Some errors occurred with running on HPC

Dear lammps users:

I got an error when running lammps on HPC, I use lammps/2021
and slurm workload manager to submit the work.

Here is the command in slurm file:

'#!/bin/bash

Request the resources required

80 cores across 2 nodes for 1 hour

#SBATCH --job-name=HOPE # Job name
#SBATCH --partition=batch
#SBATCH --ntasks=80 # Number of processor cores (i.e. tasks)
#SBATCH --nodes=2 # Number of nodes requested
#SBATCH --ntasks-per-node=40 # Tasks per node
#SBATCH --output=%j.log
#SBATCH --time=50:00:00 # walltime
#SBATCH --mail-type=ALL # send email
#SBATCH --mail-user= # email address
#load the required modules
module load lammps/2021
ulimit -s unlimited
ulimit -l unlimited
mpiexec -np $NSLOTS lmp -in ppp.lmp’Preformatted text

and here is the errors:

Loading compiler version 2021.2.0
Loading mpi version 2021.2.0
[[email protected]] i_np_fn (…/…/…/…/…/src/pm/i_hydra/mpiexec/intel/i_mpiexec_params.h:902): process count should be > 0
[[email protected]] match_arg (…/…/…/…/…/src/pm/i_hydra/libhydra/arg/hydra_arg.c:83): match handler returned error
[[email protected]] HYD_arg_parse_array (…/…/…/…/…/src/pm/i_hydra/libhydra/arg/hydra_arg.c:128): argument matching returned error
[[email protected]] mpiexec_get_parameters (…/…/…/…/…/src/pm/i_hydra/mpiexec/mpiexec_params.c:1356): error parsing input array
[[email protected]] main (…/…/…/…/…/src/pm/i_hydra/mpiexec/mpiexec.c:1749): error parsing parameters

What should I do to fix the problem?

This is not an error from LAMMPS but from the mpiexec command.

As a very general suggestion, the SLURM scheduler usually works best with the srun runner, which will almost certainly have been set up correctly by your cluster admins, instead of directly using MPI executables like mpirun or mpiexec. But follow up with your cluster admin as it is their setup which matters for your purpose.

Appreciate for your suggestion, I will go to contact with cluster admin.

Thanks for your help and reply.