[edit: I just re-read and saw you requested debug log, I’ll do that.]
Below are the log entries for the offending device (node 27) from when I added it last night. The method I used was to shut down the server, take the Z-stick to the light switch and do the battery-powered inclusion. I then returned the stick to the server and powered on.
2018-07-14 20:17:40.977 [INFO ] [age.SerialApiGetInitDataMessageClass] - NODE 27: Node found
2018-07-14 20:18:04.665 [WARN ] [tocol.commandclass.ZWaveCommandClass] - NODE 27: Unknown command class 0x91
2018-07-14 20:18:04.873 [WARN ] [wave.discovery.ZWaveDiscoveryService] - NODE 27: Device discovery could not resolve to a thingType! 001D:0501:0209::0.4
2018-07-14 20:18:04.880 [INFO ] [g.discovery.internal.PersistentInbox] - Added new thing 'zwave:device:1648c66469f:node27' to inbox.
2018-07-14 20:18:04.903 [WARN ] [l.serialmessage.SendDataMessageClass] - NODE 27: Already processed another send data request for this callback
Please let me know if I should re-try with a more verbose log setting or do the inclusion differently.
FYI: I know the device is functional, but it is older. I have a working Z-wave network I’m migrating from being controlled by a HAI Omni through serial to being controlled by OpenHAB. 17 nodes so far have been migrated successfully.
Here is a link to the log with debugging enabled for zwave binding: http://little.org/openhab/openhab_log.txt. There’s a lot of content so I hesitate to paste into thread (but will if you prefer). Let me know if there’s specific content to paste in, I’m keen to help and learn so when I add other devices I can be more efficient.
Before enabling debugging I cleared the log (log:clear) removed the device, set logging to debug and re-added (as node 29).
[edit: my server doesn’t have a .log file handler, changing extension to txt]
I’ve made a change to the database which I hope will allow the completion of the initialisation and the production of the XML. You won’t be able to upload the XML now, so please report back if it works (or not!).
This is included in the latest development version already, but you probably aren’t using this?
Thanks Chris, no, I’m not. I’m on OpenHABian because my other servers are all VMs (and Windows Hyper-V doesn’t allow USB to be attached to a VM). I may dig out an old PC and move to that… but in the mean time I’ll need to wait for the change to propagate.
If/when I get my setup moved to hardware that supports the dev version I’ll report back.
No need for that.
You just need to uninstall your current zwave binding, download the jar from this post, put it in your addons folder and restart openHAB.
Probably you need to install the serial transport dependency manually, so read the first couple of posts from the link on how to do that.
I uninstalled the old binding and added the dev binding (and I did need to feature:install openhab-transport-serial, thanks for the advance warning).
The errant device (node 29) is now recognized as the right Leviton device, but still doesn’t generate XML. In order to get good debugging info do I need to exclude and re-add the device to get richest log info?
[edit: patience may be required on my part… I restarted at 15:36 pacific time and as of 18:08 it appears OpenHAB is still interrogating the switch, iterating through classes. I’ve removed the log link below and will re-add once it looks like OpenHAB is through with the switch]
Below is a link to a new log file. I shutdown OpenHAB (via karaf), [waited], restarted OpenHAB (by entering openhab-cli console on command line), then turned on z-wave debugging as soon as I had a karaf prompt. Seems like that gets in well before z-wave gets to node 29 (it took about 6 minutes to get to node 29).
[removed incomplete log file link]
note: I trimmed all events from before the restart. I also have full versions of the log files if needed.
It’s easy enough for me to exclude and re-include the node if that would be better.
Success! It took a little over 2 hours, but the z-wave binding dutifully cranked out an XML file for node 29. I’m not sure if that length of time is typical, but between running on a Raspi2 and possibly having the binding not flood the z-wave network, I have no complaints. Thanks again @sihui and @chris for the help.
I’ll upload the XML to the device database tonight.