Monitoring 8,000 series: how many alerts are noise?
Multiple testingMediumProblem. You monitor stall rate across every combination of region, game, rig type, and client build. That is roughly 8,000 time series, checked every five minutes. Each check is a test at the 5% level. Your manager asks why the on-call is drowning. What do you tell them?
Before you reveal: say your answer out loud, as if you were in the real interview — get your reasoning across clearly first. There is no single correct answer: reading what the interviewer is really after and defending your own thinking is what makes an answer strong.