CDK v2 Example (Custom Build Image)

This example shows how to use CDK v2 and how to provision a lambda function with a custom build image.

The project is written in Typescript.

---

Table of Contents

1. Introduction
2. Bootstrap
3. Deployment
4. Staging
5. Useful Commands

---

Introduction

The example contains:

• CI/CD pipeline using CodePipeline
• passing a custom NPM token to the actual build to enable private NPM repository access for npm dependencies
• stage-support for the pipeline.
• System Manager Parameter store lookup: Look up a value from the SSM paramter store (in this example: npm token to access private npm repositories)
• custom build image using a Dockerfile
• two lambda functions using the same docker build image and ARM64 Graviton processor configuration

---

Bootstrap

Requirements

In order to build and deploy the project, you need:

• npm
• typescript: npm install -g typescript
• cdk: npm install -g aws-cdk

Install dependencies:

    npm install

Build the project

Now, you can build the project:

    npm run build

Custom Build Image

You can modify your custom build image here: docker/Dockerfile The docker image is used to build your lambda functions.

---

Deployment

After a successful build, you can start the CDK deployment:

    cdk deploy

---

Staging

A word to staging: There are several ways to achieve staging for a CDK based CI/CD pipeline:

1. add multiple stages to your CI / CD pipeline and either deploy automatically to those or use a manual confirmation to do so
2. add another stack here: /bin/cdk-v2-service.ts

While the first approach is the way to go for solo-branch repositories, I'd recommend to use multiple branches and maintain a separate CI / CD pipeline for each branch.

You can see, that I define two stacks in /bin/cdk-v2-service.ts one for each branch and stage. I pass on some values for latter use to define stack name, stage name and target branch to use.

You can use cdk deploy and define which stack you want to deploy: cdk deploy my-stack-prod.

The CI/CD pipeline will make use of the stack deployment and configure its self-mutate step accordingly to update the proper (staged) stack.

Just make sure to push your code to two branches and configure the proper branch to use in lib/pipeline.ts here:

    input: CodePipelineSource.codeCommit(Repository.fromRepositoryName(this, 'CodeRepository', 'cdk-v2-service'), 'master'),

Just change the master branch to what ever you want. Adapt the source to your source provider; in this case I use CodePipeline, but you can easily use another source provider here (e.g. GitHub).

---

Useful commands

• npm run build compile typescript to js
• npm run watch watch for changes and compile
• npm run test perform the jest unit tests
• cdk deploy deploy this stack to your default AWS account/region
• cdk diff compare deployed stack with current state
• cdk synth emits the synthesized CloudFormation template