Based on reading @Frank's experience living in a server sauna room, I thought I'd share some WTFery about my project's server room(s).
Our original office suite was in the "nicer" set of buildings, ya know, glass and steel building instead of concrete. Well, at least the outside of the building looked nice. The suite itself was probably half of the floor, so a standard L-ish configuration. Two kitchenettes with full size fridges, sinks and microwaves, a major conference room and a back door "draft" conference room, bull-pen area, not too bad at all. Sort of.
The office had two breakers, one for the front half; one for the back half. The back half was our two testing labs, server room and developer offices. So 3 full size racks of servers, about 12 user(dev/tester/whoever) laptops and about 20 test laptops once we ramped up. Nothing that should stress power too much. Technically the servers should have been on their own breaker but whatever, we didn't have any trouble.
Once we started to ramp up the project however... A few weeks in we would regularly have the breakers flip during the day. So hearing 'God-damn-son-of-a-bitch' in the middle of the afternoon got to be pretty common place because it is hard to access a SAN/VMs/Network-resources when they goes down once a day.
Turns out we'd trip the breaker around lunch when we did lose power. Remember how I mentioned the two kitchenettes? Well it was the microwave in the back half that would push the breaker overboard. Yeah I know, TRWTF is the servers weren't on their own breaker. By time we figured that out management had decided we were going to move because you can bill more if your server room is bigger or something.
So we moved into a "better" suite(ie across the street but in the same complex and shitter. It does have a dedicated breaker for the server room). Problem with the newer space? The original HVAC was apparently powered by hamsters. Since we came to our new space the HVAC has died at least 6 times, and at least 4 of them the lab team didn't find out until developers sent email complaining systems were offline. Server room had gotten to 94F+ (34.4C) in those cases. We stopped a few meltdowns because the dev test-lab was on the other side of the wall to the server room. So when we started to feel toasty we knew shit was going south.
As a result of one of those meltdowns we (only) lost(sort-of) a week of work. But as a result we did get a new HVAC and a brand new SAN. The only WTF with the SAN was we had to have an HP rep on site so the lab team could put it together, but since the HVAC has been fixed it's been mostly smooth sailing.