Book Image

Domain-Driven Design with Java - A Practitioner's Guide

By : Premanand Chandrasekaran, Karthik Krishnan
Book Image

Domain-Driven Design with Java - A Practitioner's Guide

By: Premanand Chandrasekaran, Karthik Krishnan

Overview of this book

Domain-Driven Design (DDD) makes available a set of techniques and patterns that enable domain experts, architects, and developers to work together to decompose complex business problems into a set of well-factored, collaborating, and loosely coupled subsystems. This practical guide will help you as a developer and architect to put your knowledge to work in order to create elegant software designs that are enjoyable to work with and easy to reason about. You'll begin with an introduction to the concepts of domain-driven design and discover various ways to apply them in real-world scenarios. You'll also appreciate how DDD is extremely relevant when creating cloud native solutions that employ modern techniques such as event-driven microservices and fine-grained architectures. As you advance through the chapters, you'll get acquainted with core DDD’s strategic design concepts such as the ubiquitous language, context maps, bounded contexts, and tactical design elements like aggregates and domain models and events. You'll understand how to apply modern, lightweight modeling techniques such as business value canvas, Wardley mapping, domain storytelling, and event storming, while also learning how to test-drive the system to create solutions that exhibit high degrees of internal quality. By the end of this software design book, you'll be able to architect, design, and implement robust, resilient, and performant distributed software solutions.
Table of Contents (17 chapters)
1
Part 1: Foundations
4
Part 2: Real-World DDD
12
Part 3: Evolution Patterns

Even more fine-grained decomposition

At this stage, is there any further decomposition that is required and feasible? These days, whether rightfully or otherwise, serverless architecture (specifically, functions as a service) is arguably becoming all the rage. As we pointed out in Chapter 2, Where and How Does DDD Fit?, this means that we may be able to decompose our command side in a manner that each command becomes its own independently deployable unit (hence a bounded context). In other words, LCApplicationSubmitCommand and the LCApplicationCancelCommand can be deployed independently.

But just because this is technically possible, should we do it? While it is easy to dismiss this as a passing fad, there may be good reasons to split applications along command boundaries:

  • Risk profile: Certain pieces of functionality present a higher risk when changes are made. For example, submitting an LC application may be deemed a lot more critical than the ability to cancel it. However...