Book Image

Software Architecture Patterns for Serverless Systems - Second Edition

By : John Gilbert
Book Image

Software Architecture Patterns for Serverless Systems - Second Edition

By: John Gilbert

Overview of this book

Organizations undergoing digital transformation rely on IT professionals to design systems to keep up with the rate of change while maintaining stability. With this edition, enriched with more real-world examples, you’ll be perfectly equipped to architect the future for unparalleled innovation. This book guides through the architectural patterns that power enterprise-grade software systems while exploring key architectural elements (such as events-driven microservices, and micro frontends) and learning how to implement anti-fragile systems. First, you'll divide up a system and define boundaries so that your teams can work autonomously and accelerate innovation. You'll cover the low-level event and data patterns that support the entire architecture while getting up and running with the different autonomous service design patterns. This edition is tailored with several new topics on security, observability, and multi-regional deployment. It focuses on best practices for security, reliability, testability, observability, and performance. You'll be exploring the methodologies of continuous experimentation, deployment, and delivery before delving into some final thoughts on how to start making progress. By the end of this book, you'll be able to architect your own event-driven, serverless systems that are ready to adapt and change.
Table of Contents (16 chapters)
Other Books You May Enjoy

Justifying multi-regional deployment

Multi-regional deployment is a hotly contested topic. There are many arguments for and against it. Which side of the argument you stand on depends on where you are standing. In other words, context matters. For example, it is more difficult to run traditional systems in multiple regions than it is with serverless systems. So, for our serverless systems we are more inclined to favor the multi-regional approach.Let's look at why we should run our serverless systems in more than one region. The arguments revolve around risk, cost and user satisfaction.

It's too risky not to

It is inevitable. Sooner or later a given cloud provider will experience a news-worthy regional disruption. It is not a matter of if but of when. In my experience, this happens about every 2 to 3 years or so. These major disruptions typically affect multiple cloud services and can last for hours, usually over five hours. Minor disruptions happen sporadically, impact fewer...