It's nice to see the trend unwinding, microservices are egregiously overused as a silver bullet for any problem (not just scaling). There are probably dozens (

It's nice to see the trend unwinding, microservices are egregiously overused as a silver bullet for any problem (not just scaling). There are probably dozens (/s) of companies that would benefit from going all-in with microservices. Some companies would benefit from having a few microservices; but most would be more than successful running a monolith. The most significant problems with microservices arise when they're adopted too early and when data separation is not properly handled, resulting of even the simplest requests ended up querying 3-5+ microservices. Latest anecdata point: doing exactly that (unwinding microservices) in a team of 20 engineers (15 swe/5 sre) and 400+ repositories (which isn't a "wrong", but having so many different places to update, for example, Python version is nuts). The whole project is nuts. And persuading management was not an issue, it was enough to show current AWS bills and potential savings from moving from 50+ microservices to smaller numbers of consolidated services and deprecating/refactoring some features. The real challenge was persuading the engineering team to prioritize and work on this issue - which is still ongoing.