Lab: Automating Application Delivery with AWS CI/CD Pipelines

In this lab, you will transition from manual deployments to a fully automated Continuous Integration and Continuous Delivery (CI/CD) workflow. Following the DevOps philosophy of TodoPlus Limited, you will set up a pipeline that automatically builds and deploys a Lambda function whenever code is pushed to a version control repository.

[!WARNING] Remember to run the teardown commands at the end of this lab to avoid ongoing charges. CI/CD resources often incur costs if left active.

Prerequisites

Before starting, ensure you have:

An AWS Account with administrative access.
AWS CLI installed and configured locally with aws configure.
Git installed on your local machine.
IAM Permissions: Ability to create IAM Roles, CodeCommit repositories, CodeBuild projects, and CodePipelines.

Learning Objectives

By the end of this lab, you will be able to:

Create and manage source code using AWS CodeCommit.
Define build specifications and automate artifact creation with AWS CodeBuild.
Orchestrate a multi-stage deployment workflow using AWS CodePipeline.
Implement automated triggers that respond to code commits.

Architecture Overview

The following diagram illustrates the workflow you will build. A developer pushes code to CodeCommit, which triggers CodePipeline. CodePipeline then uses CodeBuild to package the application and prepares it for deployment.

Loading Diagram...

Step-by-Step Instructions

Step 1: Create the Source Repository

We need a central place to store our application code using Git principles.

bash

aws codecommit create-repository --repository-name brainybee-lab-repo --repository-description "Lab repository for CI/CD"

▶Console alternative

Open the AWS CodeCommit console.
Choose Create repository.
Name it brainybee-lab-repo and click Create.

Step 2: Initialize Local Code and Buildspec

CodeBuild requires a buildspec.yml file to know how to compile and package your code.

Create a local folder and initialize git:

bash

mkdir cicd-lab && cd cicd-lab
git init

Create a simple Lambda function index.js:

javascript

exports.handler = async (event) => {
    return {
        statusCode: 200,
        body: JSON.stringify('Hello from the CI/CD Pipeline!'),
    };
};

Create buildspec.yml in the root directory:

yaml

version: 0.2
phases:
  install:
    runtime-versions:
      nodejs: 18
  build:
    commands:
      - echo "Building the application..."
      - zip -r deployment_package.zip index.js
artifacts:
  files:
    - deployment_package.zip

Step 3: Create the CodeBuild Project

Now we define the environment where the build happens.

bash

# Note: This assumes you have a service role. For lab simplicity, use the console to auto-generate the role.

▶Console instructions (Recommended for Role generation)

Navigate to CodeBuild > Build projects > Create build project.
Project Name: brainybee-build-project.
Source: AWS CodeCommit > brainybee-lab-repo.
Environment: Managed Image > Amazon Linux 2 > Standard > Runtime: Node.js.
Artifacts: Select No artifacts (we will configure this via CodePipeline later).
Buildspec: Use the buildspec.yml file in the root.
Click Create build project.

Step 4: Create the CodePipeline

This orchestrates the flow from CodeCommit to CodeBuild.

▶Console instructions

Navigate to CodePipeline > Pipelines > Create pipeline.
Pipeline name: brainybee-deployment-pipeline.
Source Stage: Source provider: AWS CodeCommit; Repository: brainybee-lab-repo; Branch: master (or main).
Build Stage: Build provider: AWS CodeBuild; Project name: brainybee-build-project.
Deploy Stage: For this lab, select Skip deploy stage (we are focusing on CI/Build automation).
Review and Create pipeline.

Checkpoints

Checkpoint 1: Source Connection

Verify your local files are in CodeCommit.

Action: Run git push to your CodeCommit remote.
Expected Result: CodeCommit console shows index.js and buildspec.yml.

Checkpoint 2: Pipeline Trigger

Verify the automation starts.

Action: Check the CodePipeline dashboard immediately after the push.
Expected Result: The Source stage should turn blue (In Progress) and then green (Succeeded), followed by the Build stage.

Teardown

[!IMPORTANT] To avoid charges, delete the following resources in order:

CodePipeline:
bash
aws codepipeline delete-pipeline --name brainybee-deployment-pipeline
CodeBuild Project:
bash
aws codebuild delete-project --name brainybee-build-project
CodeCommit Repository:
bash
aws codecommit delete-repository --repository-name brainybee-lab-repo
S3 Artifact Bucket: Delete the bucket created by CodePipeline (starts with codepipeline-us-east-1-...).

Troubleshooting

Error	Likely Cause	Solution
`Access Denied`	Missing IAM permissions for CodeBuild or Pipeline.	Ensure the Service Role has `AWSCodeBuildAdminAccess` and `S3FullAccess`.
`Buildspec not found`	File named incorrectly or not in root.	Ensure file is exactly `buildspec.yml` in the base directory.
`Git Push Failed`	Missing Git-Remote-CodeCommit helper.	Install `pip install git-remote-codecommit` or use HTTPS credentials.

Stretch Challenge

The "Full CD" Challenge: Modify the CodePipeline to include a Deploy Stage using AWS CodeDeploy.

Create an AWS Lambda function.
Update your buildspec.yml to produce an appspec.yml file.
Add a deployment stage to the pipeline that updates the Lambda function code using the zip file produced in the Build stage.

Cost Estimate

Service	Cost Detail
AWS CodeCommit	First 5 active users are free.
AWS CodeBuild	100 build minutes (build.general1.small) free per month.
AWS CodePipeline	1 free pipeline per month; $1.00 per active pipeline thereafter.
S3 Storage	Standard storage rates ($0.023/GB) for build artifacts.

Total Estimated Lab Cost: $0.00 (If within Free Tier limits).

Concept Review

Understanding the "DevOps Loop" in AWS:

\begin{tikzpicture}[node distance=2cm] \draw[thick, ->] (0,1) arc (90:-270:1cm); \node at (0,1.3) {Plan & Code}; \node at (1.5,0) {Build}; \node at (0,-1.3) {Deploy}; \node at (-1.5,0) {Monitor}; \draw[dashed] (0,0) circle (0.5cm); \node at (0,0) {\tinyCI/CD}; \end{tikzpicture}

Continuous Integration (CI): Automating the merge and build process. (CodeCommit + CodeBuild).
Continuous Delivery (CD): Ensuring the code is always in a deployable state. (CodePipeline).
Continuous Deployment: Automatically pushing every change to production. (CodePipeline + CodeDeploy).