[Paraview] Running ParaView on the back end of the cluster

[gklontza]

Hi all,

I have built paraview from source and run it in a small Gemini cluster 
with 2 nodes. Each node has 32 AMD CPUs and 2 NVIDIA Tesla M2090 GPGPUs.
I'm trying to run the pvserver on the back end with the script file:

# Set the shell - bash shell
#!/bin/bash

# job name (default is name of pbs script file)
#PBS -N Server_Script

# Tell pbs to use XX nodes and YY processors per node
# Note: Cluster uses nodes=XX:ppn=YY but shared memory
# machine uses ncpus=XX to specify resources
#PBS -l nodes=2:ppn=32:gpus=4

#PBS -j oe

# resource limits: amount of memory and CPU time ([[h:]m:]s).
#PBS -l walltime=00:30:00

#PBS -o pview.out
#PBS -e pview.err
#PBS -V


cd $PBS_O_WORKDIR

module load openmpi-intel/1.4.4
module load intel-cc

mpiexec ./pvserver --use-offscreen-rendering --server-port=11111


I can submit it but it can't run. Any suggestion?

Cheers,
Georgia