In recent times, there has been a strong move toward breaking applications apart and designing them as a set of services communicating via messages. This is a well-established concept for large-scale applications; the problem is finding the correct time to break the application apart and also deciding whether breaking it apart is the right thing to do. When we break an application into multiple services, we increase the complexity at this point since we now have to deal with handling communication problems crossing multiple services. We have to consider the resilience of the services, and the dependencies each service has, to provide its features.

On the other hand, when we break an application apart at a late stage, problems arise when extracting logic from the application. There is no reason why a monolith application can't be well-factored and remain easy to maintain for a long time. Breaking up an application will always cause problems, and staying with a well...