Why using docker?

There are two advantages.

1. Can be deployed beyond Lambda's capacity limits.
2. Easy version management by building Docker images.

AWS Lambda has 250MB limit storage when uploading zip file.
This limit storage also includes storage of Lambda layer.
Docker can overcome this problem by allowing you to deploy container images of up to 10GB.

Cautions : In these document assume that is conducted using AWS base image for Lambda and using node js.

Prerequisites

Aws Document link(Working with Lambda container images)

In order to deploy lambda by using docker, there are three prerequisites.

1. AWS Command Line Interface(AWS CLI)
2. Docker
3. Node js

AWS Command Line Interface is needed to access AWS ECR.
The Built image by using docker should be attached to AWS ECR.

The Docker is neeeded for building code to docker.
In these document, we will deploy lambda by using node js.

Start

1. Make Dockerfile

"Dockerfile" is set of instructions when building code to images.
This file has to be made in top path of your project.
We can make file named "Dockerfile".
This file can be recognized by installing Docker.(Please watch above Prerequisites #2)
Here are some examples of "Dockerfile"

Example 1.

FROM public.ecr.aws/lambda/nodejs:20

# Copy function code
COPY index.js ${LAMBDA_TASK_ROOT}
  
# Set the CMD to your handler (could also be done as a parameter override outside of the Dockerfile)
CMD [ "index.handler" ]

You can set runtime nodejs version by using "FROM".(Role that specifies the base image)
You can use "COPY" in order to move file in Lambda.
It is responsible for copying files or directories from the host system into the container.
"CMD" is set the entry point of Lambda.

Example 2.

# First stage
FROM public.ecr.aws/lambda/nodejs:18 as stage

# Install unzip
RUN yum install -y -q sudo unzip

ENV CHROMIUM_VERSION=1002910
# Download and extract necessary files to /opt/chrome directory
RUN curl "https://www.googleapis.com/download/storage/v1/b/chromium-browser-snapshots/o/Linux_x64%2F$CHROMIUM_VERSION%2Fchrome-linux.zip?generation=1652397748160413&alt=media" > /tmp/chromium.zip && \
    unzip /tmp/chromium.zip -d /tmp/ && \
    mv /tmp/chrome-linux/ /opt/chrome && \
    rm /tmp/chromium.zip

# Download and extract necessary files to /opt/chromedriver directory
RUN curl "https://www.googleapis.com/download/storage/v1/b/chromium-browser-snapshots/o/Linux_x64%2F$CHROMIUM_VERSION%2Fchromedriver_linux64.zip?generation=1652397753719852&alt=media" > /tmp/chromedriver_linux64.zip && \
    unzip /tmp/chromedriver_linux64.zip -d /tmp/ && \
    mv /tmp/chromedriver_linux64/chromedriver /opt/chromedriver && \
    rm /tmp/chromedriver_linux64.zip

# Second stage
FROM public.ecr.aws/lambda/nodejs:18 as base

COPY chrome-deps.txt /tmp/
RUN yum install -y $(cat /tmp/chrome-deps.txt)

# Copy /opt/chrome and /opt/chromedriver built in the previous stage to the current image
COPY --from=stage /opt/chrome /opt/chrome
COPY --from=stage /opt/chromedriver /opt/chromedriver

# Continue with the remaining tasks
COPY package.json package-lock.json ./
COPY src ./src
COPY index.js .

RUN npm install

CMD [ "index.handler" ]

This sample is conducted to downloading Chrome and Chromedriver while building code to images.
The "RUN" command allows you to apply new file system changes to layers in your Docker image and have the results reflected in the next build step.
The "ENV" command allows you to set env.
You can use this command to define environment variables, whose values you can use later.
And also you can set stage(FROM ~ as stage/base).
This can allow you to improved speed when building docker.

2. Make ECR private repository on AWS

Here is link.

