It's both the RRD values and the System Status GUI - the status page correctly shows all 8 cores, but the temperature as reported by sysctl is wildy inaccurate at the moment.
There is over 40GB free so space should not be an issue. Besides, ipmitool is giving the right result - it's only sysctl values that seem to have stopped giving the right values.
As I said, it was working perfectly for the last 9-10 months. It worked perfectly after the most recent upgrade. It was working after the last reboot for about a month.
But sometime about 3 weeks ago it suddenly stopped working and started to report 15C as the only result. (To be honest, it does give slightly different values for the other cores, as shown below) - but the graph clearly indicates that core 0 was 'reported' as 15C for the whole time - when clearly (based on previous graph values) this is not a correct number.
Code: Select all
sysctl -a | grep temper
Code: Select all
# ipmitool sensor
CPU Temp | 21.000 | degrees C | ok | 0.000 | 0.000 | 0.000 | 93.000 | 98.000 | 98.000
System Temp | 27.000 | degrees C | ok | -9.000 | -7.000 | -5.000 | 80.000 | 85.000 | 90.000
Peripheral Temp | 30.000 | degrees C | ok | -9.000 | -7.000 | -5.000 | 80.000 | 85.000 | 90.000
DIMMA1 Temp | 29.000 | degrees C | ok | 1.000 | 2.000 | 4.000 | 80.000 | 85.000 | 90.000
DIMMA2 Temp | 26.000 | degrees C | ok | 1.000 | 2.000 | 4.000 | 80.000 | 85.000 | 90.000
DIMMB1 Temp | 24.000 | degrees C | ok | 1.000 | 2.000 | 4.000 | 80.000 | 85.000 | 90.000
DIMMB2 Temp | 26.000 | degrees C | ok | 1.000 | 2.000 | 4.000 | 80.000 | 85.000 | 90.000
FAN1 | 3200.000 | RPM | ok | 300.000 | 500.000 | 700.000 | 25300.000 | 25400.000 | 25500.000
FAN2 | 1300.000 | RPM | ok | 300.000 | 500.000 | 700.000 | 25300.000 | 25400.000 | 25500.000
The only possibility I can think of is that I built a new VM in VirtualBox around 1st March and installed Ubuntu to do some testing (Joomla upgrade php5.x->php7.2), so I wonder if some microcode update could possibly have interfered with the readings from sysctl? Note that this VM was built 3 weeks prior to the issue occurring - ie:it was built around 1st March - and this issue didn't appear until the last week of March.
No other graphs appear to be affected - only CPU Temp - but it affects both the monitoring graph and the status page.
Otherwise I'm at a complete loss as to why this has happened.
Will reboot it later today to see if it recovers or is permanently broken now. (Rebooting it is a pain as the startup regenerates the jail configs - and I have one jail that requires it's own bridge otherwise routes wrongly through the host nic and that's not why I bought a 4x1GBe motherboard. Restarting the host takes barely minutes - getting the jails running and routing correctly again can take up to 2 hours of pure frustration [sigh] - vnet is (despite being there for years) still experimental, I guess).