- animted underground map of Berlin
One of the AWS serverless design principles is to share nothing, and separating your microservices to individual repositories makes it easy to not accidentally do something you shouldn’t.
Build self-containing services. Don’t build stacks that require resources from other services in order to deploy. However, you can (and should!) build microservices that depend on others to satisfy business processes through loose coupling via API calls or events.
If you’re looking to get started with *microservices*, don’t forget they are driven by domain. When looking to “break up the monolith” start by identifying the discrete pieces of your application. Find the areas of the application that have enough business logic to be on their own and iterated separately from everything else.
Also check out gopher-holes-unlimited.