Production Expert track · Modules 0 & 1 free

Docker & Kubernetes,
taught on a real cluster
failures and all.

Every terminal in this course is a genuine captured session from a live four-node Kubernetes cluster. The OOMKill, the throttled CPU, the plaintext Secret in etcd — all real, triggered on purpose, and kept in. 13 modules of black/red/blue cinematic teaching.

Enroll now Watch free preview $34.99once · lifetime
100% real captured output Interactive lab + runnable cluster Instant lifetime access
13modules
~5.5 hcinematic video
99real captured sessions
25teaching animations
Why this course is different

Most courses teach "what is a Pod."
This one shows you the cluster failing — and recovering.

The whole course runs on one realistic e-commerce stack on a live four-node cluster. We break it on purpose and read the real evidence, because the evidence is where the learning is.

Real, not illustrated

Every terminal is a captured session from a live cluster — the 97% CPU-throttle counter, the exit-code 137, the plaintext Secret in etcd, the scheduler explaining a Pending pod node by node.

Animation-first

Concepts are shown happening — namespaces slicing one kernel, cgroups throttling a container, the scheduler filtering nodes, a rolling update with zero dropped requests. 25 cinematic animations in a clean black/red/blue style.

Genuinely runnable

An interactive in-browser lab mirrors the real cluster, and shipped scripts build the exact same four-node cluster and app on any Ubuntu box with Docker. Same commands, same output, same failures.

Captured, not claimed

Real numbers from the real cluster

These aren't slide claims — they came off a live Kubernetes v1.33 cluster and are replayed verbatim in the course.

1.18 GB → 119 MB
An image cut with a multi-stage build — and another from 352 MB to 5.59 MB into a scratch image with one layer.
230 / 236
CPU periods a container spent throttled — 97% — while its utilisation graph looked calm. Read straight from the cgroup counter.
exit 137
A container OOMKilled for exceeding 64 MB by a few megabytes — triggered live, with the restart hiding the crash.
plaintext
A database password read straight out of etcd, unencrypted by default — then fixed with encryption at rest.
One application, the whole way

You carry the same stack across every module

Seven workloads — enough to produce every real failure mode, small enough to hold in your head. No fresh toy example each lesson.

WorkloadKindWhat it teaches
ordersDeployment + HPACPU throttling, OOMKilled, autoscaling, graceful shutdown from a scratch image
catalogDeployment ×3Probes, rolling updates, ConfigMap/Secret provenance, topology spread
postgresStatefulSet + PVCStable identity, headless DNS, surviving pod deletion with zero data loss
gatewayNodePortEdge routing, Ingress, the only externally reachable component
NetworkPolicy set5 policiesDefault-deny, allow-lists, and the DNS mistake everybody makes once
Curriculum · 13 modules

From "what is a container" to etcd restore

Beginner to advanced, in order, each module building on the last — mapped to CKA, CKAD and CKS domains, plus production practice no exam tests.

0

Introduction — Docker, Containers & Kubernetes, PreciselyFREE

Five things called "Docker", what Kubernetes is and isn't, when not to use it.

1

Why Containers Exist — and What One Actually IsFREE

A container is a process the kernel lies to — proven on a live machine.

2

The Kernel Underneath: Namespaces & cgroups

Build a container by hand; watch cgroups throttle and OOM-kill one.

3

Container Runtimes: Docker, containerd, runc, CRI

What removing dockershim did — and did not — change.

4

Images, Layers & Builds You Can Defend

1.18 GB → 119 MB, measured. Caching, copy-on-write, the PID 1 trap.

5

Container Networking & Persistent Data

veth, bridges, NAT — and why the writable layer is wrong for a database.

6

Kubernetes Architecture — API Server, etcd, Controllers

Read an object out of etcd; find a Secret sitting there in plaintext.

7

Pods, Controllers & Workloads in Depth

The pause container, sidecars, Deployments, StatefulSets, Jobs.

8

Scheduling, Resources, QoS & Autoscaling

Why a pod is Pending, verbatim. Throttling, OOM, HPA, PDBs, drain.

9

Kubernetes Networking: Services, DNS, Ingress, NetworkPolicy

Follow one packet across nodes; default-deny and watch it hold.

10

Storage, Configuration & Stateful Workloads

Delete the database pod, lose nothing. The subPath trap.

11

Security: RBAC, Pod Security, Secrets & Supply Chain

Escalate from one pod to cluster-admin — then close every hole.

12

Day-2: Observability, Troubleshooting, Upgrades & GitOps

A method that works under pressure; the ten symptoms and their causes.

Lifetime access

One payment. Everything, forever.

$34.99
one-time · lifetime access · no subscription
  • All 13 modules · ~5.5 hours of animated video
  • 99 real captured cluster sessions, replayed verbatim
  • Interactive lab + scripts to run the real cluster
  • Shuffled assessment mapped to CKA / CKAD / CKS
  • Word-level karaoke subtitles · resume anywhere
  • Free forever: Modules 0 & 1 — try before you buy
Enroll now — $34.99
PayPal Cards UPI · India ₹ via Razorpay
or watch Modules 0 & 1 free first →
Questions

Before you enroll

Can I try it before buying?
Yes — Modules 0 and 1 are completely free, no account and no card. Module 0 makes every term exact (and says when not to use Kubernetes); Module 1 proves what a container really is on a live machine. About 55 minutes, free. Start watching →
What makes this different from other Kubernetes courses?
Every terminal is a real captured session from a live four-node cluster — not a slide, not a re-typed example. The failures are triggered on purpose and kept in: a container OOMKilled with exit code 137, a CPU throttled 230 of 236 periods, a Secret read as plaintext out of etcd, and the scheduler explaining a Pending pod node by node.
Who is it for?
Developers, DevOps and SRE engineers, platform engineers and backend developers moving to containers and Kubernetes. It starts from what a container actually is and reaches etcd restore, RBAC hardening, encryption at rest and GitOps — beginner to advanced, in one continuous thread.
Is there real hands-on practice?
Yes. An interactive in-browser lab mirrors the real cluster, and shipped scripts build the exact same four-node cluster and application on any Ubuntu box with Docker — so you can run every command yourself and see the same output.
How do I pay, and what do I get?
A one-time $34.99 (in India, the equivalent in ₹) via Razorpay (cards / UPI / Apple Pay / Google Pay) or PayPal gives you lifetime access to all 13 modules, ~5.5 hours of video, the interactive lab and the assessment. Access is instant after checkout.

Stop watching slides. Watch the cluster.

Modules 0 & 1 are free. If it's the clearest Kubernetes teaching you've seen, the rest is $34.99, once, forever.

Enroll — $34.99 lifetime Watch free preview