Using fastAPIs to build REST API service is ideal for interrupt-driven positions on dev teams like:: DevOps Engineers, SREs, DataOps Engineers, MLOps Engineers, and Data Engineers; these are jobs that are customer interaction centric. These customers are in a constant state of needing help with local, server, or network configuration changes and issues. Further understanding of CI/CD pipelines & data pipelines and how they can leverage what is sometimes cryptic application and pipeline log messaging created by themselves, the customer, other engineering teams, and third-party vendors. Up to the recent past, these Ops Engineers and practitioners of the black arts of their designated assignments have been constrained to have a series of scripts in Python, Powershell, or bash, and other instances of heavy-lifting scripting languages to call upon from their out of their shamanistic medicine bags to solve problems in the arena of “I can’t make this go.” fastAPI is a tool to allow the wrapping of these Python scripts (Python framework) in a like its name implies quick, easy to use and high performing framework that is suited for the interrupt driven engineering position to mitigate interrupts through moving their traditional tasks towards a self-service model with customers.

• [ ] 1 POPO class1 (Plain Old Python Object - Made possible by the dataclasses built-in package released in Pythong 3.7+ and inspired by the Attr package because waiting for the Python folks to get their act together and implement a dataclasses package was just out of the question.)
• [ ] 1 Repository class
• [ ] 1 Adapter class
• [ ] 4 integration tests - (service level only, NOT CRUD) are the Minimum number of tests to cover the initial implementation of the service.
• [ ] 2 unit tests - Minimum number of tests to cover initial implementation of service.
• [ ] Wire-up to db through adapter obj.
• [ ] Migrations - Add POPO & DDL to Migration using Alembic (Alembic is DataOps stack's Liquibase - Ascend uses Liquibase for Db Migration and Versioning)
• [ ] CI/CD Build on GitLab (create a Dockerfile for the deployable)
• [ ] Deployment to EKS for automated integration testing.

Sources::

Hexagonal Architecture in Python

First Steps - FastAPI

I built my very first fastAPI REST API Service using only the two sources above.

But for additional context and easy to understand more on the why of fastAPI I found this book to use a clear and intelligible writing style. It is concise in getting the developer to where he needs to be::

Building Python Web APIs with FastAPI - By Abdulazeez Abdulazeez Adeshina

Building Python Web APIs with FastAPI: A fast-paced guide to building high-performance, robust web APIs with very little boilerplate code

At the time of writing this post the cost of the Kindle version of the book is $19.79 which is really great price for such a concise and well written book that is extremely considerate of it’s reader’s understanding in the presentation of ideas.

These are useful articles by Abdulazeez that are also extremely well written and considerate of it’s reader’s understanding in the presentation of ideas::

Abdulazeez Abdulazeez Adeshina at TestDriven.io