NUT service taking 100% CPU on 1 thread when power is cut to the UPS

[CacheMeIfYouCan]

I have an issue with NUT, the UPS monitoring service.

I've connected a Cyberpower CP1500AVRLCD3 UPS to a USB port of my TrueNAS Scale (Cobia). I've configured the NUT service, and all seems ok. Everything is fine until I pull the plug on the UPS to engage the battery. Then the NUT process runs crazy (no pun intended), using 100% on one thread, continuously and until power is recovered. This is really unproductive since this increases power draw by 16W, that's +28% in my case.

Is this the expected behavior, misconfiguration on my part or a bug?

Let me know what other info may be helpful in this investigation.

Code:

admin@serveur4[~]$ upsc cp1500
Init SSL without certificate database
battery.charge: 100
battery.charge.low: 10
battery.charge.warning: 20
battery.mfr.date: CPS
battery.runtime: 9925
battery.runtime.low: 300
battery.type: PbAcid
battery.voltage: 27.0
battery.voltage.nominal: 24
device.mfr: CPS
device.model: CP1500AVRLCD3
device.serial: BHPNW2000657
device.type: ups
driver.name: usbhid-ups
driver.parameter.pollfreq: 30
driver.parameter.pollinterval: 2
driver.parameter.port: /dev/uhid
driver.parameter.synchronous: auto
driver.version: 2.8.0
driver.version.data: CyberPower HID 0.6
driver.version.internal: 0.47
driver.version.usb: libusb-1.0.26 (API: 0x1000109)
input.voltage: 118.0
input.voltage.nominal: 120
output.voltage: 118.0
ups.beeper.status: disabled
ups.delay.shutdown: 20
ups.delay.start: 30
ups.load: 6
ups.mfr: CPS
ups.model: CP1500AVRLCD3
ups.productid: 0601
ups.realpower.nominal: 900
ups.serial: BHPNW2000657
ups.status: OL
ups.test.result: No test initiated
ups.timer.shutdown: -60
ups.timer.start: -60
ups.vendorid: 0764
admin@serveur4[~]$

 

[CacheMeIfYouCan]

Any idea on how to debug this issue?

 

[CacheMeIfYouCan]

Here's more detailed info on the situation on my TrueNAS Scale while in UPS power loss:

A process named 'upssched' uses 100% of one CPU thread (user: nut).

Code:

admin@serveur4[~]$ upsc -l
Init SSL without certificate database
cp1500
admin@serveur4[~]$
admin@serveur4[~]$ upsc cp1500
Init SSL without certificate database
battery.charge: 97
battery.charge.low: 10
battery.charge.warning: 20
battery.mfr.date: CPS
battery.runtime: 6235
battery.runtime.low: 300
battery.type: PbAcid
battery.voltage: 25.1
battery.voltage.nominal: 24
device.mfr: CPS
device.model: CP1500AVRLCD3
device.serial: BHPNW2000657
device.type: ups
driver.name: usbhid-ups
driver.parameter.pollfreq: 30
driver.parameter.pollinterval: 2
driver.parameter.port: auto
driver.parameter.synchronous: auto
driver.version: 2.8.0
driver.version.data: CyberPower HID 0.6
driver.version.internal: 0.47
driver.version.usb: libusb-1.0.26 (API: 0x1000109)
input.voltage: 0.0
input.voltage.nominal: 120
output.voltage: 120.0
ups.beeper.status: disabled
ups.delay.shutdown: 20
ups.delay.start: 30
ups.load: 9
ups.mfr: CPS
ups.model: CP1500AVRLCD3
ups.productid: 0601
ups.realpower.nominal: 900
ups.serial: BHPNW2000657
ups.status: OB DISCHRG
ups.test.result: No test initiated
ups.timer.shutdown: -60
ups.timer.start: -60
ups.vendorid: 0764
admin@serveur4[~]$

 

[CacheMeIfYouCan]

I've found that an issue was raised with upssched 2.8.0 last year on github.

upssched high CPU usage on FreeBSD · Issue #2009 · networkupstools/nut

Currently running nut 2.8.0_23 compiled from ports on FreeBSD. I noticed that any timers that I set in upssched.conf occupy ~90+% CPU usage of a single core indefinitely while a timer is running. O...

github.com

The symptoms match perfectly the issue I have noticed. As far as I can tell by running "/lib/nut/upsmon -V" it is this version 2.8.0 that is running on my TrueNAS Scale 23.10.0.1.

 

[sfatula]

It appears it matches to me as well. I think your best bet is to raise the issue as a ticket with IX and ask them to maybe include the fix in the next Scale version.

 

[CacheMeIfYouCan]

I did open a ticket yesterday (before finding the issue on GitHub) but was unsure if they would respond since I haven't bought any hardware from them, nor any service contract.

 

[Redcoat]

but was unsure if they would respond since I haven't bought any hardware from them, nor any service contract.

There is no such requirement to receive attention through the bug reporting process.

 

[CacheMeIfYouCan]

Thanks @sfatula I have found how to post a bug report ticket, it's recorded as NAS-126746.

 

[CacheMeIfYouCan]

@sfatula I've got a response from the bug report ticket. An upgrade to the latest NUT version will be part of Scale 24.04 (named Dragonfish). The Beta is due in early Ferbruary.

 

[sfatula]

At least it only matters when the power is off so limited problem.

 

[CacheMeIfYouCan]

At least it only matters when the power is off so limited problem.

On an environmental perspective, perhaps. But on a system administration and cost perspective, it does matter a lot. The power draw of my system increases by 40% thus the autonomy of the UPS is reduced by the same proportion.

 

[sfatula]

I agree of course. But it only matters if the power ever goes off. Do you really lose power that often? Ours kicks in like once a year,. Definitely reduces the runtime if it does go off. I was merely trying to say there are much worse problems you could have had.