What Is Observability?

• Observability refers to the measurable outputs a system exposes to understand its internal behavior — Instead of guessing what happens inside, we analyze the telemetry it emits — such as metrics, logs, and traces.
• Systems must be intentionally designed and instrumented to be observable — Observability does not happen automatically. Applications need proper instrumentation to expose meaningful data.
• Observability data is used for fault detection, alerting, and performance analysis — Teams use this data to detect incidents, trigger alerts, troubleshoot issues, and improve performance and reliability.
• The Cloud Native Computing Foundation (CNCF) defines three key observability signals — not "pillars" — The CNCF, the organization behind projects like Kubernetes and Prometheus, identifies metrics, logs, and traces as the core signals for understanding modern distributed systems.