Skip to content
The Internet Compass

Engineering

Observability

Observability is the degree to which a system's internal state can be understood and diagnosed from the external outputs it produces — commonly built from three data types: logs (discrete event records), metrics (numeric measurements over time), and traces (the path of a single request through a distributed system).

Observability is distinct from traditional monitoring in intent: monitoring answers known questions you set up dashboards and alerts for in advance ("is CPU usage above 80%?"); observability is meant to let you answer questions you didn't anticipate, after the fact, by exploring the data a system already emits.

In distributed, microservice-based systems, distributed tracing specifically has become essential because a single user request can touch dozens of services, and understanding where latency or errors originate requires following that request's path across all of them, not just inspecting any one service in isolation.