[Nut-upsuser] Various NUT issues/questions+NEW ONE

Fri May 12 18:55:56 UTC 2006

Scott "Tuc" Ellentuch at T-B-O-H wrote:
> 
> > > 1) I'm thinking this is a UPS issue, but when I pull the plug
> > > on my UPS, it goes from 100 to 30 in just a short amount of 
> > > time (Minutes). This causes the machine to shut down. But once
> > > it gets to 0, its beeped its little heart out for a few more
> > > HOURS providing power to the non-system items perfectly. The
> > > power supply for the Soekris is only a .35A power supply and
> > > puts out 12V 1.5A , so not like I'm shedding a huge load.
> > 
> > Possibly those non-system items are drawing very little power, causing
> > your UPS to last much longer than it would if your computer was still
> > powered up. Or possibly, the battery charge measurement is just very
> > inaccurate.
> >
> 	The computer had been drawing at most 42W/.35A . I turned on
> a 40W light bulb when the computer was halted, but not powered down.
> (Its unable to power down by itself). So effectively when I
> turned the lamp on, I was sucking MORE than previously since I
> had a 40W bulb and whatever the "at rest" draw of the computer
> was. (Not much, I realize, but still something)
> 
> 	Unfortunately, since my monitoring host is also the
> host it shuts down, this is the only logs I have :
> 
> 20060512 080139 100 NA 0 [OL CHRG] NA NA
> 20060512 080639 100 NA 0 [OL CHRG] NA NA
> 20060512 081139 100 NA 0 [OL CHRG] NA NA
> 20060512 081639 90 NA 4 [OB DISCHRG] NA NA
> 20060512 082139 74 NA 0 [OB DISCHRG] NA NA
> 20060512 082639 57 NA 0 [OB DISCHRG] NA NA
> 20060512 083139 40 NA 0 [OB DISCHRG] NA NA

That was 20 minutes from 100% to 40%. Does not look unusual. 

> > > 2) Lets say that the UPS gets low and the system shuts down,
> > > but the power comes back on... I guess that means the system
> > > stays shut down.. Whats the lowest number to safely set it
> > > to shut down? (But then I run into the problem with #1.
> > 
> > Of course not; the system should reboot when the power comes back on.
> >
> 	I guess I wasn't defining the scenerio correctly.
> 
> 	1) Power goes out
> 	2) UPS on Battery
> 	3) UPS gets to 30% battery
> 	4) System realizes it, does a "shutdown -h now"
> 	5) UPS gets to 1% battery 
> 	6) Power comes back on
> 
> 	During the entire time, there was never a power loss to the
> system

You did not configure your shutdown scripts correctly; see item 11
under "Configuring shutdowns for low battery events" in INSTALL.  Your
computer is supposed to power off (after going through the usual
shutdown procedure) by letting the UPS know to shut down its
load. This will cause the UPS to cut power to the computer, and then
sit there until power comes back on. In the event that the power
returns while the shutdown is already in progress, the UPS will cut
power to the computer for a few seconds, then come back on. This
ensures that your computer will never be in a bad state. 

> > On most (all?) PC's, there is a BIOS setting that lets you choose what
> > to do when power comes on: stay off, boot, or revert to previous
> > state. Many people use this to turn their PC on from the switch on a
> > power bar.
> >
> 	Doesn't apply here, power was never technically lost, the
> system only did a shutdown because it believed there wasn't enough
> power left in the UPS.
> >  
> > > 3) Most of the values in UPSlog are NA, such as input
> > > voltage, ups.load (Though I did see it say 4 ONCE, temp and
> > > input frequency. When I do a "upsrw" I get :
> > > [outlet.0.desc]
> > > [outlet.1.autoswitch.charge.low]
> > > [outlet.1.desc]
> > > [outlet.1.switch]
> > > [ups.delay.shutdown]
> > > 
> > > Is that the ONLY items I can get besides :
> > > 
> > > battery.charge: 0
> > > battery.charge.low: 30
> > > battery.runtime: 00000
> > > driver.name: mge-shut
> > > driver.parameter.port: /dev/cuaa0
> > > driver.version: 2.0.3
> > > driver.version.internal: 0.65
> > > outlet.0.desc: Main Outlet
> > > outlet.0.switchable: 0
> > > outlet.1.autoswitch.charge.low: 0
> > > outlet.1.desc: PowerShare Outlet 1
> > > outlet.1.id: 2
> > > outlet.1.switch: 1
> > > outlet.1.switchable: 1
> > > ups.delay.shutdown: -1
> > > ups.load: 6
> > > ups.mfr: MGE UPS SYSTEMS
> > > ups.serial: unknown
> > > ups.status: OB DISCHRG LB
> > 
> > Hard to say without comparing it to another similar device, but
> > doesn't look unusual to me. -- Peter
> > 
> 	So why isn't UPSLOG not atleast showing the "ups.load"?

I don't know. Anybody? 

> 	I also now have a NEW issue
> 
> 4) When the UPS was put back onto power, and I manually restarted
> the computer, the UPS was charging :
> 
> 20060512 100203 4 NA 0 [OL CHRG LB] NA NA
> 20060512 100703 6 NA 0 [OL CHRG LB] NA NA
> 20060512 101203 9 NA 0 [OL CHRG LB] NA NA
> 20060512 101703 11 NA 0 [OL CHRG LB] NA NA
> 20060512 102203 13 NA 0 [OL CHRG LB] NA NA
> 
> 	and then all of a sudden :
> 
> May 12 10:26:27 soekris upsd[498]: Client upsmon at 127.0.0.1 set FSD on UPS [ellip
> se]
> May 12 10:26:27 soekris upsmon[519]: UPS ellipse at localhost on battery
> May 12 10:26:27 soekris upsmon[519]: Executing automatic power-fail shutdown
> May 12 10:26:27 soekris kernel: May 12 10:26:27 soekris upsmon[519]: Executing a
> utomatic power-fail shutdown
> May 12 10:26:27 soekris upsmon[519]: Auto logout and shutdown proceeding
> 
> 	And it shut it down! It wasn't on battery, I know it.

Looks like your UPS went on battery at 10:26, about 25 minutes after
you restarted your computer. Maybe there was a short power outage, or
you yanked the cable accidentally, or there was a UPS malfunction. In
any case, since NUT saw that your battery had not yet reached 30
percent, it shut down immediately, which is what it is supposed to do
in this situation.

-- Peter