[torqueusers] problem

Gus Correa gus at ldeo.columbia.edu
Fri Aug 10 09:38:12 MDT 2012


On 08/04/2012 05:56 AM, Asd Asdi wrote:
> hello All
>
> I have one cluster that it has 4 nodes. Each nodes has 24 core but 
> when I send jobs with qsub to nodes
>
> only 18 core each nodes start to work and 6 core of each nodes is not 
> used. I do not know how to solve this problem.
>
> I was wondering if you can help me.
>
> sincerely
>
> Ahmad Bayat
>
>
> _______________________________________________
> torqueusers mailing list
> torqueusers at supercluster.org
> http://www.supercluster.org/mailman/listinfo/torqueusers
Hi Asd

Does the $TORQUE/server_priv/nodes file on the Torque server should have 
lines like this [with  your
node's names, of course]?

node01 np=24 <maybe some more parameters>
...

Does your Torque/PBS script request 24 cores/processors per node?
Something like this:

#PBS -l nodes=4:ppn=24
...
mpiexec -np 96 <other mpiexec parameters> ./my_mpi_program

I think somebody already asked you to send your 
$TORQUE/server_priv/nodes file
and the output of
qmgr -c 'p s'
A snippet of your job submission script maybe too.
These things may help to understand what is the problem.

Gus Correa




More information about the torqueusers mailing list