Implementing Domain-Driven Design

Focused Aggregates

While developing Aggregates with traditional persistence (for example, relational database without the use of Event Sourcing), development friction from introducing a new Entity into the system or enriching an existing one can be noticeable. We need to create new tables and define new mapping schemata and Repository methods. If our tendency is to resist such development overhead, it can cause us to grow Aggregates as we concentrate more state structure and behaviors on each. It can be much easier to add onto an existing Aggregate rather than to create a new one.

However, our bias can shift if Aggregates are more easily designed anew, and I assert that this is true when Event Sourcing is in use. In my experience, Aggregates ...