Exercise - Deploy Kubernetes with Azure Kubernetes Service
Fruit Smoothies wants to use Kubernetes as their compute platform. The development teams already use containers for application development and deployment, and using an orchestration platform will help them rapidly build, deliver, and scale their application.
To do this, you need to deploy the foundation of your Kubernetes environment.
In this exercise, you will:
- Create a new resource group.
- Configure cluster networking.
- Create an Azure Kubernetes Service cluster.
- Connect to the Kubernetes cluster by using
- Create a Kubernetes namespace.
You need your own Azure subscription to run this exercise, and you might incur charges. If you don't already have an Azure subscription, create a free account before you begin.
Create a new resource group
You'll first need to create a resource group for your resources to deploy into.
Sign in to Azure Cloud Shell with your Azure account. Select the Bash version of Cloud Shell.
We're going to reuse some values throughout the deployment scripts. For example, you need to choose a region where you want to create a resource group, such as East US. If you select a different value, remember it for the rest of the exercises in this module. You may need to redefine the value between Cloud Shell sessions. Run the following commands to record these values in Bash variables.
REGION_NAME=eastus RESOURCE_GROUP=aksworkshop SUBNET_NAME=aks-subnet VNET_NAME=aks-vnet
You can use the Copy button to copy commands to the clipboard. To paste, right-click on a new line in the Cloud Shell window and select Paste or use the Shift+Insert keyboard shortcut (⌘+V on macOS).
You can check each value using the
echocommand, for example,
Create a new resource group with the name aksworkshop. Deploy all resources created in these exercises in this resource group. A single resource group makes it easier to clean up the resources after you finish the module.
az group create \ --name $RESOURCE_GROUP \ --location $REGION_NAME
We have two network models to choose from when deploying an AKS cluster. The first model is Kubenet networking, and the second is Azure Container Networking Interface (CNI) networking.
What is Kubenet networking?
Kubenet networking is the default networking model in Kubernetes. With Kubenet networking, nodes get assigned an IP address from the Azure virtual network subnet. Pods receive an IP address from a logically different address space to the Azure virtual network subnet of the nodes.
Network address translation (NAT) is then configured so that the pods can reach resources on the Azure virtual network. The source IP address of the traffic is translated to the node's primary IP address and then configured on the nodes. Note, that pods receive an IP address that's "hidden" behind the node IP.
What is Azure Container Networking Interface (CNI) networking?
With Azure Container Networking Interface (CNI), the AKS cluster is connected to existing virtual network resources and configurations. In this networking model, every pod gets an IP address from the subnet and can be accessed directly. These IP addresses must be unique across your network space and calculated in advance.
Some of the features you'll use require you to deploy the AKS cluster by using the Azure Container Networking Interface networking configuration.
For a more detailed comparison, see the Learn more section at the end of this module.
Let's create the virtual network for your AKS cluster. We will use this virtual network and specify the networking model when we deploy the cluster.
First, create a virtual network and subnet. Pods deployed in your cluster will be assigned an IP from this subnet. Run the following command to create the virtual network.
az network vnet create \ --resource-group $RESOURCE_GROUP \ --location $REGION_NAME \ --name $VNET_NAME \ --address-prefixes 10.0.0.0/8 \ --subnet-name $SUBNET_NAME \ --subnet-prefixes 10.240.0.0/16
Next, retrieve, and store the subnet ID in a Bash variable by running the command below.
SUBNET_ID=$(az network vnet subnet show \ --resource-group $RESOURCE_GROUP \ --vnet-name $VNET_NAME \ --name $SUBNET_NAME \ --query id -o tsv)
Create the AKS cluster
With the new virtual network in place, you can go ahead and create your new cluster. There are two values you need to know before running the
az aks create command. The first is the version of the latest, non-preview, Kubernetes version available in your selected region, and the second is a unique name for your cluster.
To get the latest, non-preview, Kubernetes version you use the
az aks get-versionscommand. Store the value that returns from the command in a Bash variable named
VERSION. Run the command below the retrieve and store the version number.
VERSION=$(az aks get-versions \ --location $REGION_NAME \ --query 'orchestrators[?!isPreview] | [-1].orchestratorVersion' \ --output tsv)
The AKS cluster name must be unique. Run the following command to create a Bash variable that holds a unique name.
Run the following command to output the value stored in
$AKS_CLUSTER_NAME. Note this for later use. You'll need it to reconfigure the variable in the future, if necessary.
