While microservices have lots of benefits over traditional monoliths, splitting application and databases makes certain patterns harder or even impossible. Transactions are replaced by sagas, synchronous communication with events, but at what cost? Data is split among services, but we still want to consume it in a predictable way. Events are used to dispatch information, but they increase our operational cost and introduce eventual consistency. CAP theorem still applies, are our patterns up to speed? In this session we will examine how information flows in a microservice architecture and how we can update our ways of thinking that is still based in traditional monolithic design.