Storing and handling data is a complex issue in every software architecture. Wrong choices may impact scalability, performance, or maintenance costs. With microservices, there's an added complexity coming from the fact that we want the microservices to be loosely coupled.
This leads to a design pattern where each microservice connects to its own database so it is independent of any changes introduced by the other services. While this pattern adds some overhead, its additional benefit is that you can optimize the schema and indexes for each microservice individually.
Since databases tend to be pretty huge pieces of infrastructure, this approach may not be feasible, so sharing a database between microservices is an understandable trade-off.
