Design a Metrics & Monitoring System (like Datadog/Prometheus)

This lesson designs a metrics and monitoring system, the kind of service that Datadog sells and that Prometheus provides as open source. Servers report numbers about themselves every few seconds, engineers look at those numbers on charts, and the system warns people when a number goes wrong. The design has one very large path for incoming data and two small paths for reading it, and most decisions below come from that difference.

1. Problem Definition and Scope

We are designing a large metrics platform