Lecture 2 — HPC Cluster Usage & Slurm (3h)

Objectives

• Understand batch scheduling and Slurm concepts

• Write and submit Slurm job scripts; use sinfo, sbatch, squeue, scancel

• Practice interactive jobs and job arrays

Content summary

Simulate cluster usage (Docker-based environment). Explain resource requests (CPUs, memory, GPUs), job arrays and dependencies. Hands-on: write a Slurm job script to run a sample program and monitor it through the queue.