I’m new to Home Assistant and came over from HomeSeer which supports multiple z-wave networks.
I’ve installed the add-on Z-Wave JS UI and was able to connect it to one of my z-wave controllers using tcp (it’s a HomeSeer Z-Net controller), and everything works great.
I think I need to install another instance of the Z-Wave JS UI so that I can specify the network info and address for the second controller. When I try to do that, I’m simply taken to the config page of the existing instance.
I’ve searched the net and found references to multiple instances of Z-Wave JS but nothing that mentions the newer Z-Wave JS UI. Is this possible?
You can’t do it in HAOS, it doesn’t allow multiple instances of the same add-on. You’d have to install the second instance on a remote host, using Docker or bare-metal. You can also use the Z-Wave JS add-on (not UI) on the same host, for a total of two networks.
Thanks for the info and clarification.
Since I was already running HA under ProxMox, I was able to take advantage of the Z-Wave JS UI LXC
Proxmox helper script at https://tteck.github.io/Proxmox/.
This got me another instance that I could use with the other Z-Wave network I have. I was then able to add a second instance of the Z-Wave integration in HA and point it at the newly created LXC container for automatic syncing of the Z-Wave devices.
So far so good! Thanks again for pointing me in the right direction.
Trying to wrap my head around this arch, new to me.
So you are saying SSH into the ZNETs (never tried that I wonder about auth) install a Zwave JS UI and a MQTT instance on them, then add a new hub in zave integration pointing to znet IP?
Then add a new Iframe dashboard to the UI on each ZNET in HA?
I like how organized that is. If not some work to keep updated.
The gets me thinking about the auth for access, I prob should have worried about years ago too. I just keep znets on a subnet with no eternal routes so I was not careing as much.
No need to worry about auth. Just replace the SD card with a new one, install the latest RPi server OS. That way you can fall back to the “old” Z-net if desired. Once you have Z-wave-JS UI installed you would point to the Z-wave controller by device-id, no need to use the IP address (same device). Or you can clone the SD card and start from there.
Personally, I just save the Z-wave-JS UI web login in my password manager, but you could pin them to the sidebar on HA if that’s more convenient for you.
Very interesting idea to just reuse the hardware with a clean OS. Thank you.
I’m new on this, not grasping the sack. Are you saying Zwave JS/ UI on each RPi OS with a local MQTT, and set up in HA as separate MQ subs? SO I dont even need Zwave integration on HA? (I love the easy access of Zwave functions from the device page, If I just make each a MQTT device I assume I will lose all that extenability?)
Also first time learning MQTT, I have no other MQTT devices yet to test out.
Not exactly but now I know what I need to learn thank you.
I have only explored zwavejs2mqtt so far.
Question, the reason I have four ZNETs is because around 80 devices they start to get crushed, with delays and failures. I am not sure if that as because of the RPi hardware or the Zwave controller being overloaded. If I add UI and management services to the hardware do you think that can cause more load issues?
Yeah, it’s confusing. Home Assistant has a long history of confusing names.
zwavejs2mqtt is the original name of Z-wave-JS UI. Although it supports MQTT, it also supports Websockets, which is the recommended way to connect it to Home Assistant. You can disable MQTT.
So, when you see things refer to zwave2mqtt, it really means Z-wave-JS UI (today’s name).