Skip to main content

Grafana Dashboards

Grafana provides visualization dashboards backed by three data sources: Prometheus (real-time metrics), Mimir (long-term metrics), and Loki (logs).

Data Sources​

Configured Data Sources​

Data SourceTypeURL
PrometheusprometheusBuilt-in (kube-prometheus-stack default)
Mimirprometheushttp://prometheus-mimir-gateway.monitoring.svc.cluster.local/prometheus
Lokilokihttp://prometheus-loki.monitoring.svc.cluster.local:3100

Configuration​

Grafana is deployed as part of the kube-prometheus-stack with subdomain routing:

grafana:
  enabled: true
  grafana.ini:
    server:
      domain: grafana.el-jefe.me
      root_url: "https://grafana.el-jefe.me/"
      serve_from_sub_path: false
  defaultDashboardsEnabled: true
  defaultDashboardsTimezone: UTC
  resources:
    requests:
      cpu: 100m
      memory: 128Mi
    limits:
      cpu: 500m
      memory: 512Mi

The kube-prometheus-stack provides a set of default dashboards out of the box covering Kubernetes internals, node metrics, and Prometheus self-monitoring.

Key Dashboards​

Cluster Overview​

  • Total pod count by namespace
  • Node CPU and memory utilization
  • Pod restart events and crash loops
  • PersistentVolume usage

Application Health​

  • HTTP request rate per application
  • Response time percentiles (p50, p95, p99)
  • Error rate by status code
  • Active connections

Traefik Ingress​

  • Requests per second by host
  • Backend response times
  • TLS certificate expiry
  • Connection counts

Node Resources​

  • CPU utilization trends
  • Memory usage and pressure
  • Disk I/O and capacity
  • Network throughput

Ingress​

Grafana is exposed externally via Traefik with TLS, using the same Ingress resource as Prometheus:

rules:
  - host: grafana.el-jefe.me
    http:
      paths:
        - path: /
          pathType: Prefix
          backend:
            service:
              name: prometheus-grafana
              port:
                number: 80
tls:
  - hosts:
      - grafana.el-jefe.me
      - prometheus.el-jefe.me
    secretName: monitoring-tls
See Also

The Cluster Dashboard on the portfolio site provides a public view of key cluster metrics, complementing the internal Grafana dashboards. The PodRick stories in the Storybook Showcase demonstrate the dashboard components used for deployment monitoring.