Sorry.. I should have clarified that.. That is not our 'official,
complete' results.. just data that was thrown on as benchmarks
completed.
The bars represent percentage comparison to the highest scoring
result. Aka: #1 is always 100%, the next is showing a percentage of
that. The reason we did it that way was because some of the results
where in the 76-78 range and some where in the several hundred
thousand range.. We couldn't make a decent chart of all the results
without normalizing them all to a percentage rate.
Anyways.. the raw numbers represent an index from a base system that
was setup when UnixBench was written. A score of 1 = just as fast as
that system. 2 = some linear value faster.. etc. That system was a
386 if I remember correctly so most scores are much much higher than
1 any more. Things like memory performance and stuff have come light
years.
So the short of it is that higher is always better, and this is only
a 'best as I personally' could do comparison.
After our conversion things went together much smoother and now
maintenance is fairly painless. I spend all of my time setting up our
Apple cluster (The Gentoo ppc64 performance just wasn't good enough
to wow the management into using it =)
One big advantage of Gentoo is the ease in which new programs/
libraries can be installed. I have written dozens of ebuilds for all
the programs we use here in order to simplify installation and
administration.
Now to install a new program I just have to build it on a node using:
echo "emerge -b <program>" | qsub
Then I install it on everything else using:
emerge -K <program> ; pdsh -a emerge -K <program>
(Though I am so lazy that I even scripted the whole thing so emerging
a program on my cluster involves:
mass_emerge <program>
=)
On Apr 11, 2006, at 2:07 PM, Donnie Berkholz wrote:
Brady Catherman wrote:
I just converted out 134 node cluster from RedHat Enterprise Linux
v4 to
Gentoo 2005.1
Before and after the conversion we ran UnixBench as part of the
Beowulf
Performance Suite.
There is a graphical representation of the results here:
http://oceanus.ibest.uidaho.edu/~bcatherm/benchmark.png
That shows Gentoo and Redhat on x86 hardware, and Gentoo and Macos
on an
dual XServe G5.
The graph is a bit confusing because you essentially need to
compare the
two inner bars, and the two outer bars. It might be clearer with the 2
x86 setups adjacent, and the 2 ppc setups adjacent. It's also a bit
unclear to me whether high values are universally good, or low values,
or whether it varies from test to test. From what I could find, it
looked like high was good for file copy and low for the rest, but
maybe
they've been modified so high is always good.
Could you provide some more detailed hardware specs and `emerge info`?
I'd really like to put this data up somewhere on
gentoo.org/proj/en/cluster/ -- with your name etc withheld if
necessary.
It would be interesting to do some performance analysis to see
where any
slowdowns come from and try to get things up to speed.
Thanks,
Donnie
--
gentoo-cluster@gentoo.org mailing list
