AWS Deployment Methodologies: EC2, Containers, and Serverless

Learning Objectives

After studying this guide, you should be able to:

• Differentiate between mutable and immutable deployment patterns across various compute platforms.
• Select the appropriate deployment strategy (Blue/Green, Canary, Rolling) based on business resiliency requirements.
• Configure CodeDeploy components, including Deployment Groups and AppSpec files, for EC2, ECS, and Lambda.
• Analyze deployment lifecycle events to troubleshoot failed updates.
• Implement cross-Region and Multi-AZ deployment patterns to minimize downtime.

Key Terms & Glossary

• Mutable Infrastructure: Infrastructure that is updated in place. For example, SSH-ing into an EC2 instance to update software versions.
• Immutable Infrastructure: Infrastructure that is never modified after deployment. Instead, new instances or containers are launched from a fresh image (AMI or Docker image) to replace the old ones.
• Canary Deployment: A strategy where a small percentage of traffic is shifted to the new version to test stability before a full rollout.
• Blue/Green Deployment: A strategy that involves two identical environments. "Blue" is the current production, and "Green" is the new version. Traffic is toggled once Green is validated.
• AppSpec File: A YAML or JSON formatted file used by CodeDeploy to manage the deployment hooks and permissions.
• Deployment Group: A set of individual instances or a service (like ECS) targeted for a deployment.

The "Big Idea"

In the AWS DevOps Professional domain, deployment is not just about moving code; it is about orchestrating risk. By moving from mutable, manual updates to automated, immutable pipelines, teams can achieve high availability and rapid recovery. Whether you are managing long-running EC2 instances, elastic container clusters (ECS/EKS), or event-driven Lambda functions, the goal remains the same: Zero-downtime updates with automated rollbacks.

Formula / Concept Box

Hierarchical Outline

1. Deployment Foundations
   • Mutable vs. Immutable: Risks of "configuration drift" in mutable setups.
   • Artifact Management: Using S3, ECR, and CodeArtifact for versioned source control.
2. Platform-Specific Methodologies
   • Amazon EC2: Uses CodeDeploy Agent; supports In-place and Blue/Green.
   • Amazon ECS: Managed via Task Sets; primary strategy is Blue/Green via CodeDeploy.
   • AWS Lambda: Uses Traffic Shifting (Aliases); Linear vs. Canary updates.
   • Amazon EKS: Utilizes Kubernetes-native strategies (RollingUpdate) or external tools like Flux/ArgoCD.
3. The Deployment Lifecycle (CodeDeploy Hooks)
   • BeforeInstall / AfterInstall: Pre/Post scripts for dependencies.
   • ApplicationStart / ValidateService: Checking health before routing traffic.

Visual Anchors

Deployment Strategy Selection

Blue/Green Traffic Shifting Concept

Definition-Example Pairs

• Immutable Pattern
  • Definition: A deployment approach where you never patch a server, but replace it entirely.
  • Example: Using EC2 Image Builder to create a new AMI for every software update, ensuring every instance in a cluster is bit-for-bit identical.
• Traffic Shifting
  • Definition: The process of incrementally moving user requests from an old version of a Lambda function to a new version.
  • Example: Updating a Lambda function alias to point 10% of traffic to version 2 (Canary) for 10 minutes before shifting the remaining 90%.

Worked Examples

Scenario: Configuring a Lambda Canary Deployment

Goal: Deploy a new Lambda function version where 10% of traffic is sent to the new version for a 5-minute bake period.

1. Version the Function: Publish a new version of the Lambda code.
2. Create an Alias: Point an alias (e.g., prod) to the old version.
3. Define AppSpec:
   yamlCopy

   version: 0.0 Resources: - myLambdaFunction: Type: AWS::Lambda::Function Properties: Name: "MyProcessor" Alias: "prod" CurrentVersion: "1" TargetVersion: "2" Hooks: - PreTraffic: "BeforeAllowTrafficLambdaFunction" - PostTraffic: "AfterAllowTrafficLambdaFunction"
4. Deployment Configuration: Select LambdaCanary10Percent5Minutes in CodeDeploy.

Checkpoint Questions

1. Which CodeDeploy lifecycle hook is used to verify that a service is running correctly before it starts receiving production traffic?
2. What is the main advantage of an Immutable deployment over a Mutable one regarding "configuration drift"?
3. In an ECS Blue/Green deployment, what AWS service handles the actual rerouting of traffic?

[!TIP] Answers: 1. ValidateService. 2. It eliminates drift because instances are replaced rather than modified. 3. Application Load Balancer (ALB).

Muddy Points & Cross-Refs

• ECS vs. EKS Scaling: While ECS is AWS-native and integrates deeply with CodeDeploy for Blue/Green, EKS often requires Kubernetes Autoscalers (HPA/VPA) and differs in how it handles rolling updates (via Deployment manifests).
• Lambda@Edge: Note that Lambda@Edge (used for CloudFront) does not support CodeDeploy traffic shifting aliases; deployments here are managed through CloudFront distribution updates.

Comparison Tables

Platform Comparison: CodeDeploy Capabilities

Comparison: Mutable vs. Immutable