Skip to content

Latest commit

 

History

History
117 lines (85 loc) · 3.32 KB

lambda_docker_serverless.md

File metadata and controls

117 lines (85 loc) · 3.32 KB

AWS Lambda with Docker and Serverless

The architecture

Install Serverless

serverless create --template aws-python3 --path my-project

This creates a new directory with handler.py, .gitignore, serverless.yaml. You can create your Python modules in the directory and adapt the handler. Add your requirements and write the Dockerfile:

FROM public.ecr.aws/lambda/python:3.8

# Copy function code and models into our /var/task
COPY ./ ${LAMBDA_TASK_ROOT}/

# install our dependencies
RUN python3 -m pip install -r requirements.txt --target ${LAMBDA_TASK_ROOT}

# Set the CMD to your handler (could also be done as a parameter override outside of the Dockerfile)
CMD ["handler.handler"]

Add .dockerignore:

README.md
*.pyc
*.pyo
*.pyd
__pycache__
.pytest_cache
serverless.yaml
get_model.py

Run and test docker:

docker build -t my-lambda .
docker run -p 8080:8080 my-lambda

We can test the app in a separate terminal

curl --request POST \
  --url http://localhost:8080/2015-03-31/functions/function/invocations \
  --header 'Content-Type: application/json' \
  --data '{"body":"{\"context\":\"We introduce a new language representation\",\n\"question\":\"What is the GLUE score for Bert?\"\n}"}'

# the data should be a stringified json with the input to the handler you wrote for your app.
# {"statusCode": 200, "headers": {"Content-Type": "application/json", "Access-Control-Allow-Origin": "*", "Access-Control-Allow-Credentials": true}, "body": "{\"answer\": \"80 . 5 %\"}"}%

Deploy custom docker image to ECR.

aws ecr create-repository --repository-name my-lambda > /dev/null

Before pushing images to ECR, we need to log in.

aws_region=us-east-2
aws_account_id=your_account_id_numbers

aws ecr get-login-password \
    --region $aws_region \
| docker login \
    --username AWS \
    --password-stdin $aws_account_id.dkr.ecr.$aws_region.amazonaws.com

We rename our previously created image to the ECR format, which is {AccountID}.dkr.ecr.{region}.amazonaws.com/{repository-name}.

docker tag my-lambda $aws_account_id.dkr.ecr.$aws_region.amazonaws.com/my-lambda

The new image should be apparent when running docker images. If it's there, push the image to ECR.

docker push $aws_account_id.dkr.ecr.$aws_region.amazonaws.com/my-lambda

Use this ECR docker image to deploy the AWS lambda function. In serverless.yml, add the aws provider, region and function. Add the account ID numbers and the AWS Region.

service: serverless-my-lambda-docker

provider:
  name: aws # provider
  region: us-east-2 # aws region
  memorySize: 5120 # optional, in MB, default is 1024
  timeout: 30 # optional, in seconds, default is 6

functions:
  questionanswering:
    image: ACCOUNT_ID.dkr.ecr.us-east-2.amazonaws.com/my-lambda:latest #ecr url
    events:
      - http:
          path: qa # http path
          method: post # http method

Deploy with serverless

serverless deploy

Done! To test the Lambda function, use Postman or a REST client, add the json with the inputs that your handler needs.