Master Kubernetes: Core Concepts, Architecture, and Hands‑On Demo

What Is Kubernetes?

Kubernetes is an industrial‑grade container orchestration platform. Its name derives from the Greek word for “helmsman” or “pilot”, often abbreviated as K8s. It is designed to automate deployment, scaling, and management of containerized applications.

Key Functions

Service discovery and load balancing.

Scheduling containers onto appropriate nodes.

Automatic container self‑healing.

Automated rollout, rollback, and secret configuration management.

Batch job execution.

Horizontal pod autoscaling.

Three Illustrative Scenarios

1. Scheduling

The scheduler places a newly submitted container onto a node with sufficient CPU and memory resources, performing a “placement” operation.

2. Self‑Healing

Kubernetes monitors node health; when a node fails, the affected pods are automatically migrated to healthy nodes.

3. Horizontal Scaling

Based on CPU utilization or response time, the system can increase the number of pod replicas, distributing load across the new pods.

Architecture Overview

Kubernetes follows a two‑layer, client‑server model. The control plane (Master) consists of API Server, Controller Manager, Scheduler, and etcd, while each worker node runs kubelet, kube-proxy, a container runtime, and optional storage/network plugins.

API Server is the central entry point for all components. Controllers maintain desired state, Scheduler decides pod placement, and etcd stores the cluster state with high availability.

Nodes run pods via kubelet, which communicates with the API Server and delegates container execution to the container runtime. Network and storage are provided by plug‑in components such as kube‑proxy, CNI, and CSI.

Core Concepts and API

Pod

A pod is the smallest deployable unit, encapsulating one or more containers that share network and storage.

Volume

Volumes abstract storage resources and can be backed by local disks, distributed systems (e.g., Ceph, GlusterFS), or cloud services (e.g., AWS EBS, GCP Persistent Disk).

Deployment

Deployments manage a set of pod replicas, handling rollout, rollback, and scaling through the controller.

Service

Services provide a stable virtual IP that load‑balances traffic across a group of pods, supporting ClusterIP, NodePort, and LoadBalancer types.

Namespace

Namespaces isolate resources within a cluster, enabling separate environments or business units.

API Details

Kubernetes API uses HTTP+JSON (or YAML). Resources are accessed via paths such as /api/v1/namespaces/{namespace}/pods/{podName}. Objects contain metadata (name, labels, annotations), spec (desired state), and status (current state). Labels enable selector‑based queries similar to SQL WHERE clauses.

Hands‑On Demo with Minikube

Install a local sandbox using VirtualBox and Minikube ( minikube start --vm-driver=virtualbox). Verify the cluster status with kubectl get nodes, then create an nginx Deployment, perform a rolling update, and scale the replica count from 2 to 4. Finally, delete the Deployment to return the cluster to a clean state.

Conclusion

The article covered Kubernetes core concepts, architecture, API fundamentals, and a practical demo that demonstrates scheduling, self‑healing, and horizontal scaling.