cicd-vsts.md

CI/CD with Azure DevOps

This lab outlines the process to compile a Docker-based ASP.NET Core web application and deploy it to a Kubernetes cluster running on Azure Container Service (AKS) using the Azure DevOps.

What's covered in this lab

In this lab, the following tasks will be performed:

• Create an Azure SQL server

• Provision the Azure DevOps Project with a .NET Core application using the Azure DevOps Demo Generator tool

• Configure endpoints (properties) in Azure DevOps to access Azure and exisiting AKS cluster

• Configure application and database deployment using Continuous Deployment (CD) in Azure DevOps

• Modify database connection string & ACR configuration in the source code

• Initiate the build to automatically deploy the application

Lab flow

• Firstly, the source code changes are committed to the Azure DevOps git repository

• Azure DevOps will create the custom Docker image myhealth.web and push the image tagged with the build ID to the ACR. Subsequently it will publish the Kubernetes deployment YAML file as a build artifact.

• Azure DevOps will deploy mhc-front and mhc-back services into the Kubernetes cluster using the YAML file.

  mhc-front is the application hosted on a load balancer whereas mhc-back is the Redis Cache

• The Kubernetes cluster will then pull the myhealth.web image from the Azure Container Registry (ACR) into the Pods and complete the deployment file instructions

• The myhealth.web application will be accessible through a browser, once the deployment is successfully completed

Prerequisites for the lab

1. Kubernetes extension from Visual Studio Marketplace installed to the Azure DevOps account

2. You will need a Azure SQL Database. Provision a new Azure SQL Server and SQL Database in East US region before proceeding with the below steps. Make note of the SQL Server name, database name, username and password. These will be used later in the exercise.

Setting up the Azure DevOps project

1. Use the Azure DevOps Demo Generator to provision the project on your Azure DevOps account.

   Azure DevOps Demo Generator helps you create projects on your Azure DevOps account with sample content that include source code, work items, iterations, service endpoints, build and release definitions based on the template you choose during the configuration.

   Click on 'Sign In' and accpet the Authorizations.

   

2. Select your organization, provide the Project Name and make sure that "aks" is the selected Template:

   

3. After verifying that the required extensions are installed and enabled, click on the Create Project button.

   

4. Once the project is successfully provisioned, click on the button Navigate to project.

   

Exercise 1: Service Endpoint creation

Service endpoints are a bundle of properties securely stored by the Azure DevOps and is a way for Azure DevOps to connect to the external systems or services. Since the connections are not established during project provisioning, the two endpoints - Azure Resource Manager and Kubernetes need to be created manually.

1. Azure Resource Manager Service Endpoint: Defines and secures a connection to a Microsoft Azure subscription using Service Principal Authentication (SPA).

   • In the Azure DevOps, navigate to Project settings (gear icon), Pipelines, Service connections and click on the + New Service Connection button.
   • Select the Azure Resource Manager, specify the Connection name and select the Subscription from the dropdown.
   • Leave Resource Group name blank and click on the OK button.

   This endpoint will be used to connect Azure DevOps with you Azure subscription.

   You will be prompted to authorize this connection with Azure credentials. Disable pop-up blocker in your browser if you see a blank screen after clicking the OK button, and retry the step.

2. Kubernetes Service Endpoint

   • Click the + New Service Connection button, and select Kubernetes from the list. We use this endpoint to connect Azure DevOps with Azure Kubernetes Service (AKS).

     • Connection Name: Provide a custom connection name.

     • Kubeconfig: To get the KubeConfig value, run the following Azure commands in a Putty session opened for the CentOS VM

     1. Type az login in the command prompt and hit Enter. Authorize your login by accessing the url given in the prompt and enter the provided unique code to complete the authentication.

     2. Type az aks get-credentials --resource-group yourResourceGroup --name yourAKSname in the command prompt to get the access credentials for the Kubernetes cluster.

     • Execute below command to get Kube Config value

       cat .kube/config

       {apiVersion: v1, clusters: [{cluster: {certificate-authority-data: LS0tLS1CRUdJTiBDRVJUSUZJQ0FURS0tLS0t.....
       ............................................................................................................
              token: 711e57a22410bb8d06ea956d2d5bc310}}]}

     • Copy the contents of the config file and paste it in the Add a Kubernetes service connection window.

     • Check Accept Untrusted Certificates and wait until the connection was successfully verified.

     • Once succeeded, click Ok.

       

Exercise 2: Configure Build and Release definitions

Now that the connections are established, we will manually map the existing Azure endpoint, AKS and Azure Container Registry to the build and release definitions.

1. Select the Builds section under the Pipelines tab and Edit the build definition MyHealth.AKS.Build.

   

2. Navigate to Process section under the Tasks tab. In the "Agent job 1" section select Run services.

   • Select the previously created endpoints from the dropdown for Azure subscription.
   • Choose the Azure Container Registry which we had created in the previous exercises from the drop down for Azure Container Registry.
   • Repeat the same for Build services, Push services, and Lock Services Publish. Click the Save option.

