Chandra Meets CodeDeploy: My First AWS Deployment Journey

Title: Chandra's Initial AWS CodeDeploy Deployment Experience

AWS CodeDeploy is a fully managed deployment service offered by Amazon Web Services (AWS). This service enables developers to automate the deployment of application code, configuration files, scripts, and other application content to computing environments. CodeDeploy supports multiple platforms including Amazon EC2 and on-premises servers, AWS Lambda, and Amazon ECS. It can retrieve application revisions from sources like Amazon S3 and supported source-code repositories based on the deployment platform.

The purpose of AWS CodeDeploy was to simplify the deployment process that traditionally involved manual file copying and command execution, which could lead to configuration errors, inconsistent deployments, and application downtime. The service allows developers to define deployment instructions once and reuse them whenever a new application version is released.

Instead of manually performing deployment tasks, developers can let CodeDeploy handle installing the new version, executing required scripts, monitoring the deployment, and reporting the success or failure of the deployment.

CodeDeploy operates by following a simple pattern: storing the application code in a source location (such as GitHub, S3, or CodeCommit), describing the deployment steps in a file called appspec.yml that resides with the code, picking up the revision, copying it to the target instances (or updating the Lambda/ECS version), and running the lifecycle hooks defined in appspec.yml in order—stopping the old version, installing the new one, and starting it back up.

CodeDeploy can also monitor CloudWatch alarms during the rollout and automatically roll back if something appears unhealthy.

A key feature of AWS CodeDeploy is its ability to provide automated and consistent deployments. Once configured, deployments follow the same process every time, eliminating the risk of human error. Additionally, CodeDeploy supports multiple deployment strategies such as in-place deployments and blue/green deployments, allowing developers to choose between speed and safety depending on the project requirements.

The service also ensures automatic rollback on failure, which can be challenging to build reliably by hand. CodeDeploy works across various compute types, including EC2/on-premises servers, Lambda functions, and ECS containers, so developers do not need a different tool for each architecture.

For college students and small departments, this becomes particularly relevant. Most departments run small internal web apps, such as placement portals, event registration pages, club websites, or student project demos. Currently, these are deployed manually, which can be time-consuming and error-prone. An example use case is hosting student mini-project demos on a small EC2 instance, with CodeDeploy monitoring a GitHub repository.

Whenever a student pushes an update, CodeDeploy automatically redeploy the latest version to the demo server, saving time and reducing the risk of forgetting steps during a deadline. If the new version crashes, CodeDeploy can automatically roll back to the last known good version, ensuring the demo server remains operational.

A simple example of an appspec.yml file for deploying a small web app to EC2 is provided. This YAML file includes the operating system, file locations, and the hooks for BeforeInstall, AfterInstall, ApplicationStart, and ValidateService. The deployment can be triggered through the AWS CLI with a single command, and CodeDeploy manages the rest of the deployment process.

Key features of AWS CodeDeploy include automated deployment, multiple deployment strategies, automatic rollback on failure, and support across various compute types.