Create and configure an Azure Kubernetes Services (AKS) cluster to use virtual nodes in the Azure portal

To quickly deploy workloads in an Azure Kubernetes Service (AKS) cluster, you can use virtual nodes. With virtual nodes, you have fast provisioning of pods, and only pay per second for their execution time. In a scaling scenario, you don't need to wait for the Kubernetes cluster autoscaler to deploy VM compute nodes to run the additional pods. This article shows you how to create and configure the virtual network resources and an AKS cluster with virtual nodes enabled.

Important

Virtual nodes for AKS are currently in preview. Previews are made available to you on the condition that you agree to the supplemental terms of use. Some aspects of this feature may change prior to general availability (GA).

Sign in to Azure

Sign in to the Azure portal at https://portal.azure.com.

Create an AKS cluster

In the top left-hand corner of the Azure portal, select Create a resource > Kubernetes Service.

On the Basics page, configure the following options:

  • PROJECT DETAILS: Select an Azure subscription, then select or create an Azure resource group, such as myResourceGroup. Enter a Kubernetes cluster name, such as myAKSCluster.
  • CLUSTER DETAILS: Select a region, Kubernetes version, and DNS name prefix for the AKS cluster.
  • SCALE: Select a VM size for the AKS nodes. The VM size cannot be changed once an AKS cluster has been deployed.
    • Select the number of nodes to deploy into the cluster. For this article, set Node count to 1. Node count can be adjusted after the cluster has been deployed.
    • Under Virtual nodes, select Enabled.

Create AKS cluster and enable the virtual nodes

By default, an Azure Active Directory service principal is created. This service principal is used for cluster communication and integration with other Azure services.

The cluster is also configured for advanced networking. The virtual nodes are configured to use their own Azure virtual network subnet. This subnet has delegated permissions to connect Azure resources between the AKS cluster. If you don't already have delegated subnet, the Azure portal creates and configures the Azure virtual network and subnet for use with the virtual nodes.

Select Review + create. After the validation is complete, select Create.

It takes a few minutes to create the AKS cluster and to be ready for use.

Connect to the cluster

The Azure Cloud Shell is a free interactive shell that you can use to run the steps in this article. It has common Azure tools preinstalled and configured to use with your account. To manage a Kubernetes cluster, use kubectl, the Kubernetes command-line client. The kubectl client is pre-installed in the Azure Cloud Shell.

To open the Cloud Shell, select Try it from the upper right corner of a code block. You can also launch Cloud Shell in a separate browser tab by going to https://shell.azure.com/bash. Select Copy to copy the blocks of code, paste it into the Cloud Shell, and press enter to run it.

Use the az aks get-credentials command to configure kubectl to connect to your Kubernetes cluster. The following example gets credentials for the cluster name myAKSCluster in the resource group named myResourceGroup:

az aks get-credentials --resource-group myResourceGroup --name myAKSCluster

To verify the connection to your cluster, use the kubectl get command to return a list of the cluster nodes.

kubectl get nodes

The following example output shows the single VM node created and then the virtual node for Linux, virtual-node-aci-linux:

$ kubectl get nodes

NAME                           STATUS    ROLES     AGE       VERSION
virtual-node-aci-linux         Ready     agent     28m       v1.11.2
aks-agentpool-14693408-0       Ready     agent     32m       v1.11.2

Deploy a sample app

In the Azure Cloud Shell, create a file named virtual-node.yaml and copy in the following YAML. To schedule the container on the node, a nodeSelector and toleration are defined. These settings allow the pod to be scheduled on the virtual node and confirm that the feature is successfully enabled.

apiVersion: apps/v1
kind: Deployment
metadata:
  name: aci-helloworld
spec:
  replicas: 1
  selector:
    matchLabels:
      app: aci-helloworld
  template:
    metadata:
      labels:
        app: aci-helloworld
    spec:
      containers:
      - name: aci-helloworld
        image: microsoft/aci-helloworld
        ports:
        - containerPort: 80
      nodeSelector:
        kubernetes.io/role: agent
        beta.kubernetes.io/os: linux
        type: virtual-kubelet
      tolerations:
      - key: virtual-kubelet.io/provider
        operator: Exists
      - key: azure.com/aci
        effect: NoSchedule

Run the application with the kubectl apply command.

kubectl apply -f virtual-node.yaml

Use the kubectl get pods command with the -o wide argument to output a list of pods and the scheduled node. Notice that the virtual-node-helloworld pod has been scheduled on the virtual-node-linux node.

$ kubectl get pods -o wide

NAME                                     READY     STATUS    RESTARTS   AGE       IP           NODE
virtual-node-helloworld-9b55975f-bnmfl   1/1       Running   0          4m        10.241.0.4   virtual-node-aci-linux

The pod is assigned an internal IP address from the Azure virtual network subnet delegated for use with virtual nodes.

Note

If you use images stored in Azure Container Registry, configure and use a Kubernetes secret. A current limitation of the virtual nodes preview is that you can't use integrated Azure AD service principal authentication. If you don't use a secret, pods scheduled on virtual nodes fail to start and report the error HTTP response status code 400 error code "InaccessibleImage".

Test the virtual node pod

To test the pod running on the virtual node, browse to the demo application with a web client. As the pod is assigned an internal IP address, you can quickly test this connectivity from another pod on the AKS cluster. Create a test pod and attach a terminal session to it:

kubectl run -it --rm virtual-node-test --image=debian

Install curl in the pod using apt-get:

apt-get update && apt-get install -y curl

Now access the address of your pod using curl, such as http://10.241.0.4. Provide your own internal IP address shown in the previous kubectl get pods command:

curl -L http://10.241.0.4

The demo application is displayed, as shown in the following condensed example output:

$ curl -L 10.241.0.4

<html>
<head>
  <title>Welcome to Azure Container Instances!</title>
</head>
[...]

Close the terminal session to your test pod with exit. When your session is ended, the pod is the deleted.

Next steps

In this article, a pod was scheduled on the virtual node and assigned a private, internal IP address. You could instead create a service deployment and route traffic to your pod through a load balancer or ingress controller. For more information, see Create a basic ingress controller in AKS.

Virtual nodes are one component of a scaling solution in AKS. For more information on scaling solutions, see the following articles: