Provision and manage bare metal anywhere, with the same declarative, API-driven model the cloud taught us — meet Tinkerbell.

Welcome to Tinkerbell: Declarative Bare-Metal Provisioning.

Tinkerbell is a CNCF sandbox project that network-boots physical servers and runs workflows on them, bringing the declarative, controller-driven model of Kubernetes all the way down to the rack. Where provisioning bare metal once meant a fragile patchwork of DHCP, TFTP, PXE configs, and hand-run installers, Tinkerbell lets you describe a machine and a provisioning recipe as Kubernetes custom resources and reconciles the physical world toward them. This book takes the system apart component by component. We begin with the bare-metal problem and Tinkerbell's architecture, then follow a single machine through its journey: Smee answering its network-boot request, Hook booting it into an in-memory workshop, the Tink engine turning templates into containerized actions, Hegel serving it cloud-style metadata, and Rufio reaching into its management controller to control power and boot. A closing chapter shows how the pieces deploy together, integrate with Cluster API, and hold up at scale.

This title is part of the ShriIra library and is free to read in full, right here — our small contribution to making world-class knowledge easy to reach.

A note on reading it: open the Contents menu at the top of the reader to jump between chapters, use the Aa menu to set a comfortable text size, theme (light, sepia, or night), and single- or two-page layout. Your place is saved automatically, so you can always pick up where you left off.

We hope it serves you well.

— Shriira Press

Contents

1. Chapter 1 — The Bare-Metal Problem
2. Chapter 2 — The Architecture
3. Chapter 3 — Smee and the Network Boot
4. Chapter 4 — Hook, the In-Memory Workshop
5. Chapter 5 — Tink, the Workflow Engine
6. Chapter 6 — Hegel and the Metadata Service
7. Chapter 7 — Rufio and the Baseboard Management Controller
8. Chapter 8 — Tinkerbell in Practice