Azure Kubernetes Service (AKS)

Managed Kubernetes where Azure handles the control plane for free, with integrated networking, ACR, monitoring, and Azure AD RBAC.

AzureAdvanced

Why

AKS removes the operational burden of running the Kubernetes control plane (API server, etcd, scheduler) — Microsoft handles upgrades, HA, and scaling for free. You only pay for the agent (node) VMs.

It integrates natively with Azure AD for auth, ACR for images, and Monitor for logs.

When

Use AKS for microservices architectures, multi-container apps, or when you need Kubernetes features (HPA, Istio, operators). Prefer Container Apps for serverless K8s without node management, and App Service for single-container web apps.

Use AKS only if you have K8s expertise — it still requires node pool and workload management.

How

1. Create a resource group. 2. Create AKS with az aks create — choose node count, VM size, and enable Azure AD integration and monitoring. 3. Get credentials with az aks get-credentials. 4. Connect ACR so AKS can pull images without imagePullSecrets. 5. Deploy with kubectl apply or Helm. 6. Configure ingress (Application Gateway ingress controller or NGINX). 7. Set up HPA and cluster autoscaler. ```bash az aks create --name myaks --resource-group rg --node-count 3 --node-vm-size Standard_D2s_v5 --enable-managed-identity --enable-azure-rbac az aks get-credentials --name myaks --resource-group rg az aks update --name myaks --resource-group rg --attach-acr myacr ```

Example

Deploy a web app with a HorizontalPodAutoscaler scaling on CPU: ```yaml apiVersion: apps/v1 kind: Deployment metadata: { name: web } spec: replicas: 2 selector: { matchLabels: { app: web } } template: metadata: { labels: { app: web } } spec: containers: - name: web image: myacr.azurecr.io/web:v1 resources: { requests: { cpu: 100m }, limits: { cpu: 500m } } ports: [{ containerPort: 80 }] --- apiVersion: autoscaling/v2 kind: HorizontalPodAutoscaler metadata: { name: web-hpa } spec: scaleTargetRef: { apiVersion: apps/v1, kind: Deployment, name: web } minReplicas: 2 maxReplicas: 10 metrics: - type: Resource resource: { name: cpu, target: { type: Utilization, averageUtilization: 70 } } ```

Common Problems

⚠Pods stuck in ImagePullBackOff — ACR is not attached or the identity lacks AcrPull role; run az aks update --attach-acr or assign the role to the kubelet identity.
⚠Cannot connect to kubectl — credentials not merged or expired; run az aks get-credentials --overwrite-existing.
⚠Nodes not scaling — cluster autoscaler is not enabled; add --enable-cluster-autoscaler --min-count 2 --max-count 10 at create or update time.

Best Practices

✓Use a system node pool (Linux, for system pods) separate from user node pools, and enable cluster autoscaler on user pools only.
✓Enable Azure AD integration and Azure RBAC so kubectl access is governed by Entra ID groups, not shared kubeconfig files.
✓Use the Application Gateway ingress controller for production — it provides WAF, TLS termination, and cookie affinity without a separate NGINX deployment.

FAQ

What is Azure Kubernetes Service (AKS) and what problem does it solve?

Managed Kubernetes where Azure handles the control plane for free, with integrated networking, ACR, monitoring, and Azure AD RBAC. It lets developers deploy, store, and scale without self-managing infrastructure — a core reference for cloud service selection.

How do I get started with Azure Kubernetes Service (AKS) in 5 minutes?

Two steps to run it: (1) Create a resource group.; (2) Create AKS with az aks create — choose node count, VM size, and enable Azure AD integration and monitoring.. Validate on the free tier first, then wire into production.

What is the #1 pitfall with Azure Kubernetes Service (AKS)?

The most frequent issue is: Pods stuck in ImagePullBackOff — ACR is not attached or the identity lacks AcrPull role; run az aks update --attach-acr or assign the role to the kubelet identity.. Run a pre-launch checklist item by item, and add policy/config scanning in CI.

What are the best practices for Azure Kubernetes Service (AKS)?

Best practices for Azure Kubernetes Service (AKS) include: Use a system node pool (Linux, for system pods) separate from user node pools, and enable cluster autoscaler on user pools only.; Enable Azure AD integration and Azure RBAC so kubectl access is governed by Entra ID groups, not shared kubeconfig files.; Use the Application Gateway ingress controller for production — it provides WAF, TLS termination, and cookie affinity without a separate NGINX deployment.. The first one — "Use a system node pool (Linux, for system pods) separate from user node pools, and enable cluster autoscaler on user pools only." — should be a team default baked into your IaC templates.

How do I control cost and avoid bill surprises with Azure Kubernetes Service (AKS)?

Three ways to prevent overspend: (1) tag every resource with env/team and enable billing alarms; (2) set budget thresholds to trigger Slack/email alerts; (3) review idle resources in Cost Explorer regularly and clean them up. Most Azure Kubernetes Service (AKS) options offer a free tier — validate within free quota before scaling.