CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   SU2 (https://www.cfd-online.com/Forums/su2/)
-   -   Parallelizing SU2 in PBS (https://www.cfd-online.com/Forums/su2/219490-parallelizing-su2-pbs.html)

NKever July 29, 2019 09:06

Parallelizing SU2 in PBS
 
Hi,

I am trying to run the shape_optimization.py script and use all my available nodes in PBS and I don't think I am doing this right.

Here is my script:

Code:

#PBS -q devel

##PBS -q long

##PBS -q normal

#PBS -W group_list=a1489

#PBS -lselect=50:ncpus=28:model=bro,walltime=02:00:00

##PBS -lselect=80:ncpus=28:model=bro,walltime=08:00:00

##PBS -lselect=80:ncpus=28:model=bro,walltime=50:00:00

#PBS -N test_su2

#

# Request that regular output and terminal output go to the same file

#PBS -j oe

#PBS -m n

# go to where the qsub was performed.

cd $PBS_O_WORKDIR

# Running Program

export SU2_RUN="/home6/nkever/su2/install/SU2/bin"

export SU2_HOME="/home6/nkever/su2/src/SU2"

export PATH=$PATH:$SU2_RUN

export PYTHONPATH=$PYTHONPATH:$SU2_RUN

module load mpi-sgi/mpt

module load python/2.7.15

module load comp-intel/2018.0.128

./shape_optimization.py -g CONTINUOUS_ADJOINT -o SLSQP -f inv_NACA0012_basic.cfg -n 400

echo "finished mpirun"


So I think the issue may be the -n, my interpretation is that the number of partitions tells su2 the amount of nodes. How do I use all my nodes?

I've seen this thread already, but I'm not sure it addresses my question. Thank you!

jayantmukho August 8, 2019 15:39

Hey Nadeem,

Is your issue that PBS doesn't use the different nodes and just tries to start all 400 processes on 1 node?

I had this particular issue when I was trying to run using openmpi with PBS. In that case, the openmpi installation that I was using wasn't configured to work with torque/pbs (https://www.open-mpi.org/faq/?category=tm). It required a particular option to be set when configuring openmpi.

I haven't used SGI MPT before so I am not certain if there is a similar issue that would need to be rectified.

pcg August 25, 2019 04:42

Hi Nadeem,

The problem might be that all nodes are executing the shape optimization script, and only one should since it is not an mpi-ready program (I think).

The simplest way to do this kind of thing would be to wrap the SU2 execution in a script that submits an individual job per SU2 run and waits for results.

Cheers,
Pedro


All times are GMT -4. The time now is 11:16.