5.4.3. Operational Concepts
💡 First Principle: These terms name what happens while a service runs — the things that occur (events, incidents, requests) and the qualities you engineer for (reliability, observability) — and getting the incident-versus-event line right is foundational to all of service operations.
Key operational terms: an event is any change of state significant for the management of a service or component (it may be routine, warning, or exceptional — not inherently bad); an incident is an unplanned interruption to a service or a reduction in its quality (always undesirable); a service request is a request from a user for something normal and pre-approved (like access or information), not a failure.
Reliability-engineering terms: reliability is the degree to which a service performs its intended function consistently over time; Site Reliability Engineering (SRE) is a discipline that applies engineering and software practices to operations to make services more reliable and scalable; observability is the ability to understand a system's internal state from the data it emits (logs, metrics, traces) — you can't manage what you can't observe.
⚠️ Exam Trap: An event is not automatically a problem — many events are routine or informational. An incident is always an unplanned interruption or quality reduction. A service request is a normal, pre-approved ask, not an incident. Keep these three apart.
Reflection Question: Why is observability a precondition for reliability rather than a nice-to-have add-on?
