I’ve been running homeassistant / zwave for several years without any issues. About a month ago I upgraded to zwave js. Since the upgrade I have on average one node unavailable each week. I can do a network heal to correct this, but is there a permanent fix??
1 Like
I have the same problem and found your post trying once again to find if there is a fix or if other people are experiencing the same issues. Just now I reloaded the zwave js integration to see if I could figure out why my last fix on an offline device caused it to double the number of sensors where half of them retain the “renamed” name and half got the original z-wave generic name. (adding a picture to clarify)
Sometimes restarting the integration work, and forgetting and re-including the devices works, but that’s time consuming and not a solution
Right now I lost a third device trying to figure out what caused my second to fail. I’ve given up on the first one that went “offline”, as it will just be marked as dead again a short while after being brought alive.
It’s almost like certain devices will be marked as offline of they are turned off when Home Assistant is rebooted.
Had this happen to qubino, fibaro and heatit device.
Pi4 with SSD
Aeon labs z-wave stick.
I’ve run into the same issue with a node being identified as “dead” (I started a new thread before finding yours)
It only seems to be happening with one Leviton device. I’m using Zooz USB controller.
I have one node that goes unavailable periodically and if I ping it in Z-Wave JS it is fine again. As a band aid you could have automation ping the problem device on a schedule to make sure it wakes up if it drops (that is if ping fixes your issue).
Maybe you’re having USB3 interference on your ZWave?
Thanks for the tip
I found the documentation for pinging, and noticed this commentary next to sample code:
This service can be used in tandem with the node status sensor to track the node’s status and fire the command when needed. Here’s an example automation that would ping a node when the node status sensor state has changed to dead and remained dead for 30 minutes. Note that this may be useful for some devices but will definitely not be useful for all. In cases where it is not useful, all you will be doing is generating additional traffic on your Z-Wave network which could slow down communication.
trigger:
platform: state
entity_id:
- sensor.z_wave_thermostat_node_status
- sensor.z_wave_lock_node_status
to: "dead"
for: "00:30:00"
action:
- service: zwave_js.ping
target:
entity_id: ""
Should z-Wave JS be automatically trying to wake up dead devices? Is that what it means by it not being that useful and generating additional traffic?
Is there a time period in which it would NOT be unnecessary and definitely useful to attempt? (e.g. If sensor is ‘dead’ for X minutes no problem pinging to try and wake up)
I am having the exact same issue with a zooz USB controller and Leviton ZW15R. I have three, one is stable, one goes off line (Dead) occasionally (Have to power cycle to bring it back) and one that is showing turned on but is in fact still off. toggling off/on does turn on, but is the second toggle that makes it work. BOTH of the ‘bad’ ZW15R behviour is new, within the last two weeks. worked fine before that. I noticed there is an update to the zooz controller available… anyone have luck with 7.15.2 ?
@MichaelSummers which usb stick do you have? I haven’t had any issues since upgrading the Zooz usb stick firmware and upgrading the HAOS software. (Hopefully you have similar luck)
I have a zooz series zwave plus S2 usb stick.
I have yet to upgrade the firmware … should get a chance this weekend … Thank you for the information on the success of your upgrade!!!
Just want to add that I had this problem all the time until updating the firmware on my Aeotec Z-stick Gen5 to 1.2. Now, 0 problems.