My understanding is that keeping your database in a separate container from your application(s) is a more preferred topology. In theory, you could easily spin

My understanding is that keeping your database in a separate container from your application(s) is a more preferred topology. In theory, you could easily spin up/spin down application containers as needed, and they can use Docker's service discovery features to automatically become aware of which port the database is listening on.