README.md

Install serverless

npm install -g serverless

Create Service

serverless create --template aws-python3 \
  --name dp-demo-smruti
  --path pypack

This will create a serverless python3 template project at given path pypack/ with a service name dp-demo-smruti

Use virtual environment

Use virtual environment for all development

cd pypack/
virtualenv venv --python=python3
source venv/bin/activate

Sample python code

Create sample python code to generate an array(Use numpy package) Install numpy package: pip install numpy Create a requirement file: pip freeze > requirement.txt

Code to generate a sample matrix using np

handler.py

import numpy as np
import json


def main(event, context):
    a = np.arange(15).reshape(3, 5)
    arr=json.dumps({"Numpy Array": a.tolist()})
    return arr


if __name__ == "__main__":
    output = main('', '')
    print(output)

You may test the code locally.

Service Deployment

Since we need the python libraries(eg. numpy) along with code, so we have to install a plugin serverless-python-requirements

To install the serverless plugin: Create a file package.json to save the node dependencies then install the plugin

npm init
npm install --save serverless-python-requirements

Edit the serverless.yml file, add the plugins

Also you need to have Docker installed to be able to set dockerizePip: true or dockerizePip: non-linux. Alternatively, you can set dockerizePip: false, and it will not use Docker packaging. But, Docker packaging is essential if you need to build native packages that are part of your dependencies like Psycopg2, NumPy, Pandas, etc.

The service name is dp-demo-smruti

• serverless.yml file

service: dp-demo-smruti

provider:
  name: aws
  runtime: python3.7
  region: eu-central-1

functions:
  dp-demo-test:
    handler: handler.main

plugins:
  - serverless-python-requirements

custom:
  pythonRequirements:
    dockerizePip: non-linux

Now, we can deploy the code -

sls deploy -v

Then invoke the function -

serverless invoke -f sls invoke -f dp-demo-test