Computational Thermodynamics -Thursday, June 17, 2004 -11:08 pm-
A week and 2 days ago, my AC went out. Again. For the third time this year, the fifth time overall. Management finally decided to give me a new condensor/compressor unit. I thought this would be a good thing. It really wasn't. It meant I spent a week and 2 days totally without AC. Now, think back over the last 9 days and remember the weather. Ick. I had to sleep on the couch some. When it gets humid, my heart rate and blood pressure rise dangerously, so I couldn't sleep on the top floor. This is also why I had to come home from Kentucky early. Sorry if I offened anyone, but I was starting to ahve trouble staying alive and it was making me cranky. I wanted to leave before I snapped at someone and pissed people off. I slept for most of the ride home and then for the next 14 hours. Anyway. As rough of a time as I was having, my server was also greatly stressed. The CPU was running HOT. Hotter than I've ever seen it, and I've run it without the heat sink on (don't ask). I had a thermal monitoring program running the whole time, with a script running to shutdown -h now the computer if it reached 55C.
The new AC arrived today, thank god. Once I had cooled off, I took a look at the historic graph of CPU temp that had accumulated. It cracked my ass up. The time scale in the first image is 7 days. You can see that when I tried to do a video import, I almost killed the computer. Then, you can see the day/night cycles as the temperature in the room rose and fell with the sun. Finally, you can see a crash in the temperature when the new AC arrives and starts working.
The second image only contains six hours. You can see a very clear downward trend when the AC starts to function and then a cyclical pattern as the house reaches its target temperature and begins to cycle the AC on and off. Finally, you can see the spike in temperature as I finally feel safe using the computer again.
