A team is deciding what to monitor for a newly instrumented service. Which set of signals gives the most complete picture of its health?
Select an answer to reveal the explanation.
Short Explanation
You want the four signals that show how slow, how busy, how broken, and how full the service is. Everything else you might monitor is either a detail underneath one of those or not about health at all.
Full Explanation
The correct answer is the set of latency, traffic, errors, and saturation, because these four cover the distinct ways a service degrades. Latency shows response time, traffic shows demand, errors show failures, and saturation shows how close resources are to exhaustion - the leading indicator that the other three are about to worsen.
The option about hostnames is incorrect because hostnames are inventory. Knowing where a service runs tells you nothing about whether it is working.
The option about the most recent deployment date is incorrect because deployment date is useful context when correlating a change to a regression, but it is not a health measurement - a service can be broken on old code or healthy on new.
The option about dashboard colour themes is incorrect because dashboard styling is presentation with no bearing on service behaviour.
For the exam, name all four signals and be able to say what each one detects.