Version 2.2.0 of the backend for the Kubernetes Monitoring solution in Grafana Cloud introduces significant enhancements to cluster observability through updated alert and recording rules. Supported by the kubernetes-mixin, the update improves alert accuracy and reduces noise by adding cluster labels to alerts, introducing new alert rules for common failure modes, and refining existing alerts for better performance. The update also removes misleading alerts to reduce false positives and implements more robust recording rules that ensure compatibility with different Prometheus versions and support all workload types. Highlighting a shift toward more stable metrics collection, the transition from the irate to a rate-based calculation for CPU usage is emphasized, reflecting the community's preference for accuracy and performance. These updates enhance monitoring capabilities for both small and large Kubernetes environments, promising improved visibility and control, and users are encouraged to explore the full changelog for technical details.