3 ms·
Well, the idea of microservices would be to build some degree of flexibility into the setup, so that you can handle problems you don't have now, but you might h
by karles 3y ago
Well, the idea of microservices would be to build some degree of flexibility into the setup, so that you can handle problems you don't have now, but you might have later.
I understand that the Monolith - if done right - is a viable option as well.
As a consultant, our customers rarely know what they want, and they can't really describe their vision, no matter how many workshops we throw at them. Microservices give us the opportunity to say "We might not know this now, but with this architecture, we won't be locked into a certain pattern or meet certain technical limitations by a monolithic approach".
It seems that what you are really paying for with microservices is flexibility and maneuverability "down the line".
Solving only the problems you have "here and now" with the simplest solution will also lead to technical debt.
- geewee 3y agoI don't think I agree with your thoughts. Microservice setups are generally less flexible than monolithic setups. Particularly if you get the boundaries wrong it's notoriously difficult to refactor multiple microservices, particularly if you need full uptime. For monoliths you can refactor the whole thing with editor support, and then deploy a new version. So microservices will cost more, give you less flexibility and maneuverability, at more upfront cost. It seems like the worst of both worlds.