if that's not the last word on
how a system will perform (as others said, no one uses a default setup
for their servers)
And, if anyone else suggests a tuned-system benchmark ... I'm ok with
that too as long it's done properly and with guidance of a person of the
community that can give proper advice.

I'm a regular reader of Phoronix because it's basically easier than
being subscribed to a couple of mailing lists, blogs and such just to
get latest information on developments.
And the site is fairly popular, so you hold a pretty big responsibility
in there.

I got really disappointed on how a bad benchmark could impact on the
reputation of FreeBSD.
Things that are not true that people will be repeating (not-so-long-ago
I was a moderator on a Linux forum and I saw that misinformation in
action, it's terrible !)

Sorry if the words sound strong ... but I'm glad that these benchmarks
can be re-done.

Thanks for reading !


Received on Thu Jan 05 2012 - 13:26:23 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:22 UTC