Kubernetes Fundamentals #2
Kubernetes High-Level Architecture
Kubernetes has two major parts that constantly work together: Control Plane and Worker Nodes.
Control Plane
The Control Plane manages everything in the cluster.
It decides where workloads run, keeps checking health and ensures the desired state you define is always maintained.
It is made up of several components, each playing a key role:
1. API Server
- Acts as the front door of Kubernetes
- Receives commands (kubectl, UI, automation)
- Validates and updates the cluster’s state
2. etcd
- It's a reliable key-value database
- Stores all cluster configuration and desired state
- The memory of Kubernetes
3. Scheduler
- Decides which worker node will run each Pod
- Chooses the best fit based on resources and rules
4. Controller Manager
- Continuously checks the cluster state
- Fixes any mismatches automatically (for example, if a Pod dies, it creates a new one)
5. Cloud Controller Manager
Handles cloud-specific tasks like load balancers, storage and node lifecycle (in cloud environments)
Together, these components plan, manage and heal the system.
Worker Nodes
These are the machines that run your application Pods.
Each node includes:
1. kubelet - talks to the Control Plane and ensures containers are running correctly
2. container runtime - pulls images and starts containers (like containerd)
3. networking component - ensures Pods can communicate across the cluster
Worker nodes are the execution layer that follows the Control Plane’s instructions.
Networking — One Cluster, One Network
Kubernetes gives every Pod its own IP address.
Applications communicate with each other without caring where Pods are running.
Even when Pods restart or move to another node, the network layer keeps communication seamless.
This makes the entire infrastructure feel like a single connected system.
In Short
Control Plane = decides and manages
Worker Nodes = run the applications
Networking = keeps everything talking smoothly
#Kubernetes #devopsdeveloper #SoftwareEngineering