On 08/28/2012 03:39 PM, W.J.M. Nelis wrote:
Hello,
Hi, I've written a quick addon to collect the temperature of each HDD in my system, and report the data back to Xymon like this:
Fri Aug 24 02:33:11 EST 2012 Temp GOOD (27 degrees)sda : 27 sdb : 27
Also, hobbit is putting this data in to an RRD file called temp.rrd:
........ An approach is not to monitor the temperatures of the individual disks, but to monitor the temperature of the group of disks on a single host. The rationale of this approach is that normally one is interested in the range of temperature values, combined with an alert if at least one temperature is out of range.
I've used this approach to monitor 'the temperature' of a switch. The number of temperature sensors per switch varies from 1 to 29 (!). The returned value is the minimum, average and maximum temperature, thus always three values per switch.
I wanted to monitor each individual disk for the following reasons:
- To see if one drive is failing, maybe it will have a much higher temp than the others
- Since each drive is in a different position, it is nice to see the effect that this has on the individual drive temp
I can see why it would be pointless to get 29 different temp readings from a single switch, I can't imagine that the temp would vary much between all of those, and it isn't going to have individual parts that can be replaced like you can with individual drives in a server.
Regards, Adam