[SOLVED] Z-Wave association groups, best practice for lifeline, other groups?

Hi all, I’m seeking some advice from someone wiser than me.

I’ve noticed recently that some of my devices aren’t updating their status when changed manually, and believe I’ve pinpointed it to the reported issue where they lose their lifeline. Restoring the lifeline to “Controller” in Paper UI seems to fix the issue, at least for now. Now I need to do this for all the misbehaving devices in my network, but before I do, would like to improve my understanding of how things should be.

I have a pretty robust mesh, with a bunch of devices. Geometrically, the devices are stretched out from the hub, spanning a decent distance. It’s dispersed widely enough that there are peripheral devices that don’t have a direct connection to the controller (according to the network viewer in HABmin).

Here are my questions:

  1. I this situation, do I still set all of my lifelines to “Controller”, even if it takes multiple hops to get back to it? In particular, I’m wondering about the devices on the far outside edge of the mesh.

  2. If I’m only using OpenHAB rules to control devices, should I leave all of my other groups blank? For example, with a GE 14294, “Basic group” and “Scene group”.

  3. Finally, have we determined what is causing this to happen? Are there known behaviors to avoid?

Thank you!

Yes. It’s the controller your device needs to talk to. Transmission across multiple hops is handled on the lower ZWave layers. Like a TCP endpoint you talk to over multiple IP hops.

Depends on whether you want to make use of the functionality they provide. Don’t know your device and needs so can’t tell.

I am not aware if developers (@chris) are aware of it. Have you raised a Github issue ?

Then again that’s difficult to pinpoint right.
PaperUI changes all ZWave parameter in one rush, habmin is more selective. Probably a good start to use the latter, but clearly not an analysis when this is happening and why.

1 Like

This is great, thank you @mstormi. Answers to questions like this are one of many reasons I appreciate this project and its community.

I believe @chris is aware, based upon the mega-thread from the original post. The most likely culprit at the time was a UI issue where the lifeline was not populating correctly in HABmin, resulting in incomplete config data being pushed to the device. This would wipe out the configured lifeline. One fix, proposed by @chris, appears to be “Update the binding so we don’t push any changes to the device that result in an empty lifeline.”

I’ve been using HABmin to push config changes, so it could very easily be the root cause of my issue, but I wanted to ask just in case any other causes or fixes have emerged.

I’m not sure from this post what version you are using? If you use a recent version, then it should not be possible to remove the lifeline associations - the binding should always ensure that they are configured.

Hi @chris,

I’m on the 2.4.0 release build.

If that’s the case, then maybe there’s something else going on. I’ll fix them, see if I lose my lifelines again, and then follow up.

This version doesn’t have this feature - it was introduced in 2.5 so is in the latest milestone release.

1 Like

That’s helpful to know. Thanks!