Prasanta Kumar Behera
Edge Kubernetes & Zero-Touch Provisioning
Learning Hub / Bare Metal & Edge

Edge Kubernetes & Zero-Touch Provisioning

Advanced → Architect12 lessonsAvailable

Zero-touch provisioning for edge Kubernetes using open-source tooling: Redfish and PXE to boot, Tinkerbell or Metal3 to provision, EKS Anywhere or RKE2 to form the cluster, and GitOps to manage the fleet, all without internet access.

You'll meetzero-touchair-gappedRedfishPXE / iPXETinkerbellMetal3cloud-initEKS AnywhereRKE2GitOps fleet
Start lesson 01 →

What you'll be able to do

  • Design a ZTP flow from BMC to a Ready node
  • Build immutable OS images and serve them offline
  • Operate a fleet of small clusters with GitOps

Before you start

Linux — Level by Level (Level 2), Kubernetes Administration Level 2.

How it works

Each lesson: plain-language idea → how it really works → hands-on. Each section ends with a cheat sheet & self-check.

Curriculum

Lessons marked “Read” are ready; the rest are on the way.

Foundations

  1. 01Why ZTP for edgeTruck rolls, scale, consistencyRead →
  2. 02Architecture for edge ZTPManagement cluster, sites, trust boundariesRead →
  3. 03Cluster sizing for edgeSingle-node, 3-node compact, small HARead →
  4. 📋Cheat sheet & self-checkEvery command from this section on one page, then 9 questions to check yourself.Open →

Provisioning

  1. 04Bare-metal provisioningRedfish, IPMI, PXE/iPXE, DHCPRead →
  2. 05Tinkerbell & Metal3Workflows, hardware inventory, BareMetalHostRead →
  3. 06OS image creationImmutable images: ISO, RAW, QCOW2, cloud-initRead →
  4. 07Air-gapped deliveryHarbor mirrors, offline charts, signed bundlesRead →
  5. 📋Cheat sheet & self-checkEvery command from this section on one page, then 12 questions to check yourself.Open →

Cluster & Fleet

  1. 08EKS Anywhere on bare metalCluster spec, lifecycle, upgradesRead →
  2. 09Rancher + RKE2 provisioningHardened defaults, SUSE SLE MicroRead →
  3. 10GitOps fleet managementOne repo, hundreds of sitesRead →
  4. 11Edge networking, storage & securityThe constraints that differ at the edgeRead →
  5. 12DR & compliance at scaleRebuild a site from GitRead →
  6. 📋Cheat sheet & self-checkEvery command from this section on one page, then 15 questions to check yourself.Open →

Real-world scenarios

Work through each one: symptom → misleading signal → evidence → root cause → prevention.

Nodes PXE-boot, then hang waiting for network

DHCP, VLAN tagging or the boot image? Isolate the layer.

Same image, two sites, different behaviour

Firmware and BIOS drift across hardware batches.

The site came back after a power loss; the cluster didn't

Boot ordering, etcd quorum and time sync.

This site is a public version of my personal engineering knowledge hub. It intentionally excludes confidential company information and internal operational details.