LearningTree · AWS · DevOps & Developer Tools
DevOps & Developer Tools —
Automate Everything from Code to Cloud
CloudFormation, CDK, CodePipeline, CodeBuild, CodeDeploy, CodeCommit — infrastructure as code and CI/CD services that automate the path from code commit to production deployment.
01
Chapter One
Why DevOps on AWS?
DevOps on AWS means automating every step from infrastructure provisioning (IaC) to application deployment (CI/CD). The goal: push code → automatically build, test, and deploy to production — with rollback if anything fails.
The Two Pillars of AWS DevOps
Infrastructure as Code (IaC)
- Define infrastructure in code (templates or programs)
- Version-controlled, repeatable, testable
- Create identical environments on demand
- Delete a stack → all resources cleaned up
- Services: CloudFormation, CDK
CI/CD Pipeline
- Automate build, test, and deploy on every commit
- Catch bugs early with automated testing
- Deploy with blue/green, canary, or rolling strategies
- Rollback automatically on failure
- Services: CodePipeline, CodeBuild, CodeDeploy
The DevOps Spectrum — Define to Deploy
← Infrastructure Application deployment →
CloudFormation
IaC (YAML)
CDK
IaC (Code)
CodeBuild
Build & test
CodeDeploy
Deploy
02
Chapter Two
Services
Core DevOps Services
Infrastructure as Code
AWS CloudFormation
Declarative IaC using YAML/JSON templates. Define stacks of AWS resources, deploy with automatic rollback, detect drift, and manage multi-account environments.
Deep dive →
Infrastructure as Code
AWS CDK
Programmatic IaC in Python, Java, TypeScript. Write infrastructure with loops, classes, and unit tests — CDK synthesizes to CloudFormation and deploys.
Deep dive →
CI/CD
CodePipeline · CodeBuild · CodeDeploy · CodeCommit · CodeArtifact
End-to-end CI/CD — source control, build automation, artifact management, and deployment with blue/green, canary, and rolling strategies.
Deep dive →
Services at a Glance
| Service | Category | What It Does | Best For |
|---|---|---|---|
| CloudFormation | IaC | Declarative templates (YAML/JSON) → AWS resources | All AWS teams; exam context; simple-to-medium complexity |
| CDK | IaC | Programmatic IaC (Python/Java/TS) → synthesizes to CFN | Dev teams; complex logic; reusable constructs; unit testing infra |
| CodePipeline | CI/CD | Orchestrates end-to-end release pipelines | Automating source → build → test → deploy workflows |
| CodeBuild | CI/CD | Fully managed build service (compile, test, package) | Running builds without managing Jenkins/build servers |
| CodeDeploy | CI/CD | Automated deployments to EC2, ECS, Lambda | Blue/green, canary, and rolling deployments with rollback |
| CodeCommit | Source | Managed Git repositories | Private repos within AWS (being deprecated — prefer GitHub) |
| CodeArtifact | Artifacts | Managed artifact repository (npm, Maven, pip) | Caching and hosting private packages |
03
Chapter Three
Decision Guide
When to Use What
| If you need… | Use… |
|---|---|
| Define infrastructure in a simple declarative template | CloudFormation |
| Write infrastructure with loops, conditions, and unit tests | CDK |
| Orchestrate a multi-stage release pipeline | CodePipeline |
| Compile code, run tests, produce artifacts | CodeBuild |
| Deploy to EC2/ECS/Lambda with blue/green or canary | CodeDeploy |
| Host private Git repositories in AWS | CodeCommit (or GitHub) |
| Cache npm/Maven/pip packages privately | CodeArtifact |
| Multi-cloud IaC (AWS + Azure + GCP) | Terraform (not AWS-native) |
CloudFormation vs CDK — Quick Decision
| Scenario | CloudFormation | CDK |
|---|---|---|
| Simple stacks, small teams | ✅ Simpler to start | Overkill |
| Complex logic (loops, inheritance) | Limited | ✅ Full language power |
| Unit testing infrastructure | Not possible | ✅ pytest / JUnit / Jest |
| IDE autocomplete & type safety | Limited (YAML) | ✅ Full IDE support |
| Reusable components across teams | Nested stacks (meh) | ✅ Construct libraries |
| AWS exam context | ✅ Primary exam focus | Secondary |
04
Chapter Four
Pipeline Flow
How the Services Connect — End-to-End Pipeline
Source
CodeCommit / GitHub
→
Build
CodeBuild
→
Artifact
S3 / CodeArtifact
→
Deploy
CodeDeploy
→
Production
EC2 / ECS / Lambda
CodePipeline orchestrates this entire flow — connecting source, build, and deploy stages into an automated release pipeline that triggers on every code push.
IaC + CI/CD Together
IaC Pipeline
- Template changes pushed to Git
- CodePipeline triggers
- CloudFormation Change Set created
- Manual approval → execute
- Infrastructure updated safely
App Pipeline
- Application code pushed to Git
- CodeBuild compiles + tests
- Produces deployable artifact
- CodeDeploy rolls out (blue/green)
- Auto-rollback on health check failure
👉 Key Takeaway
AWS DevOps = IaC (CloudFormation/CDK) + CI/CD (CodePipeline/CodeBuild/CodeDeploy). Together they automate the entire lifecycle: infrastructure and application code flow from Git to production with testing, approval gates, and automatic rollback.