Slurm Cluster Management

Help developers submit, manage, and troubleshoot GPU-accelerated workloads on SRP's Slurm clusters. Supports training, inference, and data processing jobs using Apptainer containers.

When to Use This Skill

Use this skill when:

• Submitting GPU training or inference jobs to Slurm clusters
• Managing running or queued jobs
• Monitoring cluster resources and job status
• Debugging job failures or performance issues
• Writing Slurm job scripts with Apptainer containers
• Checking GPU availability and utilization

SRP Slurm Clusters

Oracle OKE Cluster (H100 GPUs)