On Tuesday (04/09/2002 at 04:47PM -0400), Mike Snitzer wrote: > > Funny you should say that, I implemented a script that pushes out the cpu > temps through /etc/cron.d/ (will add fan speed shortly).. I've attached > the script in case others would like it. (the cut et al was based on a > previous list post).
Alright! My scripts are only half-baked at this point but I chose to read the /proc/sys/dev/sensors files directly rather than parse 'sensors' output. I then have a mapping between the various sensor files and generic ganglia metric names. We'll see how well this works out. > These scripts should start to find a place of their own in the ganglia > project; I could see users contributing some extremely useful gmetric > scripts, Matt? > > > > > Might this be because I currently have only one node sending in these > > values via gmetric? I was expecting to still see mins and maxes even > > if they are only for one node... > > That's exactly why; if you send the metric via gmetric on multiple hosts > all is well with high-low range indicator. ah ha! OK. I'll have to enlist more nodes for the experiment then. Thanks for the clue. Chris -- Chris Elmquist mailto:[EMAIL PROTECTED] Unlimited Scale, Inc. http://www.unlimitedscale.com St. Paul, MN
