It is always beneficial for a project to define a clear software architecture. But how can you fight growing deviations between the planned architecture and the physical code base? How can you avoid expensive redesigns and refactoring phases? How can you achieve an outstanding technical quality of your code base? The session explains the basics concepts of architecture management for Java projects and introduces a couple of simple rules, that help you to keep your project on track.

* how to cut your application into layers, vertical slices and subsystem and define the allowed dependencies between these elements.

* how to map your code base to these logical entities.

* some useful software metrics, that help you to fight local complexity.

* strategies to automatically enforce dependency restrictions and metric rules

I will use our Java Architecture Management tool SonarJ to show how to implement these concepts in real life projects.