[stater]

Hi everybody,

I set up a small computing cluster to run calculations on OpenFoam, but I can't run a job on multiple nodes. I am using the following sbatch file:

#!/bin/sh
#SBATCH --partition=vip
#SBATCH -N 2
#SBATCH --ntasks=2
#SBATCH --ntasks-per-node=1
##SBATCH --time=5-00:00:00
#SBATCH --output=log-%j.out
#SBATCH --error=log-%j.err
#SBATCH --job-name=model
#SBATCH --mail-type=ALL

mpirun --mca pml ob1 --mca btl tcp,self --mca btl_tcp_if_include 192.168.0.0/24 chtMultiRegionFoam -parallel > log.run

I can run this job in parallel on a single node with two cores, but when I run it on two different nodes, nothing happens in the "processor" directories. I checked on the nodes on the nodes the processes actually launch, but I have no results on the sharing directory.


I am using OpenFoam 22.12 with openmpi 4.1.2 and slurm 21.08.5.
My network is Ethernet 1Gb/s. I checked the NFS server and write permissions, everything is OK

Could anyone help me!!!

Thank you!!!