Skip to content

Appendix E — Learning paths

Ordered ways through this guide. Every link is a real chapter in this repository; the guide is standalone, so these map its chapters — they do not assume any other notes. Four paths: a one-week Fast track, the full Zero-to-production arc, an Exam-oriented mapping to CKAD/CKA/CKS, and a Platform/SRE/ops track. Each path lists ordered chapter links, a one-line "why this order", what it honestly skips, and ends at the same finish line: the end-to-end Bookstore (the right finish for Parts 00 - 09; readers continuing into the production-reality material then cross into the Bookstore Platform v2 grand capstone after Parts 10 - 12).

Reference material — no nine-section anatomy. Keep Appendix A — cheatsheet and Appendix B — glossary open while you work through any path.


Path 1 — Fast track (~1 week): deploy & operate an app

The minimum coherent path to deploy a real app and keep it running. It deliberately skips internals depth, scheduling theory, the full security/ delivery arc, and day-2 — you can deploy and debug, not yet harden or GitOps-deliver.

# Chapter Why here
1 00-foundations/01 — Why Kubernetes The problem it solves; declarative vs imperative
2 00-foundations/03 — Architecture overview The parts you'll talk to (skip 04/05 deep dives for now)
3 00-foundations/06 — The declarative API model The one mental model everything else is a variation of
4 00-foundations/07 — Local cluster setup A real local cluster + the everyday kubectl verbs
5 01-core-workloads/01 — Pods The unit you actually run
6 01-core-workloads/02 — Health and lifecycle Probes — without these nothing is production-real
7 01-core-workloads/04 — ReplicaSets and Deployments Self-healing, rolling updates, rollback
8 02-networking/02 — Services Stable in-cluster endpoints
9 02-networking/04 — Ingress Reach the app from outside
10 03-config-and-storage/01 — ConfigMaps Externalize config
11 03-config-and-storage/02 — Secrets Credentials (and why base64 ≠ encryption)
12 03-config-and-storage/04 — Persistent storage The app's database needs durable storage
13 08-day-2-operations/03 — Troubleshooting playbook When it breaks: the method + kubectl debug
14 09-end-to-end-bookstore/01 — Bookstore end-to-end Stand the whole thing up (skim the GitOps/DR sections)

Why this order: model → real cluster → run a workload → expose it → configure it → persist data → debug it → see it whole. Skips (be honest): control-plane/node internals, StatefulSets/DaemonSets/Jobs depth, scheduling, the full security hardening, observability/autoscaling, Helm/Kustomize/GitOps, and most of day-2 (troubleshooting is included; cluster upgrades/backup-DR/multi-tenancy/operators are skipped). Do the full arc (Path 2) before running anything you care about in production.


Path 2 — Zero-to-production (the full arc)

Parts 00→09 in order. This is how the guide is meant to be read: the Bookstore manifests are cumulative, each chapter adds one field/primitive to the same app. Nothing is skipped. Time estimates assume the hands-on is actually run on a local cluster.

Part Chapters Est. Why this order
00 Foundations 01 · 02 · 03 · 04 · 05 · 06 · 07 6–9 h Containers → architecture → internals → the declarative model → a real cluster, before any workload
01 Core Workloads 01 · 02 · 03 · 04 · 05 · 06 · 07 · 08 8–12 h Pod → health → resources → controllers; stateful/daemon/batch; how to roll changes
02 Networking 01 · 02 · 03 · 04 · 05 · 06 6–9 h Model → Services → DNS → edge (Ingress/Gateway) → segmentation
03 Config & Storage 01 · 02 · 03 · 04 · 05 5–8 h Config → secrets → volumes → PV/PVC → operating stateful data
04 Scheduling 01 · 02 · 03 3–5 h How placement is decided → steering it → priority/preemption
05 Security 01 · 02 · 03 · 04 6–9 h Identity/RBAC → pod hardening (PSA) → supply chain → secrets & cluster hardening
06 Production Readiness 01 · 02 · 03 · 04 · 05 · 06 7–10 h See it (metrics/logs/traces) → scale it → keep it reliable → afford it
07 Delivery 01 · 02 · 03 · 04 · 05 7–10 h Package (Helm/Kustomize) → CI/CD → GitOps → progressive delivery
08 Day-2 Operations 01 · 02 · 03 · 04 · 05 6–9 h Lifecycle/upgrades → backup/DR → troubleshooting → multi-tenancy → operators
09 Capstone 01 2–4 h Compose every part: GitOps up, observed, autoscaled, hardened, DR-drilled

Why this order: it is the dependency order of the system itself — you cannot secure or deliver a workload you cannot yet run, observe, or place. Skips: nothing — this is the complete path. Total ≈ 55–85 hours of focused, hands-on study (less if you skip the under-the-hood and In production: sections; those are the production-relevant parts, so don't skip them if you're going to production).


