This interactive workshop focuses on scheduling your first HPC jobs using Bash and is designed to help researchers understand their Bash environment and how it ties into running work on a cluster. It walks through the structure of a SLURM job script, explains key parameters, resources, time limits, partitions or queues, output handling, and shows how those settings impact job behavior.
The session also covers common troubleshooting and debugging techniques, including reading output and error logs, checking job status and resource usage, and gathering the right details to streamline support and submit a well-prepared ticket when issues arise.
This session is part of the CUIT Research Services HPC Training Series, designed to onboard researchers to the shared high-performance computing (HPC) resources at Columbia. Led by HPC engineers, these webinars will introduce participants to Columbia’s shared on-premise HPC cluster (Insomnia) and provide foundational skills for effectively leveraging its resources. Whether you're new to HPC or looking to refine your expertise, this series will guide you through essential tools and workflows for running computational research efficiently.
You may take one or all of the courses, depending on your needs, though we recommend attending all sessions for the best experience.
