API Observability: Logs, Metrics, and Tracing

Suppose your order-details API becomes slow after a deployment, but the usual checks offer no obvious explanation. CPU usage is normal, most responses still return 200, and several services handle each request. To investigate, you need to find out whether the services spent time waiting for a database connection, retrying shipping requests, or building the response.

Observability is the ability to investigate a system's behavior through the evidence it produces. Logs, metrics, and traces provide different views of that behavior.

This chapter explains how to design them together so an API team can move from a user-visible symptom to a useful explanation.