Design a High-Performance Computing Cluster

Functional Requirements:

• The system should allow users to submit computational jobs to the cluster for processing.
• Jobs must be scheduled and distributed efficiently across available compute nodes.
• The cluster should support both CPU and GPU workloads.
• Users should be able to monitor the status and progress of their jobs.
• The system should provide basic resource management (CPU, memory, GPU allocation) per job.
• The cluster should support horizontal scaling by adding more compute nodes as needed.
• The system should handle job failures gracefully and reschedule jobs if a node fails.
• Users and administrators should have secure access to submit jobs and manage the cluster.

Non-Functional Requirements:

• High performance: The system should minimize job wait and execution times.
• Reliability: The cluster should be resilient to node failures and minimize downtime.
• Scalability: The system should efficiently handle increased workloads by adding more nodes.
• Security: Only authorized users can submit jobs or access cluster resources.
• Usability: Provide a simple interface (CLI or web) for job submission and monitoring.
• Maintainability: The system should be easy to update and maintain.
• Availability: The cluster should be available for job submission and processing most of the time (e.g., 99.9% uptime).