Skip to content

Efficiency🔗

Building on Essentials

Essentials gave you the primitives and the objects you already touch as an app dev. Efficiency is where the platform underneath them stops being a black box — the runtime that actually runs your containers, the scheduler that places them, and the networking that gets real traffic to them.

Essentials treated kubectl apply as something that just works once you understand the objects. Efficiency asks the next question: what makes it work? kubelet doesn't run your container directly — it hands the job to a runtime through a well-defined interface. The scheduler doesn't place Pods randomly — it filters and scores every Node before picking one. And exposing a Service to the internet is a chain of independent, composable pieces, not a single Ingress rule.

This tier treats you as the platform engineer now responsible for these systems, not just a consumer of them.

Start Here: Architecture & Scheduling🔗

The mechanics underneath every Pod you've already deployed.

  • The CRI

    How kubelet actually talks to containerd and CRI-O — and the dockershim history behind why this interface exists at all.

  • The Scheduler

    Filtering and scoring: how a Pod picks a Node, and why affinity should target a type of Node, never one specific Node.

Putting It on the Internet🔗

The chain of pieces between a running Pod and a real domain name.

  • Gateway API

    The standard front door — Traefik as the Gateway API implementation, replacing the aging Ingress model.

  • cert-manager

    Certificates as cluster resources — automated issuance and renewal, no more manually copied .pem files.

  • external-dns

    Pointing your domain at the cluster automatically, kept in sync as Services and Ingresses change.

  • Advanced Workloads (coming soon)

    StatefulSets for stable identity, DaemonSets for node-level agents, and the rollout controls that keep them healthy.

What You'll Take Away🔗

By the end of Efficiency you'll be able to:

  • Trace a container from kubectl apply through kubelet, the CRI, and the runtime to a running process
  • Explain why a Pod landed on the Node it did — and predict where the next one will go
  • Stand up a real front door for a cluster app: routing, TLS, and DNS, all reconciled automatically

What's Next?🔗

Start with The CRI — the layer between the object you applied and the process actually running.

After Efficiency, the Mastery tier takes you into production operations: storage, scheduling depth, security, and observability for clusters you're responsible for.