[nut-upsdev] Belkin and newhidups
Peter Selinger
selinger at mathstat.dal.ca
Thu Sep 1 01:00:24 UTC 2005
zaid_aj at telus.net wrote:
>
> Hi Peter,
>
> Thanks for clarifying the OL CHRG message.
> About the experiments, I don't have ups.timer.* but ups.delay.* which I
> assumed was the same thing so I started writing values to
> ups.delay.shutdown & ups.delay.restart and here are the results:
Yes, that's what I meant. They are called ups.timer.* in the belkinunv
driver; I'll submit a patch to fix that.
> Permenant shutdown:
> upsrw -s ups.delay.shutdown=60
> I turn my system off, and after a few seconds the UPS goes offline
> and doesn't come back until I manually press the power button.
>
> Timed shutdown:
> upsrw -s ups.delay.restart=2
> upsrw -s ups.delay.shutdown=60
> I turn my system off, and after a few seconds the UPS goes offline,
> then after a few more seconds it comes back online (doesn't matter
> if AC was present or not). By the way, for some reason, it seemed
> that the UPS restarted in less than 2 minutes, something like
> ~4 seconds..
The behavior of the timers is a bit strange. First, the restart timer
decrements at 60-second intervals, but the timing of these intervals
is determined by a fixed clock, independently of when the variable was
set. So the decrease of this variable from 2 to 1 could occur any time
from 0 seconds to 60 seconds after the variable was set.
Further, the countdown of the ups.delay.restart timer is independent
of the ups.delay.shutdown timer. I.e., the restart interval does not
start at shutdown time, but as soon as you set the timer.
It is actually dangerous to do
upsrw -s ups.delay.restart=2
upsrw -s ups.delay.shutdown=60
because if you are extremely unlucky, the first timer could go down
from 2 to 1 just before you set the second timer. The first timer will
then reach 0 just before the second timer reaches 0, and your UPS will
never restart. Safer would be:
upsrw -s ups.delay.restart=2
upsrw -s ups.delay.shutdown=55
assuming you wait less than 5 seconds between the two commands.
> Soft shutdown:
> I'm not sure how to produce this :)
>
> I'm having a bad feeling that this UPS is another brain-dead one from
> Belkin..
:( -- Peter
More information about the Nut-upsdev
mailing list