Alerts that page for symptoms, and the limits of one Prometheus
Write symptom-based alerts with recording rules and error-budget burn rates, route them through Alertmanager, and learn where a single Prometheus server runs out of room. Part 3 of a series.