Start Here

In a large, distributed environment such as an HPC cluster, a scheduler is required to request compute resources and dispatch user processes (called "jobs") on them. There are several available for HPC systems but the most common one is called Simple Linux Utility for Resource Management or "Slurm" for short [Wikipedia]. The Hyak clusters use Slurm so any general documentation you can find on the internet is valid, including directly from the developer [SchedMD]. While SchedMD documentation is the most authoritative, this section has some common examples you might find useful with sample arguments specific to the klone cluster of Hyak.