Kubernetes Monitoring Helm Chart

Why use this chart?

There are many great Helm charts for individual pieces of the observability stack: Grafana Alloy, the OpenTelemetry Collector and Operator, kube-prometheus-stack, kube-state-metrics, Node Exporter, and more. Each is excellent at its job, but wiring them together into a complete, correct, and consistent monitoring solution can be a significant amount of work.

This chart exists to do that wiring for you. Instead of assembling and maintaining a collection of separate charts, you describe what you want to monitor and where it should go, and the chart generates the collector configuration and deploys the supporting components to make it happen.

• One chart for all telemetry signals. Metrics, logs & events, traces and profiles are all covered by a single, coordinated deployment rather than a patchwork of charts that have to be kept in sync.

• Feature-oriented configuration, not collector plumbing. The chart is organized into features that enable telemetry gathering based on outcomes like node and cluster metrics, node and pod logs, application observability, auto-instrumentation and more. The chart translates those choices into the underlying collector configuration, so there is no need to hand-write and debug collector pipelines.

• Pluggable destinations. Send data to Grafana Cloud or to any Prometheus-, Loki-, OTLP-, or Pyroscope-compatible backend. The same configuration can fan out to multiple destinations at once, and switching backends is a configuration change rather than a re-architecture.

• Collection only, no bundled database. Unlike all-in-one charts such as kube-prometheus-stack, this chart does not deploy a storage backend, Grafana, or Alertmanager. It focuses solely on collecting and shipping telemetry to the destination of your choice, which keeps it lightweight yet scalable enough for testing, development, and production environments.

• Batteries included, but not required. Supplemental components such as kube-state-metrics, Node Exporter, Windows Exporter, and OpenCost are deployed and pre-wired when the relevant features are enabled.

• Opinionated, curated defaults. Sensible scrape configs, relabeling, and metric tuning come out of the box to keep cardinality and cost under control, while still allowing for overrides.

• Validated and tested. The generated configuration is validated at render time to catch mistakes early, and the chart is exercised by extensive unit, integration, and platform test suites across Kubernetes distributions.

If you only need one signal from one component, a single-purpose chart may be all you need. If you want a complete monitoring solution on Kubernetes that ties those components together and sends everything to the backend of your choice, this chart is built for that.

Maintainers

Usage

Helm must be installed to use the chart. Please refer to Helm's documentation to get started.

Install from the Helm repository

helm repo add grafana https://grafana.github.io/helm-charts
helm install k8s-monitoring grafana/k8s-monitoring --values values.yaml

Install from the OCI registry

helm install k8s-monitoring oci://ghcr.io/grafana/helm-charts/k8s-monitoring --values values.yaml

Refer to the Chart Documentation to learn more, including about chart installation instructions.

Office Hours

We hold office hours on the 4th Friday of the month. Meeting times and recordings will be posted here:

Contributing

See our Contributing Guide for more information.

Links

• Kubernetes Monitoring on Grafana Cloud
• Grafana Alloy
• Kube State Metrics
• Node Exporter
• OpenCost