I’m using Home Assistant 82.1 and this is how I have successfully removed a dead node using the GUI
Restart Home Assistant and verify the Dead Node has been successfully removed
Solved it!
Neither restarting HA nor rebooting the server helped. I had to power down the box and unplug the USB zwave stick (it was still flashing when the server was shut down!). The dead zwave device was gone after HA was up and running again.
1 Like
Hello,
I have tried most of the solutions, put the node to dead and failed state but all solutions ends with “remove dead node” button. Every time I press it the log reads:
2019-01-13 11:38:32.299 Warning, contrlr, WARNING: Received reply to FUNC_ID_ZW_REMOVE_FAILED_NODE_ID - Remove process busy
2019-01-13 11:38:32.299 Detail, Node139, Expected callbackId was received
2019-01-13 11:38:32.299 Detail, Node139, Message transaction complete
2019-01-13 11:38:32.299 Detail,
2019-01-13 11:38:32.299 Detail, contrlr, Removing current message
2019-01-13 11:38:32.300 Detail, Notification: ControllerCommand - Failed
Any ideas? I have tried restart, reboot etc.
Update: the device re-appeared in Home Assistant even it was not online at all (it was unplugged).
I then used ozwcp (open zwave control panel) to remove it from the zwave controller (even not supported though) and it did not re-appear so far.
Different but related question. I’m just getting started. I successfully added my first node. I also tested that I could remove a node via the config GUI.
It all seemed to work fine as far as Home Assistant is concerned. However, when I added the node back again, it was assigned Node ID 3 in the network. It was originally Node ID 2.
Is node id 2 now dead in the controller?
I’m using Aeotec Zstick Gen 5 running HA 0.100.2 on Hassbian.
Yes, but no.
The controller assigns node id’s in sequential order without re-using any that have already been used. So what you are seeing is normal.
But after you get to 255 nodes (I think that’s the limit) it will start over again using nodes it had skipped previously. At least that’s my understanding since I haven’t gotten to the limit yet.
But I have seen the node skipping behavior and I know that part is correct. 
1 Like
Not dead, but still listed in your zwcfg_.xml file.
To get rid of it so the number doesn’t increase, you would need to find the node in the zwcfg file and delete it, pretty much everything from <Node id=“#” to “Node” just before the next <Node id=“#”, or reboot HA after removing the node before re-adding it, hope this make sense.
Thanks @finity and @Coolie1101 for the replies.
There is no Node id="2" in my zwcfg_*.xml. There are only <node id="n"> tags for the devices HASS shows as active in the GUI and various drop downs. So all looks ok there.
Doing a little more digging, in OZW_Log.txt, a mention of Node002 shows up during a Heal. I don’t really understand what I’m looking at there. It appears that it never sends anything to Node 2, but it prints these lines, which I interpret as having receiving some sort of response that looks like it comes from Node 2? (There are no other references in the log to Node002 or [nN]ode 2.)
2019-10-27 16:38:40.119 Detail, Node002, Expected reply was received
2019-10-27 16:38:40.120 Detail, Node002, Message transaction complete
Also when adding more switches, the inclusion for one of them seemed to fail. Worked on the second try, but skipped Node 8, so the new device should up as Node 9.
I also see similar lines for Node033 which I don’t have as far as I know. (My max node number is currently 10 – I have no idea where Node033 came from – anyone have any clues?)
2019-10-27 16:38:40.192 Detail, Node033, Expected callbackId was received
2019-10-27 16:38:40.192 Detail, Node033, Expected reply was received
2019-10-27 16:38:40.192 Detail, Node033, Message transaction complete
So I’m hoping it is as @finity mentioned – if/when it comes to it, and node numbers wrap around 2, 8, etc. will be available for assignment.
I have been having issues with a Aeotec Smart Switch 6 that constantly goes dead (Dead (Complete) ) in my Z-Wave interface. I have tried resetting it and even got a replacement and even that one goes dead eventually.
I am using an Aeotec ZW090 Z-Stick Gen5 that is probably within 20 feet from where the plug is located. I also have 4 other 5 Z-Wave devices (4 water leak detectors and 1 motion sensor) and none of them give any issues.
It has got to the point of where I am ready to chuck the Smart Switch 6 in the trash and go back to WiFi. Where I am sure it would work there.
If its any use I managed to get dead nodes removed that would not work with changing the xml file, the HA instance I’m using on one of the Pi’s is 0.93.1. Went into configuration settings, chose customise entities at the bottom, chose the dead node, edited the ‘is failed’ field from ‘false’ to ‘true’ and saved this. Went to zwave netwrok options, chose this node, I now had option to remove this failed node. It took a reboot to removed from Zwave nodes in the drop down for the zwave network configuration.
2 Likes
I tried to set the is_failed to true several times, but no success with the Remove Failed Node (error message: FUNC_ID_ZW_REMOVE_FAILED_NODE_ID - Node not found). Instead, I edited the zwcfg_xxxxxx.xml file and deleted the failing node. That worked for me 
.
I can’t find any zwcfg_xxxxxx.xml file in my whole /usr/share/hassio/ dir, I just found ozwcache_0xe5bb791d.xml, and there are the nodes, but I cannot see any is_failed attribute. I tried removing the nodes there, from <Node id=" to </Node>, but the devices are still there, even after restarting the OpenZWave addon or home assistant. I am on v0.114.4, perhaps they are listed in another xml file, but I cannot find them anywhere…
[EDIT] ok, I deleted them from theozwcache_0xe5bb791d.xml and they are still there. If I grep for example “General Thermostat V2”, I can see them in here:
Binary file ./homeassistant/home-assistant_v2.db matches
Binary file ./addons/data/core_zwave/mosquitto.db matches
./addons/data/core_zwave/ozw/config/device_classes.xml: <Specific key="0x06" label="General Thermostat V2" command_classes="0x40,0x43,0x72,0x86" basic="0x40"/>
Binary file ./addons/data/core_mosquitto/mosquitto.db matches
Are you using open zwave? Lots of things are different for open zwave.
I had a dead node that drove me crazy for over a year. Here’s what finally worked for me:
-
Under Configuration, I opened the Integration for Z-Wave and hit the Configure text item. This opens a new page with a bunch of Zwave controls.
-
In the ‘Z-Wave Node Management’ section, I found the drop down list box for Nodes, and found the dead node listed there.
-
Once the dead node was selected in the drop down, a set of text items appeared below, including one for “Remove Failed Node”. I hit this button. I noted on the Z-wave log that the command reportedly failed because the node could not be found, but I persisted anyway.
-
I then restarted Home Assistant using Server Controls/Server management/restart.
-
Once HA has restarted and stabilized, the dead node was gone and has not reappeared.
Good luck! (YMMV)
Thanks Jeff, really appreciated. My Raspi went dead the other day and I am in the process of setting everything up from scratch. Usinf ZWave JS now and removing a dead node works flawlessly now.
I know this is an old series but the question still exists for me… My question is simple
Usinf ZWave JS now and removing a dead node works flawlessly now.
how??? ahhh duuuh. the 'the new gui of the month" had me lost but i found it finallty. The menu is accessed with the the dot menu next to configure under the z-wave item