A game studio pages on-call when an instance is both high on CPU and failing its status check. Separate alarms already exist for each condition, and CPU-only spikes are waking people overnight. What should send the page?
Select an answer to reveal the explanation.
Short Explanation
Two smoke detectors in the same hallway should not each call the fire department. You wire them so the page goes out only when both are shouting. Averaging a CPU percent with a status check is mixing a thermometer with a yes-or-no light.
Full Explanation
A composite alarm watches other alarms and can require all of them to be in ALARM before it acts, which is the page rule the studio described. The child alarms stay in place for people who want the separate signals. An average of CPU percent and a status-check metric is not a meaningful unit and will not implement AND logic. An EventBridge rule on every EC2 API call pages on ordinary control-plane activity, not on the two symptoms named here. A dashboard widget helps a human look; without an action it never pages. Caveat: a composite alarm does not solve a child alarm stuck in INSUFFICIENT_DATA, so confirm both children publish the metric you think they do. Operational check: put one child in ALARM in a test and confirm the composite stays OK, then put both in ALARM and confirm the notification fires.