Kubernetes is one of the most powerful container orchestration platforms, but its architecture can feel overwhelming at first. Let’s break it down into Master (Control Plane) and Worker Nodes so it’s easier to understand.
🧠 Kubernetes Master Components (Control Plane)
The Control Plane is the brain of your Kubernetes cluster. It manages cluster state, makes scheduling decisions, and responds to events.
🔹 kube-apiserver
- Acts as the front end of the Kubernetes control plane.
- Exposes the Kubernetes API, which is how everything talks to Kubernetes.
- Who interacts with it?
- Users (via kubectl or API calls)
- Other master components (scheduler, controller manager, etcd)
- Worker node components like Kubelet
🔹 etcd
- A consistent and highly available key-value store.
- Stores all cluster data: configs, state of master and worker nodes, secrets, etc.
- Think of it as Kubernetes’ source of truth.
🔹 kube-scheduler
- Responsible for distributing Pods across nodes.
- Watches for new Pods with no assigned node and picks a suitable node for them.
🔹 kube-controller-manager
Runs different controllers to keep the cluster healthy:
- Node Controller → Detects & reacts when nodes go down.
- Replication Controller → Ensures the right number of Pods are running.
- Endpoints Controller → Joins Services & Pods by updating Endpoints objects.
- Service Account & Token Controller → Creates default accounts & API access for new namespaces.
Basically, controllers are the “autopilot” of Kubernetes.
🔹 cloud-controller-manager
- Embeds cloud-specific control logic (only runs in cloud environments).
- On-prem clusters don’t have this.
- Responsibilities include:
- Node Controller → Checks with cloud provider if a node is deleted.
- Route Controller → Sets up routes in the cloud infrastructure.
- Service Controller → Manages cloud load balancers.
- Other controllers depend on your cloud provider’s Kubernetes integration.
⚙️ Kubernetes Worker Node Components
Worker nodes are where your applications actually run. Each worker node has components that keep Pods running and make networking possible.
🔹 Container Runtime
- The underlying software that runs containers.
- In GKE, the default is containerd.
- Other options: Ubuntu with containerd, Ubuntu with Docker, Windows nodes.
🔹 Kubelet
- The agent that runs on every node.
- Ensures containers are running in Pods as expected.
- Talks with the kube-apiserver and reports node status.
🔹 Kube-Proxy
- A network proxy that runs on each node.
- Maintains network rules to allow traffic in/out of Pods.
- Ensures services inside the cluster are accessible from inside and outside.
🚀 Wrapping Up
Kubernetes architecture is built around a Control Plane (Masters) that makes decisions, and Worker Nodes that run workloads.
- Control Plane = brains (API server, etcd, scheduler, controllers).
- Worker Nodes = muscles (container runtime, kubelet, kube-proxy).
Together, they make Kubernetes a self-healing, scalable, and powerful orchestration system for running your applications.
🌟 Thanks for reading! If this post added value, a like ❤️, follow, or share would encourage me to keep creating more content.
— Latchu | Senior DevOps & Cloud Engineer
☁️ AWS | GCP | ☸️ Kubernetes | 🔐 Security | ⚡ Automation
📌 Sharing hands-on guides, best practices & real-world cloud solutions
Top comments (0)