[Nut-upsuser] Unexpected shutdown begins for WOW-500U

[Arnaud Quette]

2011/7/24 Sergei V. Dubrov <dubrov at inp.nsk.su>

> G'day!
>
> On 23 Jul 2011 at 11:26, Sergei V. Dubrov wrote:
>
> To clarify some points...
>
> > >
> > > What I see after pulling the plug:
> > >
> > > ./upsc xen-home battery.charge
> > > 72
> > >
> > > ./upsc xen-home ups.status
> > > OB DISCHRG
> > >
> > > OB status is correct but why is there DISCHRG status?! This flag, as I
> > > think, cause immediately begin of shutdown although the battery is not
> > > discharged and has 72 percent of charge.
> > >
> > > Is DISCHRG exactly from UPS or is it an inaccurate interpretation of
> > > obtained status from WOW-500U?
> > >
> > > DISCHRG is a normal value:
>
> Yes, it's absolutely true. My mistake was that I  thought that DISCHRG
> status means DISCHARGED (low) battery. But really this mean DISCHARGING,
> that is quite natural immediately after switching to battery.
>
> >
> > Start test of battery:
> ...
> > ups.status: OB DISCHRG
>
> I used a test battery instead of turning off from the outlet, because the
> behavior of NUT in these cases coincides. After switching UPS to battery
> and obtaining status OB (instead OB+LB) NUT immediately starts shutdown
> (after FINALDELAY - 5 sec by default). My workaround - is significantly to
> increase the delay (FINALDELAY=90 sec for me) but it's not quite right.
>
> Confirmation that the shutdown should not begin on the status OB (instead
> critical OB+LB) I found in manual page:
>
> http://www.networkupstools.org/docs/man/optiups.html
>
> "fake_lowbatt
>
> This forces the low battery flag true. Without it, if you want to test your
> UPS, you will have to unplug it and wait until the battery drops to a
> low/critical voltage level before NUT will respond and power down your
> system. With the flag, NUT should power down the system soon after you pull
> the plug. When you are done testing, you should remove this flag."
>
> And once again want to emphasize - UPS does not go to state 'data stale'
> (loss of communication) after switching to battery since we can immediately
> read its status, for example, with the command 'upsc'.
>

but since upsmon launches the system shutdown, there is something triggering
it!

fake_lowbatt doesn't exist in usbhid-ups.
you're probably facing a data mapping mismatch.

the best would be to monitor this a bit more closely.
prefer to remove your PC from the UPS for those tests!
set SHUTDOWNCMD to empty or something not hurting (gedit) and restart upsmon
and the driver in debug mode (ie "usbhid-ups -DDD -a ..."), in separate
consoles.

then test a power failure, and send us back the results, along with your
configuration (passwd trimmed!)

cheers,
Arno
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/nut-upsuser/attachments/20110730/36103ece/attachment.html>