The entire service going down for 24 hours due to a reboot is not a consequence of poor engineering on the part of the user. A production service which people rely on for critical data failing on the _textbook_ example of running a live service is poor engineering on the services part.