Hi,

Am 19.11.2011 um 10:54 schrieb mahbube rustaee:

> Hi,
> I run a job with command via qsub script:
> #!/bin/sh
> #$ -N Det2-200core
> # -o $HOME/mypvfs/Det2

above is a $ missing.


> #$ -cwd
> #$ -l h_vmem=500M,mem_free=10M
> #$ -j y
> #$ -pe mpifill 200
> #cd $HOME/mypvfs/Det2
> . ~/.intelbash
> . /var/mpi-selector/data/openmpi_intel_qlc-1.4.2.sh

What's the above script doing?


> mpirun -n $NSLOTS  a.out

So, it's Intel MPI? You started the daemons for Intel MPI beforehand:

http://arc.liv.ac.uk/SGE/howto/mpich2-integration/mpich2-integration.html


> and exit-status is zero , but the output is empty.
> and with following command in the script:
> mpirun -np $NSLOTS --hostfile $TMPDIR/machines a.out
> 
> job is running for a long time  and the output is:
> error: executing task of job 919 failed: execution daemon on host "amd-9-2" 
> didn't accept task

This looks like you specified machines which aren't in the list, or you 
oversubscribed a machine.

$ qconf -sp MPI2
...
job_is_first_task  FALSE

is set?

-- Reuti
_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users

Reply via email to