Run the following
az aks createcommand to create the AKS cluster running the latest Kubernetes version. This command can take a few minutes to complete.
az aks create \ --resource-group $RESOURCE_GROUP \ --name $AKS_CLUSTER_NAME \ --vm-set-type VirtualMachineScaleSets \ --node-count 2 \ --load-balancer-sku standard \ --location $REGION_NAME \ --kubernetes-version $VERSION \ --network-plugin azure \ --vnet-subnet-id $SUBNET_ID \ --service-cidr 10.2.0.0/24 \ --dns-service-ip 10.2.0.10 \ --docker-bridge-address 172.17.0.1/16 \ --generate-ssh-keys
Let's review the variables in the previous command:
$AKS_CLUSTER_NAMEspecifies the name of the AKS cluster.
$VERSIONis the latest Kubernetes version you retrieved earlier.
$SUBNET_IDis the ID of the subnet created on the virtual network to be configured with AKS.
Note the following deployment configuration:
--vm-set-type: We're specifying that the cluster is created by using virtual machine scale sets. The virtual machine scale sets enable you to switch on the cluster autoscaler when needed.
--node-count: We're specifying that the cluster is created with two nodes. The default node count is three nodes. However, if you're running this exercise using a free trial account, cluster creation may fail due to quota limits if left at the default setting.
--network-plugin: We're specifying the creation of the AKS cluster by using the CNI plug-in.
--service-cidr: This address range is the set of virtual IPs that Kubernetes assigns to internal services in your cluster. The range must not be within the virtual network IP address range of your cluster. It should be different from the subnet created for the pods.
--dns-service-ip: The IP address is for the cluster's DNS service. This address must be within the Kubernetes service address range. Don't use the first IP address in the address range, such as 0.1. The first address in the subnet range is used for the kubernetes.default.svc.cluster.local address.
--docker-bridge-address: The Docker bridge network address represents the default docker0 bridge network address present in all Docker installations. AKS clusters or the pods themselves don't use docker0 bridge. However, you have to set this address to continue supporting scenarios such as docker build within the AKS cluster. It's required to select a classless inter-domain routing (CIDR) for the Docker bridge network address. If you don't set a CIDR, Docker chooses a subnet automatically. This subnet could conflict with other CIDRs. Choose an address space that doesn't collide with the rest of the CIDRs on your networks, which includes the cluster's service CIDR and pod CIDR.
Test cluster connectivity by using
kubectl is the main Kubernetes command-line client you use to interact with your cluster and is available in Cloud Shell. A cluster context is required to allow kubectl to connect to a cluster. The context contains the cluster's address, a user, and a namespace. Use the
az aks get-credentials command to configure your instance of kubectl.
Retrieve the cluster credentials by running the command below.
az aks get-credentials \ --resource-group $RESOURCE_GROUP \ --name $AKS_CLUSTER_NAME
Let's take a look at what was deployed by listing all the nodes in your cluster. Use the
kubectl get nodescommand to list all the nodes.
kubectl get nodes
You'll see a list of your cluster's nodes. Here's an example.
NAME STATUS ROLES AGE VERSION aks-nodepool1-29333311-vmss000000 Ready agent 2m36s v1.17.9 aks-nodepool1-29333311-vmss000001 Ready agent 2m34s v1.17.9
Create a Kubernetes namespace for the application
Fruit Smoothies want to deploy several apps from other teams in the deployed AKS cluster as well. Instead of running multiple clusters, the company wants to use the Kubernetes features that let you logically isolate teams and workloads in the same cluster. The goal is to provide the least number of privileges scoped to the resources each team needs.
What is a namespace?
A namespace in Kubernetes creates a logical isolation boundary. Names of resources must be unique within a namespace but not across namespaces. If you don't specify the namespace when you work with Kubernetes resources, the default namespace is implied.
Let's create a namespace for your ratings application.
List the current namespaces in the cluster.
kubectl get namespace
You'll see a list of namespaces similar to this output.
NAME STATUS AGE default Active 1h kube-node-lease Active 1h kube-public Active 1h kube-system Active 1h
kubectl create namespacecommand to create a namespace for the application called ratingsapp.
kubectl create namespace ratingsapp
You'll see a confirmation that the namespace was created.
In this exercise, you created a resource group for your resources. You created a virtual network for your cluster to use. You then deployed your AKS cluster, including the Azure CNI networking mode. You then connected to your cluster with kubectl and created a namespace for your Kubernetes resources.
Next, you'll create and configure an Azure Container Registry (ACR) instance to use with your AKS cluster and store your containerized ratings app.