While some code happily lives on a single server forever, most big projects will have to cross the boundary into running both their application and storing their data across multiple systems. The basic strategies are well-known, but we’ll take a look at what to do as you cross the painful threshold where you can’t run your app as a monolith or store everything on a single database server. Among other things, we’ll look at how to split up business logic and application code to run on different servers, how to scale to handle different kinds of web traffic (read-heavy, write-heavy, and long-connections/WebSockets), when and how to make parts of your code not run inline with HTTP processing, strategies for storing data across multiple machines, and how to structure your engineering team to best cope with all these changes. We’ll also look at a few apparently innocuous decisions and the spiral of bad performance they lead to, and how to recognise some of these common problems so you can avoid them yourself in future.