Kubernetes Installation

Deploy Grafana on Kubernetes using manifests or Helm charts

Kubernetes Installation

This guide covers deploying Grafana on Kubernetes using Kubernetes manifests or Helm charts.

Before You Begin

Ensure you have:

• A Kubernetes cluster (local or cloud-based)
  • Local: minikube, kind, or Docker Desktop
  • Cloud: GKE, EKS, or AKS
• kubectl CLI configured to access your cluster
• For Helm installation: Helm 3.x installed

System Requirements

Minimum hardware requirements per pod:

• CPU: 250m (0.25 cores)
• Memory: 750 MiB
• Disk: 1 GB persistent storage

Ensure port 3000 is accessible in your network environment.

Installation with Kubernetes Manifests

Deploy Grafana using native Kubernetes manifests for full control over the deployment.

Installation with Helm

Helm simplifies Kubernetes deployments using packaged charts.

Helm Configuration

Enable Persistent Storage

Create or download the values.yaml file from the Grafana Helm Charts repository.

Edit values.yaml to enable persistence:

persistence:
  type: pvc
  enabled: true
  storageClassName: default

Apply the configuration:

helm upgrade my-grafana grafana-community/grafana -f values.yaml -n monitoring

Install Plugins

Add plugins to values.yaml:

plugins:
  - grafana-clock-panel
  - grafana-simple-json-datasource

Apply the configuration:

helm upgrade my-grafana grafana-community/grafana -f values.yaml -n monitoring

Custom Admin Password

Set a custom admin password in values.yaml:

adminUser: admin
adminPassword: your-secure-password

Apply the configuration:

helm upgrade my-grafana grafana-community/grafana -f values.yaml -n monitoring

Deploy Grafana Enterprise

To deploy Grafana Enterprise on Kubernetes:

Update Deployment

Perform rolling updates to change the Grafana version:

kubectl edit deployment grafana --namespace=my-grafana

Change the image version:

image: grafana/grafana:12.1.0

Verify the rollout:

kubectl rollout status deployment grafana --namespace=my-grafana

Rollback Deployment

View rollout history:

kubectl rollout history deployment grafana --namespace=my-grafana

Rollback to a previous revision:

kubectl rollout undo deployment grafana --to-revision=1 --namespace=my-grafana

Troubleshooting

View logs

kubectl logs --namespace=my-grafana deploy/grafana

For multi-container pods:

kubectl logs --namespace=my-grafana deploy/grafana -c grafana

Enable debug logging

Create a ConfigMap with custom configuration:

echo '[log]
level = debug' > grafana.ini

kubectl create configmap grafana-config \
  --from-file=grafana.ini \
  --namespace=my-grafana

Cleanup

Remove the Grafana deployment:

kubectl delete -f grafana.yaml --namespace=my-grafana

For Helm installations:

helm uninstall my-grafana -n monitoring

Delete the namespace:

kubectl delete namespace my-grafana

Next Steps

• Configure data sources using provisioning
• Set up persistent storage for production
• Configure high availability with multiple replicas
• Integrate with ingress controllers for external access