Design the Control Plane for a Compute Cluster: Host Monitoring + Job Dispatch

Company: NVIDIA

Role: Software Engineer

Category: System Design

Difficulty: medium

Interview Round: Technical Screen

Quick Answer: This question assesses the ability to design a distributed control plane that tracks live host state across a large compute fleet under high write concurrency. It is commonly used in system design interviews to evaluate reasoning about data storage schema choices, scalability as fleet size grows, and failure detection and availability when a central coordinator goes down. It represents a practical, applied system design scenario rather than a purely conceptual one.