[Nut-upsuser] Can't get 'load.off' to turn off power to the PC: CyberPower 1200AVR

[Arjen de Korte]

Mark E. Hansen wrote:

> Thanks for the help in getting the current development sources (to Charles
> Lepple as well). I've downloaded and configured/installed nut-2.3.0-r1114.
> 
> However, I don't see any difference in how the setting of some variables
> appears to be ignored.

In that case, chances are that your UPS doesn't support changing these
values. If it doesn't accept any changes to these values, there is
nothing we can do about it.

> For example, I used upsrw to view the variables available on my UPS and
> see that this list includes battery.charge.low, which is set to 10. When
> I set this to another value, an entry is written to /var/log/messages which
> indicates that the value was successfully set, but the next upsrw run shows
> that the value is still 10.

That's something that certainly needs improvement. Currently it is
set-and-forget, while it really should be set-and-get instead.

> I tried many different values, 20, 50, 90, 25, etc., but all acted the same
> way.

See above. Your UPS may not support changing these values.

> I tried setting battery.runtime.low, which is the number of seconds of battery
> runtime left when the UPS switches to LB, in seconds, but was not able to set
> this above 600. When looking at the current status of the UPS, I see that the
> existing runtime is 4702, so 600 seconds is about 13%.
>
> What I really want to be able to do is get the UPS to decide to shut down the
> PC when it is only a few minutes into the power failure. I know the UPS must
> be able to support this because I can do this using the Windows-based
> configuration tool (I have the exact UPS running on a Windows/2000 machine as
> well).

This may be an action that is not triggered by the UPS, but rather a
timer that is started by the PC software as soon as the UPS is on
battery. We have 'upssched' to do something similar. See 'man 8
upssched'. I assume you have read and understood the reasons in the FAQ
why we generally recommend against doing that.

> Perhaps there's just some magic in setting these values which I've not yet
> found?

I doubt it.

Best regards, Arjen