Distributed read-write lock
Design a distributed read-write (shared/exclusive) lock service for protecting a large mutable config blob read by ~40k services continuously but updated rarely: many concurrent readers must proceed in parallel, a writer needs exclusive access, and readers must not see a half-applied write. Constraints: readers should almost never block, writer starvation must be bounded, and a crashed reader or writer must not deadlock the resource. Describe lock modes, how concurrent readers and a writer coordinate, and crash recovery.
What a strong answer looks like
Clarify scale and constraints first. Propose a clean component breakdown, then go deep on the hard parts (data model, bottlenecks, consistency, failure modes) and name the trade-offs you are making.
Clarify5:00 left
Estimate5:00 planned
Design15:00 planned
Deep dive12:00 planned
Failure8:00 planned
Which questions mattered is sealed until you submit. Telling you now would just be handing over the edge cases.
Run or narrate your approach, then ask the coach.