=============================
Compute
=============================






SLURM Usage Guide
======================

This document provides a clean, modular reference for basic SLURM usage, including writing a job script, submitting jobs, checking job status, and deleting jobs.


  **1. Basic SLURM Job Script**
  
  
  A minimal example job script (`job.sh`) using 10 CPUs:

  .. code-block:: bash

    #!/bin/bash
    #SBATCH --job-name=myjob
    #SBATCH --error=myjob_%j.err
    #SBATCH --time=01:00:00
    #SBATCH --partition=compute
    #SBATCH --ntasks=1
    #SBATCH --cpus-per-task=10

    # Load modules as required
    # module load <executable>
    
    # Run your program
    ./<executable> <file>


  **2. Submitting a Job**

  Submit a SLURM job using ``sbatch``:

  .. code-block:: bash

    sbatch job.sh

  SLURM will return a job ID, for example:

  .. code-block:: text

    Submitted batch job 123456


  **3. Checking Job Status**

  Use ``squeue`` to check the status of your job:

  .. code-block:: bash

    squeue -j <jobid>

  Example:

  .. code-block:: bash

    squeue -j 123456

  You can also see all your jobs:

  .. code-block:: bash

    squeue -u $USER


  **4. Deleting / Canceling a Job**

  To cancel a job that is queued or running:

  .. code-block:: bash

    scancel <jobid>

  Example:

  .. code-block:: bash

    scancel 123456


  **5. Additional Useful Commands**

  - View detailed job information:

  .. code-block:: bash

      sacct -j <jobid>

  - See SLURM partitions:

  .. code-block:: bash

      sinfo