HON95-wiki/config/hpc/slurm.md

---

title: Slurm Workload Manager breadcrumbs:

• title: Configuration
• title: IoT --- {% include header.md %}

Usage

Basics

• Cluster information:
  • Show partitions: scontrol show partition [-a]
  • Show partition/node usage: sinfo [-a]
  • Show node capabilities: sinfo -o "%20N %8c %10m %25f %10G" (example)
• Accounting:
  • Show accounts for user: sacctmgr show assoc where user=<username> format=account
  • Show default account for user: sacctmgr show user <username> format=defaultaccount
• Job and job queue information:
  • Show job queue: squeue [-u <user>] [-t <state>] [-p <partition>]
  • Show job details: scontrol show jobid -dd <jobid>
• Job handling:
  • Create a job (overview): Make a Slurm script, make it executable and submit it.
  • Using GPUs: See example Slurm-file, using --gres=gpu[:<type>]:<n>.
  • Submit batch/non-blocking job: sbatch <slurm-file>
  • Start interactive/blocking job: srun <job options> [--pty] <bash|app>
  • Cancel specific job: scancel <jobid>
  • Cancel set of jobs: scancel [-t <state>] [-u <user>]

Example Slurm-File

#!/bin/sh

#SBATCH --partition=<partition>
#SBATCH --time=03:00:00
#SBATCH --nodes=2
# #SBATCH --nodelist=compute-2-0-[17-18],compute-5-0-[20-21]
#SBATCH --ntasks-per-node=2
# #SBATCH --exclusive
# #SBATCH --mem=64G
#SBATCH --gres=gpu:V100:2
#SBATCH --job-name="xxx"
#SBATCH --output=log.txt
## SBATCH --mail-user=user@example.net
# #SBATCH --mail-type=ALL

# Run some program on all processors using mpirun
mpirun uname -a

{% include footer.md %}