Continuous Integration and Continuous Delivery With Fargate

About this lab

Delivering new iterations of software at a high velocity is a competitive advantage in today’s business environment. The speed at which organizations can deliver innovation to customer and adapt changing markets is a standard to determined organizations will success or failure. That’s why continuous integration and continuous delivery are important.

Nowadays, many software can help you to make a DevOps flow. But you need to waste many times to maintain those infrastructures, and sometimes will occur some human errors. It not only increase cost, but also decrease your development times.AWS provide a set of flexible DevOps tool. It can easily to integrate with another AWS service and also third party DevOps software. With those tool, you can quickly and easily to implement an automated deployment and integration pipeline for applications. All those tools are fully management.

In this tutorial, we will use AWS Cloud9 to create and co-work environment, AWS codepipeline to create an automated deployment and integration pipeline, and finally deploy on AWS ECS – Fargate type.

Prerequisites

1. Sign-in a AWS account, and make sure you have select N.Virginia region.

2. Download the source file from github.

Lab tutorial

Create a repository with AWS CodeCommit

1.1. Open AWS Manage Console and Sign in.

1.2. On the Service Menu, choose CodeCommit service under Developer Tools.

1.3. If you are first time use this service, click Get started, if not, click Create Repository.

1.4. Enter workshop-commit as the Repository name.

1.5. Click Create Repository.

1.6. Click Skip.

1.7. Notice the clone command under Steps to clone your repository.

1.8. Click Close.

Create a cloud-based IDE with AWS Cloud9.

1.9. On the Service Menu, choose Cloud9 service under Developer Tools.

1.10. Click Create environment.

1.11. Type workshop-cloud9 as the Name.

1.12. Click Next step.

1.13. Let all configure option default, click Next step.

1.14. Click Create environment.

1.15. Now you will see the Cloud9 IDE is open, wait until it done.

Integrate Cloud9 With CodeCommit and Push First Code

1.16. At the bottom of page, you will see the console tab.

1.17. Copy the follow command and paste to console.

git config --global credential.helper '!aws codecommit credential-helper $@'
git config --global credential.UseHttpPath true

