Alert coverage and noise
Compare incidents, suppressed signals, thresholds, batching and missing heartbeats.
Count incident coverage separately from notification noise
Use incident identities to quantify duplicate messages and missed incidents.
Test whether an alert suppression hides another service
Compare global and service-scoped suppression against supplied fault cases.
Compare alert thresholds by missed cases and false alarms
Evaluate two rules against labelled cases under an explicit missed-incident limit.
Check alert batching against a maximum wait
Calculate individual delays for fixed dispatch boundaries rather than average delay alone.
Treat a missing heartbeat as unknown rather than healthy
Apply inclusive freshness boundaries while separating monitoring absence from confirmed service failure.