Skip to content

Slurm Usage

Slurm is used to submit jobs on the different partitions from the Monolithe frontend. The available partitions are listed in the description page (see the "Slurm Partition" column in the summary table).

Frontend Connection

It is recommended to add some lines to your ~/.ssh/config file as explained in the SSH access section. Then, to connect to the frontend from your computer you only have to do:

ssh front.mono.lip6
Or, if your are not on the LIP6 network:
ssh front.mono.lip6.ext

Basic Slurm Commands

Here are some useful command to start using Slurm:

  • sinfo -l lists the available partitions (= nodes in our case)

    $ sinfo -l
    Mon Mar 18 10:57:44 2024
    PARTITION AVAIL  TIMELIMIT   JOB_SIZE ROOT OVERSUBS     GROUPS  NODES       STATE NODELIST
    xu4          up   infinite 1-infinite   no       NO        all      1        idle vroum
    tx2          up   infinite 1-infinite   no       NO        all      1        idle tegrax2c
    xagx         up   infinite 1-infinite   no       NO        all      1        idle tegraagx
    brub         up   infinite 1-infinite   no       NO        all      1        idle brubeck
    xnano        up   infinite 1-infinite   no       NO        all      1        idle jetson-nano1
    rpi4         up   infinite 1-infinite   no       NO        all      1        idle selfix
    xnx          up   infinite 1-infinite   no       NO        all      1        idle tegranx-1
    m1u          up   infinite 1-infinite   no       NO        all      1        idle m1ultra
    onx          up   infinite 1-infinite   no       NO        all      1        idle orinnx
    oagx         up   infinite 1-infinite   no       NO        all      1       mixed orinagx
    onano        up   infinite 1-infinite   no       NO        all      1        idle orinnano
    opi5         up   infinite 1-infinite   no       NO        all      1        idle orangepi5
    

  • srun -p [partition] command runs a command on a partition

    $ srun -p m1u hostname
    m1u
    
    Submission of a job that execute the hostname command on the m1u partition.

  • srun -p [partition] --pty bash -i runs a interactive session on a partition

    srun -p opi5 --cpus-per-task=8 --pty bash -i
    
    Interactive job on the Orange Pi 5, all the 8 cores are used in this session (by default, if --cpus-per-task is not specified, only one core is allocated).

    Note

    An easier way to connect interactively to the nodes is to use a custom ~/.ssh/config file as detailed in the SSH Access page.

  • sbatch [script] runs a Slurm script on the cluster

  • squeue -l allows you to view current submitted jobs on the cluster

    $ squeue -l
    Mon Mar 18 10:58:20 2024
                 JOBID PARTITION     NAME     USER    STATE       TIME TIME_LIMI  NODES NODELIST(REASON)
                  1702      oagx     bash galveze-  RUNNING      47:10 UNLIMITED      1 orinagx
    
    For instance, here one job from the galveze user is running on the oagx partition.

  • scancel [jobid] cancels a job

  • scancel -u [user] cancels all the jobs for a given user