kube-prometheus-stack bundles Prometheus, Grafana, Alertmanager, and supporting components into a single package for Kubernetes monitoring. It includes the Prometheus Operator for declarative configuration, Node Exporter for host metrics, and kube-state-metrics for cluster object states. The stack provides pre-configured scraping through ServiceMonitor and PodMonitor resources, alert management via PrometheusRule definitions, and built-in Grafana dashboards. Production deployments require persistent storage, resource tuning based on time-series cardinality, and strategies for handling high-cardinality metrics through relabeling, aggregation, and recording rules. For multi-cluster visibility and long-term retention, tools like Thanos or managed solutions extend the base stack's capabilities.
Table of contents
What kube-prometheus-stack Does for YouThe Monitoring ArchitectureWhat Metrics You Actually Have Access ToConfigure What Prometheus ScrapesAlerts Drive Action in PrometheusWhere All That Metric Data LivesWrite PromQL Queries You Use Every DayHow Grafana Fits Into the StackRun kube-prometheus-stack in ProductionManage High-Cardinality MetricsFinal ThoughtsSort: