[Ranjan]

Hello

I am trying to compile and run SU2 V. 7.0.2 on a cluster that uses qsub to submit jobs. While I was able to compile and run SU2 on the head node, Jon submission using qsub keeps failing. I am currently in the process of working this out with the cluster admin but I wanted to make sure that the qsub file is indeed correct :

#!/bin/bash
#PBS -l walltime=00:05:00
#PBS -l nodes=1pn=16
#PBS -N openmpi_job
#PBS -q secondary
#PBS -j oe
###PBS -o openmpijob.out
###PBS -e openmpijob.err
###PBS -m be
#
#####################################

# Load MPI module (Enable MPI in user environment)
module load openmpi/3.1.1-intel-18.0
module load vim/8.1
module load boost/1.71.0
module load git/2.19.0
module load python/3
module load anaconda/2019-Oct/3
module unload python/2
module load /projects/eng/shared/kaijames/modulefiles/swig/2.0.10
module load gcc
# Change to the directory from which the batch job was submitted
export PBS_JOBDIR=/projects/eng/shared/kaijames/SU2/test/Turbulent_2D_Constrained_RAE2822/
# Run the MPI code
shape_optimization.py -n 16 -g DISCRETE_ADJOINT -o SLSQP -f turb_SA_RAE2822.cfg > log.solver &

I understand that I need to make changes to shape_optimization.py to make it compatible with qsub but I am not sure how to do so.

Any help is highly appreciated.

Thank you

[bigfootedrockmidget]

what does 'failing' mean? Can you share the output from the log files?
Does a primal computation using mpirun work?

[Ranjan]

Hello

Thank you for your reply. So it seems like my there was an error in my .pbs file which would result in an exception after submission.

I am able to run direct flow simulations using qsub using mpiexec SU2_CFD <case.cfg>. However, I am not sure how to use the python wrapper for shape optimization problems.

To be exact, I am not sure how to execute shape_optiization.py -n -$j ... in the .pbs file when the number of cores are specified earlier in the file.

My qsub file is attached herewith. Any help will be greatly appreciated!

Thank you in advance.