GraphQL — Part 3: Advanced Topics & Interview Q&A (Senior/Staff)

What is this?

We've ordered à la carte (Part 1) and watched the kitchen cook it without sending a runner to the fridge fifty times (Part 2). Now we run the restaurant — at scale, with strangers walking in. That introduces problems the home cook never has: a customer who orders a meal so elaborate it ties up every station for an hour (query cost), the fact that you can't just photocopy yesterday's plates because everyone orders something different (caching is hard), and a dozen chefs who each insist on owning part of the menu (federation).

The one-line idea: at scale, GraphQL's flexibility becomes its liability — the same "ask for anything" power that delights clients lets an attacker craft a ruinous query and defeats the easy HTTP caching REST takes for granted, so you trade some flexibility back for safety and speed.

These are exactly the trade-offs a senior interview probes. We'll cover the four big ones, then drill them in mock-interview format.