with some battery devices I got sometimes the problem they loose connection to the stick, so they went in inclusion mode again. Some others, like sensative stripes has to exchange against new if battery is empty.
Two things I had seen and have questions to the profis:
If I mark those ghost devices as failed, that doesn’t work with openHAB (already saw in OH 2.5)
If I do with zensys tool or zwave controller, it works. (Stick is aeotec 5)
with zwave pc controller I’m able to use same node to reinclude if earlier set as failed. So Node will be the Same. Very convenient and no free nodes in row (didn’t know what will happen if 253 is reached, but lower like 5-20 are free again)
So 1) is a bug? Or do I make something wrong?
would be a great feature, but maybe I cannot use, because failed devices never shown as failed in OH (they still come to inbox and always shown as online)
Maybe someone can help or explain if I make some mistakes
Most problems with danfoss LC-13 (have had between 15 and 35 routing neighbors always on)
10 of them, 1 running since beginning, 1 stuck about all 3 month, 2 others if battery fully empty, 1 today stucked with full batteries, had been included ca. 2019, one of the oldest.
But when such dead nodes I got traffic problems, so until not set to failed always need to stop OH, take usb to win pc to mark nodes as failed (and then delete) before plug to server and restart OH3
Well, it’s shown as offline. And due to a failure in device I think: device lost connection and is reseted on its own. I can directly reinclude as new device.
If it’s shown offline and I use the function „set device as failed“ in openhab: it isn’t set to failed and went to online again for some time, then to offline again.so this doesn’t work, but looks like implemented
I have some devices marked with zwave pc controller as failed, those are in inbox of openHAB and if added: shown online for some time instead of „failed“ then offline.
So maybe the status „set as failed“ missed in openHAB?
The other point to use node of failed device for new inclusion would be a feature request, but only makes sense if the function: set device as failed would work.
Is this just one device? Your first message seems to indicate that this is happening with multiple devices, and it seems very strange it multiple devices are failing. I would suspect there may be another reason.
It only works under some circumstances. So it’s not missing, but it not reliable. You can try to use it if you like.
The binding doesn’t support this and it’s not something I’m likely to add. In the 9 years I’ve been writing the binding, this has never been needed.
Sure, but these things have a 10 year battery life - right? So this isn’t very common thing to do, and given these devices have only been on the market around 5 years (from memory) this shouldn’t really be needed. You can always include again if needed and if it uses a different node id this really shouldn’t make a lot of difference - you just need to change the node id in the thing definition once every 10 years or so.
If you get to node 252, then it will use unused nodes again - this is not a problem.
Ok, fair point. But even so, if you need to do this after 2, 3, 4… years, that’s still not very often. I thought from your original message that you had to rejoin devices quite often, so at least that’s not the case .
No - you can’t change the ID - this comes from the controller. What I meant is that if you already have your thing definition with a certain node, then you just need to change the node id to the new node within your thing definition in the OH config.
in real I had to change first after 2 years, the others show 10-25% after 3 years.
.