Once you have more than a handful of containers (03-docker), you need something to schedule them across machines, restart them when they die, scale them, and route traffic to them. That something is an orchestrator — and in practice it means Kubernetes. Part of the DevOps curriculum.

[reference] — covered from roadmap.sh and primary docs; the honest next step is running a real cluster (even a local kind/minikube one).

Reading order

  1. Kubernetes — [Advanced] — the architecture (control plane vs nodes), the core objects (pods, deployments, services, ingress), the declarative model, networking, storage, and config/secrets
  2. The Orchestration Landscape — [Advanced] — Docker Swarm, the managed offerings (GKE/EKS/AKS), OpenShift, and Nomad — what each trades off against raw Kubernetes
  • Docker — the containers this section orchestrates
  • CI/CD — how images get built and deployed into a cluster
  • GitOps — declaratively driving a cluster from git
  • Service Mesh — the networking layer on top of k8s
↑ Back to DevOps