I once tried to help a company with similar infrastructure insanity recover from a massive failure. Absolutely brutal.

When my team finally got services up and running (barely) after ~18 hours of non-stop work, the CTO demanded that we not go home and get some sleep until everything was exactly as it had been before the failure.

Not my happiest day.