The Trade Desk Interview Question

Distributed cache design Implement a low latency read APIs supporting millions QPS