00: Kubernetes Architecture Overview

00: Kubernetes Architecture Overview

Overview

This exercise introduces the fundamental architecture of Kubernetes and how Azure Kubernetes Service (AKS) implements it as a managed platform. You will learn about control plane components, worker node components, and AKS-specific features before diving into hands-on exercises.

Objective

• Understand the high-level architecture of a Kubernetes cluster
• Learn how AKS manages the control plane on your behalf
• Familiarize yourself with key Kubernetes concepts and terminology
• Get oriented with the training map and exercise structure

Theory

Kubernetes Architecture

A Kubernetes cluster consists of two main parts: the control plane and worker nodes.

Control Plane Components

Worker Node Components

Cluster Architecture Diagram

graph TB
    subgraph ControlPlane["Control Plane (Managed by Microsoft in AKS)"]
        API[kube-apiserver]
        ETCD[etcd]
        SCHED[kube-scheduler]
        CM[kube-controller-manager]
        CCM[cloud-controller-manager]
    end

    subgraph WorkerNode1["Worker Node 1"]
        K1[kubelet]
        KP1[kube-proxy]
        CR1[containerd]
        P1[Pod A]
        P2[Pod B]
    end

    subgraph WorkerNode2["Worker Node 2"]
        K2[kubelet]
        KP2[kube-proxy]
        CR2[containerd]
        P3[Pod C]
        P4[Pod D]
    end

    API --> ETCD
    SCHED --> API
    CM --> API
    CCM --> API
    K1 --> API
    K2 --> API
    KP1 --> API
    KP2 --> API

AKS Specifics

Managed Control Plane

In AKS, Microsoft manages the entire control plane for you. This means:

• You do not have access to the control plane VMs or etcd directly.
• Microsoft handles upgrades, patching, and high availability of control plane components.
• You interact with the cluster exclusively through the kube-apiserver endpoint.
• The control plane is free in the Free tier; you only pay for worker node VMs.

AKS Pricing Tiers

Azure Linux 3.0 as Default Node OS

Starting in recent AKS versions, Azure Linux 3.0 (previously known as CBL-Mariner) is the default node operating system. Key points:

• Optimized for container workloads with a minimal footprint.
• Faster boot times and reduced attack surface compared to Ubuntu.
• Microsoft-maintained and security-patched.
• You can still choose Ubuntu if needed via --os-sku flag.

AKS Managed vs Customer-Managed Components

graph LR
    subgraph Microsoft["Managed by Microsoft"]
        direction TB
        CP[Control Plane]
        ETCD2[etcd]
        API2[API Server]
        UP[Upgrades & Patches]
        HA[High Availability]
    end

    subgraph Customer["Managed by Customer"]
        direction TB
        NP[Node Pools & VMs]
        APP[Application Workloads]
        NET[Virtual Network Config]
        MON[Monitoring & Logging]
        SEC[RBAC & Security Policies]
        REG[Container Registry]
    end

    Microsoft --- Customer

kubectl and kubeconfig

kubectl is the command-line tool for interacting with Kubernetes clusters. It communicates with the kube-apiserver using REST API calls.

kubeconfig is a configuration file (default location: ~/.kube/config) that contains:

• Clusters — API server addresses and CA certificates.
• Users — Authentication credentials (tokens, certificates, or Azure AD integration in AKS).
• Contexts — A combination of cluster + user + namespace. You switch contexts to target different clusters or namespaces.

# Download AKS credentials into your kubeconfig (or receive kubeconfig from instructor)
az aks get-credentials --resource-group <RG> --name <CLUSTER> --overwrite-existing

# View current context
kubectl config current-context

# List all contexts
kubectl config get-contexts

# Set default namespace for the current context
kubectl config set-context --current --namespace=student-XX

Training note: In this training, the instructor provides you with a kubeconfig for the shared cluster. Each participant works in their own pre-created namespace student-XX.

Component Communication Flow

sequenceDiagram
    participant User as kubectl / User
    participant API as kube-apiserver
    participant ETCD as etcd
    participant Sched as kube-scheduler
    participant Kubelet as kubelet (node)

    User->>API: kubectl apply -f pod.yaml
    API->>ETCD: Store Pod spec
    API->>Sched: Notify: unscheduled Pod
    Sched->>API: Assign Pod to Node
    API->>ETCD: Update Pod with node assignment
    API->>Kubelet: Pod scheduled to your node
    Kubelet->>Kubelet: Pull image, start container
    Kubelet->>API: Report Pod status: Running

Glossary of Terms

Training Map

This training consists of 11 modules and 31 exercises (numbered 00 through 30).

Summary

Kubernetes provides a robust platform for container orchestration, and AKS simplifies operations by managing the control plane. Throughout this training, you will progress from exploring the cluster architecture to deploying, configuring, scaling, and troubleshooting real workloads in your student-XX namespace.

Proceed to Exercise 01 to explore the AKS cluster hands-on.