Last week we reviewed three examples in which overheating caused severe malfunction to data centers. In the case of Microsoft, the outlook.com endured a 16-hour long outage on March 14th of 2013, and this unexpected accident brought on trouble for users. Similar downtimes took place in other data centers as well. Regardless of the sizes and services of their data centers, no company is immune to the constant threat of overheating.
Meanwhile, overheating causes more damage than immediate failures. An increase from 68 °F to 86 °F can lower the long-term reliability of the equipment by as much as 50%. Professor Robert G. Brown from Duke University pointed out that excessive heat “weakens electronic components like power supplies, motherboards, and memory chips, so even if they don’t fail immediately, they become more susceptible to failure over time.” In his blog article “Skimp on Server Room Air Conditioning? At Your Peril." he writes, “power supplies, motherboards, memory chips -- that kind of heat weakens components so that forever afterward they are more susceptible to failure, not just during the event. The overheating can just occur one time, for a few minutes, and you’ll be cursing and bitching for months and months later dealing with all the stuff that got almost-damaged, including the stuff that isn’t actually broken, just bent out of spec so that it fails, sometimes, under load.”
According to Professor Brown, one time a server room got drastically overheated for an extended period, it was followed by node crashes and a series of hardware failures over the next three months, “some immediate and obviously due to overheating, some a week later, two weeks later, four weeks later.”
Even though some individual servers come with built-in temperature monitoring devices, which automatically send alerts if the temperature around the server is too high, it is not enough to ensure safe operating temperatures, because data center temperatures vary from one area to another.
Please consider the following guidelines in order to protect data centers from varying environmental conditions:
- Temperature and humidity sensors should be installed on or near individual racks and critical devices
- Measurements should be logged and graphed over time to help administrators spot trends, such as temperature spikes during peak operating hours or fluctuations
- Sensors should be placed:
- On top, middle, and bottom of individual racks to measure the heat being generated by equipment
- At the air conditioning system’s intake and discharge vents to measure efficiency
- Probes should be placed around critical devices because the temperature inside a rack-mounted device could be as much as 20° higher than the surrounding area
- A probe near the thermostat can help check what the thermostat is “seeing” while controlling the room temperature
To prevent disasters or correct them before any damage is done, Temperature@lert has designed an array range of
low-cost automated products that protect your precious assets 24/7 while minimizing or even eliminating the amount of
time, effort and expertise needed to install them, as well as the amount of time needed to manually check on current
environmental conditions. Temperature@lert products are simple and reliable – simple in that we only include the
features you need. Reliable because most of our products represent 4th or 5th generation hardware, so you know
they’re battle tested.
Ivory Wu, Sharp Semantic Scribe
Traveling from Beijing to Massachusetts, Ivory recently graduated with a BA from Wellesley College in Sociology and Economics. Scholastic Ivory has also studied at NYU Stern School of Business as well as MIT. She joins Temperature@lert as the Sharp Semantic Scribe, where she creates weekly blog posts and assists with marketing team projects. When Ivory is not working on her posts and her studies, she enjoys cooking and eating sweets, traveling and couch surfing (12 countries and counting), and fencing (She was the Women's Foil Champion in Beijing at 15!). For this active blogger, Ivory's favorite temperature is 72°F because it's the perfect temperature for outdoor jogging.