☁️ Free practice test question from: Kubernetes Architecture (KCNA)

In a traditional event-driven (edge-triggered) system, a component reacts once when an event occurs (the “edge” or state transition). In Kubernetes, controllers employ a level-triggered model: they continuously compare the actual state of objects with the desired state and take action as long as a discrepancy exists—even if no “new” event has occurred. This means that if the system remains in an undesired state, the controller will keep reapplying corrections until the state matches the specification.

****

A) This statement is partially true in that Kubernetes uses "watching" and avoids constant polling, but it doesn't describe the main difference between the system types in terms of event/state handling. Traditional event-driven systems usually don't rely on polling; they use push notifications instead.

C) Saying Kubernetes components “passively watch” may imply that controllers do nothing until a notification arrives. In reality, Kubernetes controllers continuously reconcile actual vs. desired state, even if no new events are fired.

D) This statement is completely incorrect. Kubernetes is based on events (changes in object states). Communication happens asynchronously and is driven by object states.