Rediger

Tutorial: Scale applications in Azure Kubernetes Service (AKS)

  • Artikkel

If you followed the previous tutorials, you have a working Kubernetes cluster and you deployed the sample Azure Voting app. In this tutorial, part five of seven, you scale out the pods in the app and try pod autoscaling. You also learn how to scale the number of Azure VM nodes to change the cluster's capacity for hosting workloads. You learn how to:

  • Scale the Kubernetes nodes.
  • Manually scale Kubernetes pods that run your application.
  • Configure autoscaling pods that run the app front-end.

In the upcoming tutorials, you update the Azure Vote application to a new version.

Before you begin

In previous tutorials, you packaged an application into a container image, uploaded the image to Azure Container Registry, created an AKS cluster, and deployed the application to the AKS cluster.

If you haven't completed these steps and would like to follow along with this tutorial, start with the first tutorial, Prepare an application for AKS.

This tutorial requires Azure CLI version 2.0.53 or later. Run az --version to find the version. If you need to install or upgrade, see Install Azure CLI.

Manually scale pods

When you deployed the Azure Vote front end and Redis instance in the previous tutorials, a single replica was created.

  1. See the number and state of pods in your cluster using the kubectl get command.

    kubectl get pods

    The following example output shows one front-end pod and one back-end pod:

    NAME                               READY     STATUS    RESTARTS   AGE
azure-vote-back-2549686872-4d2r5   1/1       Running   0          31m
azure-vote-front-848767080-tf34m   1/1       Running   0          31m

  2. Manually change the number of pods in the azure-vote-front deployment using the kubectl scale command. The following example command increases the number of front-end pods to five:

    kubectl scale --replicas=5 deployment/azure-vote-front

  3. Verify the additional pods were created using the kubectl get pods command.

    kubectl get pods

                                    READY     STATUS    RESTARTS   AGE
azure-vote-back-2606967446-nmpcf    1/1       Running   0          15m
azure-vote-front-3309479140-2hfh0   1/1       Running   0          3m
azure-vote-front-3309479140-bzt05   1/1       Running   0          3m
azure-vote-front-3309479140-fvcvm   1/1       Running   0          3m
azure-vote-front-3309479140-hrbf2   1/1       Running   0          15m
azure-vote-front-3309479140-qphz8   1/1       Running   0          3m

Autoscale pods

Kubernetes supports horizontal pod autoscaling to adjust the number of pods in a deployment depending on CPU utilization or other select metrics. The Metrics Server is automatically deployed into AKS clusters with versions 1.10 and higher and provides resource utilization to Kubernetes.

  • Check the version of your AKS cluster using the az aks show command.

    az aks show --resource-group myResourceGroup --name myAKSCluster --query kubernetesVersion --output table

Note

If your AKS cluster is on a version lower than 1.10, the Metrics Server isn't automatically installed. Metrics Server installation manifests are available as a components.yaml asset on Metrics Server releases, which means you can install them via a URL. To learn more about these YAML definitions, see the Deployment section of the readme.

Example installation:

kubectl apply -f https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml

To use the autoscaler, all containers and pods must have defined CPU requests and limits. In the azure-vote-front deployment, the front-end container requests 0.25 CPU with a limit of 0.5 CPU.

These resource requests and limits are defined for each container, as shown in the following condensed example YAML:

  containers:
  - name: azure-vote-front
    image: mcr.microsoft.com/azuredocs/azure-vote-front:v1
    ports:
    - containerPort: 80
    resources:
      requests:
        cpu: 250m
      limits:
        cpu: 500m

Autoscale pods using kubectl autoscale

  • Autoscale pods using the kubectl autoscale command. The following command autoscales the number of pods in the azure-vote-front deployment with the following conditions: if average CPU utilization across all pods exceeds 50% of the requested usage, the autoscaler increases the pods up to a maximum of 10 instances and a minimum of three instances for the deployment:

    kubectl autoscale deployment azure-vote-front --cpu-percent=50 --min=3 --max=10

Autoscale pods using a manifest file

  1. Create a manifest file to define the autoscaler behavior and resource limits, as shown in the following example manifest file azure-vote-hpa.yaml:

    Note

    If you're using apiVersion: autoscaling/v2, you can introduce more metrics when autoscaling, including custom metrics. For more information, see Autoscale multiple metrics and custom metrics using v2 of the HorizontalPodAutoscaler.

    apiVersion: autoscaling/v1
kind: HorizontalPodAutoscaler
metadata:
  name: azure-vote-back-hpa
spec:
  maxReplicas: 10 # define max replica count
  minReplicas: 3  # define min replica count
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: azure-vote-back
  targetCPUUtilizationPercentage: 50 # target CPU utilization

---

apiVersion: autoscaling/v1
kind: HorizontalPodAutoscaler
metadata:
  name: azure-vote-front-hpa
spec:
  maxReplicas: 10 # define max replica count
  minReplicas: 3  # define min replica count
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: azure-vote-front
  targetCPUUtilizationPercentage: 50 # target CPU utilization

  2. Apply the autoscaler manifest file using the kubectl apply command.

    kubectl apply -f azure-vote-hpa.yaml

  3. Check the status of the autoscaler using the kubectl get hpa command.

    kubectl get hpa

# Example output
NAME               REFERENCE                     TARGETS    MINPODS   MAXPODS   REPLICAS   AGE
azure-vote-front   Deployment/azure-vote-front   0% / 50%   3         10        3          2m

    After a few minutes, with minimal load on the Azure Vote app, the number of pod replicas decreases to three. You can use kubectl get pods again to see the unneeded pods being removed.

Manually scale AKS nodes

If you created your Kubernetes cluster using the commands in the previous tutorials, your cluster has two nodes. If you want to increase or decrease this amount, you can manually adjust the number of nodes.

The following example increases the number of nodes to three in the Kubernetes cluster named myAKSCluster. The command takes a couple of minutes to complete.

  • Scale your cluster nodes using the az aks scale command.

    az aks scale --resource-group myResourceGroup --name myAKSCluster --node-count 3

    Once the cluster successfully scales, your output will be similar to following example output:

    "agentPoolProfiles": [
  {
    "count": 3,
    "dnsPrefix": null,
    "fqdn": null,
    "name": "myAKSCluster",
    "osDiskSizeGb": null,
    "osType": "Linux",
    "ports": null,
    "storageProfile": "ManagedDisks",
    "vmSize": "Standard_D2_v2",
    "vnetSubnetId": null
  }

Next steps

In this tutorial, you used different scaling features in your Kubernetes cluster. You learned how to:

  • Manually scale Kubernetes pods that run your application.
  • Configure autoscaling pods that run the app front end.
  • Manually scale the Kubernetes nodes.

In the next tutorial, you learn how to update applications in Kubernetes.

Update an application in Kubernetes