Microservices with FastAPI

Prepared for Calgary / Edmonton Python Bootcamp 2020

Prerequisites

• Knowledge of Python basics
  • variables, naming convention, loops, functions, classes, object-oriented programming, testing, best practices
  • Attendance at sessions 1 and 2 of the bootcamp is satisfactory
• Understanding of RESTful web APIs
  • What are RESTful APIs? How do we start thinking about them in Python?
  • Provided by Zero To API

Environment

• Virtual environment with Python 3.7+ (I will probably use conda)
• Jupyter notebook or Google Colab
• IDE for writing code (I will probably use sublime)
• Terminal for executing code (likely this is possible from the IDE, but I will just use basic mac terminal)
• RESTful client (I usually just use Jupyter notebooks and Python requests)

Note: The setup from Zero To API should be satisfactory for this presentation as well, as long as Python is 3.7+.

Overview

• We will start by covering a few additional prerequisite topics for understanding FastAPI
  • Access notebook in Google Colab
  • Declarative programming with dataclasses and pydantic
  • Decorators
  • Async with asyncio
• Next, we will then talk about microservices
  • Pros, cons, comparison to monolith and realistic architectures
  • No coding for this part!
• Finally, FastAPI!
  • What is it? Why is it great?
  • Building an example application together
  • We will consume the NHL API
  • Notebook to access API here
    • Note: This will not actually work in Google Colab, you can only access localhost locally.