Run OpenFOAM codes on Curnagl
You are using OpenFOAM on your computer and you need more ressources. Let’s go on Curnagl!
OpenFOAM is usually using MPI. Here is a bash script to run your parallelized OpenFOAM code. NTASKS should be replaced by the number of processors you want to use into your OpenFOAM code.
#!/bin/bash -l
#SBATCH --job-name openfoam
#SBATCH --output openfoam.out
#SBATCH --partition cpu
#SBATCH --nodes 1
#SBATCH --ntasks NTASKS
#SBATCH --cpus-per-task 1
#SBATCH --mem 8G
#SBATCH --time 02:00:00
module purge
module load gcc/10.4.0 mvapich2/2.3.7 openfoam/2206
# RUN YOUR OpenFOAM CODE HERE
# For instance:
# decomposepar ...
# snappyHexMesh ...
Please note that running your parallelized OpenFOAM code should not be performed via mpirun
but srun
. For a complete MPI overview on Curnagl, please refer to compiling and running MPI codes wiki.
How do I transfer my OpenFOAM code to Curnagl ?
Once you are connected on Curnagl, you can upload your OpenFOAM code thanks to FileZilla (see FileZilla wiki) or copy and paste data to the cluster thanks to the `scp` command.
Example: I want to copy test.py to Curnagl
`scp test.py username@curnagl.dcsr.unil.ch:/YOUR_PATH_ON_CURNAGL`
Where YOUR_PATH_ON_CURNAGL is something like /users/username/work/.......
In these commands, do not forget to change `username` with yours.