It looks like it is partly working to me. I can see you sending commands (strangely, a lot of the same command!) and the binding is sending these out and getting responses -:
The first thing to note here is that you send 5 commands to set node 69 to 1!
However, a bunch of commands are sent here and they are all acknowledged by the devices, so this works. However I certainly see other areas in the log where there are problems, and the reason I’ve picked on this part of the log is it has an interesting feature…
In the above, we generally see the response times getting slower and slower. 72, 93, 136, 116, 293 ms. After this the responses are all delayed, so something strange is happening. Unfortunately it’s hard to know what might be happening on the network - maybe there is a node that is misbehaving and this is congesting the network - I can’t say really as there’s insufficient information to know what could be happening at the network level - the only way to see that is with a sniffer
The 5 commands to set node 69 to 1 is likely a poorly constructed rule combination. I will fix that. The system is basically not working. So, it’s not a serial problem.
I also get karaf message:
Failed uninstalling ‘openhab-ui-dashboard’: Feature named ‘openhab-ui-dashboard/0’ is not installed.
It’s not anything to do with the restdocs being moved. I could tear the whole zwave network down and rebuild it, but Ive done this 3 or so times. Should I do it again? If this some node polluting the system, is there some smoking gun I could look for? Say I added 5 nodes. Checked. Add 5 more…
My girlfriend is starting to have her doubts. LOL!
That dashboard is only used when initially setting up OH and is then uninstalled. That is a condition that should be detected and the nuisance error killed.Unfortunately, that is expected due to lazy coding IMO.
So. I think what you are saying is that the hab system (minus my bad rule) and the zwave binding are fine. The problem is in the zwave network itself. If that’s the case, then I guess I need to see inside of that network or find some way to using Hab and the bindings, seek and destroy the offending node(s).
Is there a definitive guide on the habmin network viewer. The habmin document does not elaborate on the syntax of the colors, arrows, etc. Posts seem to give opinions. Did the author write something on this?
Well, until you told me you’d done it three times already, this would have been my recommendation. To me it looks like something on the networks is upset - the increasing delay for responses kind of indicates that, although there could be other explanations for it as well (eg the nodes being more hops away - but I can’t tell that).
If you’ve completely reset everything, including the controller, 3 times, and the problem is the same, then it will likely be the same again. How many devices do you have? The nodes I see are around the 69 mark, but I also see node 9 and node 91, so do you have around 100 devices in your network? Or is this indicative that you’ve excluded and re-included devices and now could have ghost nodes?
There are a few ghost nodes, but I have about 100. Mostly switches (Homeseer and GE)
Have some Aeotek light/motions (120v) and some aeoteck motion/humidty/temp (Battery)
4 of the Aeotek strip lights and a high power switch (for basic sound amps) So… Lots of stuff.
I have perhaps 10 switches I havent added yet because I cant control the ones I have. When I did rebuild the system, it worked and then it got slow to unusable over some time. Cant give you any precision here.
Have alot of items, about 1800 lines of rules.
The other bindings (russound, openweathermap, astro all work great) Telsa binding seemed flaky but I am spending too much time on zwave to get to the Tesla thing.
Maybe pair down the rules to some very simple macro like stuff (lighting scenes only perhaps) and add 20 nodes at a time until it gets flacky?
Ghost nodes can cause routing issues on the network. Devices sometimes try routing through the non-existent nodes.
With HABmin, have you tried discovering the ghost nodes, marking as failed, and then removing the ghost device from the controller? Delete from OH and see if it gets rediscovered.
It is supposed to work, but people have had mixed results.
Sometimes I have put my stick into a Windows PC and used other software to delete ghost nodes from the controller.
Wes, it feels like you have a flakey node. One bad node can ruin the party. And rebuilding 3 times has almost surely left a few ghost nodes further complicating things. Look for battery powered devices with low batteries. Don’t worry about the network map in habmin, it tells you very little. You can use habmin to exclude bad nodes as Bruce suggested. If you have a Windows pc around you can load free zwave PC controller software. I’m on phone right now somebody provide link or search
Burning it all down again. Will rebuild in small bites. Start with basic main level lighting. Will add stuff until it starts flaking out. Thanks everyone. Will know more as week progresses.
Yes. The devices are configured for the old network, similar to a WiFi router being hard reset and the SSID changing. The device also needs to be configured or reset to get to the new network.