How would you design a rate limiter that supports burst traffic and quick lookups?

Overview

Evaluates understanding of sliding window, token bucket, and data structure choices.

Example Answer

• Use a token bucket stored in Redis with atomic increments and expirations.
• Sliding window counters can be approximated with two buckets for simplicity.

Tips

• Discuss token bucket vs. sliding window log/counter.
• Highlight per-user keying and TTL.
• Talk about distributed consistency if needed.