Book Image

Building Python Microservices with FastAPI

By : Sherwin John C. Tragura
3 (2)
Book Image

Building Python Microservices with FastAPI

3 (2)
By: Sherwin John C. Tragura

Overview of this book

FastAPI is an Asynchronous Server Gateway Interface (ASGI)-based framework that can help build modern, manageable, and fast microservices. Because of its asynchronous core platform, this ASGI-based framework provides the best option when it comes to performance, reliability, and scalability over the WSGI-based Django and Flask. When working with Python, Flask, and Django microservices, you’ll be able to put your knowledge to work with this practical guide to building seamlessly manageable and fast microservices. You’ll begin by understanding the background of FastAPI and learning how to install, configure, and use FastAPI to decompose business units. You’ll explore a unique and asynchronous REST API framework that can provide a better option when it comes to building microservices. After that, this book will guide you on how to apply and translate microservices design patterns in building various microservices applications and RESTful APIs using the FastAPI framework. By the end of this microservices book, you’ll be able to understand, build, deploy, test, and experiment with microservices and their components using the FastAPI framework.
Table of Contents (17 chapters)
1
Part 1: Application-Related Architectural Concepts for FastAPI microservice development
6
Part 2: Data-Centric and Communication-Focused Microservices Concerns and Issues
11
Part 3: Infrastructure-Related Issues, Numerical and Symbolic Computations, and Testing Microservices

Implementing coroutines

In the FastAPI framework, a thread pool is always present to execute both synchronous API and non-API transactions for every request. For ideal cases where both the transactions have minimal performance overhead with CPU-bound and I/O-bound transactions, the overall performance of using the FastAPI framework is still better than those frameworks that use non-ASGI-based platforms. However, when contention occurs due to high CPU-bound traffic or heavy CPU workloads, the performance of FastAPI starts to wane due to thread switching.

Thread switching is a context switch from one thread to another within the same process. So, if we have several transactions with varying workloads running in the background and on the browser, FastAPI will run these transactions in the thread pool with several context switches. This scenario will cause contention and degradation to lighter workloads. To avoid performance issues, we apply coroutine switching instead of threads.

...