SLURM Compute Cluster
Overview
Slurm is an open source cluster management system and job scheduler.
CSAIL’s Slurm cluster comprises partitions of systems belonging to individual research groups, as well as a partition of systems available for general lab use.
If you would like to have your group’s jobs scheduled with Slurm, please inquire at help@csail.mit.edu.
Join the Slurm-Announce email list. Downtime and impactful changes will be announced here.
Getting started with Slurm
How to request resources for your job
Explanation of required parameters --partition and --qos
Where to store your files
Basic commands for using Slurm
Summary of Shared Resources
How priority is determined in the queue
Using SSH with Slurm
Frequently Asked Questions about SLURM
Regularly scheduled downtime
How to join your nodes to the CSAIL Slurm Cluster