Tasks, actors and futures; raylets that schedule locally and spill back; owners that count references and rebuild lost values from lineage; a GPU platform on EKS with KubeRay and Kueue: one-hour boards and the theory behind them.
Infrastructure, platform and reliability engineers.
Your approach: Use diagrams or prose to explain responsibilities, state, capacity and failure behavior. Show the calculations requested by the question.
Design a Distributed Compute Framework like Ray. Tasks, actors and futures; raylets that schedule locally and spill back; owners that count references and rebuild lost values from lineage; a GPU platform on EKS with KubeRay and Kueue: one-hour boards and the theory behind them. Work from the scoping questions below. State assumptions for any unspecified load, guarantee or target, then trace your design end to end. Explain one difficult case and a credible alternative; the worked example is a reference, not a required implementation.
Resolve the scoping questions for a Distributed Compute Framework like Ray. Separate stated behavior from assumptions, and identify what is outside your design.
Declare relevant volume, latency, freshness, quality or cost targets with units. Show calculations or an evaluation plan that can test them; unspecified targets are your assumptions, not hidden pass criteria.
Explain how your guarantees hold in a difficult case relevant to this subject. Address: Autoscaling? Ordering?
Identify users, required behavior and exclusions. Answer: Workloads? Scale?
Define the information owned by the system and the inputs, outputs and errors at its boundaries. Resolve: Task sizes? Who uses the cluster?
Estimate the dominant workload and resource demand with units and explicit assumptions. For a learned system, also state how quality is measured and what data is available.
Draw or describe the responsibilities needed for a Distributed Compute Framework like Ray. Trace a representative request, event or job from its input to a visible result; identify durable state owners.
Walk through a difficult case step by step, including detection and recovery. Consider: What must survive? Autoscaling? Ordering?
Compare a credible alternative using your chosen workload and guarantees. Explain a remaining risk, a signal to watch and when you would change the design.