I turned off a breaker for about a day while doing some work on my house. That stopped power to three of my Z-wave devices - a wall switch that’s close to my Zwave radio and two plug-in lamp controllers.
This morning, about a third of my Z-wave network is reporting dead. What’s the best fix?
I’m running Z-wave JS and my radio is a ZWA-2.
I’ve tried:
Grr. I can’t remove the failed nodes to re-add them either!
And it appears that an interview of the node works, but pinging after interview fails, marking the node as dead.
2026-05-01 15:09:26.536 INFO Z-WAVE: [Node 039] Value added: 119-0-name => Noise Maker
2026-05-01 15:09:26.536 INFO Z-WAVE: [Node 039] Value added: 119-0-location =>
2026-05-01 15:09:26.537 CNTRLR [Node 039] Beginning interview - last completed stage: None
2026-05-01 15:09:26.537 CNTRLR [Node 039] new node, doing a full interview...
2026-05-01 15:09:26.538 INFO Z-WAVE: [Node 039] Interview started
2026-05-01 15:09:26.538 CNTRLR » [Node 039] querying protocol info...
2026-05-01 15:09:26.545 CNTRLR « [Node 039] received response for protocol info:
basic device class: Routing End Node
generic device class: Binary Switch
specific device class: Binary Power Switch
node type: End Node
is always listening: true
is frequent listening: false
can route messages: true
supports security: false
supports beaming: true
maximum data rate: 40000 kbps
protocol version: 3
2026-05-01 15:09:26.546 INFO Z-WAVE: [Node 039] Interview stage PROTOCOLINFO completed
2026-05-01 15:09:26.546 CNTRLR [Node 039] Interview stage completed: ProtocolInfo
2026-05-01 15:09:26.547 CNTRLR » [Node 039] pinging the node...
2026-05-01 15:09:30.211 CNTRLR [Node 039] The node did not respond after 1 attempts, it is presumed dead
2026-05-01 15:09:30.211 CNTRLR [Node 039] The node is dead.
2026-05-01 15:09:30.211 INFO Z-WAVE: [Node 039] Is dead
2026-05-01 15:09:30.212 CNTRLR [Node 039] ping failed: The node did not acknowledge the command (ZW0204)
2026-05-01 15:09:30.212 CNTRLR [Node 039] Interview attempt (1/5) failed, node is dead.
Could this be a Zwave-JS issue?
2026-05-01 15:22:51.001 INFO Z-WAVE: Calling api removeFailedNode with args: [ 39, [length]: 1 ]
2026-05-01 15:22:51.003 CNTRLR » [Node 039] pinging the node...
2026-05-01 15:22:54.660 CNTRLR [Node 039] The node did not respond after 1 attempts, it is presumed dead
2026-05-01 15:22:54.661 CNTRLR [Node 039] ping failed: The node did not acknowledge the command (ZW0204)
2026-05-01 15:23:08.725 DRIVER Dropping message with invalid payload
2026-05-01 15:23:19.896 INFO Z-WAVE: The removal process could not be completed (ZW0360) removeFailedNode undefined
2026-05-01 15:23:25.114 INFO Z-WAVE: [Node 093] Value updated: 50-0-value-66049 9.5 => 9.5
If anybody has advice about either fixing this or how I can remove and re-add the nodes it would be much appreciated.
Are you sure the devices didn’t fry? I have ~70 zwave switches and breakers tripping have killed 4 of them over the past 10 years. Sometimes the dead ones physically work at the switch, sometimes they don’t. But the zwave portion is always dead.
The one that was disabled via the breaker is working fine now that power has returned. The dead ones are on breakers that didn’t cycle.
if they don’t have power they will show up as dead and you need to cycle the breaker to get power back.
Otherwise, my point above remains… Are you sure the devices themselves have not fried?
The dead devices never lost power. For the plug-in items, I have tried unplugging them. At some point overnight, about 20 devices stopped working. I guess there could’ve been a whole house surge, but the only things affected are my Zwave devices.
I assume they still operate manually with a button on the device.
Odd that you have a number of devices that have failed.
I think I’ve only had one node that failed similar to this. IIRC, I did a hard-reset on the device and then included it on a spare controller just to see that it worked. I then excluded it and then re-included it on my main controller. Not sure which step there made a difference (if any), but having the spare controller can help to isolate where the issue is.
Zwave doesn’t work great with partial power interruptions as it causes routes that have been working and reliable to fail which causes the use of less optimal routes. Assuming those devices are working, likely the were using the nodes that were turned off as their route and have no backup route and have given up on using that route since it’s been dead too long.
I don’t like excluding / including except as a last resort.
a) shutdown the HA host, remove power from it and the zwave stick, start it all back up.
b) any zwave devices that are dead should be power cycled
c) rebuild routes one by one for the powered nodes
d) do b again
In my system my host is on a UPS and I have a backup generator. When backup power come online all of the line powered devices restart but diffent devices may take longer. So there’s alot of broken routes and I see alot of timeouts for the next couple of days and then rebuild routes manually for the devices with the most timeouts.
If those dead devices can be moved closer to the stick I’d try that as they should always have a route to the stick. If that works then rebuild routes and move them back.
Ugh. Looks like I was chasing the wrong issue. The reason I had a breaker turned off was because I was having my screened-in porch repainted. I threw the breaker so that I could remove the ceiling fan.
Before painting, the painters power-washed. They clobbered an old Z-wave motion sensor that I forgot to remove from the porch. I found it this morning. I popped the battery out of it and removed the node.
After doing that, I was able to manually rebuild routes to my dead nodes.
Looks like that motion sensor was jamming some of my devices. Mostly older devices (lower node number), incidentally.
Thanks for the advice!