Design a Task Scheduler

Design a service that schedules tasks for execution. Explain how tasks become due, how workers claim them, and how the system handles failures without losing accepted work. ### Constraints & Assumptions The reported system-design prompt is a task scheduler without detailed scaling requirements. Start with a modest durable design and state whether tasks are one-time or recurring; do not assume an unreported global scale. ### Clarifying Questions What timing precision and lateness are acceptable? Are tasks recurring? What execution duration, retry, cancellation, and duplicate-effect guarantees are required? ### What a Strong Answer Covers A task model, durable scheduling state, atomic claims, worker leases, retries, and observable delivery semantics. ### Follow-up Questions How do you avoid two schedulers dispatching the same due task? What happens on a crash after dispatch? How would you add recurring schedules or partition the service if load grows?