Slurm is a workload manager that manages the various competing demands on computing resources and allocates them to let everyone run their jobs optimally.
This section explains how to use it; a more exhaustive list and description of its capabilities can be found in its official online documentation. Detailed descriptions of its commands can also be found in its man pages, accessed by typing man <command_name>
slurm can be run from any terminal session, from any location in the cluster, either interactively or by invoking a batch script. Sometimes it's easier to use a detachable terminal session but it isn't necessary.
⚠️ All work must be submitted through slurm ⚠️ NEVER run jobs on the login node ⚠️