Skip to content

Latest commit

 

History

History
193 lines (147 loc) · 11.5 KB

02-build_vsts_ci.md

File metadata and controls

193 lines (147 loc) · 11.5 KB

Build Out VSTS CI Pipeline

In this portion of the lab we will be creating the Continuous Integration (CI) pipeline in VSTS.

A CI Pipeline allows us to define a Build Stage, which will allow us to automate the required steps to package our application code into their preferred runtime environments.

In this lab we will be pushing in new code to our Git Repository, which will then trigger the packaging of three (3) Applications - a Web frontend, an REST API backend and a Database Server - into three (3) separate Docker Container Images.

Upon a successful build for each application, we will push the newly created Docker Image to an Azure Container Registry which will allow us to leverage these Images from a central location into multiple target environments (e.g. Dev, Test, Production).

build_workflow

Exercise 1 - Create New Project in VSTS

In this section we will be creating a new Git-based Project in VSTS. This will be used as source control for our project as well as building out the CI and CD Pipelines.

  1. Navigate to your VSTS Account
  • It will look similar to the following depending on what you setup in Lab 1: http://<VSTS_ACCOUNT_NAME>.visualstudio.com
  1. Create a New Project
  • Click on the New Project button.
  • Enter akspipeline for the Project name.
  • Make sure Version control is set to Git, should be the default.
  • Make sure Work item process is set to Agile, again, should be the default.
  • Click Create button. empty process
  1. Note Git Endpoint
  • Once the Project has been created you will see a getting started page.
  • Note the Git Url, it should look like: http://<VSTS_ACCOUNT_NAME>.visualstudio.com/_git/akspipeline

Exercise 2 - Add App Code to VSTS Project

Please note it is possible to setup VSTS to leverage other source control repositories, that is not the main focus of this exercise. For the purposes of this lab we will be pushing existing code into VSTS to keep things simple.

  1. Note Getting Started Options
  • As you can see there are a variety of ways to get started. You can clone to your computer, push existing, import, initialize or build code from external repository.
  • As noted above, we are going to push an existing repository. Click on the or push an existing repository from command line link to see the steps.
  1. Create New Local Repo
  • We will be using the Justice League App folder from the following repository: http://github.com/azure/blackbelt-aks-hackfest
  • If you do not already have the repository cloned, please clone it to your local computer. 
    cd ~/Documents
git clone [email protected]:Azure/blackbelt-aks-hackfest.git
  • Create a new Repo directory on your computer called justiceleagueapp. 
    cd ~/Documents
mkdir justiceleagueapp
cd justiceleagueapp
  • Copy the app and helper-files folders from the Repo above into the newly created justiceleagueapp folder. 
    cd ~/Documents/justiceleagueapp
cp ~/Documents/blackbelt-aks-hackfest/linux-container/app .
cp ~/Documents/blackbelt-aks-hackfest/linux-container/helper-files .
  • Open that newly created directory using VS Code.
    • Change directory into the new justiceleagueapp folder (or whatever you named that directory)
    # presuming you are in the directory ~/Documents/justiceleagueapp

code .
# (=> code "dot"/"period") to launch current app directory in VSCode to start editing files
  • Add a .gitignore file to the Repo. Copy and paste the contents from here.
  • We will be adding one new file to the Repo. It is a databaseload.sh script file to help us load the Mongo DB database as part of our Release pipeline. Copy and paste the content of the databaseload.sh file from here into a new file called databaseload.sh in the helper-files directory.
  • Using the Source Control navigation option in VS Code on the left-hand side, Initialize the local Repo.
  • empty process
  • Once initialized, add all the files to the Repo by entering a Comment such as Initial commit. and then click on the checkmark at the top of the window.
  • You now have a local Git Repo, next step is to push it up to VSTS.
  • empty process
  1. Push Local Repo to VSTS
  • Now that you have a local Git Repo, we need to add VSTS as a Remote origin. To do this we will open the VS Code Terminal Window. This is done by hitting **Ctrl + ~** on teh keyboard.
  • Once at the command prompt, you are going to use the or push an existing repository from command line commands from Exercise 1 above.
git remote add origin https://<VSTS_ACCOUNT_NAME>.visualstudio.com/_git/akspipeline
git push -u origin --all
  • At this point you should have all of your code in VSTS Source Control. Click on the Code navigation item near the top of the VSTS window to see your checked in code. empty process

