Bug 16302
-
Posting here as there was a bug that was rather hastily rejected on redmine that I was in fact able to reproduce.
I would have also rejected it as it was marked high ... and well the sensor graphs are certainly never a high concern. Sure, thermals can be problems, but graphing them over 6 months are more isn't a critical item. Noting that it is sitting at 130 degrees celcius currently...well that might be critical and that part works fine.
However it wasn't rejected for criticality rating; please open it back up and put it on the back burner.
Confirmed on 25.07.b.20250628.0006
No patch notes or redmine activity in the vicinity lately; will bump to latest built today shortly.
Confirmed still on 25.07.b.20250707.2346 -
MMMM.. I have several boxes running in a room with broken (no) AC running that were before the latest upgrades running just about where they are now.. fairly warm.
Not discounting.. (much) but what is your setup.. environmentals.. ect??
Intel(R) Core(TM) i5-4590T CPU @ 2.00GHz running at 57c before the last upgrade and now. on a 1u box..
Not seeing any changes here. Maybe driver differences?
-
@chpalmer
yeah I'm fairly sure this confusion is the source of the rather quick bug closure.It isn't about the temperatures themselves.
It isn't about the sensors themselves.
It is about the graph function listed in bug 16302.There is a graph function with a handful of dropdown options. It defaults to CPU; and you can pick right now, a week, a month, whatever of data. A year of cpu is whatever it is, but the dates are correct at the bottom of the graph. Flip it to thermal sensors, and the graphs have a pretty whack time range despite having the exact same 1 year or 6 month setting or whatever.
To your query; mine is sitting ambient room temperature between 65-74F (18-23C and the thermals on the firewall itself usually report between 91-95F (32-35C).
Again...it isn't the sensors...it is the specific graph function. -
This looks like what happens if the actual thermal sensors available changes and it stops logging thermal data.
On systems hitting that you would see at boot something like:
Jul 9 16:18:50 php-cgi 582 rc.bootup: The command '/usr/bin/nice -n20 /usr/local/bin/rrdtool update /var/db/rrd/system-sensors.rrd N:U:U:U:U:U' returned exit code '1', the output was 'ERROR: /var/db/rrd/system-sensors.rrd: found extra data on update argument: U'If you remove /var/db/rrd/system-sensors.rrd and reboot it will recreate it with the correct number of data fields.
-
This might be of interest:
After upgrading to 23.01 Status/Monitoring Thermal Sensors not updating
-
Yup, nice. If you want/need to keep the thermal data you can use the rrd tools to modify the data set to match the new sensor list.
-
Clearing it does in deed fix up the graph dates.
I don't remember any hardware thermal changes...but sure. :)
-
Most commonly new sensors appeared because an update had new drivers that exposed them. The chipset PCH sensor was pulled into base for 23.09 (I think!) for example.
