AWS re:Invent 2023 - The pragmatic serverless Python developer (OPN305)

Question 1

What is Pydantic?

Answer 1

It is data validation library for Python.

Question 2

How structure the folder?

Answer 2

Business domain: lambda handlers code
Business logic
integrations code

Ex:
- Product
- crud
- handlers
- schemas (pydantic models)
- tests
- e2e
- integration
- unit
- pyproject.toml
- IaC: CDK, AWS SAM, Terraform, …

Ex 2:
- stream_processor
- domain_logic
- handlers
- integrations
- __init__.py
- constants.py

Question 3

What is Architectural Layers?

Answer 3

Is a way to split responsibility.

Handler:
- No domain code
- Concise responsibilities
-> Configuration > Env. vars > Validation > Call domain > Serialize output

Domain:
- Business logic only
- May be shared by N handlers
- Isolated tests

Integration:
- Adapter pattern
- Contains API/DB code

API Gateway > Handler > Domain > Integration > DynamoDb

Question 4

Explain how test a serverless application?

Answer 4

• Unit test (locally)
  
  • test schema validations (Pydantic)
  • test isolated integration classes
  • no deployment required
  • fast and repeatable
• Integration tests
  
  • requires deployment
  • Run locally & on AWS
  • Test runner (pytest) > test data > call > lambda handler (AWS) > assertion
• E2E test
  
  • required deployment
  • Run on AWS
  • Test runner (pytest) > test data > (call) > Product API (Api Gateway endpoint + lambda) > assertion
  • Simulate bad input
  • API security tests (invalid authorization / invalida authentication)