[Rocks-Discuss] How to test my Rocks Cluster
Pascal Weißhaupt
i have installed a Rocks Cluster frontend with 2 nodes. Now i want to test
my clusters cpu performance.
Is there a how-to?
Comfreak
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://lists.sdsc.edu/pipermail/npaci-rocks-discussion/attachments/20101223/5adf89b0/attachment.html
Gus Correa
The very basic functionality test used to be
(I'm not sure if it still is):
tentakel -g compute hostname
where all compute nodes should answer by printing their names.
And perhaps:
tentakel -g frontend hostname
to see the frontend answer.
"man tentakel" may help you out.
**
For more serious wrangling with MPI, assuming your
thing is number crunching and parallel applications:
0) Install the Torque roll (my favorite) OR the SGE roll (official),
but *not both*. Having one of these two
resource managers / queue systems
is peace of mind when it comes to launch and manage parallel jobs.
**
1) If you have Ethernet only, download and install MPICH2 and/or OpenMPI
in /share/apps/mpich2 or /share/apps/openmpi
Use the --prefix= option
when you configure each of them to choose the installation directory,
otherwise it goes to /usr/local, which is not the best choice.
http://www.open-mpi.org/
http://www.mcs.anl.gov/research/projects/mpich2/
MPICH2 has Install and User Guides, OpenMPI has a decent README file,
and lots of help in their FAQ web page.
**
2) If you have Infinband instead, choose among OpenMPI and MVAPICH2:
http://mvapich.cse.ohio-state.edu/index.shtml
MVAPICH also has a User/Install guide.
**
3) Here's a good tutorial to MPI from Lawrence Livermore Natl. Lab:
https://computing.llnl.gov/tutorials/mpi/
**
4) If you want to go beyond, solve PDEs, etc, and want a
toolbox, try PETSc:
http://www.mcs.anl.gov/petsc/petsc-as/
**
I hope this helps,
Gus Correa
Gus Correa
PS -
5) To compare your cluster to the Top500 list, and check CPU performance
(well, the network impact is not negligible at all), try the HPL
benchmark (a.k.a. Linpack benchmark),
which you can compile with Goto BLAS:
http://www.tacc.utexas.edu/tacc-projects/gotoblas2/
http://www.netlib.org/benchmark/hpl/
**
6) Other benchmarks:
Intel:
http://software.intel.com/en-us/articles/intel-mpi-benchmarks/
NAS (from NASA):
http://www.nas.nasa.gov/Resources/Software/npb.html
**
Gus Correa