Previous

Next

Exercise - Deploy a sample application to the cluster with GitOps and Flux

Completed

In this exercise, we deploy a containerized Linux application to your Arc-enabled AKS Edge Essentials cluster using GitOps and Flux.

Fork the demo application GitHub repository

If you don't have a GitHub account already, create one now.

1. Navigate to the Azure Arc Jumpstart Apps repository and fork it to your own GitHub account by selecting Fork and then Create fork.

   

2. Confirm that the fork was successfully created by navigating to your account repositories and selecting the azure-arc-jumpstart-apps repository. The repository url should look like:

   https://github.com/<your-github-username>/azure-arc-jumpstart-apps
   

Deploy the application

In the Azure portal, navigate to your Arc-enabled AKS Edge Essentials cluster and select GitOps under Settings. Let's create a cluster-level configuration and a namespace-level configuration.

1. Select Create and use the following values for the cluster-level configuration:

   Attribute	Input
   Basics
   Configuration name	config-nginx
   Namespace	ingress-nginx
   Scope	Cluster
   Type	Flux v2
   Source
   Source kind	Git repository
   Repository URL	<URL of your fork>
   Reference type	Branch
   Branch	main
   Repository type	Public
   Sync interval	1
   Sync timeout	10
   Kustomizations
   Kustomizations	Select Create
   Instance name	nginx
   Path	./nginx/release
   Sync interval	10
   Sync timeout	10
   Prune	Enabled
   Force	Not enabled

   Note

   Wait until the config-nginx has successfully been created and visible on your Azure portal GitOps before creating the namespace-level configuration (you can move on to creating the next configuration if the compliance is in a pending state).

2. Select Create and use the following values for the namespace-level configuration:

   Attribute	Input
   Basics
   Configuration name	config-helloarc
   Namespace	hello-arc
   Scope	Namespace
   Type	Flux v2
   Source
   Source kind	Git repository
   Repository URL	<URL of your fork>
   Reference type	Branch
   Branch	main
   Repository type	Public
   Sync interval	1
   Sync timeout	10
   Kustomizations
   Kustomizations	Select Create
   Instance name	app
   Path	./hello-arc/releases/app
   Sync interval	10
   Sync timeout	10
   Prune	Enabled
   Force	Not enabled

3. Refresh your configuration table and wait for the configurations to be installed and compliant.

   

4. In your VM, use kubectl to check that the service is running:

   kubectl get svc -n ingress-nginx
   kubectl get pods -n hello-arc
   

   The following example output shows the service is running:

   PS C:\akseeLearn> kubectl get svc -n ingress-nginx
   NAME                                 TYPE           CLUSTER-IP      EXTERNAL-IP   PORT(S)                      AGE
   ingress-nginx-controller             LoadBalancer   10.43.136.240   192.168.0.4   80:31838/TCP,443:30386/TCP   18m
   ingress-nginx-controller-admission   ClusterIP      10.43.11.51     <none>        443/TCP                      18m
   PS C:\akseeLearn> kubectl get pods -n hello-arc
   NAME                         READY   STATUS    RESTARTS   AGE
   hello-arc-7c66864f8d-662b7   1/1     Running   0          12m
   hello-arc-7c66864f8d-fdbkc   1/1     Running   0          12m
   hello-arc-7c66864f8d-prb2v   1/1     Running   0          12m
   

5. To view the application, open a browser and navigate to the EXTERNAL-IP address assigned to the ingress-nginx-controller service. In the previous example, the IP address assigned to the service is 192.168.0.4.

   

Update the application

We configured GitOps and Flux to watch for changes in the main branch of your forked repository. Let's make a change to the application and see how GitOps responds.

1. In your fork of the azure-arc-jumpstart-apps repository, navigate to hello-arc > releases > app > hello-arc.yaml.

2. Make a change to this YAML file by selecting Edit. Change the replicaCount to 5. Change the value to Deploying to AKS Edge Essentials GitOps!.

3. Commit the change by selecting Commit changes... and then Commit changes.

   

4. Use kubectl to see the old pods terminating and new pods coming online:

   kubectl get pods -n hello-arc -w
   

   The following example output shows the old pods terminating and new pods coming online:

   PS C:\akseeLearn> kubectl get pods -n hello-arc -w
   NAME                         READY   STATUS    RESTARTS   AGE
   hello-arc-699dff9888-5mcjs   1/1     Running   0          19m
   hello-arc-699dff9888-qrsfw   1/1     Running   0          19m
   hello-arc-699dff9888-xm82r   1/1     Running   0          19m
   hello-arc-699dff9888-mpdxt   0/1     Pending   0          0s
   hello-arc-699dff9888-mpdxt   0/1     Pending   0          0s
   hello-arc-699dff9888-cvkgf   0/1     Pending   0          0s
   hello-arc-699dff9888-cvkgf   0/1     Pending   0          0s
   hello-arc-699dff9888-mpdxt   0/1     ContainerCreating   0          0s
   hello-arc-699dff9888-cvkgf   0/1     ContainerCreating   0          0s
   hello-arc-699dff9888-cvkgf   0/1     Running             0          1s
   hello-arc-699dff9888-mpdxt   0/1     Running             0          1s
   hello-arc-699dff9888-mpdxt   1/1     Running             0          2s
   hello-arc-699dff9888-cvkgf   1/1     Running             0          2s
   

   Note

   Because we set the sync interval to 1 min when creating the configuration, Flux pulls down changes from GitHub every minute.

5. Refresh your application to see this change reflected as a rolling update.

   

Clean up your Azure resources

In the course of this module, you created Azure resources. If you don't expect to need these resources in the future, delete the resource group and Service Principal by running az group delete and az ad sp delete commands in Azure Cloud Shell:

az group delete --name "aksedge-training" --force-deletion-types Microsoft.Compute/virtualMachines
az ad sp delete --id "<your-service-principal-ID>"

Important

To avoid accruing unwanted charges, you must remove the Azure resources and Service Principal that you used in this module.

Continue