The reason we all know to be careful with rm -r is because of that one time we weren't. Me, it was the time I rm r'ed the MySQL data directory for my company's customer service management system. Oops. Thankfully we had a backup from the month prior, but I learned two things that day: a) be really careful with rm and b) take it on yourself to make sure IT is backing up stuff that you're messing with. You gotta hedge…
So, I did a simple /etc/init.d/clearcase stop (not sure that is the exact name) and:
# rm -rf /opt/clearcase
(hmm... that seems to be taking a little too long to run)
Panic - then ctrl-C - it was too late, /opt/clearcase/foo was NFS mounted 127.0.0.1:/ and it had already trashed /bin, /sbin, /etc, /var, and most of /usr.
Luckily I had good backups, but we did spend the rest of the day rebuilding the source repository while the developers couldn't check in any code.