Kubernetes Architecture & Chaos
How Kubernetes actually works under the hood, from API server request lifecycle to etcd Raft to the scheduler framework, paired with chaos engineering reasoning that turns architectural knowledge into operational confidence. Built for the interview question "walk me through what happens when you create a pod" and the production question "how do we test resilience without breaking customers?"
What you'll learn
Curriculum
12 modules · 36 lessonsThe Kubernetes Architecture Mental Model
The clean separation of control plane and data plane, the API server as universal bus, and reconciliation loops as the universal pattern.
API Server Internals
Request lifecycle, admission controller pipeline, and the performance characteristics that determine how the apiserver scales.
etcd Architecture
Raft consensus mechanics, the watch stream model, and the failure modes that take clusters offline.
The Scheduler
The scheduling framework, the predicate/priority pipeline, and the design choices that make the scheduler work past 10,000 pods.
Controllers and Operators
The controller pattern that makes Kubernetes a platform, the built-in controllers that run your workloads, and the operator pattern for everything else.
kubelet Deep Dive
The pod lifecycle from kubelet's perspective, the CRI shim model, and the kubelet failure modes that take nodes offline.
Networking Internals
How Services, CNI, and DNS actually work at the iptables/eBPF layer, with the scaling characteristics that matter at thousands of services.
Failure Domains
What survives which kind of failure. The bounded-impact reasoning that turns architecture knowledge into resilience design.
Chaos Engineering Fundamentals
The discipline of breaking things on purpose, why it works, and the hypothesis-driven approach that separates chaos engineering from random pod-killing.
Chaos Engineering for Kubernetes
The concrete experiments at the pod, node, and cluster level: what each one tests, what bugs it surfaces, and how to scope it.
Tools in Practice
The tooling landscape, the game day playbook, and the CI/CD integration that turns chaos from quarterly drill into continuous practice.
Interview-Ready Architecture Reasoning
The big interview questions answered with the reasoning frameworks the rest of the course built up.
About the Author
Sharon Sahadevan
AI Infrastructure Engineer
Building production GPU clusters on Kubernetes. H100s, large-scale model serving, and end-to-end ML infrastructure across Azure and AWS.
10+ years designing cloud-native platforms with deep expertise in Kubernetes orchestration, GitOps (Argo CD), Terraform, and MLOps pipelines for LLM deployment.
Author of KubeNatives, a weekly newsletter read by 3,000+ DevOps and ML engineers for production insights on K8s internals, GPU scheduling, and model-serving patterns.