alt Hacker NewsOne of the major differences between like, your PC, and a high-availability server is that high-availability servers require those things as well—and get them, piecewise, while still running. It's like doing a complete overhaul while the plane is flying.
IBM mainframes, famously, phone home if they detect a faulty component. Service personnel will be on site, same day, to replace it before you even knew it was there, let alone had the opportunity to ask.
I wonder how many datacenters bother with those sorts of things anymore, given how many workloads are no longer tied to individual nodes. Seems like it would just be easier to accept that a certain percentage of nodes will be down at any one time, and go through and repair/replace them as needed.