Norwegian Research Infrastructure Services · NRIS
Courses/ HPC Fundamentals/ HPC Scheduling — Slurm Basics
◆ Beginner · self-paced

HPC Scheduling — Slurm Basics

Go from never having touched a cluster to confidently submitting, monitoring and cancelling batch jobs with Slurm — on Olivia, Betzy, Saga or Fram.

★★★★★ 4.9 (312 ratings) ◷ ~75 min ▦ 7 sections ◆ Certificate of completion ⟳ Updated Mar 2026
N
Created by Sigma2 / NRIS Training Team · Norwegian Research Infrastructure Services

What you'll learn

  • Explain what a scheduler does and why HPC systems need one
  • Write a Slurm batch script with the right #SBATCH directives
  • Submit jobs with sbatch and monitor them with squeue
  • Inspect output and error files to debug failed jobs
  • Request CPUs, memory, GPUs and walltime appropriately
  • Cancel jobs and start interactive sessions

Course content

7 sections · 13 lessons · 6 quizzes · 1 certificate · ~75 min total
0 Welcome & orientation · 1 lesson
Course objectives & prerequisites5 min
1 What is a scheduler? · 3 lessons
Why clusters need a scheduler (video)6 min
Slurm & the restaurant-queue analogy7 min
Quick check: scheduler concepts3 Q
2 Submitting your first job · 2 lessons
Writing a batch script & sbatch10 min
Lab: submit hello-world (Cloze)4 Q
3 Inspecting output & status · 2 lessons
Reading slurm-*.out & squeue8 min
Quick check: job states3 Q
4 Reusable SBATCH scripts · 2 lessons
#SBATCH directives & accounts9 min
Short answer: fix the broken script4 Q
5 Requesting the right resources · 2 lessons
CPUs, memory, GPUs & walltime9 min
Quick check: resource flags3 Q
6 Cancelling & interactive jobs · 2 lessons
scancel & salloc interactive sessions8 min
Quick check: managing jobs3 Q
7 Final assessment & certificate · 2 lessons
Final Assessment — pass mark 50%7 Q
Certificate of Completion — unlocks at 50%PDF

Your instructor

Sigma2 / NRIS Training Team

Norwegian Research Infrastructure Services

The team that operates Norway's national supercomputers and the NIRD storage infrastructure. Every course is written and kept current by the same engineers who run the systems — so what you learn here matches what you'll meet on the clusters.

24courses 4.8★avg rating 1,900+certificates 4national systems

Learner reviews

MK
Marit K.
PhD candidate · UiO
★★★★★

"Finally a Slurm intro that doesn't assume I already know what a node is. Submitted my first real job the same afternoon."

JL
Jonas L.
Research engineer · NTNU
★★★★★

"The Cloze labs make you actually type the commands instead of just reading. The certificate was a nice touch for our group's onboarding records."

AH
Aisha H.
Postdoc · UiB
★★★★☆

"Clear and to the point. Would love a follow-up specifically on GPU jobs on Olivia."