Strumenti Utente

Strumenti Sito

Traccia:
calcoloscientifico:cluster:softwareapplicativo:fds

Differenze

Queste sono le differenze tra la revisione selezionata e la versione attuale della pagina.

Link a questa pagina di confronto

Entrambe le parti precedenti la revisione Revisione precedente
Prossima revisione 		Revisione precedente
calcoloscientifico:cluster:softwareapplicativo:fds [24/05/2020 16:13]
fabio.spataro 		calcoloscientifico:cluster:softwareapplicativo:fds [24/05/2020 16:38] (versione attuale)
fabio.spataro
Linea 9: Linea 9:
 === Job FSD OpenMP (single node) === === Job FSD OpenMP (single node) ===
  
-Example script ''slurm-fds-omp.sh'' to run FSD on 1 node using 16 cores and at most 4 GB of memory:+Example script ''slurm-fds-omp.sh'' to run FSD on 1 node of the ''skl'' partition using 16 cores and at most 4 GB of memory:
  
 <code bash slurm-fds-omp.sh> <code bash slurm-fds-omp.sh>
 #!/bin/bash #!/bin/bash
-#SBATCH --job-name=fds+#SBATCH --job-name=Modello_R04_SC1_R02
 #SBATCH --output=%x.o%j #SBATCH --output=%x.o%j
 #SBATCH --error=%x.e%j #SBATCH --error=%x.e%j
Linea 29: Linea 29:
  
 ulimit -s unlimited ulimit -s unlimited
-ulimit -s 
  
 WORKING_DIR="$(pwd)/${SLURM_JOB_NAME}.d${SLURM_JOB_ID}" WORKING_DIR="$(pwd)/${SLURM_JOB_NAME}.d${SLURM_JOB_ID}"
Linea 36: Linea 35:
 cd "$WORKING_DIR" cd "$WORKING_DIR"
  
-mpirun -np 1 fds "${SLURM_JOB_NAME}.fds"+mpirun fds "${SLURM_JOB_NAME}.fds"
 </code> </code>
  
-Submit:+=== Job FSD MPI OpenMP (single node) ===
  
-<code bash> +Example script ''slurm-fds-mpi-omp.sh'' to run FSD on 1 node of the ''skl'' partition using 4x4=16 cores and at most 1 GB of memory:
-sbatch --job.name=Modello_R04_SC1_R02 slurm-fds-omp.sh +
-</code> +
- +
-=== Job FSD MPI+OpenMP (single node) === +
- +
-Example script ''slurm-fds-mpi-omp.sh'' to run FSD on 1 node using 4x4=16 cores and at most 8 GB of memory:+
  
 <code bash slurm-fds-mpi-omp.sh> <code bash slurm-fds-mpi-omp.sh>
 #!/bin/bash #!/bin/bash
-#SBATCH --job-name=fds+#SBATCH --job-name=lapse_rate
 #SBATCH --output=%x.o%j #SBATCH --output=%x.o%j
 #SBATCH --error=%x.e%j #SBATCH --error=%x.e%j
Linea 58: Linea 51:
 #SBATCH --cpus-per-task=4 #SBATCH --cpus-per-task=4
 #SBATCH --partition=skl #SBATCH --partition=skl
-#SBATCH --mem=8G +#SBATCH --mem=1G 
-#SBATCH --time=0-01:00:00+#SBATCH --time=0-00:05:00
 ##SBATCH --account=<account> ##SBATCH --account=<account>
  
Linea 67: Linea 60:
  
 ulimit -s unlimited ulimit -s unlimited
-ulimit -s 
  
 WORKING_DIR="$(pwd)/${SLURM_JOB_NAME}.d${SLURM_JOB_ID}" WORKING_DIR="$(pwd)/${SLURM_JOB_NAME}.d${SLURM_JOB_ID}"
Linea 77: Linea 69:
 </code> </code>
  
-Submit: +<note> 
- +This example uses a modest amount of memory and lasts a few seconds. For a realistic calculation it is necessary to significantly increase the required memory and the expected calculation time. 
-<code bash> +</note>
-sbatch --job.name=lapse_rate slurm-fds-mpi-omp.sh +
-</code>+
  
 === SMV === === SMV ===
calcoloscientifico/cluster/softwareapplicativo/fds.1590329634.txt.gz · Ultima modifica: 24/05/2020 16:13 da fabio.spataro

Strumenti Pagina

Donate Powered by PHP Valid HTML5 Valid CSS Driven by DokuWiki