Esegui un MPI lavoro multinodo con Slurm in AWS PCS

Diventa l'utente predefinito. Passa alla sua home directory.

sudo su - ec2-user
cd ~/

Crea codice sorgente nel linguaggio di programmazione C.

cat > hello.c << EOF
// * mpi-hello-world - https://www.mpitutorial.com
// Released under MIT License
// 
// Copyright (c) 2014 MPI Tutorial.
//
// Permission is hereby granted, free of charge, to any person obtaining a copy
// of this software and associated documentation files (the "Software"), to 
// deal in the Software without restriction, including without limitation the 
// rights to use, copy, modify, merge, publish, distribute, sublicense, and/or 
// sell copies of the Software, and to permit persons to whom the Software is 
// furnished to do so, subject to the following conditions:
// The above copyright notice and this permission notice shall be included in 
// all copies or substantial portions of the Software.
//
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 
// IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 
// FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
// AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 
// LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING 
// FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER 
// DEALINGS IN THE SOFTWARE.

#include <mpi.h>
#include <stdio.h>
#include <stddef.h>

int main(int argc, char** argv) {
  // Initialize the MPI environment. The two arguments to MPI Init are not
  // currently used by MPI implementations, but are there in case future
  // implementations might need the arguments.
  MPI_Init(NULL, NULL);

  // Get the number of processes
  int world_size;
  MPI_Comm_size(MPI_COMM_WORLD, &world_size);

  // Get the rank of the process
  int world_rank;
  MPI_Comm_rank(MPI_COMM_WORLD, &world_rank);

  // Get the name of the processor
  char processor_name[MPI_MAX_PROCESSOR_NAME];
  int name_len;
  MPI_Get_processor_name(processor_name, &name_len);

  // Print off a hello world message
  printf("Hello world from processor %s, rank %d out of %d processors\n",
         processor_name, world_rank, world_size);

  // Finalize the MPI environment. No more MPI calls can be made after this
  MPI_Finalize();
}
EOF

Caricate il MPI modulo Open.

module load openmpi

Compila il programma C.

mpicc -o hello hello.c

Scrivi uno script di invio di lavori Slurm.

cat > hello.sh << EOF
#!/bin/bash
#SBATCH -J multi
#SBATCH -o multi.out
#SBATCH -e multi.err
#SBATCH --exclusive
#SBATCH --nodes=4
#SBATCH --ntasks-per-node=1

srun $HOME/hello
EOF

Passa alla directory condivisa.

cd /shared

Invia lo script del lavoro.

sbatch -p demo ~/hello.sh

Utilizzatelo squeue per monitorare il lavoro fino al termine.

Controlla il contenuto dimulti.out:

cat multi.out

L'output è simile a quello riportato di seguito. Nota che ogni rank ha il proprio indirizzo IP perché è stato eseguito su un nodo diverso.

Hello world from processor ip-10-3-133-204, rank 0 out of 4 processors
Hello world from processor ip-10-3-128-219, rank 2 out of 4 processors
Hello world from processor ip-10-3-141-26, rank 3 out of 4 processors
Hello world from processor ip-10-3-143-52, rank 1 out of 4 processor