Last month I wrote about modular monoliths and the value of modern monolithic architecture. One of the more interesting discussions that came out of that article (and video) is the inverse discussion: when is it right to still pick microservices?

Like any design choice, the answer is subjective and depends on many things. But there are still general rules of thumb and global metrics we can use. Before we get into these problems, we need to understand what it means to have a microservice architecture. Then we can gauge the benefit and price of having such an architecture.