Microservices : Small or smaller?

In a recent project i'm working on we are working with microservices and domain driven design approach. There was quite some discussion on how small or large microservices should be, as always every engineer has their own opinion and arguments :-)

In the project there are multiple teams and plans to scale the project up to even more teams, which is why a microservice approach was chosen in the first place. Interestingly the teams have chosen a different approach for how small or large their microservices should be. So i'm going to explore the consequences of those choices here.

In this blog i'm not going to replicate the arguments for small or large microservices, there is already excellent articles on that subject, i recommend reading for example this article.

Let's look at the choices the teams made and the consequences:

Obviously the above is biased by my personal opinion and not completely factual. I believe though that the choice to make microservices very small means (initially) development of product functionality is slowed down.

Reason for the choice to make microservices so small was to lay a foundation so that the team can be split easily, so bascially frontloading some of the work. When the team with the larger microservice will be split, we'll have to do that work at that time.

I'm going conclude this article with my recommendation to be aware of the consequences of whatever choice you make. If you have personal stories or anecdotes, please comment, i'm interested you hear your experiences and opinions!