Who Should Read This Book?

If any of the following are true for you, you’ll find this book valuable:

• You’re a software engineer, software architect, or technical manager who needs to make decisions about the architecture of the systems you work on—for example, you need to choose tools for solving a given problem and figure out how best to apply them. This applies especially to backend systems.
• You’re a data engineer who wants to understand the wider context of the systems you deal with, or a cloud engineer who wants insights into the underpinnings of the systems you’re using. You will find that even though modern distributed systems hide a lot of complexity from you, understanding their underlying principles is extremely useful for performance optimization and debugging.
• You want to learn how to make data systems scalable (e.g., to support apps with millions of users), highly available (minimizing downtime), operationally robust, and easier to maintain in the long run (even as they grow and as requirements and technologies change).
• You are preparing for a “system design” job interview in which you will be asked to sketch an architecture for an application, and you need to learn the principles for good data architectures.
• You are curious to find out what goes on behind the scenes at major websites and online services, and inside various databases and data processing systems—especially if you like to dig deeper than buzzwords to gain a technically accurate and precise understanding of various technologies and their trade-offs.

This book assumes that you already have some experience building web-based applications and that you are familiar with relational databases and SQL. A high-level understanding of common network protocols like TCP and HTTP is helpful. Your choice of programming language or framework makes no difference for this book.

We have also added discussions of sync engines and local-first software, workflow engines and durable execution, formal methods and randomized testing, GraphQL, and various other technologies that are worth knowing about. We have included a bit of legal context as well, by exploring the impact of the EU General Data Protection Regulation (GDPR) and related law. We’ve also taken a few things away—for example, as MapReduce is now largely obsolete, we have rewritten the batch processing chapter accordingly, and we sadly decided to drop the Tolkien-style maps.

A few discussions have been restructured, and the chapter numbering has changed. Some chapters required only a light edit, while others (such as Chapter 10, on consistency and consensus) were almost completely rewritten to make them clearer. Overall, the second edition is about 60 pages longer than the first.