Phase 4 — Monitoring Stack
Visibility into cluster health, node metrics, and application performance.
Stack
| Tool | Purpose |
|---|---|
| Prometheus | Metrics collection & storage |
| Grafana | Dashboards & visualization |
| Loki | Log aggregation (phase 11) |
| Alertmanager | Alerts & notifications |
Install kube-prometheus Stack
kubectl apply -f https://github.com/prometheus-operator/prometheus-operator/releases/latest/download/bundle.yaml
This deploys the Prometheus Operator which manages Prometheus, Grafana, and Alertmanager via CRDs.
Access Grafana
kubectl port-forward svc/grafana -n monitoring 3000:3000
Open: http://localhost:3000
Default credentials: admin / admin
Key Dashboards to Import
| Dashboard | Grafana ID |
|---|---|
| Kubernetes cluster overview | 7249 |
| Node exporter full | 1860 |
| k3s cluster monitoring | 13770 |
Done When
✔ Prometheus scraping all nodes
✔ Grafana dashboards showing metrics
✔ Node CPU, RAM, disk visible