FastService

Ready-to-go micro-service template based on FastAPI

Just set configs, add endpoints and deploy using docker

Local deployment for development

Deploy FastService in local development environment using docker-compose in just a few steps:

1. Modify environment configs in local_config directory as required.
2. Start DB and other external services as required.
3. Build service using sudo docker-compose build.
4. Start service using sudo docker-compose up.
5. Add test cases for your endpoints in app/app/tests/test_api.py.
6. Start adding API endpoints in app/app/src/public_api.py and app/app/src/private_api.py.
7. Service automatically refreshes for any changes made in the code in the development environment.

Dockerfile and docker-compose.yml are used as configuration files for local deployment.

Production deployment

Dockerfile_Prod and fastservice_stack.yml are used as configuration files for production deployment.

1. A sample CI/CD config file for gitlab .gitlab-ci.yml is provided. Replace with configuration of CI/CD provider of your choice.
2. Add any service prestart script in app/prestart.sh as required.
3. Create configs and secrets for your service in your deployment environment.
4. Push and deploy to your environment.

Some features and considerations of FastService

• Authentication and Authorization

  • Authentication and authorization modules here use Auth0. Make relevent changes for using any other auth service.
  • All auth related code are in app/app/dependency/authentication.py
  • requires_auth() function is used for validating access token for every request coming to all private endpoints.
  • requires_scope() and requires_permission() functions are used to check authorization for protected endpoints. Their usage in given as sample endpoints in app/app/src/private_api.py

• Private API endpoints - Any endpoints defined in app/app/src/private_api.py using @privateAPI decorator is accessible only with a valid access token in the request header

• Public API endpoints - Any endpoints defined in app/app/src/public_api.py using @publicAPI decorator is accessible without any authentication

• Context - Context available throughout the request lifecycle along with a request id generated at very start of an incoming request

• User id - User id is automatically extracted from access token and added to context for private API endpoints

• Logging - Two log cofiguration files app\app\dependency\app_logger\log_config_file.json and app\app\dependency\app_logger\log_config_stdout.json are provided to configure logging to rotating file and stdout respectively. Select the logging configuration to use using LOG_TYPE config.

• DB Connection - Database connection is provided using encode/databases. Sample connection is given in app/app/src/public_api.py

• Test Framework - Test are defined using pytest. Sample test cases present in app/app/tests/test_api.py

• Healthcheck - A heartbeat endpoint is used for checking health of service instance.

• Access token for M2M calls - To retrieve access token for M2M calls to private endpoints of other services use the get_access_token() function in app/app/dependency/access_token.py

References

• FastAPI
• Starlette
• Pydantic
• Starlette-Context
• Python-Jose
• Python-Dotenv
• Requests
• Databases
• Asyncpg