Path 3 — Exam-oriented (CKAD / CKA / CKS)

This guide is not an exam cram — it teaches the why and goes deeper than the exams in places (internals, GitOps, operators) and lighter in others (raw cluster install/etcd ops, exam-specific muscle memory). The maps below show which guide chapters cover which exam domain, honestly flagged [deeper] / [lighter] / [≈] vs the exam's depth. The official curricula are the authority for exact, current domain weights: https://github.com/cncf/curriculum. Drill speed with Appendix A's CKAD speed tips.

Methodology: each exam domain is mapped to the guide chapter(s) that cover its objectives. [deeper] = the guide goes well beyond exam depth; [lighter] = the guide covers the concept but the exam demands more hands-on speed or a topic the guide treats as note-only (e.g. raw kubeadm install, etcd backup/restore by hand); [≈] = roughly exam-aligned. Not overclaimed: chapters are mapped only where they genuinely cover the objective.

CKAD — Certified Kubernetes Application Developer

Focus: building, deploying, configuring, and observing applications.

CKAD domain (approx.) Guide chapters Depth vs exam
Application design & build (multi-container, init/sidecar, Jobs/CronJobs, volumes) 01-core-workloads/01, 07; 03-config-and-storage/03 [≈]
Application deployment (Deployments, rolling updates/rollback, Helm, blue-green/canary) 01-core-workloads/04, 08; 07-delivery/01, 02 [deeper] (Helm/Kustomize/GitOps beyond exam)
Application observability & maintenance (probes, logs, kubectl debug, deprecated APIs) 01-core-workloads/02; 06-production-readiness/02; 08-day-2-operations/03; appendix C [≈]
Application environment, config & security (ConfigMaps/Secrets, SA, resources, securityContext) 03-config-and-storage/01, 02; 01-core-workloads/03; 05-security/01, 02 [≈]
Services & networking (Services, NetworkPolicy basics) 02-networking/02, 03, 04, 06 [≈]

CKAD-targeted order: Path 1 chapters 1–12, then 01-core-workloads/03 · 07 · 08 · 03-config-and-storage/03 · 05-security/01 · 02 · 02-networking/06capstone. Exam reality the guide under-drills: raw command speed under time pressure — practice the imperative→manifest workflow in Appendix A §2.

CKA — Certified Kubernetes Administrator

Focus: operating and administering a cluster.

CKA domain (approx.) Guide chapters Depth vs exam
Cluster architecture, installation & configuration (control plane, kubeadm, RBAC, upgrades) 00-foundations/03, 04, 05; 05-security/01; 08-day-2-operations/01 [lighter]kubeadm init/kubeadm upgrade and manual certificate/kubeconfig management are covered conceptually + as In production:; the exam wants them performed from scratch (rehearse on a kind/kubeadm cluster)
Workloads & scheduling (Deployments, scaling, scheduling, affinity/taints) 01-core-workloads/04; 04-scheduling/01, 02, 03; 06-production-readiness/04 [deeper]
Services & networking (Services, Ingress, Gateway, CoreDNS, NetworkPolicy, CNI) 02-networking/01, 02, 03, 04, 05, 06 [≈]/[deeper]
Storage (StorageClass, PV/PVC, access modes, reclaim) 03-config-and-storage/03, 04, 05 [≈]
Troubleshooting (cluster/node/app, logs, events, networking) 08-day-2-operations/03; 06-production-readiness/01, 02 [≈]
Cluster maintenance & backup (etcd backup/restore, drain/cordon, version skew) 08-day-2-operations/01, 02 [lighter] on bare-hand etcdctl snapshot restore (covered, but exam wants it fast & manual)

CKA-targeted order: Part 00 (all) → 01-core-workloads/04 → Part 02 (all) → Part 03 (all) → Part 04 (all) → 05-security/01 → Part 08 01/02/03capstone. Exam reality the guide under-drills: hand-running kubeadm upgrades and etcdctl snapshot save/restore on a vanilla cluster — rehearse those against a kind cluster using the official docs linked in Appendix D Part 08.

CKS — Certified Kubernetes Security Specialist

Focus: securing a cluster and its workloads (CKA is a prerequisite).

CKS domain (approx.) Guide chapters Depth vs exam
Cluster setup (NetworkPolicy, CIS benchmark, ingress TLS, restricting metadata/dashboard) 02-networking/06; 02-networking/04; 05-security/04 [≈]
Cluster hardening (RBAC least privilege, restrict API access, upgrade discipline) 05-security/01; 05-security/04; 08-day-2-operations/01 [deeper] (RBAC treated end-to-end)
System hardening (least-privilege OS, seccomp/AppArmor, reduce attack surface) 05-security/02; 00-foundations/02 (distroless) [≈]
Minimize microservice vulnerabilities (PSA, securityContext, secrets, mTLS concept, sandboxing) 05-security/02; 03-config-and-storage/02; 05-security/04 [≈]
Supply chain security (image footprint, signing, scanning, SBOM, admission/Kyverno) 05-security/03; 00-foundations/02; 07-delivery/03 [deeper]
Monitoring, logging & runtime security (audit logs, behavioral analytics, threat detection) 05-security/04 (audit logging) + 06-production-readiness/02 [lighter] — metrics ([06/01]) and tracing ([06/03]) are not CKS-tested; focus on audit logs (05-security/04) + the official-curriculum Falco/runtime supplement (note-only in this guide)

