Summary
The Jeeves bot is a typical web app that interacts with a database and a few backend services. The only unusual feature is that it receives most of its workload from one endpoint. Building this application with a monolithic architecture has allowed us to quickly iterate over several choices and get a prototype that works well during development and low-volume usage.
From our discussions about the actions, it should be clear that there are good candidates for migration to a microservice. If we run this bot for dozens or hundreds of Slack workspaces, then we may find that one component is used much more than others, or has reliability issues that are difficult to address in the current architecture. How should the application scale up? What happens when an external service it relies upon is encountering errors? What happens if one of our own components—our database or message broker—goes down?
In the next chapter, we will look at these issues: how to change...