job file for mpi application

job file for mpi application

by Latife Atar -
Number of replies: 3

I believe I implemented mpi in my cpp file correctly. 

When I submit the job with 1 node with the different configuration of ntasks and cpus per task I got this error: Running mpianalyzewave with 2 processors

--------------------------------------------------------------------------

There are not enough slots available in the system to satisfy the 2

slots that were requested by the application:


  ./mpianalyzewave


Either request fewer slots for your application, or make more slots

available for use.

I couldnt find how we organize our job file for the mpi application. 

Thank you 

Have a good friday

In reply to Latife Atar

Re: job file for mpi application

by Ramses van Zon -
Could you share all the #SBATCH lines in your jobscript? That's likely where the error is.
In reply to Ramses van Zon

Re: job file for mpi application

by Latife Atar -
Sure!
#!/bin/bash
#SBATCH --job-name=mpianalyzewave
#SBATCH --nodes=1
#SBATCH --cpus-per-task=16
#SBATCH --ntasks=1
#SBATCH --time=01:00:00
#SBATCH --output=mpianalyzewave%j.out
In reply to Latife Atar

Re: job file for mpi application

by Ramses van Zon -
Thanks. In SLURM, ntasks must be (at least) as big as the number of mpi processes. cpus-per-tasks would be the number of omp threads inside each process, so that should be 1 in this case.