Observability

Observability is being able to work out what a running system is doing from the outside, without attaching a debugger. In practice it rests on three things: logs of what happened, metrics of how much and how fast, and traces following a single request across every service it touched.

The distinction from monitoring is worth keeping. Monitoring answers questions you thought to ask in advance — is it up, is it slow. Observability is what lets you answer the question you did not anticipate, which is invariably the one an incident presents.

The practical test is whether a report of "it was slow for one customer this morning" can be investigated at all. If the answer is to ask them to try again and watch, the system is not observable, and every incident will be diagnosed by guesswork.