How to Set Up CI/CD on AWS

📘 Chapter 4: Building a Complete Pipeline with CodePipeline

🔍 Overview

AWS CodePipeline is a fully managed continuous integration and delivery (CI/CD) service that automates the steps required to release software changes. It connects your source repository, build process, test scripts, and deployment mechanism into a visual and programmable pipeline.

In this chapter, you’ll learn how to:

• Create a multi-stage pipeline using CodePipeline
• Integrate CodeCommit, GitHub, CodeBuild, and CodeDeploy
• Visualize pipeline executions and manage versions
• Troubleshoot pipeline errors
• Secure and optimize your pipeline for real-world usage

🧠 1. What is AWS CodePipeline?

CodePipeline automates the workflow for:

• Source Code Retrieval
• Compilation and Testing
• Deployment to Development/Staging/Production

✅ Benefits

🧱 2. Pipeline Structure & Components

A typical CodePipeline has 3 core stages:

1. Source – Pulls code from GitHub/CodeCommit
2. Build – Compiles and tests using CodeBuild
3. Deploy – Deploys artifacts to EC2/Lambda/ECS using CodeDeploy

🧩 Pipeline Workflow

text

CodeCommit / GitHub

        ↓

   CodePipeline

     ↓       ↓

Build (CodeBuild)  →  Deploy (CodeDeploy)

🛠️ 3. Create a Pipeline via Console (Step-by-Step)

✅ Step 1: Open AWS Console → CodePipeline → Create Pipeline

• Enter pipeline name: MyAppPipeline
• Choose New Service Role (auto-managed IAM)
• Enable Artifact Store: S3 bucket is created automatically

✅ Step 2: Configure Source Stage

Authorize GitHub if needed and select your repository.

✅ Step 3: Add Build Stage (CodeBuild)

• Provider: AWS CodeBuild
• Region: (same as pipeline)
• Project name: Select existing or create new
• Input artifacts: SourceArtifact
• Output artifacts: BuildArtifact

✅ Step 4: Add Deploy Stage (CodeDeploy)

• Deployment provider: AWS CodeDeploy
• Application name: MyApp
• Deployment group: MyAppDG
• Input artifact: BuildArtifact

✅ Step 5: Review and Create

Click Release change to trigger your first execution.

⚙️ 4. Creating CodePipeline via CLI

bash

aws codepipeline create-pipeline --cli-input-json file://pipeline.json

📄 Sample pipeline.json

json

{

  "pipeline": {

    "name": "MyAppPipeline",

    "roleArn": "arn:aws:iam::123456789012:role/AWS-CodePipeline-Service",

    "artifactStore": {

      "type": "S3",

      "location": "my-codepipeline-bucket"

    },

    "stages": [

      {

        "name": "Source",

        "actions": [{

          "name": "SourceAction",

          "actionTypeId": {

            "category": "Source",

            "owner": "AWS",

            "provider": "CodeCommit",

            "version": "1"

          },

          "outputArtifacts": [{ "name": "SourceArtifact" }],

          "configuration": {

            "RepositoryName": "MyWebAppRepo",

            "BranchName": "main"

          }

        }]

      },

      {

        "name": "Build",

        "actions": [{

          "name": "BuildAction",

          "actionTypeId": {

            "category": "Build",

            "owner": "AWS",

            "provider": "CodeBuild",

            "version": "1"

          },

          "inputArtifacts": [{ "name": "SourceArtifact" }],

          "outputArtifacts": [{ "name": "BuildArtifact" }],

          "configuration": {

            "ProjectName": "MyBuildProject"

          }

        }]

      },

      {

        "name": "Deploy",

        "actions": [{

          "name": "DeployAction",

          "actionTypeId": {

            "category": "Deploy",

            "owner": "AWS",

            "provider": "CodeDeploy",

            "version": "1"

          },

          "inputArtifacts": [{ "name": "BuildArtifact" }],

          "configuration": {

            "ApplicationName": "MyApp",

            "DeploymentGroupName": "MyAppDG"

          }

        }]

      }

    ]

  }

}

🔁 5. Automating Release Triggers

• Use webhooks (enabled automatically for GitHub)
• Manual release: Click “Release Change”
• EventBridge: Automate scheduled or chained triggers

👮‍♂️ 6. Securing Your Pipeline

• Use IAM roles with least privileges
• Encrypt artifacts in S3 with KMS
• Store credentials securely using Secrets Manager
• Approve deployments manually in critical environments

✅ Manual Approval Example

Add a “Manual Approval” action before deploy:

json

{

  "category": "Approval",

  "provider": "Manual",

  "configuration": {

    "CustomData": "Approve production deployment"

  }

}

🧪 7. Monitoring, Debugging & Logs

✅ Pipeline Monitoring Tools

✅ Common Issues

📋 Summary Table – Pipeline Workflow