CKS-targeted order (do CKA first): Part 05 (all, in order) → 02-networking/0600-foundations/0207-delivery/0306-production-readiness/01/02capstone §(h) "Verify security posture". Exam reality the guide under-drills: a runtime-security sensor (e.g. Falco) and gVisor/kata sandboxing hands-on — the guide explains the threat model and PSA/securityContext thoroughly but treats runtime IDS as note-only; supplement from the official CKS curriculum.


Path 4 — Platform / SRE / Ops track (Parts 04–08 emphasis)

For engineers who already run apps and need the operate-at-scale skills: scheduling, security posture, observability, delivery automation, and day-2. Assumes Parts 00–03 fluency (skim them; do not skip 00-foundations/06 — SSA/GitOps hangs off it).

# Chapter Why here
1 00-foundations/04 — Control plane deep dive You operate these components; know them cold
2 00-foundations/05 — Node components kubelet/CRI/CNI — where node incidents live
3 04-scheduling/01 — Scheduler and nodes Placement is a platform lever
4 04-scheduling/02 — Affinity, taints, topology Spread, isolation, node pools
5 04-scheduling/03 — Priority and preemption Capacity contention policy
6 05-security/01 — Authn, authz, RBAC Least-privilege is platform table stakes
7 05-security/02 — Pod security PSA restricted as the cluster floor
8 05-security/04 — Secrets and cluster hardening Encryption, audit, hardening
9 06-production-readiness/01 — Observability: metrics You cannot operate what you cannot see
10 06-production-readiness/02 — Logging Centralized logs
11 06-production-readiness/03 — Tracing Cross-service latency/failure
12 06-production-readiness/04 — Autoscaling HPA/VPA/KEDA/Cluster Autoscaler
13 06-production-readiness/05 — Reliability and disruptions PDBs, SLOs, disruption budgets
14 06-production-readiness/06 — Capacity and cost Right-sizing, bin-packing, FinOps
15 07-delivery/01 — Packaging with Helm The platform's release unit
16 07-delivery/02 — Packaging with Kustomize Env overlays without a DSL
17 07-delivery/03 — CI/CD pipeline Build→scan→push→deploy
18 07-delivery/04 — GitOps with Argo CD The cluster is the repo
19 07-delivery/05 — Progressive delivery Metric-gated rollouts
20 08-day-2-operations/01 — Cluster lifecycle Upgrades, version skew, node maintenance
21 08-day-2-operations/02 — Backup and DR etcd/PV backup, a rehearsed restore
22 08-day-2-operations/03 — Troubleshooting playbook The incident method + distroless debug
23 08-day-2-operations/04 — Multi-tenancy and namespaces Many teams, one cluster
24 08-day-2-operations/05 — Operators and CRDs Automating stateful day-2
25 09-end-to-end-bookstore/01 — Bookstore end-to-end The whole platform exercised: GitOps + observe + scale + harden + DR

Why this order: internals → placement → security floor → observability → scale/reliability/cost → delivery automation → day-2 → compose it all. Skips (be honest): application-author depth (multi-container patterns, Jobs/CronJobs detail, app config ergonomics) — that is Path 1 / CKAD; this track assumes you can already write a workload and focuses on running fleets of them.


Choosing a path

 Just need to deploy something soon ............ Path 1 (Fast track)
 Learning Kubernetes properly, end to end ...... Path 2 (Zero-to-production) ← the default
 Sitting CKAD ................................. Path 3 → CKAD map
 Sitting CKA .................................. Path 3 → CKA map
 Sitting CKS (after CKA) ...................... Path 3 → CKS map
 Platform/SRE building & operating clusters ... Path 4 (Platform/SRE/ops)

Every path finishes at the end-to-end Bookstore — the Bookstore stood up from zero with GitOps, observed, autoscaled, hardened, and DR-drilled. If you can do that end-to-end and explain why each step is there, the path worked. Readers who then take Parts 10 - 12 (cloud / advanced production / ML) close on the Bookstore Platform v2 grand capstone — the production reality: N tenants across three regions with the day-2 runbook, on-call playbook, DR drill, and chaos game-day that turn the platform's capabilities into a working system.


See also: Appendix A — kubectl cheatsheet · Appendix B — Glossary · Appendix C — YAML & API conventions · Appendix D — Further reading (the book chapters behind each Part, and the CNCF-landscape pointer). Exam curricula (authoritative, current): https://github.com/cncf/curriculum.