Deployments & Managing Infrastructure at Scale

Deployments & Managing Infrastructure at Scale

CloudFormation Overview

• What it is:
  • AWS service for Infrastructure as Code (IaC).
  • Declarative way of describing AWS resources (you state what you want, not how to create it).
  • Supports almost all AWS resources.
• Example use case:
  • Define: Security Group, 2 EC2 instances using that SG, an S3 bucket, a Load Balancer.
  • CloudFormation will automatically create all resources in the correct order with the exact configuration.

Benefits of CloudFormation

1. Infrastructure as Code
   • No manual creation of resources.
   • Changes reviewed via code review (improves control and governance).
2. Cost Advantages
   • Resources in a stack are automatically tagged for easier cost tracking.
   • Can estimate costs before deployment.
   • Enables cost-saving automation, e.g., delete stacks at 5 PM, recreate at 8 AM.
3. Productivity
   • Easily destroy and recreate infrastructure.
   • Generates diagrams of resources and their relationships.
   • Declarative programming: CloudFormation figures out dependencies (e.g., DynamoDB before EC2).
4. Reusability
   • Use existing templates from AWS docs or community.
   • Almost everything in AWS is supported.
   • If not, use Custom Resources.

Visualization

• Infrastructure Composer can visualize stacks.
• Example: WordPress stack diagram shows ALB Listener, DB, Security Groups, Launch Config, etc.
• Helps understand relations between components.

Exam Tips (AWS CCP)

• CloudFormation = Infrastructure as Code.
• Used for:
  • Repeating architectures across environments (dev, prod).
  • Deploying to multiple regions or AWS accounts.
• Remember: Declarative, automated, cost-efficient.

Key takeaway: CloudFormation is the core AWS service for Infrastructure as Code, making deployments reproducible, scalable, and cost-efficient.

AWS Cloud Development Kit (CDK)

What is CDK?

• AWS Cloud Development Kit (CDK) is a tool to define cloud infrastructure using familiar programming languages.
• Alternative to writing raw CloudFormation templates (YAML/JSON).
• Supported languages: Python, JavaScript, TypeScript, Java, .NET.

How It Works

1. Write infrastructure code in your chosen language (e.g., Python, TypeScript).
2. CDK compiles it into a CloudFormation template (JSON or YAML).
3. The template is then deployed via CloudFormation to provision resources.

Benefits of CDK

• Infrastructure as Code but in a programming language you already know.
• Type safety and familiar constructs.
• Can use loops, conditionals, reusable functions.
• Easier and faster to develop compared to raw YAML/JSON.
• Can deploy infrastructure and runtime code together.
  • Useful for Lambda functions, ECS/EKS containers, etc.

Example Use Case

• Write a CDK app in Python:
  • Define a Lambda function, DynamoDB table, other services.
  • CDK app → compiled into CloudFormation template → deployed.
• Example (TypeScript/JavaScript CDK snippet):
  • Define a VPC.
  • Define an ECS cluster.
  • Define an Application Load Balancer (ALB) with a Fargate service.
  • CDK compiles this into a CloudFormation template for deployment.

Exam Tips (AWS CCP)

• CDK is built on top of CloudFormation.
• Lets you use familiar languages instead of YAML/JSON.
• Still results in a CloudFormation template under the hood.
• Best for developers who want:
  • Reusable code patterns.
  • Type safety.
  • To deploy infra + app code together.

Key takeaway: CDK = CloudFormation, but written in real programming languages, giving you more flexibility, reusability, and speed.

AWS Elastic Beanstalk

Context and Motivation

• Most web apps on AWS follow a 3-tier architecture:
  • Load Balancer (multi-AZ for HA).
  • EC2 instances in an Auto Scaling Group.
  • Database (Amazon RDS).
  • Optional cache (ElastiCache for sessions/cached data).
• This can be built manually or via CloudFormation.
• But developers usually don’t want to manage infrastructure, they just want to deploy code and let AWS handle scaling, load balancing, and provisioning.

What is Elastic Beanstalk?

