Skip to content

Monitoring and Performance in k3s

Published by bobo on May 27, 2025

with Prometheus, Grafana, and Homarr

As your home Kubernetes cluster grows, so does the need to keep tabs on its health, performance, and resource usage. In this post, we’ll walk through setting up a robust monitoring stack using Prometheus and Grafana, with bonus integration into the Homarr dashboard.

Stack Overview

We’ll be using:

  • Prometheus: For collecting metrics
  • Grafana: For visualizing metrics
  • Kube State Metrics and Node Exporter: For detailed cluster data
  • MetalLB: To expose the Grafana UI
  • Homarr: To view widgets from Prometheus

Step-by-Step Instructions

1. Add the Prometheus Helm Repo

SSH into your jumpbox and run:

helm repo add prometheus-community https://prometheus-community.github.io/helm-charts
helm repo update

2. Install kube-prometheus-stack

This will install everything we need:

helm upgrade --install monitoring prometheus-community/kube-prometheus-stack \
  --namespace monitoring --create-namespace \
  --set prometheus.prometheusSpec.maximumStartupDurationSeconds=300

This includes:

  • Prometheus
  • Grafana
  • AlertManager
  • Kube State Metrics
  • Node Exporter

3. Expose Grafana with MetalLB

Create a grafana-service.yaml:

apiVersion: v1
kind: Service
metadata:
  name: grafana-lb
  namespace: monitoring
spec:
  type: LoadBalancer
  selector:
    app.kubernetes.io/name: grafana
  ports:
    - port: 80
      targetPort: 3000
      protocol: TCP
      name: http
  loadBalancerIP: 10.0.0.40

Apply it:

kubectl apply -f grafana-service.yaml

4. Get the Grafana Admin Password

kubectl get secret --namespace monitoring monitoring-grafana -o jsonpath="{.data.admin-password}" | base64 --decode ; echo

Default user: admin

Access Grafana at http://10.0.0.40

5. Add Prometheus Widgets to Homarr

If you already have Homarr installed, add widgets by editing your config and using Prometheus as a data source:

- type: prometheus
  title: CPU Usage
  endpoint: http://monitoring-kube-prometheus-prometheus.monitoring.svc.cluster.local:9090
  query: sum(rate(container_cpu_usage_seconds_total[1m]))
  format: short

Note: You can expose Prometheus via MetalLB or a Traefik ingress if needed.

Useful Dashboards to Import in Grafana

  • Kubernetes Cluster Monitoring (ID: 315)
  • Node Exporter Full (ID: 1860)
  • Longhorn Dashboard (search on Grafana.com)

📅 Recap

With Prometheus and Grafana, you now have:

  • Live monitoring of CPU, memory, pods, and nodes
  • Dashboard access via MetalLB IP
  • Visibility from your Homarr homepage

This closes the loop on not just deploying workloads in k3s, but operating them confidently.

Up next: we’ll explore setting up Alerts and Notification hooks using AlertManager.

Published inKubernetes