Kubernetes (K8s) Architecture: Control Plane & Worker Nodes

Lab 15 of 20 · K8s Architecture · 5 hands-on activities · bilingual (English + සිංහල)

පිටුව 15: Kubernetes (K8s) ගෘහ නිර්මාණ ශිල්පය: Control Plane සහ Worker Nodes

Docker තනියම මදි වුණේ ඇයි? Kubernetes Cluster එකේ මොළය (Control Plane: API Server, etcd, Scheduler) සහ අතපය (Worker Nodes: Kubelet, Kube-proxy).

Architectural Foundation & Systems Diagram

Docker is incredible for running containers on a single laptop or server. But when you have 5,000 containers running across 200 cloud servers, you need an automated orchestrator: Kubernetes (K8s).

• 🔄 Declarative Desired State — In K8s, you never say "run this command now". You write a YAML declaration stating: "I desire 5 copies of Nginx running". Kubernetes continuously loops to ensure actual state matches desired state!

• 🛡️ Self-Healing Automation — If a worker node physical hardware crashes, or a container dies from an OOM error, the Kubernetes Controller Manager detects the mismatch and instantly reschedules new pods on surviving nodes!

• 🎮 kubectl CLI Tool — The universal command-line interface for communicating with the kube-apiserver. Every DevOps engineer and SRE must master kubectl.

Step-by-step practical lab

Step 01: Verify Kubernetes Cluster Control Plane Health

සිංහලෙන් — Kubernetes Cluster එකේ Control Plane (Master API සහ CoreDNS) තත්ත්වය පරීක්ෂා කරමු.

Query cluster control plane endpoints to confirm kube-apiserver and CoreDNS services are responsive.

kubectl cluster-info

• cluster-info : Kubernetes master API server සහ cluster add-on endpoints පෙන්වයි.

Expected output

Kubernetes control plane is running at https://172.31.0.1:6443
CoreDNS is running at https://172.31.0.1:6443/api/v1/namespaces/kube-system/services/kube-dns:dns/proxy

Step 02: Query Cluster Worker Nodes & OS Kernels

සිංහලෙන් — Cluster එකේ ඇති Worker Nodes, Ready status සහ Kernel අනුවාද බලමු.

Display all cluster worker nodes with Ready status, OS image, kernel release, and internal IP address.

kubectl get nodes -o wide

• -o wide : Internal IPs, OS Images සහ Kernel Versions සහිත පුළුල් output එකක් පෙන්වයි.

Expected output

NAME           STATUS   ROLES           AGE   VERSION   INTERNAL-IP    OS-IMAGE             KERNEL-VERSION
k8s-master-01  Ready    control-plane   14d   v1.30.2   172.31.10.10   Ubuntu 24.04 LTS     6.8.0-45-generic
k8s-worker-01  Ready              14d   v1.30.2   172.31.10.11   Ubuntu 24.04 LTS     6.8.0-45-generic
k8s-worker-02  Ready              14d   v1.30.2   172.31.10.12   Ubuntu 24.04 LTS     6.8.0-45-generic

Hands-on practice activities

Work through these yourself — tick them off as you go.

• Activity 71: Verify Kubernetes Cluster Health (Drill) — Query cluster control plane endpoints (Kubernetes master API and CoreDNS).
• Activity 72: Query Cluster Worker Nodes Status (Drill) — Display all cluster worker nodes with Ready status, OS image, kernel, and internal IP.
• Activity 73: The Declarative Reconciliation Loop (Checkpoint) — Understand that K8s controllers run in an infinite loop: Observe -> Analyze Difference -> Act to reach desired state.
• Activity 74: etcd Disaster Recovery Principle (Checkpoint) — etcd stores the complete cluster state. Always automate automated etcd snapshots before performing cluster upgrades!
• Activity 75: Kubernetes Database Component Quiz (Quiz) — Which component acts as the consistent, distributed key-value store for all Kubernetes cluster state? Answer: etcd.

Keep going: ← Lab 14 · Lab 16 → · Track your progress in the Self Guide