[Nut-upsuser] What is the meaning of ups.delay.*

[Greg Troxel]

Eyal Lebedinsky via Nut-upsuser <nut-upsuser at alioth-lists.debian.net>
writes:

> When shut down the ups reported 38% charge. When it booted the ups reported zero charge.
> ups capacity is logged every minute:
> 	20260822141100 100
> 	20260822141200 97	wall power lost
> 	...
> 	20260822143200 40
> 	20260822143300 38	machine shuts down
> 	20260822152600 0	machine booted
> 	20260822152700 1

This is messy. There's an unclear relationship between battery voltage
and %time remaining.  It's highly impacted by discharging vs charging,
to the point where you need totally separate estimations.   You are
observing that either

  the machine shut down at 38 but the ups powered the not-running
  computer or other things and didn't shut down and drained the battery

or

  the estimate of charge when charging afer boot is pessimistic


As they always say "further research is needed".

> What is the relationship between the battery charge and the timing of the shutdown.
> Maybe it is related to the UPS not sending data, which triggers the shutdown?
>
> The status of the UPS now (over 3.5h since charging started) is:
>
> $ date ; upsc ups
> Sat Aug 22 19:00:00 AEST 2026
> battery.charge: 87
> battery.charge.low: 20

.low is a clue that shutdown happens at 20, maybe.