Google
advanced
Design a Globally Distributed Rate Limiter Supporting 100M Requests Per Second
Key Architectural Insight: Centralizing counters in a single global Redis creates unacceptable cross-region WAN latency (100ms+). The optimal design uses local in-memory token buckets synchronized asynchronously via distributed sliding-window consensus or token batching.
Technical Round 3 (System Design) · 2025
Priya Sharma
Verified Google
92Practice