1.18. Paste the command that you notice at step 1.8.

1.19. If you clone successful, you will see a folder name workshop-codecommit at left navigation panel.

1.20. At the workshop-commit folder layer create a New File, and named app.py.

1.21. Copy code in app.py which you download into Cloud9’s app.py.

1.22. Create a New File, and named requirements.txt.

1.23. Copy code in requirements.txt which you download into Cloud9’s requirements.txt.

1.24. Create a New File, and named Dockerfile.

1.25. Copy code in Dockerfile which you download into Cloud9’s Dockerfile.

1.26. Create a New File, and named buildspec.yml.

1.27. Copy code in build/buildspec.yml which you download into Cloud9’s buildspec.yml.

1.28. Select File in Tool bar, and click Save all.

1.29. In console, enter the follow code to push your code to CodeCommit.

cd workshop-commit
git add .
git commit -m "First Commit "
git push

1.30. Go to AWS CodeCommit page, and you will see those code you push previously.

Create a Cluster, Task Definition, and Service Using Fargate

1.31. On the Service Menu, choose Elastic Container Service under Compute.

1.32. Click Get started.

1.33. Click Cancel.

1.34. If you see this page, it means that you can do next.

1.35. Click Create Clusters.

1.36. Select Networking only, and click Next step.

1.37. Type workshop-ecs as the Cluster name.

1.38. Click Create VPC.

1.39. Create your own CIDR which not affect others VPC.

1.40. Click Create.

1.41. When it created successfully, click View Clusters.

1.42. If you see the cluster’s status is ACTIVE, you can do next step.

1.43. Click Repositories under Amazon ECR at left panel.

1.44. If you are first created, click Get started, else, click Create repository.

1.45. Type workshop-ecr as the Repository name.

1.46. Click Next step.

1.47. Click Done.

1.48. Note the Repository URI and copy it.

1.49. Go back to Cloud9 IDE to replace the Repository URI.

1.50. Use Git command push code to CodeCommit.

1.51. Back to ECS page, click Task Definitions at left panel.

1.52. Click Create new Task Definition.

1.53. Select Fargate, and click Next step.

1.54. Type workshop-task as the Task Definition Name.

1.55. Select None as Task Role

1.56. Under the Task Size, select 0.5GB as Task memory, 0.25vCPU as Task CPU.

1.57. Click Add container.

1.58. Type workshop-container as Container name.

1.59. Paste the Repository URI you noted previously into Image field.

1.60. At the Port mappings part, Type 80 as Container port.

1.61. Click Add.

1.62. Click Create.

1.63. Go back to Cloud9 IDE to replace the Container name.

1.64. Use Git command push code to CodeCommit.

1.65. Back to ECS page, click Action, and Select Create Service.

1.66. Choose Fargate as Launch type.

1.67. Type workshop-service as the Service name.

1.68. In Number of tasks, enter 1.

1.69. Click Next step.

1.70. Select VPC which you create previously.

1.71. At Subnets, select all two subnets.

1.72. Under Load balancing, choose Application Load Balancer.

1.73. You will see a warning which mean you need to create ALB, click EC2 Console.

1.74. Select Create under Application Load Balancer.

1.75. Type workshop-ALB as Name.

1.76. Under Availability Zones, choose the VPC you created previously, and select both.

1.77. Click Next: Configure Security Settings.

1.78. Click Next: Configure Security Groups.

1.79. Select Create a new security group.

1.80. Type workshop-ALB-SG as Security group name.

1.81. At Type part, Select HTTP.

1.82. At Source part, Select Anywhere.

1.83. Click Next: Configure Routing.

1.84. Type workshop-TG as the name.

1.85. At the Target type, select ip.

1.86. Click Next: Register Target.

1.87. Click Next: Review.

1.88. Click Create.

1.89. When it created successfully, click Close.

1.90. Back to ECS create service page.

1.91. Click Reload.

1.92. Select workshop-ALB as Load balancer name.

1.93. Under Container to load balance, click Add to load balancer.

1.94. At the Target group name, select workshop-TG.

1.95. Click Next step.

1.96. Let option default, click Next step.

1.97. Click Create Service.

1.98. When it created Successfully, click View Service, and you can do next step.

Create a Pipeline Use AWS CodePipeline

1.99. In the AWS Management Console, on the service menu, click CodePipeline.

1.100. Click Get started.

1.101. Type workshoppipeline as the pipeline name.

1.102. Click Next step.

1.103. Select AWS CodeCommit as Source provider.

1.104. Select workshop-commit as Repository name.

1.105. Select master as Branch name.

1.106. Click Next step.

1.107. Select AWS CodeBuild as Build provider.

1.108. Choose Create a new build project.

1.109. Type imagebuild as Project name.

1.110. Select Ubuntu at Operating system, select Docker, and select aws/codebuild/docker:17.09.0.

1.111. Click Save build project.

1.112. Click Next step.

1.113. Select Amazon ECS as Deployment provider.

1.114. Select workshop-ecs as Cluster name.

1.115. Select workshop-service as Service name.

1.116. Type images.json as Image filename.

1.117. Click Next step.

1.118. At AWS Service Role, Click Create role.

1.119. Click Allow.

1.120. Click Next step.

1.121. Click Create pipeline.

Edit the CodeBuild IAM

1.122. After you created pipeline, it will start running, but fail in build stage. You need to add some policies to access ECS resources.

1.123. In the AWS Management Console, on the service menu, click IAM.

1.124. Click Role at the left panel.

1.125. Find code-build-imagebuild-service-role and click it.

1.126. Click Attach policy.

1.127. Type AmazonEC2ContainerRegistryPowerUser in filter.

1.128. Select it, and click Attach policy.

1.129. Back to AWS Codepipeline page, click Retry.

Try to See First Deploy and Test First Change

1.130. When all Stage Success, In AWS Manage Console, click service, and click EC2.

1.131. Click Load Balancers at left panel.

1.132. Click workshop-ALB, and find the DNS name.

1.133. Open a new web page and paste it in URL line, and press enter, you will see the web which you build by Cloud9 IDE.

1.134. Now Back to Cloud9 IDE, try to change app.py.

1.135. In app.py, change the version to Second Fargate Version.

1.136. Save it and use Git command to push it to CodeCommit.

1.137. When it push to CodeCommit successfully, it would trigger Codepipeline to automate integration and delivery, go back to CodePipeline page to see.

1.138. When all stage success, reload the web page, and you can see the new version is deployed.

Try to Add Test Stage and Config CodeBuild

1.139. Go Back to Cloud9 IDE.

1.140. At left panel, workshop-commit layer, create a new folder named build.

1.141. Drag the buildspec.yml and drop it to build folder.

1.142. At workshop-commit layer, create another folder named test folder.

1.143. In test folder, create a new file name buildspec.yml.

1.144. Copy the code in test/buildspec.yml and paste it to Cloud9’s test/buildspec.yml.

1.145. Go back to CodePipeline page, and click Edit.

1.146. Click + Stage between Source and Build.

1.147. Type Test as Stage name.

1.148. Click + Action.

1.149. Select Test as Action category.

1.150. Type CodeTest as Action Name.

1.151. Select AWS CodeBuild as Test provider.

1.152. Click Create a new build project.

1.153. Type codetest as Project name.

1.154. Select Ubuntu as Operating system, select Python, select aws/codebuild/python:2.7.12.

1.155. Click Save build project.

1.156. In the Input artifacts #1, select MyApp.

1.157. Click Add action.

1.158. Click Save pipeline changes.

1.159. Click Save and continue.

1.160. In AWS Manage Console, click service, click CodeBuild.

1.161. Select codetest, click Actions, click Update.

1.162. Under Environment: How to build, click Update build specification.

1.163. Type test/buildspec.yml as Buildspec name.

1.164. Click Update.

1.165. Click Back.

1.166. Select imagebuild, click Actions, click Update.

1.167. Under Environment: How to build, click Update build specification.

1.168. Type build/buildspec.yml as Buildspec name.

1.169. Click Update.

1.170. Back to Cloud9 IDE, change the version to Third Fargate Version, and use Git command to push to CodeCommit.

1.171. Wait all stage success, you can reload the web page to see new version.

• For this Test stage, it will run the code first before to build docker image, if code correct, it will continue run the pipeline, if code have the bug, the pipeline will stop at Test stage, this method can avoid build the bug docker images.

Try Some Wrong Code

1.172. In AWS Cloud9 IDE, try to make some bug in app.py.

1.173. Use Git command to push code and see the pipeline workflow.

1.174. In the AWS Manage Console, click service, and click Codebuild.

1.175. Click codetest.

1.176. Click the latest build history.

1.177. At the Build logs, you can see the test process, and realize how it was fail.

1.178. You can also see the reason why this stage would fail in codepipeline by click Details.

Conclusion

Congratulations! You now have learned how to:

• Code by using AWS Cloud9.
• Integrate AWS Cloud9 with AWS CodeCommit.
• Create Amazon ECS using Fargate type.
• Use AWS Codepipeline, AWS CodeBuild, and Amazon ECS to implement an automatic CI/CD workflow with Container.