[MrMatt]

Been in this exact situation myself a couple of times - one instance that has stuck with me was when >300 sites crashed within around 5 minutes of each other due to a time-sensitive bug.

This was my first job out of University, so I had the unpleasant pleasure of seeing both sides of the coin; my manager panicked, but the senior developer didn't. I got roped into a pointless panic-driven hunt for viruses / corrupted installs / whatever by my manager (at one point we were analysing the assembly code in DOS debug, for some reason), while the senior dev calmly reverse-engineered the crash conditions, reproduced the problem and found the root (a negative index into an empty array, written by the MD of the company).

It was on that weekend I decided to renege on my old panicky habits, and become the calm center of reason in the face of a crisis.