A more expensive event that happened on various sites was in the days of mainframe "large" exchangeable disks (8MB - 200MB).
The operator had a failure of the system disk - so he moved it to each of the drives in turn - and they all failed. He then loaded the reserve copy on a drive - and that failed. Finally he loaded the master copy on a drive - and that failed.
The original fault had been a head crash which ruined the heads on the first drive - and also scratched the first disk's surface. Moving that disk ruined the heads on all the other drives - which in turn ruined the reserve and master disk's surfaces.