[Nut-upsuser] 2 different APC UPS using USB get in power off recoverable state

[Charles Lepple]

On Jan 3, 2012, at 8:03 AM, Bendtsen, Jon wrote:

> Hi
> 
> Yesterday I had 2 different APC UPS using USB get in power off recoverable state.
> 
> dmesg said:
> 
[...]
> [7706011.117044] usb 2-1: new low speed USB device using uhci_hcd and address 9
> [7706011.529059] usb 2-1: device not accepting address 9, error -71
> [7706011.529104] hub 2-0:1.0: unable to enumerate USB device on port 1

Unfortunately, there is nothing that NUT will be able to do to recover from that state, since the kernel has not fully recognized the USB device at that point.

It is possible that we are sending an unexpected command, though. Did this happen as a result of a power-down command from NUT, or did they just drop off the bus on their own? If it was a commanded power-down, do you have anything in your shutdown scripts other than the normally suggested NUT commands (i.e. upsdrvctl shutdown, or calling the driver with "-k")?

> The machine is a guruplug server plus, running debian lenny on armtel. Linux dkplugbab9 2.6.32-00007-g56678ec #1 PREEMPT Mon Feb 8 03:49:55 PST 2010 armv5tel GNU/Linux
> 
> I have hand installed Network UPS Tools upsstats 2.4.3 from source.
> 
> ups.conf:
> [apc1500]
>  driver = usbhid-ups
>  port = auto
>  serial=AS0535132782
>  # powers switches and routers
> 
> [Smart-UPS_3000]
>  driver = usbhid-ups
>  port = auto
>  serial=JS1030003421
> 
> [R3000XR]
>  port = /dev/ttyUSB0
>  driver = bcmxcp

Interesting that the USB-to-serial unit (R3000XR) was not affected. But I imagine that the units monitored by usbhid-ups are fairly different inside.

> The R3000XR did work, but neither apc1500 nor Smart-UPS_3000. I tried restarting nut, I tried in other linux servers (just plugging in and seeing dmesg).

Did you happen to try on any non-ARM servers? (I'm generally suspicious of the USB controllers on ARM hardware, but I have nothing substantial to back that up.)

> Nothing helped. So I scheduled downtime and powered off the UPses. That did help, and since that nut has been able to monitor the UPSes.
> 
> 
> 
> JonB
> 
> _______________________________________________
> Nut-upsuser mailing list
> Nut-upsuser at lists.alioth.debian.org
> http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/nut-upsuser

-- 
Charles Lepple
clepple at gmail