Exercise 3 - Create Build Pipeline

  1. Create Empty Build Pipeline
  • Hover over the Build and Release navigation item near the top of the VSTS window. You will see a list of drop-down options, click on Builds.
  • Click on + New definition to create a new Build.
  • All the defaults should be correct. empty process
  • Once verifying the defaults, click Continue.
  • Select Empty Process. empty process
  1. Set Correct Agent Queue
  • Note the default pipeline name, akspipeline-CI, you can choose to keep the default or name it somethign new.
  • Under the Agent queue, be sure the Hosted Linux Preview option is selected from the drop-down menu. This represents the type of build machine that is being used. build_agent
  1. Validate Sources and add two (2) Docker Tasks
  • If you click on Get sources on the left you will notice that it is pointing to your VSTS Project Repo. If not, you have done something wrong, please re-visit previous steps. get sources
  • Next, click on the Phase 1 link on the left-hand side. You will notice that this Build Agent Phase automatically inherits the type of Build Agent from the Parent Build Process.
  • You can change the Display name or leave it as is, up to you.
  • The next step is to add Docker Build and Docker Push steps to the pipeline. This is done by clicking on the + next to Phase 1, entering docker in the search box, hover over the Docker taks, then hit the Add button twice.
    • This will add two "Docker Tasks" to your "Phase 1" Agent Phase add docker tasks
  1. Configure Docker Build and Push DB Tasks

In this step we will be configuring the automated build of the Container Image for the database portion of the application.

  • Link Existing Azure Service Principal to VSTS so drop-down boxes will be automatically filled in.

    • Click on the first of the two (2) Docker Tasks, go to the Azure Subscription and click on the Manage link.
    • empty process
    • Next, create a new Azure Resource Manager (ARM) Service Endpoint.
    • empty process
    • Next, click on the full version option at the bottom of the screen.
    • empty process
    • Lastly, fill in the blanks using the Service Principal credential information provided as part of the lab creation.
    • empty process

  • Click on the first of the two (2) Docker Tasks, and modify the following fields and values:

    • Display name: Build API Image
    • Container Registry Type: Azure Container Registry
    • Azure Subscription: Choose Azure Subscription
    • Azure Container Registry: Select from drop-down
    • Action: Build an image
    • Docker File: Choose file using the ... select menu on the right
    • Build Arguments: (LEAVE BLANK)
    • Use Default Build Context: Ensure it is Checked
    • Image Name: <ACR_NAME>.azurecr.io/azureworkshop/rating-api:$(Build.BuildId)
    • Qualify Image Name: Ensure it is Checked
    • Additional Image Tags: (LEAVE BLANK)
    • Leave the rest as the defaults.

  • Click on the second of the two (2) Docker Tasks, and modify the following fields and values:

    • Display name: Push API Image
    • Container Registry Type: Azure Container Registry
    • Azure Subscription: Choose Azure Subscription
    • Azure Container Registry: Select from drop-down
    • Action: Push an image
    • Image Name: <ACR_NAME>.azurecr.io/azureworkshop/rating-api:$(Build.BuildId)
    • Qualify Image Name: Ensure it is Checked
    • Additional Image Tags: (LEAVE BLANK)
    • Leave the rest as the defaults.

  • After completing the above, click on the Save & queue button near the top right. A window will pop-up, leave the defaults and hit the save & queue button to start the build.

  1. Validate Successful Build
  • After clicking on the save & queue button above, you will see a message near the top of the page that a new build has been kicked off with a number. Clicking on that number will take you to the build run and show you the progress.
  • A successful run is green. If there is red it means the build failed. Use the build run output to troubleshoot errors, fix the errors, and then run another build.
  1. Repeat Steps 3 and 4 Above for Web and API Containers
  • Hint: You will need to add 4 more Docker Tasks, 2 each for API and Web.
  • Hint: Remember to select the correct Docker File associated to each part of the app.
  • Hint: Don't forget to ensure you change the Image Name associated to each part of the app.
  1. Add Artifacts to Build Output
  • The next step is to add Publish Output steps to the pipeline. This is done by clicking on the + next to Phase 1, entering publish in the search box, hover over the Publish Build Artifacts task, then hit the Add button.
  • Click on the Publish Artifact task and update the following:
    • Display name: Publish Yaml
    • Path to Publish: Use the ... (elipsis) on the right-hand side to select the helper-files folder
    • Artifact Name: yaml
    • Artifact publish location: Visual Studio Team Services/TFS empty process
  • After completing the above, click on the Save & queue button near the top right. A window will pop-up, leave the defaults and hit the save & queue button to start the build.
  1. Enable Continuous Integration
  • To enable Continuous Integration, meaning that everytime your code is checked in a build will run, click on Triggers just above the Process bar.
  • Click the Enable continuous integration checkbox.
  • Click the Save & queue button in the top right and select Save from the drop-down to save your changes.
  • You are done, now wasn't that easy. empty process

Summary

At this point you should have a fully functioning CI process that builds all 3 Container Images (Web, API, DB) as well as pushing all 3 Container Images (Web, API, DB) to Azure Container Registry. Now you are ready to build the Release Pipeline and release the build to AKS.