| 1 |
Brady Catherman wrote: |
| 2 |
> I just converted out 134 node cluster from RedHat Enterprise Linux v4 to |
| 3 |
> Gentoo 2005.1 |
| 4 |
> |
| 5 |
> Before and after the conversion we ran UnixBench as part of the Beowulf |
| 6 |
> Performance Suite. |
| 7 |
> |
| 8 |
> There is a graphical representation of the results here: |
| 9 |
> http://oceanus.ibest.uidaho.edu/~bcatherm/benchmark.png |
| 10 |
> |
| 11 |
> That shows Gentoo and Redhat on x86 hardware, and Gentoo and Macos on an |
| 12 |
> dual XServe G5. |
| 13 |
|
| 14 |
The graph is a bit confusing because you essentially need to compare the |
| 15 |
two inner bars, and the two outer bars. It might be clearer with the 2 |
| 16 |
x86 setups adjacent, and the 2 ppc setups adjacent. It's also a bit |
| 17 |
unclear to me whether high values are universally good, or low values, |
| 18 |
or whether it varies from test to test. From what I could find, it |
| 19 |
looked like high was good for file copy and low for the rest, but maybe |
| 20 |
they've been modified so high is always good. |
| 21 |
|
| 22 |
Could you provide some more detailed hardware specs and `emerge info`? |
| 23 |
I'd really like to put this data up somewhere on |
| 24 |
gentoo.org/proj/en/cluster/ -- with your name etc withheld if necessary. |
| 25 |
|
| 26 |
It would be interesting to do some performance analysis to see where any |
| 27 |
slowdowns come from and try to get things up to speed. |
| 28 |
|
| 29 |
Thanks, |
| 30 |
Donnie |
Archives
