The OP's statement makes the context clear though: It's not about knowing everything. It's about knowing a couple of abstraction layers under you (and, as I arg

The OP's statement makes the context clear though: It's not about knowing everything. It's about knowing a couple of abstraction layers under you (and, as I argued elsewhere, above you too). It's useful to understand a few classic networking problems that a backend engineer might face — e.g. weird latency issues caused by Nagle's algorithm, or TCP CLOSE_WAIT leading to ephemeral port exhaustion between a proxy and an application node. It's useful to understand why we mostly moved to event loop-based servers instead of thread-per-connection servers as a way to handle c10k. A backend engineer doesn't necessarily have to be an expert on any of these things, but they should be able to follow along if an expert explains that sort of problem.