Skip to main content
Version: 1.22.X

Grafana

Grafana is an open-source data visualization and graph composer platform for numeric time-series data with Prometheus integration.

Image repository and tag​

Requirements​

  • Kubernetes >= 1.21.0
  • Kustomize = v3.5.3

Configuration​

Fury distribution Grafana is deployed with the following configuration:

  • Replica number: 1
  • Anonymous authentication enabled
  • Admin role for unauthenticated users
  • Resource limits are 200m for CPU and 200Mi for memory
  • Listens on port 3000
  • Prometheus configured as the data source

Add new dashboards​

You can create a ConfigMap/Secret in your namespace with a JSON of a grafana dashboard and then labeling it with the label key grafana-sighup-dashboard, the value of the label is up to you. Labeling it, the sidecar k8s-sidecar will take care of it and inject it into a shared volume where grafana does a lookup and discover it. Look at the dashboards folder kustomization.yaml for an example.

Add new datasource​

Additionally, you can create a ConfigMap/Secret in your namespace with a datasource definition then labeling it with the label key grafana-sighup-datasource, the value of the label is up to you. Labeling it, the sidecar k8s-sidecar will take care of it and inject it into a shared volume and send an API reload signal to grafana itself. Look at the datasources folder kustomization.yaml for an example.

Deployment​

You can deploy Grafana by running the following command:

kustomize build | kubectl apply -f -

Accessing Grafana UI​

You can access Grafana Dashboard by port-forwarding on port 3000:

kubectl port-forward svc/grafana 3000:3000 --namespace monitoring

Grafana will be available on http://127.0.0.1:3000 from your browser.

Adding/Removing Dashboards​

To learn how to add or remove dashboards to Grafana please see the examples folder.