four golden signals
Latency, traffic, errors and saturation: the four measurements every service should emit.
In more detail
Split latency by success and failure, because fast errors can make latency look better during an outage. Saturation is the one that predicts trouble.
Related terms
- latency — The time between asking for something and getting the answer back.
- error rate — The share of requests that fail, reported alongside latency so a fast result is not mistaken for a healthy one.
- saturation — How full a system’s most constrained resource is, such as its CPU, memory, connections or queue.
- metric — A number measured over time, such as requests per second or error count, that shows how a system behaves in aggregate.
Taught in
- Observability As A Design Input — Scalable System Design & Distributed Architecture