A service's error rate and latency both look normal, but a team wants earlier warning of trouble. Which signal best serves that purpose?
Select an answer to reveal the explanation.
Short Explanation
Saturation is your early-warning signal. When a pool fills up or memory creeps toward its ceiling, you can see trouble coming, while latency and errors only tell you it has already arrived.
Full Explanation
Saturation measures how full a constrained resource is—connection pools, memory, thread pools, queues. It rises before the effects become visible to users, which is exactly what gives a team time to act. The traffic option is incorrect because traffic measures demand, and it usually rises rather than falls before saturation-driven trouble; a drop in traffic more often indicates a failure upstream that has already happened. The deployment-frequency option is incorrect because release cadence is a process metric. It may correlate with risk across an organisation but does not indicate the state of a running service. The dashboard-count option is incorrect because dashboard count measures team activity, not service behaviour. For the exam, associate saturation with early warning and latency and errors with realised impact.