CodeDeploy pour EC2/Lambda

Deploying applications can be tricky. Manual steps are error-prone and can cause downtime. AWS CodeDeploy automates this process for you!

• It helps deploy applications to EC2 instances, AWS Lambda, and on-premises servers.
• It minimizes downtime during deployments.
• It provides options for easy rollbacks if something goes wrong.

CodeDeploy organizes your deployments using a few core components:

• Application: A unique name to identify the application you want to deploy.
• Deployment Group: A set of individual instances or a specific Lambda function to deploy to.
• Revision: The source code, web pages, executables, and scripts you want to deploy, packaged into a file (e.g., zip, tar).

CodeDeploy supports different strategies to update your application:

• In-place Deployment: Stops the application on each instance, deploys the new version, and restarts it. This might cause brief downtime.
• Blue/Green Deployment: A new set of instances (Green) is provisioned with the new version. Traffic is then shifted from the old instances (Blue) to the Green ones. This minimizes downtime and risk.

For EC2 deployments, CodeDeploy needs an agent installed on your instances. This agent handles the deployment tasks.

You define how the deployment happens using an appspec.yml file. This file tells CodeDeploy what files to copy, where to put them, and what scripts to run.

Here's a simple example of an appspec.yml for an EC2 deployment. It tells CodeDeploy to copy files and run a script.

Hooks are scripts that CodeDeploy runs at specific points during a deployment to an EC2 instance. They allow you to customize the deployment process:

• BeforeInstall: Tasks before files are copied.
• AfterInstall: Tasks after files are copied, like dependency installation.
• ApplicationStart: Start your application service.
• ValidateService: Check if the deployed application is working correctly.

CodeDeploy also simplifies deployments for AWS Lambda functions. It allows you to safely shift traffic between different versions of your function.

This is especially useful for canary deployments or gradual rollouts, ensuring your new function works as expected before taking all traffic.

For Lambda, the appspec.yml focuses on defining the function and how traffic should be shifted. Here's a basic example:

CodeDeploy automatically monitors deployments. If a deployment fails (e.g., a hook script exits with an error), it can automatically roll back to the previous stable version.

• You can also manually initiate a rollback.
• It integrates with CloudWatch to monitor deployment health and trigger alarms.

Which CodeDeploy deployment strategy involves provisioning a completely new set of instances for the new application version, then shifting traffic?

Great job! You've learned about AWS CodeDeploy, a powerful service for automating application deployments.

• It manages deployments to EC2, Lambda, and on-premises servers.
• You can choose between in-place or blue/green strategies.
• The appspec.yml file defines deployment steps and hooks.
• It supports automatic rollbacks and integrates with monitoring.

CodeDeploy helps ensure your applications are updated reliably and with minimal disruption.