• Elastic Beanstalk (EB) = Platform as a Service (PaaS) on AWS.
• Provides a developer-centric view for deploying apps.
• Under the hood it still uses:
  • EC2, ASG, ELB, RDS, etc.
• You focus only on application code.
• AWS handles provisioning, scaling, monitoring, health checks, and deployment strategies.

Key Features

• Managed Service (free to use, pay for underlying resources).
• Handles:
  • EC2 OS setup and configuration.
  • Deployment strategy (configurable).
  • Auto Scaling and Load Balancing.
  • Health monitoring and metrics (via CloudWatch).
• Developer responsibility = only application code.

Beanstalk Architectures

1. Single instance
   • Best for development/test environments.
2. Load balanced + Auto Scaling group
   • Best for production web apps.
3. Worker environments
   • For non-web apps (background jobs, workers).

Supported Platforms

• Programming languages and runtimes:
  • Go, Java, .NET, Node.js, PHP, Python, Ruby.
• Container support:
  • Docker (single, multi, preconfigured images).
• Many application stacks are supported, making it flexible.

Health Monitoring

• Beanstalk includes a health agent on each EC2 instance.
• Pushes metrics to CloudWatch.
• Beanstalk console provides:
  • Application health checks.
  • Monitoring dashboards.
  • Health event logs.

Cloud Service Models Recap

• IaaS (Infrastructure as a Service) → EC2, RDS, etc.
• PaaS (Platform as a Service) → Beanstalk.
• SaaS (Software as a Service) → Services where you only use the software, no infra management (e.g., WorkMail, Chime).

Exam Tips

• Elastic Beanstalk = PaaS.
• Developer responsibility = only app code.
• AWS handles infra (provisioning, scaling, monitoring).
• Free to use, but pay for underlying resources.
• Three deployment models: Single Instance, LB + ASG, Worker.
• Provides built-in health monitoring.

AWS CodeDeploy

What is CodeDeploy?

• A service to automatically deploy applications.
• More flexible and independent compared to Elastic Beanstalk or CloudFormation.
• You can deploy to:
  • EC2 instances
  • On-Premises servers

Why CodeDeploy?

• Helps upgrade applications (e.g., v1 → v2) seamlessly.
• Works in hybrid environments:
  • AWS Cloud (EC2).
  • On-premises data centers.
• Useful for organizations migrating to AWS, since it keeps a consistent deployment method.

Key Features

• Hybrid deployment service (EC2 + On-Premises).
• Supports any type of server, but:
  • You must provision servers yourself.
  • You must install the CodeDeploy Agent on those servers.
• Provides a single interface to manage deployments across environments.

Deployment Flow

1. Prepare your servers (EC2 or On-Premises).
2. Install the CodeDeploy Agent.
3. Define your application deployment (AppSpec file).
4. CodeDeploy handles automatic upgrade from one version to another.

Benefits

• Simplifies application updates across many servers.
• Ensures consistency in deployment process between cloud and on-premises.
• Helps in cloud migration by standardizing deployments.

Exam Tips

• CodeDeploy = automated app deployment service.
• Works with both EC2 and On-Premises servers (hybrid).
• Requires CodeDeploy Agent installed on target servers.
• You are responsible for server provisioning and configuration.
• Great for migrations and hybrid deployments.

AWS CodeCommit

What is CodeCommit?

• Fully-managed source control service by AWS.
• Based on Git (G-I-T).
• Similar to GitHub, GitLab, or Bitbucket, but hosted within AWS.

Purpose

• Store your application code securely in AWS.
• Acts as a version control repository.
• Used before pushing/deploying application code to servers.

Benefits

• Versioning: Tracks code changes automatically, with rollback capability.
• Collaboration: Makes it easy for developers to work together on projects.
• Fully managed: No need to manage infrastructure.
• Scalable and highly available: Runs on AWS infrastructure.
• Secure: Resides inside your AWS account, not a public platform.
• AWS integration: Works seamlessly with CodeBuild, CodeDeploy, CodePipeline, etc.

Exam Tips

• CodeCommit = Git-based repo service in AWS.
• Alternative to GitHub or other Git hosting platforms.
• Key value: private, secure, scalable, AWS-integrated source control.
• Great for organizations that want full control of their code inside AWS.

