One VM causes fans spinning up in intervals

[Asimov]

Hi all,

I am running a single node ProxMox 6.04 with a bunch of VMs in it and with very minimum load as you can see below:



I recently imported a new VM which was originally a VMWARE's VM. I followed the documentation. It all worked well. However, when I turn on that VM, after about 5 min the host fans star spinning up and down. They go up for about 10 seconds (very loudly) and then they go down for about five minutes. The process then starts over again and again. Now, if I shout down the VM it does not make any difference, fans keep going up and down. The only workaround I have been able to find is to reboot the host itself making sure that VM does not start up on boot.

When the fans are up, I cannot see any metrics either on the host or the VM itself going up as well.

I deleted the VM, re-created the media (migration process from zero) and created it again new. Same behavior.

Any idea where I could start troubleshooting this? Any useful log somewhere?

Thank you!

 

[Stoiko Ivanov]

Any idea where I could start troubleshooting this? Any useful log somewhere?

First thing to check is for any messages in the system's journal (and dmesg):
when the fans start running you can start watching the log:
* `journalctl -f` (keeps the journal open and outputs any lines which get written)
* `dmesg` (outputs the kernel message buffer)

I hope this helps!

 

[Asimov]

Thank you @Stoiko Ivanov . I'll try this out.

 

[James Crook]

I'm gonna just randomly say, are you crossing NUMA nodes ?

or loading up the second CPU cause of RAM being used on the other cpu, which wouldn't really show extra usage, just the CPU coming out of low power c-state adding to extra heat in the case?

Not sure on your case/chassis but i've seen some HPs not throttle back fans to super low levels after ramping up.

 

[Asimov]

I'm gonna just randomly say, are you crossing NUMA nodes ?

or loading up the second CPU cause of RAM being used on the other cpu, which wouldn't really show extra usage, just the CPU coming out of low power c-state adding to extra heat in the case?

Not sure on your case/chassis but i've seen some HPs not throttle back fans to super low levels after ramping up.

Hi @James Crook , thanks for the reply. how do I know if I am crossing NUMA nodes? (sorry I am not familiar with NUMA). I have a Lenov D30 workstation as host.

 

[Dark26]

try tu launch in consoel a watch sensors to see your temp

watch sensors


k10temp-pci-00c3
Adapter: PCI adapter
temp1:        +19.4°C  (high = +70.0°C)
                       (crit = +80.0°C, hyst = +77.0°C)

it8728-isa-0228
Adapter: ISA adapter
in0:          +1.27 V  (min =  +0.00 V, max =  +3.06 V)
in1:          +1.64 V  (min =  +0.00 V, max =  +3.06 V)
in2:          +2.00 V  (min =  +0.00 V, max =  +3.06 V)
in3:          +2.03 V  (min =  +0.00 V, max =  +3.06 V)
in4:          +2.03 V  (min =  +0.00 V, max =  +3.06 V)
in5:          +1.22 V  (min =  +0.00 V, max =  +3.06 V)
in6:          +2.22 V  (min =  +0.00 V, max =  +3.06 V)
3VSB:         +3.31 V  (min =  +0.00 V, max =  +6.12 V)
Vbat:         +2.95 V 
fan1:         749 RPM  (min =    0 RPM)
fan2:        17763 RPM  (min =    0 RPM)
fan3:           0 RPM  (min =    0 RPM)
fan4:        2657 RPM  (min =    0 RPM)
fan5:           0 RPM  (min =    0 RPM)
temp1:        +33.0°C  (low  = +127.0°C, high = +127.0°C)  sensor = thermistor
temp2:        +36.0°C  (low  = +127.0°C, high = +127.0°C)  sensor = thermal diode
temp3:        +18.0°C  (low  =  +0.0°C, high = +90.0°C)  sensor = Intel PECI
intrusion0:  ALARM

fam15h_power-pci-00c4
Adapter: PCI adapter
power1:       79.90 W  (crit =  94.99 W)

and see if one temp increase rapidly.

 

[Asimov]

try tu launch in consoel a watch sensors to see your temp

Thanks @Dark26 , I've been keeping an eye on this but oddly enough I cannot catch any sizable change, even fan RPM remains quite stable at around 4300 RPM.

I'll keep monitoring this to try catching what the heck is going on.

Thanks again.

 

[James Crook]

Hi @James Crook , thanks for the reply. how do I know if I am crossing NUMA nodes? (sorry I am not familiar with NUMA). I have a Lenov D30 workstation as host.

This post has some more info, but i'm using an old i5 so no numa for me ...

https://forum.proxmox.com/threads/best-practice-for-numa.25934/