Overview
Appliance based Kubernetes
RuckOS Linux is a purpose-built, immutable Kubernetes node OS. It boots a small signed operating system, exposes a narrow management API, and hands the node to Kubernetes without SSH, package managers, or mutable server drift.
The platform pairs RuckOS with Cilium networking, AD-rooted PKI,
GitOps delivery, and the rctl operator CLI.
Getting started
Operator workflow
Operators boot a node from the RuckOS image, discover it with
rctl scan, apply machine configuration, then bootstrap
Kubernetes control plane, worker, storage, and platform services.
rctl scan
rctl -n <node-ip> get-config --insecure
rctl -n <node-ip> bootstrap init --site config/site.yaml
System requirements
Baseline
- UEFI-capable VM or bare-metal host.
- Network with DHCP for first boot or static config via rctl.
- Persistent state disk for Kubernetes and container runtime data.
- AD DNS and ADCS for the enterprise integration path.
- Registry access for workload images.
Quickstart
From boot to cluster
- Boot the node from the RuckOS image.
- Find the node with
rctl scan. - Apply hostname, role, and network configuration.
- Generate and push Kubernetes bootstrap files.
- Install Cilium, storage, registry trust, and platform services.
Deploy your first workload
Descriptor to GitOps
Application teams describe intent in a small deployment descriptor.
CI builds the image, pushes it to the platform registry, renders
Kubernetes manifests with rctl render, and commits the
result to the GitOps repo for Argo CD to reconcile.
rctl render -f .kubeos/deployment.yaml --environment dev
git add apps/dev/<app>
git commit -m "Deploy <app> to dev"