/python-exercise

Python technical exercise

Primary LanguagePython

Python Exercise

Usage

Docker has been used for development and deployment to a production environment can be done in two ways:

  • traditional, run on a server with a cron job to run the script every weekday at 9:00 AM
  • serverless with either aws lambda functions or GCP cloud functions, using triggers initialize the script at 9:00 AM
    • currently the app is only intended to be run on AWS

Development

  • To run the app in development run the following commands in the same directory as the docker-compose.yml file:
    • docker-compose up -d
    • docker-compose run app alembic upgrade head
    • docker-compose run app python main.py
  • For further development and running unit tests:
    • docker-compose run app nosetests

Production deployments

  • see https://docs.aws.amazon.com/lambda/latest/dg/python-programming-model-handler-types.html for details on handlers and lambda functions
  • run docker-compose run app pip install -r /src/requirements.txt -t /src
    • this will install dependencies into the root level directory, the contents of the src directory should then be zipped up
  • using the AWS console or cli set the environment variables appropriately:
    • FIXER_API_KEY
    • FIXER_URL
    • DATABASE_URL
  • provision a postgres (or other SQL) database on your chosen platform
  • run alembic migrations against the db, this may require provisioning an EC2 instance and checking out the code to run. the important command to run against the db is:
    • alembing upgrade head
  • create the lambda function, in this case using the aws cli: 
    •   --region region \
  --function-name stor_fixer_data \
  --zip-file fileb://deployment-package.zip \
  --role arn:aws:iam::account-id:role/lambda_basic_execution  \
  --handler main.handler \
  --runtime python3.6 \
  --timeout 60 \
  --memory-size 512```
  • the cron expression:
    • aws events put-rule --schedule-expression "cron(0 9 * * MON-FRI *)" --name MyRule1
    • set this as a trigger for the lambda function using the aws console or whatever method you are most familiar with

Monitoring

AWS cloudwatch can be used for monitoring of the app, at its simplest it can be configured to store output of the logger. more advanced would be to setup SMS alerts