Your organization processes millions of requests per second across hundreds of microservices, and your Prometheus instance is hitting performance limits. How would you architect a scalable observability platform?

Sample answer preview

Scaling Prometheus beyond a single instance requires a thoughtful architecture that addresses storage limits, query performance, high availability, and multi-cluster federation. There are several proven approaches, each with different trade-offs.