Deploying our sample application to AKS Cluster

After provisioning your AKS cluster (including the resource group, virtual network, and other prerequisites), the next step is to connect via Azure CLI and deploy a Dockerized sample application. This guide will walk you through each step—from authentication to exposing your service—using PowerShell examples that apply equally to macOS and Linux terminals.

Note

All Azure CLI commands shown here work in Azure Cloud Shell, Windows PowerShell, macOS, and Linux terminals. Replace any Windows-specific paths or syntax as needed on other platforms.

1. Connecting to Your AKS Cluster

You can authenticate and configure kubectl in multiple environments:

Connection Method	Platform	Pre-installed Tools
Azure Cloud Shell	Web Browser	kubectl, Azure CLI, PowerShell modules
Local PowerShell	Windows	kubectl, Azure CLI
Terminal	macOS / Linux	kubectl, Azure CLI

Log in to Azure CLI
```
PS C:\Users\msadmin> az login
```

Verify your active subscription

PS C:\Users\msadmin> az account show
{
  "environmentName": "AzureCloud",
  "homeTenantId": "...",
  "id": "36b0e8a-2743-4be7-891c-1e2b3a83ad1b",
  "isDefault": true,
  "name": "Sub_global",
  "state": "Enabled",
  "tenantId": "...",
  "user": {
    "name": "[email protected]",
    "type": "user"
  }
}

Set a default resource group

PS C:\Users\msadmin> az configure --defaults group=RG1-KodeKloud-AKS

Download and merge AKS credentials

PS C:\Users\msadmin> az aks get-credentials --name AKS1-KodeKloudApp
Merged "AKS1-KodeKloudApp" as current context in C:\Users\msadmin\.kube\config

Confirm the current Kubernetes context

PS C:\Users\msadmin> kubectl config current-context
AKS1-KodeKloudApp

2. Inspecting the Cluster State

Before deploying, it’s good practice to confirm that your cluster has no existing workloads:

Command	Purpose
`kubectl get nodes`	List all nodes and their status
`kubectl get deployments`	Show existing deployments in default ns
`kubectl get pods`	Display running pods in default ns

PS C:\Users\msadmin> kubectl get nodes
NAME                              STATUS   ROLES   AGE   VERSION
aks-agentpool-18097611-vmss000000 Ready    agent   45m   v1.24.10

PS C:\Users\msadmin> kubectl get deployments
No resources found in default namespace.

PS C:\Users\msadmin> kubectl get pods
No resources found in default namespace.

As expected, there are no deployments or pods at this point.

3. Deploying the Sample Application

We previously built and pushed the Docker image hpranav/kodekloudappcs:v1 to Docker Hub. Deploy it to AKS with a single replica:

PS C:\Users\msadmin> kubectl create deployment kodekloudapp `
   --image=hpranav/kodekloudappcs:v1 `
   --replicas=1
deployment.apps/kodekloudapp created

Verify that the deployment and pod are running:

PS C:\Users\msadmin> kubectl get deployments
NAME            READY   UP-TO-DATE   AVAILABLE   AGE
kodekloudapp    1/1     1            1           2m

PS C:\Users\msadmin> kubectl get pods
NAME                              READY   STATUS    RESTARTS   AGE
kodekloudapp-67ff275c85-mm42w     1/1     Running   0          2m

You can also view the pod status in the Azure portal under your AKS resource.

4. Exposing the Deployment via Load Balancer

To allow external access, create a Service of type LoadBalancer. Azure will provision a cloud load balancer and assign a public IP:

PS C:\Users\msadmin> kubectl expose deployment kodekloudapp `
   --type=LoadBalancer `
   --port=80 `
   --target-port=80
service/kodekloudapp exposed

Retrieve the service details and public IP:

PS C:\Users\msadmin> kubectl get service kodekloudapp
NAME            TYPE           CLUSTER-IP     EXTERNAL-IP     PORT(S)        AGE
kodekloudapp    LoadBalancer   10.0.152.169   20.50.123.456   80:30541/TCP   1m

Warning

Replace 20.50.123.456 with your actual EXTERNAL-IP. It can take a minute for Azure to provision the public IP.

Open your browser to http://20.50.123.456 to verify that the sample application is running on your AKS cluster.

Links and References

Watch Video

Watch video content