Simple definition
Observability helps engineers understand a system's internal state through signals such as metrics, logs, and distributed traces.
Purpose and use
Observability explains internal service behavior from signals produced during operation.
How does it work?
Metrics, logs and traces offer complementary views; correlation helps follow a request and its dependencies.
Example and verification
For a slowdown, connect latency measurements to relevant traces and component events using consistent timestamps.
Points to watch
Collecting large volumes does not guarantee useful explanations. Sensitive data, cardinality and sampling affect security, cost and diagnostic quality.
In short
Instrumentation should answer operational questions and support testing incident hypotheses.