   

   Tasks	Usage
   Run services	prepares the suitable environment by restoring required packages
   Build services	builds the docker images specified in a docker-compose.yml file with registry-qualified names and additional tags such as $(Build.BuildId)
   Push services	pushes the docker images specified in a docker-compose.yml file, to the container registry
   Publish Build Artifacts	publishes the myhealth.dacpac file to Azure DevOps

3. Navigate to the Releases section under the Pipelines tab, click on MyHealth.AKS.Release release definition. Edit the release definition MyHealth.AKS.Release and select Tasks.

   

   

4. In the Dev environment, under the DB deployment phase, update the Azure Subscription value from the dropdown for Execute Azure SQL: DacpacTask task.

   

5. In the AKS deployment phase, under the Create Deployments & Services in AKS task, update the Kubernetes Service Connection value from the dropdown.

   • Expand the Secrets section and update the parameters
     • Azure subscription and
     • Azure Container Registry with the endpoint components from the dropdown.

6. Repeat similar steps for Update image in AKS task. Click Save to save the release definition

   

   • Create Deployments & Services in AKS will create the deployments and services in AKS as per the configuration specified in mhc-aks.yaml file. The Pod, for the first time will pull the latest docker image.

   • Update image in AKS will pull the appropriate image corresponding to the BuildID from the repository specified, and deploys the docker image to the mhc-front pod running in AKS.

7. Click on the Variables section under the release definition, update ACR and SQL server values for Process Variables with the details noted earlier while configuring the environment.

   • Click on the Save button.
   • Also update the DatabaseName and the SQLAdmin Login and Password based on the values you provided while creating the Azure SQL Database.

   

Exercise 3: Update Connection String & ACR URL in the manifest file

We will update the database connection string for the .NET Core application and ACR URL in the manifest YAML file.

IMPORTANT It's usually not necessary to change the __ACR__ variable. Azure DevOps will replace this variable as part of the build process.

1. Click on the Files tab in the Repos section and navigate to the below path AKS/src/MyHealth.Web to edit the file appsettings.json

   Scroll down to the line number 9 and provide the database server name as given in the step 7 of the previous exercise and manually update the User ID and Password based on the credentials you provided while creating the Azure SQL DB. Initial Catalog will be the database name. Click on the Commit button.

"DefaultConnection": "Server=<YOUR_SQLSERVER_NAME>.database.windows.net,1433;Initial Catalog=<Name>;Persist Security Info=False;User ID=<User ID>;Password=<Password>"

2. Navigate to the AKS path to edit the file mhc-aks.yaml. This YAML manifest file contains configuration details of deployments,services and pods which will be deployed in Kubernetes.

   Scroll to the line number 93. modify the value __ACR__ with your <ACR Login server>.azurecr.io which was noted earlier while setting up the environment.

   Also add the ImagePullSecrets under spec section. Provide the secret key as mysecretkey (this is the secret name - don't change it!) as seen below

spec:
  imagePullSecrets:
  - name: mysecretkey
  containers:
  - name: mhc-front

Click on the Commit button.

Exercise 4: Trigger a Build and deploy application

In this exercise, let us trigger a build manually and upon completion, an automatic deployment of the application will be triggered. Our application is designed to be deployed in the pod with the load balancer in the front-end and Redis cache in the back-end.

1. Click on the Builds section under the Pipelines tab, click on the build definition MyHealth.AKS.Build and then click on the Queue new build... button.

   

2. On the Queue Build for MyHealth.AKS.Build windows, click on Queue

3. Once the build process starts, navigate to the Builds tab. Click on the build number to see the build in progress.

   

   

4. The build will generate and push the docker image to ACR. After the build completes, you will see the build summary. To view the generated images in the Azure Portal, click on the Azure Container Registry and navigate to the Repositories.

   

5. Switch back to the Azure DevOps portal. Click on the Releases section on the Pipelines tab, and click on the latest release. Click on the Logs section to see the release summary.

   

   

6. Once the release is complete, open the putty session to the Cent OS VM and run the below command to see the pods running in AKS:

   kubectl get pods

   

   The deployed web application is running in the displayed pods.

7. To access the application, run the below command. If you see that External-IP is pending, wait for sometime until an IP is assigned.

   kubectl get service mhc-front --watch

   

8. Copy the External-IP and paste it in the browser and hit the Enter button to launch the application.

   

   Access AKS through the browser Type the below command in a command prompt from your base machine to access the AKS through the browser.

   az aks browse --resource-group <yourResourceGroup> --name <yourAKSname>

   

   AKS Dashboard: Once the AKS dashboard is launched, the following details will be displayed.

   

Summary

AKS reduces the complexity and operational overhead of managing a Kubernetes cluster by offloading much of that responsibility to Azure. With the Azure DevOps and Azure Container Services (AKS), we can build DevOps for dockerized applications by leveraging docker capabilities enabled on Azure DevOps Hosted Agents.

Attribution:

Content originally created by Srivatsa, Sachin Raj and sriramdasbalaji et al. from this ALM Doc. Modified it to work with exisiting K8S cluster

Return back to BootCamp Table of Contents (Main Page)