Compare Paxos and Raft
Company: Uber
Role: Software Engineer
Category: System Design
Difficulty: hard
Interview Round: Onsite
Explain how Paxos and Raft achieve consensus over a replicated log. Compare leader election, quorum requirements, log replication, membership changes, and behavior under partitions. Walk through failure scenarios (leader crash, message delays, disk loss) and recovery. Given goals of operational simplicity and debuggability, choose one for a new service and justify the trade-offs.
Quick Answer: This question evaluates understanding of distributed consensus protocols, replicated logs, linearizability, fault tolerance, and operational concerns such as leader election, quorum semantics, log replication, membership changes, and failure recovery.