Slurm Cluster Deployment on Chameleon Using Infrastructure Manager (IM)

📌 NOTE: THIS PROJECT IS A TROVI-COMPATIBLE REPRODUCIBLE EXPERIMENT. RUN THIS NOTEBOOK INSIDE CHAMELEON'S JUPYTERLAB ENVIRONMENT FOR BEST RESULTS.

This repository provides a simple, fully reproducible experiment that demonstrates how to deploy a Slurm cluster on Chameleon using the Infrastructure Manager (IM), a tool that deploys virtual infrastructures.

✅ Reproduce this experiment via: deploy_slurm_cluster.ipynb

This experiment:

• Deploys a virtual Slurm cluster in OpenStack on Chameleon using the IM.
• Submits a basic job to the cluster.
• Captures the output of the Slurm job.
• Destroys the infrastructure to free up resources.

---

⚙️ Requirements

• A Chameleon account and access to the KVM@TACC site.

---

🚀 Steps in the Notebook

1. Install the IM client.
2. Configure access to the KVM@TACC OpenStack site.
3. Deploy infrastructure using a provided TOSCA recipe.
4. Run a Slurm job to demonstrate usage.
5. Destroy infrastructure to clean up the space.