Matt sent over a link that describes a recent problem Level3 ran into in a London Datacenter. The problem arose when one of their cooling units failed on a hot day. Below is a graphic display of what happened during the course of the day:
When the temperatures reached a certain level, servers began automatically going critical and shutting down to prevent hardware failure. As many of you might have guessed, 1U-2U servers really are not built to sustain operation in a remotely warm room, and running without proper air conditioning can cause some pretty serious problems!
This is a perfect example of why WebHostingBuzz uses the N+1 redundant cooling method to significantly reduce the chances of something like this ever happening to our clients.
