Fwiw, I would only ask your DI question as a trick question. The correct answer is to explain why DI isn’t actually useful on balance and causes far more problems than it helps.

But setting that aside… it mostly depends on the domain. A good one to reach for regardless of domain is to ask the candidate about a particular area of your technical surface area, high enough level and broad enough that they already fully understand its purpose and existence just from knowing what your company does, and ask them to explain how they think it works. Probing questions from there. For bonus points start dropping info bits on pieces which are fundamentally done badly and ask them what sorts of consequences that might have and how they would fix it.

At a lower level of abstraction, like your DI question, it would depend very greatly on the domain. For example, I’ve asked infrastructure leaders to explain to me their opinions on mono- vs polyrepo setups, not to verify their biases match mine, but to see if they’ve actually thought it through (most haven’t, even with absolutely sparkling resumes btw). Another good thing to probe (for a highly available multi-regional system) is global ingress and how that interacts with user state. At the opposite end of the spectrum, questions about testing philosophy and (for senior candidates) how best to cultivate it in an organization and (for junior-ish candidates), how to best deploy it when writing new code, or refactoring old code.

The well is deep. 🙂 Just look for a good technical conversation with someone you hope is a peer (or beyond you in some ways).