probably because proper architecture (clustering, HA etc) and planning would have never made this an issue. This is still an extremely risky operation, hot swapping power and switching interfaces on the fly all while sitting on a cart in a corridor. In any disruptive work there is never a guarantee of no downtime for affected assets. I know the OP came in as a consultant, but If I was the MSP tech, I would have demanded a paper trail a mile long to cover my ass if this went sideways and If i was the account manager for the client, I would have refused the work. Its not good business to agree to do work where you know there is a better than good chance there will be an outage and your client is saying they wont pay if there is an outage. Even agreeing to it puts you in a bad spot for future work. I guess as an outside consultant, bewilderment is a better reaction than ranting, but this is the kind of shit that drives ops folks crazy