A Deployment's rolling update strategy has maxUnavailable: 0 and maxSurge: 1 with 3 replicas. What happens during an update?
- A All 3 pods restart simultaneously
- B One new pod is created before an old one is removed, repeated until complete (correct answer)
- C 3 new pods start, then all old pods stop
- D The update fails because maxUnavailable cannot be 0
Shortcut: maxUnavailable: 0 forbids dropping below the desired count, and maxSurge: 1 permits one extra pod. The controller therefore adds one, waits for readiness, removes one, and repeats.