On Tue, 21 Jul 2009, Jean Delvare wrote: > Given the specific timing of your project, I suspect it's too late, > but maybe for next time... It's actually working out well. We haven't run a system through the thermal chamber yet, so this info will come in handy. I also haven't asked the motherboard manufacturer yet; I'm 80% sure that the first few rounds will consist of "Just use the handy Windows monitoring utility" and I don't feel like explaining it at the moment. :) > Out of curiosity, which monitoring chip is it? Not all chips require > compute statements. sensors-detect says "Winbond W83627EHF/EHG Super IO Sensors" at 0x290, driver w83627ehf . (Note that this is lm-sensors 2.x; I have to use the packaged version that is already on the machine.) I haven't opened up the case to see what chip is actually there, but this driver provides plausible results so far. > Grabbing the raw values from /sys/class/hwmon has the advantage that > it is much easier to feed into a spreadsheet or scripts for > post-processing. The output of "sensors" is more tedious to parse. I already worked up a really simple script to grep out what I wanted from the "sensors" output and hack it into a one-line-per-reading CSV format, but it may not have been very robust in the face of things like alarm warnings. I have enough disk that I _could_ just log the straight "sensors" output to timestamped files, or to one big file with timestamps in it, and then process the whole thing later. Thanks! Matt Roberds
