I've had to clean up after some incidents like that. There was one where an accidental recursive delete was running for several hours traversing a bunch of moun

I've had to clean up after some incidents like that. There was one where an accidental recursive delete was running for several hours traversing a bunch of mounts. root was mapped to nobody so it could only delete from world-writable directories but that was bad enough. It was on Netapp filers so it wasn't too hard to restore from snapshots but it took us a while to find the culprit. There was a lot of traffic from many machines so it wasn't trivial to figure out where the REMOVE calls in question were coming from.