Docker vs. Kubernetes: Runtime Engines vs. Cluster Orchestrators

Quick Answer: Docker builds and runs containers on a single machine; Kubernetes orchestrates containers across many machines. They solve different problems and are complementary, not competing.

Detailed Answer:

Docker is a container runtime and build tool — it provides the docker build, docker run, and docker push workflow for creating and running containers locally. Kubernetes is an orchestration platform — it decides where containers run across a cluster, how many run, and what happens when they fail. In 2020, Kubernetes deprecated Docker as its runtime in favor of the Container Runtime Interface (CRI), and most clusters now use containerd or CRI-O directly. This caused confusion, but it doesn't affect developers: you still build images with Docker, push to a registry, and reference them in Kubernetes manifests — K8s just uses containerd to pull and run them.

Real-World Implication for You:

On EKS today, nodes run containerd, not Docker. If you SSH onto a node, docker ps won't work — use crictl ps instead.

# On an EKS node — use crictl, not docker
sudo crictl ps
sudo crictl images
sudo crictl logs <container-id>

Key Takeaway: Docker builds images; Kubernetes runs them at scale. They're separate tools that work together in the same pipeline.