Scheduling

Anunna uses the SLURM scheduler to share compute resources fairly. This section covers how to submit, monitor, and manage jobs.

In this section

• Scheduler Overview (Slurm) — how scheduling works
• Partitions / Queues — the available partitions
• Choosing a node (constraints) — targeting particular hardware
• Batch Jobs — submitting a batch script
• Interactive Jobs — an interactive shell on a compute node
• Array Jobs — running many similar jobs at once
• Monitoring Jobs — checking on your jobs
• Cancelling Jobs — stopping a job
• Reservations — reserving nodes
• Fair Use Policy — using shared resources considerately