Steve, >> Several hundred sensors? >> >> Most data centers will run out of floor space before you can >> put that many racks in the room! >> >> We haven't seen much "value" in a sensor per machine.. the >> air is so mixed by the fans in the box that you really can't >> predict much about what is going on with a given machine in the rack. >> >> You don't want to put sensors inside the machines, because >> you will block, a possibly "significant", amount of airflow >> with the sensor. >> >> Our best luck has been a sensor per rack and a "grid" of >> sensors on the ceiling.
I should clarify: I was only planning two sensors per rack. One under the floor to measure incoming temperatures and one at the top of the rack. Plus two sensors per air handler (same configuration). We have over 175 racks now and are getting ready to begin construction to double the floor space of our data center. Last year, our supercomputer was #3 on the top 500 list. We've dropped to #16 now: http://www.top500.org/sublist/Site.php?id=1391 As far as I know, we will be the first supercomputing center to utilize such an extravagant temperature monitoring system. >> With a couple of hundred machines we only have about 40 racks. We have in the neighborhood of 1200 machines now. Our primary supercomputer has 968 nodes alone plus over a dozen racks of network infrastructure. It sounds like you have "been there, done that" in this application; I have a lot to learn from you. Where in the rack did you mount your sensor? I'd planned on the top of the rack, but maybe that isn't the best place for it. I'm also interested in your equipment configuration. How many sensors do you have? Which interface are you using? Which sensors? My biggest hangup was cabling; how did you wire your sensors up? I'm trying to avoid the time consuming job of soldering up custom cables for every sensor. >> We can easily scan our entire collection in less than a >> minute.. and we log temps every 5 minutes. (RRD handles the >> average duties.) How many sensors and are you using parasitic power? I'd still prefer the ability to get nearly instant readings. I'd planned on coding a web interface where one could query a group of sensors (or the whole room) and get rapid feedback. I suppose I could simply have the web interface query the database; it wouldn't be "real time", but should be close enough. >> Try instrumenting one rack "all over the place" and another >> just at the exhaust and see if you can convince yourself >> that you can see something of value in any given reading >> change before you commit to putting dozens of sensors in every rack. Nope - never thought there would be any value in this. I just have a big data center. (Actually, in my world our data center is tiny... The really big ones span acres.) >> Our most useful reading for alarms is the deltaT on the >> airhandler.. it's where the first indication of an AC >> problem shows up. Absolutely; that was one of the first things on my list to monitor. -___----------------------------------------------------------------- | , | ___ ___ _ -- Ryan P. Wright | _/__ __/ _ \ | | | -- MSCF Operations | \ \/ / /_\ \| | | | -- [EMAIL PROTECTED] |_|\_\ /_/---\_\_|___| -- (509) 376-3502 -----|_|------------------------------------------------------------- ------------------------------------------------------- This SF.Net email is sponsored by: NEC IT Guy Games. How far can you shotput a projector? How fast can you ride your desk chair down the office luge track? If you want to score the big prize, get to know the little guy. Play to win an NEC 61" plasma display: http://www.necitguy.com/?r _______________________________________________ Owfs-developers mailing list Owfs-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/owfs-developers