AWS CodeBuild

What is CodeBuild?

• Fully managed build service on AWS.
• Purpose: compile source code, run tests, and produce deployment-ready packages.
• Output = artifacts that can be deployed (e.g., by CodeDeploy).

How it Works

1. Developer pushes code → stored in CodeCommit (or another repo).
2. CodeBuild retrieves code.
3. Runs build scripts (defined by you, usually in buildspec.yml).
4. Compiles, tests, packages.
5. Produces artifacts ready for deployment.

Features

• Fully managed & serverless: no servers to manage.
• Continuously scalable: handles multiple builds concurrently.
• Highly available & secure: built on AWS infra.
• Pay-as-you-go pricing: only pay for build time.
• Automatic build triggers: runs every time code is updated in CodeCommit (or other source).

Benefits

• Automates the build and test process in CI/CD pipelines.
• Integrates tightly with CodeCommit, CodePipeline, and CodeDeploy.
• Lets developers focus on coding, while AWS handles build infrastructure.

Exam Tips

• CodeBuild = build service (compile, test, package).
• Works with CodeCommit (source) and CodeDeploy (deployment).
• Serverless, scalable, pay-as-you-go.
• Key output = artifacts ready for deployment.

AWS CodePipeline

What is CodePipeline?

• A fully managed orchestration service for automating the flow of code changes.
• Connects CodeCommit, CodeBuild, and CodeDeploy into a pipeline.
• Ensures that code is automatically built, tested, and deployed to production.

Use Case Example

1. Developer pushes code to CodeCommit.
2. CodeBuild compiles code and runs tests.
3. CodeDeploy (or Elastic Beanstalk) deploys the application to servers.
4. Orchestration is handled by CodePipeline.

Key Features

• CICD (Continuous Integration & Continuous Delivery):
  • Every code push triggers build, test, and deploy.
• Fully Managed: No servers to manage.
• Integrations: Works with AWS services and third-party tools.
  • AWS: CodeCommit, CodeBuild, CodeDeploy, Elastic Beanstalk, CloudFormation.
  • External: GitHub, custom plugins, third-party services.
• Benefits:
  • Fast delivery, rapid updates.
  • Central to AWS CICD ecosystem.

Exam Tip

• Anytime you see orchestration of pipelines in AWS exam questions, think AWS CodePipeline.

AWS CodeArtifact

What is CodeArtifact?

• A fully managed artifact management service by AWS.
• Stores and retrieves software packages (dependencies).
• Secure, scalable, and cost-effective alternative to running your own artifact system.

Why Do We Need It?

• Software projects often rely on code dependencies (e.g., libraries, frameworks).
• Traditionally:
  • Set up custom artifact management on Amazon S3 or EC2.
  • This adds complexity and maintenance overhead.
• CodeArtifact removes that burden by providing a ready-to-use solution.

Key Features

• Integrates with common dependency managers:
  • Java: Maven, Gradle
  • JavaScript: npm, Yarn
  • Python: pip, twine
  • .NET: NuGet
• Central secure repository for teams.
• Works seamlessly with CodeCommit and CodeBuild:
  • CodeBuild can fetch dependencies directly from CodeArtifact.

Benefits

• Fully managed by AWS (no servers to manage).
• Secure by default.
• Scales with your project.
• Pay-as-you-go pricing.

Exam Tip

• If a question mentions artifact management system or storing dependencies → answer is AWS CodeArtifact.

AWS Developer Tools Comparison

AWS Systems Manager (SSM)

• AWS Systems Manager (SSM) helps manage EC2 instances and on-premises servers at scale.
• A hybrid AWS service (works for both AWS and on-prem).
• Provides operational insights and a suite of 10+ tools.

Key Features

• Hybrid support: Works for AWS and on-premises servers/VMs.
• Automation: Run commands across the fleet.
• Patch Management: Automated patching for compliance.
• Configuration Management:
  • Store configurations with SSM Parameter Store.
  • Apply consistent configurations across servers.
• Cross-platform support: Linux, Windows, macOS, Raspberry Pi.

