hard20 pts · first answer countsdistributed-systemsconcurrency

Why a lock alone is not enough

Worker A takes a Redis lock with a 10-second TTL, then pauses for 15 seconds due to a long GC pause. The lock expires, Worker B takes it and writes to storage. Then A wakes up and also writes. What prevents A from corrupting data?

A.Making the TTL 60 seconds solves it in all casesB.Using a Redis cluster instead of a single Redis nodeC.A fencing token: an increasing number issued with each lock, which the storage checks and rejects older tokensD.Having A check the clock before writing

Discussion 0

Join free to post in the discussion.

Why a lock alone is not enough

Discussion 0

Related questions