IDRIS Jean-Zay cluster JEAN-ZAY/std

Description

This plugin deals with specific environment characteristics of the Jean-Zay supercomputer at IDRIS (France). Jean-Zay documentation is available at <http://www.idris.fr/eng/jean-zay/index.html>`__

When submiting jobs on GPU partitions the genereated batch job file header follows this template:

#!/usr/bin/bash

#SBATCH --job-name={job_name}
#SBATCH --output=%x.%j.out
#SBATCH --error=%x.%j.out

#SBATCH --account={project}@{allocation}
#SBATCH --C={constraint}         # Only if 'constraint' is used
#SBATCH --partition={partition}  # Only if 'partition' is used
#SBATCH --qos={qos}

#SBATCH --nodes={nodes}
#SBATCH --ntasks-per-node=1
#SBATCH --cpus-per-task={cores}
#SBATCH --gres=gpu:1
#SBATCH --hint=nomultithread

#SBATCH --time={walltime}

YAML arguments

The following arguments are used to configure the plugin. pyCIF will return an exception at the initialization if mandatory arguments are not specified, or if any argument does not fit accepted values or type:

Optional arguments

python : str, optional, default “python”

the python command used to run sub-instances of pyCIF

python_venv : str, optional

path to the python virtual environment to use

python_module : str, optional, default “python/3.11.5”

the python module to load

project : str, optional

project on which to submit sbatch jobs. --account=[project]@[allocation] in sbatch.

allocation : “cpu” or “v100” or “a100”, optional

hour allocation type to use in sbatch jobs. --account=[project]@[allocation] in sbatch.

partition : “cpu_p1” or “prepost” or “visu” or “archive” or “compil” or “compil_h100” or “gpu_p13” or “gpu_p2” or “gpu_p2s” or “gpu_p2l” or “gpu_p5” or “gpu_p6”, optional

partition on which to submit sbatch jobs. --partition=[partition] in sbatch.

constraint : “v100-16g” or “v100-32g” or “a100” or “h100”, optional

constraint to use in sbatch jobs. --C=[constraint] in sbatch.

nodes : int, optional, default 1

number of nodes allocated in sbatch jobs. --nodes=[nodes] in sbatch.

cores : int, optional, default 1

number of cores per node allocated in sbatch jobs. --cpus-per-task=[cores] in sbatch.

qos : “qos_cpu-t3” or “qos_cpu-t4” or “qos_cpu-dev” or “qos_gpu-t3” or “qos_gpu-t4” or “qos_gpu-dev” or “qos_gpu_a100-t3” or “qos_gpu_a100-dev” or “qos_gpu_h100-t3” or “qos_gpu_h100-t4” or “qos_gpu_h100-dev”, optional

Quality of Service (QoS) used in sbatch jobs. --qos=[qos] in sbatch.

walltime : str, optional, default “01:00:00”

maximum walltime of sbatch jobs, in the format ‘hh:mm:ss’. --time=[walltime] in sbatch.

submit_sbatch : bool, optional, default True

Submit the job with sbatch. If false simply run it within the same instance

Requirements

The current plugin requires the present plugins to run properly:

Requirement name	Requirement type	Explicit definition	Any valid	Default name	Default version
model	Model	True	True	None	None

YAML template

Please find below a template for a YAML configuration:

platform:
  plugin:
    name: JEAN-ZAY
    version: std
    type: platform

  # Optional arguments
  python: XXXXX  # str
  python_venv: XXXXX  # str
  python_module: XXXXX  # str
  project: XXXXX  # str
  allocation: XXXXX  # cpu|v100|a100
  partition: XXXXX  # cpu_p1|prepost|visu|archive|compil|compil_h100|gpu_p13|gpu_p2|gpu_p2s|gpu_p2l|gpu_p5|gpu_p6
  constraint: XXXXX  # v100-16g|v100-32g|a100|h100
  nodes: XXXXX  # int
  cores: XXXXX  # int
  qos: XXXXX  # qos_cpu-t3|qos_cpu-t4|qos_cpu-dev|qos_gpu-t3|qos_gpu-t4|qos_gpu-dev|qos_gpu_a100-t3|qos_gpu_a100-dev|qos_gpu_h100-t3|qos_gpu_h100-t4|qos_gpu_h100-dev
  walltime: XXXXX  # str
  submit_sbatch: XXXXX  # bool