How It Works

1. SSM Agent must be installed on instances/servers.
   • Runs in the background, communicates with SSM.
   • Preinstalled on Amazon Linux AMI and some Ubuntu AMIs.
2. The agent reports back to the SSM service in AWS.
3. Once installed:
   • Run commands across all servers.
   • Patch the fleet.
   • Apply consistent configurations.

Exam Tips

• Keywords: fleet management, patching, hybrid environment, run command.
• If asked how to patch or run commands across EC2 + on-prem servers → Answer: SSM.
• If an instance isn’t managed by SSM → likely an agent issue.

Benefits

• Simplifies management at scale.
• Hybrid ready.
• Improves compliance and security.
• Centralized operations for multiple environments.

AWS Systems Manager – Parameter Store

Purpose

• Central place to store configuration and secrets securely.
• Examples: API keys, passwords, application configs.
• Serverless, scalable, durable, secure.

Key Features

1. Storage Options
   • String: plain text value.
   • StringList: list of values.
   • SecureString: encrypted using AWS KMS (for sensitive data like passwords).
2. Security
   • Access controlled by IAM policies (who can read/write specific parameters).
   • Optional encryption with KMS for secure parameters.
3. Versioning
   • Every time you edit a parameter, a new version is created.
   • Useful for tracking and rollback.
4. Integration
   • Applications, EC2, and Lambda functions can retrieve parameters directly.
   • Reduces need to hardcode secrets in code or environment variables.
5. Tiers
   • Standard (free): basic usage.
   • Advanced (paid): higher throughput, larger sizes, more features.

Hands-On Example (from transcript)

• Create parameter named demo parameter.
• Choose type: String, StringList, or SecureString.
• Store value like "my configuration parameter".
• Retrieve it anytime via console, CLI, or SDK.
• Parameter versions are maintained automatically.
• Can delete parameters anytime.

Benefits

• Centralized configuration management.
• Secure storage of sensitive values (encrypted with KMS).
• Easier auditing and tracking with version history.
• Simplifies deployments by removing secrets from code.

Exam Tip: Use Parameter Store when you need a secure, serverless, versioned, and IAM-controlled place to store configuration and secrets.

AWS Deployment and Developer Services

Deployment Services

• CloudFormation
  • AWS-only tool for Infrastructure as Code (IaC).
  • Works with almost all AWS resources.
  • Uses templates to deploy infra in multiple Regions and accounts.
  • Enables repeatable and automated deployments.
• Elastic Beanstalk
  • PaaS (Platform as a Service).
  • Limited to supported languages and Docker.
  • Deploy apps consistently with known architecture, e.g. Load Balancer + EC2 + RDS.
  • AWS manages scaling and environment details.
• CodeDeploy
  • Deploys and updates apps onto servers (EC2 or on-premises).
  • Hybrid service (works AWS + on-prem).
  • Automates deployments at scale.
• Systems Manager
  • Hybrid service (AWS + on-premises).
  • Manage servers at scale (patch, configure, run commands).
  • Useful for fleet management.

Developer Services

• CodeCommit
  • Private Git-based repo in AWS.
  • Secure, version-controlled code storage.
• CodeBuild
  • Serverless build service.
  • Compiles, tests, and packages code.
• CodeDeploy
  • Repeated here because it bridges deployment + developer services.
  • Deploys app code to servers.
• CodePipeline
  • Orchestration tool for CI/CD.
  • Automates flow: Source → Build → Test → Deploy → Provisioning.
• CodeArtifact
  • Secure, managed storage for software packages & dependencies.
  • Supports multiple ecosystems (npm, Maven, pip, etc).
• AWS CDK (Cloud Development Kit)
  • Define infra using real programming languages (TS, JS, Python, Java, etc).
  • Compiles into CloudFormation templates for deployment.

Exam Tips:

• CloudFormation vs CDK: CDK uses code, CloudFormation uses YAML/JSON.
• Elastic Beanstalk: Quick PaaS deployment with limited flexibility.
• CodeDeploy: Hybrid deployment (on-prem + AWS).
• CodePipeline: End-to-end automation of CI/CD.
• CodeArtifact: Think npm registry but inside AWS.
• Systems Manager: Manage fleet of servers, both AWS + on-prem.