Comment by wheresmycraisin

One thing that has helped me grow over the last few years building startups is: microservices software architecture and microservice deployment are two different things.

You can logically break down your software into DDD bounded contexts and have each own its data, but that doesn't mean you need to do Kubernetes with Kafka and dozens of tiny database instances, communicating via json/grpc. You can have each "service" live in its own thread/process, have it's own database (in the "CREATE DATABASE" sense, not the instance sense), communicate via a simple in-memory message queue, and communicate through "interfaces" native to your programming language.

Of course it has its disadvantages (need to commit to a single softare stack, still might need a distributed message queue if you want load balancing, etc) but for the "boring business applications" I've been implementing (where DDD/logical microservices makes sense) it has been very useful.