Mar 13
Loading video player...
I cloned the kube-prometheus-stack repo. It has over 20 YAML files. I was not going to write all of that by hand. In this video, we discover Helm - Kubernetes' package manager - and use it to deploy a full monitoring stack (Prometheus + Grafana) to a 6-node Raspberry Pi 4 cluster running Talos Linux. One command replaces 20+ manifests. Then we hit the wall: Talos's security model blocks node-exporter from running, and we debug our way through four separate fixes to get everything green. This is the video where the cluster stops being a black box. What you'll learn: - What Helm is and why it exists (shown, not just explained) - How to deploy kube-prometheus-stack with a single command - Customizing charts with values.yaml - Navigating Talos PSS restrictions - and fixing them step by step - Exploring pre-built Grafana dashboards showing live cluster telemetry - Why every future video in this series references these dashboards My Setup: - 6-node Raspberry Pi 4 cluster running Talos Linux - Pi 5 NAS providing NFS persistent storage - Gitea already deployed (Video 3) - shows up as a live workload in Grafana Previous Videos: - Video 1: Kubernetes Cluster Setup https://youtu.be/Qt4FfbWn55o - Video 2: NFS Persistent Storage https://youtu.be/UxZlCho7-KI - Video 3: Deploying Gitea with Raw Manifests https://youtu.be/UK5siIFChqk Next Video: - Video 5: Ingress - Proper URLs Instead of Port-Forwarding Everything #kubernetes #homelab #raspberrypi #helm #prometheus #grafana #monitoring #talos #selfhosted