Katkat
Back

Design delayed job scheduler (LLD)

AmazonSoftware Engineer · Onsite
HardSoftware EngineerOnsiteSystem Design

Last updated

This question evaluates system design and distributed-systems competencies, including scheduling algorithms, durable persistence and recovery, concurrency control, delivery semantics, and operational concerns, and it belongs to the System Design category.

Design a system that schedules a job to execute X seconds in the future (LLD). Define APIs (e.g., schedule(job, delaySeconds) -> jobId, cancel(jobId)), data structures (e.g., time wheel, min-heap by due time), persistence for durability, worker execution model, handling of restarts and clock drift, idempotency/exactly-once vs at-least-once semantics, concurrency limits, and time complexity. Provide class diagrams and discuss how you would test it.