Happy coding!

I've successfully migrated my AWS infrastructure for the fake-crypto-webapp application from manual CLI commands to Terraform! This upgrade improves consistency, reliability, and reduces operational overhead. 🚀 The new infrastructure is highly automated and scalable.

🔧 My Terraform configuration orchestrates a seamless CI/CD pipeline:

• 📂 CodeCommit Repository (fake-crypto-webapp): The application's source code is version-controlled here.

• 🏗️ CodeBuild Project (codebuild-fake-crypto-webapp): Changes in CodeCommit trigger CodeBuild, which builds a Docker image and pushes it to...

• 📦 ECR Repository (fake-crypto-webapp): The built Docker image is stored securely in this registry.

• 🔄 CodePipeline (crypto-app): This pipeline orchestrates the entire deployment process, starting with CodeCommit and deploying to...

• ☁️ ECS Cluster (khalil-lab-cluster): The application's containers run within this managed cluster, leveraging an autoscaling group for dynamic scaling.

• ⚙️ ECS Service (fake-crypto-webapp): This service manages deployment and scaling of containers, incorporating a deployment circuit breaker for resilience.

• 🔗 Application Load Balancer (khalil-lab-alb): The load balancer distributes traffic across healthy ECS service instances, ensuring high availability. Security is managed by the khalil-lab-sg security group, and various IAM roles provide secure access.

After diving deep into AWS services like ECR (Elastic Container Registry), CodeCommit, CodeBuild, and IAM roles, I’m excited to share my journey in developing a streamlined CI/CD pipeline! Below is a step-by-step outline of my process:

1. AWS Access Creation:
   Generate an AWS Access ID and secret access key to utilize AWS CLI commands.

2. Create ECR Repository:
   Set up an ECR repository for Docker images.

3. Code Storage in CodeCommit:
   Use CodeCommit to store code from my GitHub repository.

4. Build Process with CodeBuild:
   Integrate CodeBuild with the CodeCommit repository using a buildspec.yml file to define the build process.

5. Build Execution:
   Initiate the build and ensure the Docker image is successfully pushed to ECR. I aim to minimize console use by leveraging the AWS CLI.

To put theory into practice, I utilized the KodeKloud playground and implemented all AWS CLI commands in a bash script. You can find the script in my GitHub repository: CryptoAppAWS_CICD.

1. CodeCommit Credentials Issue:
   Faced hurdles with AWS CodeCommit as the KodeKloud playground doesn’t allow the creation of Git credentials, complicating code transfer between GitHub and CodeCommit.

   🔗 Check my experimental script: Issue 1 Script
   

2. Policy Permissions Error:
   Encountered access denied errors due to insufficient IAM user permissions for AWS CodeCommit when using my SSH public key.

   🔗 Check my experimental script: Issue 2 Script
   

3. Build Project Creation Error:
   Received an AccessDeniedException while creating the CodeBuild project due to the lack of iam:PassRole permission for the required service role.

   🔗 Check my experimental script: ./aws_scripts/issue_2.sh

To tackle these challenges, I adapted my strategy by manually creating the CodeBuild project. My script now retrieves the CodeBuild role and assigns the AmazonEC2ContainerRegistryPowerUser policy, enabling the buildspec.yml to successfully push the Docker image to ECR.

🔗 Check my AWS CICD script: ./aws_scripts/start.sh
📹 Watch the results: Demo Video

I am committed to overcoming obstacles to successfully implement a robust CI/CD pipeline using AWS services. Thank you for your support as I continue this exciting learning journey!

Feel free to connect or reach out if you have any questions or insights to share!

#AWS #CICD #DevOps #CloudComputing #LearningJourney