How to Create Elastic Container Repository(ECR) in AWS

3. Set Git Action.

Example 1.

name: Deploy Lambda Function

on:
  push:
    branches:
      - main

jobs:
  deploy:
    name: build and deploy lambda
    strategy:
      matrix:
        node-version: [18.x]
    runs-on: ubuntu-latest

    steps:
      - uses: actions/checkout@v1
      - name: Use Node.js ${{ matrix.node-version }}
        uses: actions/setup-node@v1
        with:
          node-version: ${{ matrix.node-version }}

      - name: npm install and build
        run: |
          npm ci
          npm run build --if-present
        working-directory: ./
        env:
          CI: true

      - name: tsc
        run: tsc

      - name: Install AWS CLI
        run: |
          sudo apt-get update
          sudo apt-get install -y awscli

      - name: Configure AWS Credentials
        uses: aws-actions/configure-aws-credentials@v1
        with:
          aws-access-key-id: ${{ secrets.AWS_ACCESS_KEY_ID }}
          aws-secret-access-key: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
          aws-region: ca-central-1

      # Uncomment the following section to log in to AWS ECR
      - name: AWS ECR Login
        run: aws ecr get-login-password --region ${{ secrets.AWS_REGION }} | docker login --username AWS --password-stdin ${{ secrets.AWS_REGISTRY_URL }}

      - name: Build Docker Image
        run: docker build --platform linux/x86_64 -t your-build-name .
        
      - name: Tag Docker Image
        run: docker tag your-build-name:latest ${{ secrets.AWS_REGISTRY_URL }}:latest

      - name: Push Docker Image to ECR
        run: docker push ${{ secrets.AWS_REGISTRY_URL }}:latest

      - name: Update Lambda Function
        run: aws lambda update-function-code --function-name your-lambda-function-name --image-uri ${{ secrets.AWS_REGISTRY_URL }}:latest

      - name: Delete Untagged Images In ECR
        run: |
          # Get untagged image digest
          UNTAGGED_IMAGES=$(aws ecr describe-images --repository-name your-ecr-repository-name --query 'imageDetails[?imageTags==null].imageDigest' --output json)

          # Delete untagged images
          if [ -n "$UNTAGGED_IMAGES" ]; then
            for IMAGE_DIGEST in $(echo "$UNTAGGED_IMAGES" | jq -r '.[]'); do
              aws ecr batch-delete-image --repository-name your-ecr-repository-name --image-ids imageDigest=$IMAGE_DIGEST
            done
          fi

In this example, the project is set to typescript. and .gitignore is set to not push node_modules. Here are some details.

1. This is for install AWS CLI. We need to install AWS CLI for accessing ECR.

- name: Install AWS CLI
        run: |
          sudo apt-get update
          sudo apt-get install -y awscli

2. This is for setting AWS Credentials.

- name: Configure AWS Credentials
        uses: aws-actions/configure-aws-credentials@v1
        with:
          aws-access-key-id: ${{ secrets.AWS_ACCESS_KEY_ID }}
          aws-secret-access-key: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
          aws-region: ${{ secrets.AWS_REGION }}

3. This is for build Docker and update Lambda function.

# Uncomment the following section to log in to AWS ECR
      - name: AWS ECR Login
        run: aws ecr get-login-password --region ${{ secrets.AWS_REIGON }} | docker login --username AWS --password-stdin ${{ secrets.AWS_REGISTRY_URL }}

      - name: Build Docker Image
        run: docker build --platform linux/x86_64 -t your-build-name .

      - name: Tag Docker Image
        run: docker tag your-build-name:latest ${{ secrets.AWS_REGISTRY_URL }}:latest

      - name: Push Docker Image to ECR
        run: docker push ${{ secrets.AWS_REGISTRY_URL }}:latest

      - name: Update Lambda Function
        run: aws lambda update-function-code --function-name your-lambda-function-name --image-uri ${{ secrets.AWS_REGISTRY_URL }}:latest

      - name: Delete Untagged Images In ECR
        run: |
          # Get untagged image digest
          UNTAGGED_IMAGES=$(aws ecr describe-images --repository-name your-ecr-repository-name --query 'imageDetails[?imageTags==null].imageDigest' --output json)

          # Delete untagged images
          if [ -n "$UNTAGGED_IMAGES" ]; then
            for IMAGE_DIGEST in $(echo "$UNTAGGED_IMAGES" | jq -r '.[]'); do
              aws ecr batch-delete-image --repository-name your-ecr-repository-name --image-ids imageDigest=$IMAGE_DIGEST
            done
          fi

This process should be set Login>Build>Tag>Push>Update.
By tagging, you can handle latest version of Docker image which is built.
"secrets.AWS_REGISTRY_URL" should be set the uri of ECR private repository.