Skip to content
Learni
View all tutorials
Cloud Computing

How to Deploy an App on Azure Kubernetes Service (AKS) in 2026

12 minBEGINNER
DevOpsAzureKubernetesConteneursAKS
Lire en français

Introduction

Azure Kubernetes Service (AKS) is Microsoft's managed service on Azure for orchestrating Kubernetes containers. In 2026, AKS simplifies cluster management by automating updates, scaling, and security, letting you focus on your applications. This beginner tutorial walks you through creating an AKS cluster, deploying a simple web app (Nginx), and exposing it via a LoadBalancer. Why does it matter? Kubernetes dominates container orchestration (80% of enterprises use it), and AKS cuts setup time by 70% compared to a vanilla cluster. By the end, you'll have a production-ready, horizontally scalable deployment. We use Azure CLI for infrastructure and kubectl for workloads, with complete, tested YAML manifests. Estimated time: 30 minutes. Ready to containerize?

Prerequisites

  • A free Azure account (create one at portal.azure.com)
  • Azure CLI installed (version 2.65+)
  • kubectl installed (version 1.31+)
  • Docker Desktop for local testing (optional)
  • Basic terminal and YAML knowledge

Install and Connect to Azure CLI

terminal
# Install Azure CLI (on Ubuntu/Debian, adapt for Windows/macOS)
curl -sL https://aka.ms/InstallAzureCLIDeb | sudo bash

# Check the installation
az --version

# Log in to Azure
az login

# Set the default subscription (replace with your ID)
az account set --subscription "your-subscription-id"

# List available locations
az account list-locations --output table

These commands install Azure CLI, authenticate your session, and set up the environment. Use az login to open a browser for authentication. Avoid errors by checking your subscription with az account show. On Windows, use the official PowerShell script.

Create a Resource Group

A resource group is a logical container for your Azure resources. It streamlines management, billing, and cascading deletion.

Create the Resource Group

terminal
# Create a resource group (West Europe region, adjust as needed)
az group create --name rg-aks-demo --location "westeurope"

# Verify creation
az group show --name rg-aks-demo --query location -o tsv

This creates rg-aks-demo in West Europe to minimize latency. Pick a region close to your users. Pitfall: Don't forget --location, or it fails. Use az group delete for cleanup.

Create the AKS Cluster

terminal
# Create an AKS cluster with 2 nodes (system nodepool)
az aks create \
  --resource-group rg-aks-demo \
  --name aks-demo-cluster \
  --node-count 2 \
  --enable-addons monitoring \
  --generate-ssh-keys \
  --node-vm-size Standard_D2_v2

# Wait 5-10 min, then check
az aks show --resource-group rg-aks-demo --name aks-demo-cluster --output table

Creates a managed cluster with monitoring enabled and auto-generated SSH keys. Standard_D2_v2 is cost-effective for demos (~$0.10/hour). Creation time: ~7 min. Avoid tiny VMs for real workloads.

Configure kubectl for AKS

kubectl is the Kubernetes CLI tool. Configure it to point to your AKS cluster.

Connect kubectl to the Cluster

terminal
# Get kubectl credentials
az aks get-credentials --resource-group rg-aks-demo --name aks-demo-cluster

# Verify connection
kubectl get nodes

# List namespaces
kubectl get ns

get-credentials updates ~/.kube/config. kubectl get nodes confirms 2 ready nodes. Pitfall: Multiple clusters? Check with kubectl config current-context.

Deploy the Nginx Application

nginx-deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx-deployment
  labels:
    app: nginx
spec:
  replicas: 3
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx:1.27-alpine
        ports:
        - containerPort: 80
        resources:
          requests:
            memory: "64Mi"
            cpu: "50m"
          limits:
            memory: "128Mi"
            cpu: "100m"
---
apiVersion: v1
kind: Service
metadata:
  name: nginx-service
spec:
  selector:
    app: nginx
  ports:
    - protocol: TCP
      port: 80
      targetPort: 80
  type: LoadBalancer

This manifest deploys 3 Nginx pods with resource requests/limits and a LoadBalancer Service for external access. Alpine image keeps it lightweight. Copy-paste into a file and apply.

Apply the Deployment

terminal
# Save the previous YAML to nginx-deployment.yaml

# Apply the manifest
kubectl apply -f nginx-deployment.yaml

# Check pods
kubectl get pods -l app=nginx

# Check service and get external IP
kubectl get svc nginx-service

# Test (wait for IP)
curl http://<EXTERNAL_IP>

apply -f creates the Deployment + Service atomically. get svc shows EXTERNAL-IP (~5 min provisioning). Scale with kubectl scale deployment nginx-deployment --replicas=5. Public access ready.

Scale and Monitor

terminal
# Scale horizontally
kubectl scale deployment nginx-deployment --replicas=5

# Auto-scaler (HPA)
kubectl autoscale deployment nginx-deployment --cpu-percent=50 --min=3 --max=10

# Pod logs
kubectl logs deployment/nginx-deployment

# Describe a pod for debugging
kubectl describe pod <POD_NAME>

Manual scaling or HPA based on CPU. Logs and describe help with debugging. Pitfall: No resources defined? HPA ignores it. Check with kubectl get hpa.

Best Practices

  • Always set resource requests/limits: Prevents noise and optimizes costs.
  • Use namespaces: kubectl create ns prod to isolate environments.
  • Enable Azure Monitor: Built into AKS, auto-dashboards in portal.azure.com.
  • Secrets for credentials: kubectl create secret generic db-pass --from-literal=password=secret.
  • Helm for complex apps: Move beyond raw YAML after this beginner tutorial.

Common Errors to Avoid

  • Cluster not ready: Wait for kubectl get nodes all Ready before deploying.
  • LoadBalancer IP stuck : Check quotas/subnet in Azure Portal.
  • Lost kubectl context: Refresh with az aks get-credentials.
  • Hidden costs: ~$0.10/hour/node + LoadBalancer; delete with az group delete --name rg-aks-demo --yes.

Next Steps

Master advanced AKS: Learni DevOps & Kubernetes Training. Resources: Official AKS Docs, Kubernetes.io Basics. Try NGINX Ingress or CI/CD with GitHub Actions.

How to Deploy a Web App on Azure AKS in 2026 | Learni