Slurm Cloud Bursting: Watch Your HPC Jobs Scale Instantly into Azure (Lab 4)
The grand finale! Lab 4 validates the complete end-to-end hybrid workflow. We start by ensuring consistent user identity (UID/GID), then submit a large workload using sbatch that exceeds the local on-prem capacity. Watch in real-time as the Slurm scheduler hands off the overflow jobs to CycleCloud, which automatically provisions new cloud compute nodes in Azure. We'll monitor job progress using squeue, track VM creation in the Azure Portal, and finally, validate the automatic node teardown once the job finishes. Experience the true elasticity of Hybrid HPC!
#CloudBurstingDemo #SlurmJobSubmission #DynamicScaling #HPCWorkflow #AutoTeardown #squeue #Lab4