Z-wave battery nodes won't initialize

After upgrading openHAB 1.6.1 to 1.7.1 and resetting persistence- and z-wave files some of my battery powered nodes won’t initialize.
I left openHAB running for 2 days without any luck. Manually waking up a node a couple of times didn’t help either.
I then updated the z-wave binding to a cloudbees version (2015-11-02), still no luck.
The devices do work (motion is detected, contacts update their state), so the nodes are still linked to the controller.

I have a relatively small log file logging a few minutes of z-wave activity (including a couple of manual wakeup’s) on DEBUG level.
Once I figure out how to attach files on this new system I’ll add it to this post… :sweat:

Here’s a temp upload for now: http://we.tl/1qjByjSHJG

I would guess that the wakeup is not configured for the devices - I see no wakeup messages and this is quite indicative of wakeup not being configured (ie the sensor information works, but configuration doesn’t complete).

What nodes are battery nodes in your log (so I know what I’m looking at)?

All my battery nodes have a wakeup period of 6 hours, unless this can be
reset by a z-wave binding update this should still be the case.
During the the debug log period I did at least 3 manual wakeups on node 31
(fibaro motion sensor, triple clicking the button inside).
All nodes > 14 are battery powered nodes with the exception of node 23.

Meanwhile I downgraded back to my 1.6.1 install (with the old zwave XML
files) and kept it running for ~ 24 hours.
All nodes are now initialized, all green except for 3 yellow and 2 red

I’m now doing a network heal, hoping it’s a routing issue (but i doubt it,
I don’t live in a big mansion or anything).
I’ll give it another 24 hours to complete and then switch back to 1.7.1
(with 1.8 z-wave binding).
Do you recommend copying the zwave XML files to the new install?

I didn’t notice earlier, but your log is only very short time period, so it’s impossible to tell what’s happening. There’s no wakeup, but then with a short log it is normal - really I’d need to see a longer log where the device actually wakes up to see what’s happening.

The devices at the moment are stuck in PING stage as they have received no wakeup messages…

You cant copy the XML files over as they are slightly different - I suspect that you are doing this as I see errors in the log. Can you post the XML files somewhere?

Right, so like I said, I reverted back to my previous stable install.
After the initial network heal I changed the wakeup period for all my battery operated devices from 6 hours to 2 hours.
After all the nodes had picked up on this change I did 2 more network heals (over a longer period of time) to be on the safe side.

I just deleted my /logs/* and /etc/zwave/* from my 1.7.1 install.
It’s starting up now, will give it a day and see what happens before I take up more of your time Chris.

My server needed another reboot, but since then all has been fine and dandy.
Sorry for wasting everyone’s time…

I think a reboot fixed all my issues.
After my last switch from 1.6.1 to 1.7.1 even the mains powered devices wouldn’t initialize.
I then did a reboot of the entire server, and at the time of posting all my nodes have come online like they should, even the battery operated ones.

Once again, sorry for the trouble and thanks for your time.