Kubernetes Architecture Overview
You'll learn how Kubernetes' control plane and worker nodes collaborate to deploy and manage containers at scale.
The Big Picture
Kubernetes is a distributed system with two main parts: the control plane (the brain) and worker nodes (the workforce). Think of it like a construction company: the control plane is the project management office that decides what to build and where, while worker nodes are the construction sites where the actual building happens.
Control Plane Components
The control plane runs several coordinated processes:
API Server – The central hub. Every request (deploy a container, scale up, check status) goes through the API server. It's the only component that talks directly to the datastore (etcd).
Scheduler – Watches for newly created Pods (groups of containers) that have no assigned node, then selects an appropriate worker node based on resource availability and constraints.
Controller Manager – Runs multiple controllers that continuously monitor the cluster's state. If you requested 3 replicas of a container but only 2 are running, a controller detects this and creates the third.
Worker Node Components
Each worker node runs:
Kubelet – An agent that ensures containers are running as specified. It receives Pod specifications from the API server and manages the containers on its node using the container runtime.
Container Runtime – The software that actually runs containers (like Docker or containerd).
How They Work Together
When you deploy an application: API Server accepts your request → Scheduler picks nodes → Kubelet on each node pulls images and starts containers → Controller Manager ensures your desired state is maintained → API Server provides status updates. It's a continuous reconciliation loop.
Key Takeaway: Kubernetes separates decision-making (control plane) from execution (worker nodes), with the API server as the central coordinator ensuring your declared desired state matches reality.