[Nut-upsuser] I/O errors with usbhid-ups and Tripp Lite SMART1500LCDT

Fri Oct 23 12:42:55 UTC 2015

[please use reply-all to include the list - the NUT lists do not add a reply-to header.]

On Oct 23, 2015, at 2:03 AM, Null <null at wuffleton.com> wrote:
> 
> Hi List,
> 
> I recently picked up a Tripp Lite SMART1500LCDT while they were on sale to pair with my home server, and have tried getting it set up with NUT, since it seems to be supported. I can get the whole NUT stack up and running, but the driver seems to only stay up for a couple minutes before crashing out.
> 
> I've uploaded the level 6 debug output of the driver (with my serial number removed) here: https://wuffleton.com:10101/paste/VKP5olfp#IaqkLAyzbMDnrervAofVxqggDv3KP2B-CoUT1w+486a

Here's the relevant portion:

38.788121     ==================================================
38.788129     = device has been disconnected, try to reconnect = 
38.788137     ==================================================
38.788203     Checking device (1D6B/0002) (002/001)
38.788239     Failed to open device, skipping. (Permission denied)
38.788249     Checking device (1D6B/0001) (008/001)
38.788267     Failed to open device, skipping. (Permission denied)
38.788277     Checking device (09AE/3016) (007/003)
38.823794     - VendorID: 09ae
38.823813     - ProductID: 3016
38.823821     - Manufacturer: Tripp Lite
38.823829     - Product: TRIPP LITE UPS
38.823837     - Serial Number: <snip>
38.823845     - Bus: 007
38.823853     - Device release number: 0002
38.823860     Trying to match device
38.823882     Device matches
38.823898     failed to claim USB device: Device or resource busy
38.823916     failed to detach kernel driver from USB device: No such file or directory
38.823927     failed to claim USB device: Device or resource busy
38.823938     failed to detach kernel driver from USB device: No such file or directory
38.823949     failed to claim USB device: Device or resource busy
38.823960     failed to detach kernel driver from USB device: No such file or directory
38.823971     failed to claim USB device: Device or resource busy
38.823982     failed to detach kernel driver from USB device: No such file or directory
38.823991     Can't claim USB device [09ae:3016]: No such file or directory
38.824002     upsdrv_cleanup...

(Those messages are printed at debug level 2, so that's probably all you need for future logs.)

I am curious to see if there is a "sweet spot" in terms of when the driver reconnects (not too early vs. not too late). At one point, I think we had a report of an UPS that needed to see a USB connection within a certain amount of time, or it wouldn't connect. This was in the context of boot times.

> From what I can gather - it runs into an I/O error, and when it tries to reconnect it's unable to claim the device because it's still busy. Are there any settings I can set in ups.conf to workaround this or any places I should start looking for issues in my configuration? This doesn't seem to be a permissions thing, since I'm seeing the same issue if I run usbhid-ups as root. I've also explicitly given the 'ups' user that I'm running NUT as permissions to the UPS device via udev.

For reconnections at startup, we have the following options:

   http://www.networkupstools.org/docs/man/ups.conf.html#_global_directives

We might want to consider applying a similar delay to the auto-reconnect code in the drivers.

Does `dmesg` show any USB messages around the time of the "No such file or directory" errors?

This seems like a similar set of symptoms:

   http://article.gmane.org/gmane.comp.monitoring.nut.user/8662

> Environment for reference:
> NUT Version: v2.7.3.r128.g96c93fb (GIT); v2.7.3 stable exhibits the same behavior
> OS: Arch Linux x86_64
> Kernel Version:  4.2.4.201510222059-1-grsec

So the interesting part is that I was testing the master branch yesterday (same commit) on an older Debian box (3.16 kernel). The driver had trouble starting, but once it got going, it has been running continuously.

I'm not really sold on this approach (due to the testing needed on other hardware), but this pull request reworks the USB communication so that the interface is only claimed while it is being polled:

   https://github.com/networkupstools/nut/pull/122

Even though Github is reporting a potential merge conflict, it should be possible to check out that branch, and see if the driver is more stable in that version.

If nothing else, there are some value scaling fixes for this model in that pull request. The HID descriptor in this UPS is a bit of a mess. (6kHz Hz power would be... interesting.)

> Would really love to get this working, and any help would be greatly appreciated!
> 
> _______________________________________________
> 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