[Nut-upsuser] Liebert PSA UPS problem.

[Charles Lepple]

On Sep 10, 2013, at 9:30 AM, Paul Smith wrote:

> Hi all,
>   Sorry for the delay - three business trips abroad and LOTS of head
> scratching with USB Keys and DVD's of SuSE and FreeBSD...
> 
> The question about the USB reports from Windows is still unanswered (I am
> abroad again !), but *think* the Liebert driver had been stopped.

I'll be honest, I don't really want to speculate on all of the possible things that might be going on in the background in Windows.

A potentially useful trace (and I can't guarantee it will prove anything) would be to disable the Liebert driver, shut everything down (including powering off the UPS, and unplugging it from the wall for a few seconds to clear any potential saved state), and trace what Windows does at boot time. Then we would compare that to what the driver does (probably one of the older traces). My concern is that there is something being sent to the UPS in Windows, but not in FreeBSD.

Given that the virtual environments are not yielding correct results, I am not sure exactly how we compare these traces to FreeBSD.

> I can report the following updates for my linux setups; SuSE/FreeBSD/FreeNAS
> [remember here I am trying to make nut work for a physical machine running
> FreeNAS]:
> 
> * SuSE Live CD will successfully obtain descriptors. Hurrah! 
> * I am not sure how being able to build NUT for a SuSE environment may help,
> but I think I am in a situation where I can do it. - I now have a 'Live' USB
> stick and a partition on it that is persistent. This also reads the
> descriptors.

I think this was to make sure that we didn't break anything after the 2.6.5 release, but also to possibly add in any extra USB code to emulate what Windows is doing.

> * my NAS box has now been updated to FreeNAS 9.1 ... but there is no change
> to UPS behaviour - still can't obtain descriptors.
> * I now also have FreeBSD 9.1 (not FreeNAS, but the 'base' of it) running as
> a VM, not as a 'Live' install - this includes nut sources which I believe I
> can build for my NAS box.
> * Both implementations of nut : FreeBSD 9.1 in a VM environment and FreeNAS
> 9.1 as a physical standalone machine behave in the same way...i.e. can't
> obtain the descriptors properly, even though the same SuSE version of nut
> can.

Unfortunately, this doesn't narrow things down much. I am at a loss here.

> * I am now stuck again - my FreeBSD "ports" collection for nut does not seem
> to be the latest source - it reports "Network UPS Tools - Generic HID driver
> 0.37 (2.6.4)" ... my NAS box and SuSE implementations both report 2.6.5

These page indicate that the ports tree is now at 2.6.5, since November 2012:

http://www.freshports.org/sysutils/nut/

http://svnweb.freebsd.org/ports/head/sysutils/nut/

There are several tools for updating the ports tree, after which you would need to rebuild the port.

> * Finally my FreeBSD VM output is only as follows - I am not sure why it is
> not finding the UPS:
>       0.000000	debug level is '5'
>       0.024036	upsdrv_initups...
>       0.025984	No appropriate HID device found
>       0.026251	No matching HID UPS found
> Network UPS Tools - Generic HID driver 0.37 (2.6.4)
> USB communication driver 0.31
> 
> I was hoping to be able to report back that my FreeBSD implementation was
> working in the same way as my FreeNAS, but this seems to have a more
> fundamental reason for not wanting to play.


I really need to collect all of our *BSD/USB wisdom in one spot, but off the top of my head, I think this is a permissions issue. If NUT (libusb, really) doesn't have permission to scan the USB bus (which I think requires write permission to *several* /dev nodes), it won't report any USB devices at all.

This is, of course, assuming the USB device for the UPS has been passed through to the VM guest.

You can get a list of the basic /dev nodes with 'usbconfig' run as root.

More info:

http://lists.alioth.debian.org/pipermail/nut-upsuser/2006-May/001111.html

-- 
Charles Lepple
clepple at gmail