Slurm.md

Slurm

We use SLURM (https://slurm.schedmd.com/overview.html ) as a workload manager to schedule jobs onto compute resources. Via SLURM we can ensure that each user gets a fair share of the limited compute resources and that multiple users do not interfere with each other when e.g. running benchmarks.

Important: You can only access a node via SSH when you have a SLURM allocation of that node.

Other resources:

• Slurm Tutorial

Basics

IMGW special commands

There are currently a few extra commands that can be used on the Jet Cluster to facilitate usage of the nodes.

Tools:

• jobinfo
• jobinfo_remaining
• nodeinfo
• queueinfo
• watchjob