Book Image

Python Architecture Patterns

By : Jaime Buelta
Book Image

Python Architecture Patterns

By: Jaime Buelta

Overview of this book

Developing large-scale systems that continuously grow in scale and complexity requires a thorough understanding of how software projects should be implemented. Software developers, architects, and technical management teams rely on high-level software design patterns such as microservices architecture, event-driven architecture, and the strategic patterns prescribed by domain-driven design (DDD) to make their work easier. This book covers these proven architecture design patterns with a forward-looking approach to help Python developers manage application complexity—and get the most value out of their test suites. Starting with the initial stages of design, you will learn about the main blocks and mental flow to use at the start of a project. The book covers various architectural patterns like microservices, web services, and event-driven structures and how to choose the one best suited to your project. Establishing a foundation of required concepts, you will progress into development, debugging, and testing to produce high-quality code that is ready for deployment. You will learn about ongoing operations on how to continue the task after the system is deployed to end users, as the software development lifecycle is never finished. By the end of this Python book, you will have developed "architectural thinking": a different way of approaching software design, including making changes to ongoing systems.
Table of Contents (23 chapters)
2
Part I: Design
6
Part II: Architectural Patterns
12
Part III: Implementation
15
Part IV: Ongoing operations
21
Other Books You May Enjoy
22
Index

Subdividing tasks

It's entirely possible to generate more tasks from an initial one. This is done by creating the right event inside a task and sending it to the right queue.

This allows a single task to distribute its load and parallelize its action. For example, if a task generates a report and sends it by email to a group of recipients, the task can first generate the report and then send the emails in parallel by creating new tasks that will focus only on creating the emails and attaching the report.

This spreads the load over multiple workers, speeding up the process. Another advantage is that individual tasks will be shorter, which makes them easier to control, monitor, and operate.

Some task managers may permit the creation of workflows where tasks are distributed, and their results are returned and combined. This can be used in some cases, but in practice it is less useful than it initially appears, as it introduces extra waiting and we can end up...