Can I extend z-wave network to another house

Configuration Z-Wave
1

I’m using latest HA and have several z-wave components in my main house. Now I would like to control a floor thermostat in another house of mine. I don’t have z-wave coverage from my current z-wave controller in this house. There is 30 meters between the houses and I have IP network in the other house.

I tried to search for this situations in the forum but did not find any obvious answers.

I have thought that it would be possible to put a raspberry pi with another z-wave controller in the other house and then use MQTT from my HA to the raspberry pi.
Another nice solution would be to have a IP based “satellite” that acts as a virtual z-wave network from the HA perspective.

Are there anyone who has suggestions for how to solve this situation in a good way?
I gues I’m not the first person who has this kind of problem.

2

A second home? Have you considered setting up a discreet HA install to manage it and then link the HA instances with something like Remote HA available in HACS?

1 Like
3

I suggest this link: Linking two or more Home Assistant instances · Issue #246 · home-assistant/architecture · GitHub
Have you considered MQTT Eventstream or Statestream?

1 Like
4

Thanks! Remote HA sounds very interested. I have not heard about it earlier. I will check.

5

Thanks! I will read in your linked text.
I will also look at eventstream and statestream.

6

You can if you use an ethernet based z-wave hub connected via your router at each site connected to the Internet. Just be careful on your device naming so you can easily identify which devices are home and which are remote.

7

Yes do this, just cut out the MQTT part. If they are all on the same LAN you can just load as 2nd zwavejs integration in HA to connect to the 2nd zwavejs2mqtt running on the Pi.

Example:
image

Just load docker onto the pi in the other house and fire up zwavejs2mqtt, that’s all you need, nothing extra.

2 Likes
8

Got to get me some of that Z-Wave with Special Sauce :rofl: :joy: :rofl: :joy:

1 Like
9

That seems like a simple solution. I have ordered the missing HW and will try it out.

1 Like
10

Now the hardware is in place and I have started up the container running zwavejs2mqtt.
I have activated Home Assistant JS support. In the logs I get
Error: listen EADDRNOTAVAIL: address not available 192.168.44.39:3000
192.168.44.39 is the local ip to my HA docker instance. I think it means that port 3000 is not exposed outside my HA.
I do not understand how to expose port 3000 on my HA. Or, am I on the wrong track?

11

Totally wrong track, you are only to install RaspberryPi OS, then docker and docker-compose, you should only have 1 instance of HA. That being on your main device.

12

Well, I have HA installed on a Home Assistant Blue with my one z-wave stick using z-wake JMS. I have an rpi 2 with a second z-wave stick. On this I have installed docker and the zwavejs2mqtt container. I have configured it so it sees the z-wave stick. Now I want it to connect to my HA. That is the setup where I see the message above.
In what way am I on the wrong track?

13

I suppose you have a rpi, plus a zwave controller, with pi OS and docker and zwavejs2mqtt in container all setup up ready in your another house, where IP network is reachable from the main house, where your HA (192.168.44.39) sits…?

If yes to all of the above, replace 192.168.44.39 with the IP of the rpi in your remote house, in the HA config.

1 Like
14

Apologies, your previous post seemed like you had installed HA on both devices.

Can you post your docker-compose yaml?

EDIT: looks like @k8gg’s on the right track here

15

Yes to all questions.

I really feel stupid since I don not understand what you want me to do on HA config.
The error message I get on the rpi tells me that rpi fails to connect to HA. Or do I miss understand the message.
As I wrote above I have z-wave stick already connected to the HA blue an integrated with the z-wave js integration.
I understood an earlier post as if I can install a second remote z-wave stick. This is what fails for me.

16

You need the IP address of the 2nd Pi

17

I will post when I comes home.

18

This is where I feel stupid. I know the ip for the pi, 192.168.44.44. My problem is that I don’t understand where to put it. Especially after seeing the error in the post above on the rpi zwavejs2mqtt log.

I can guess that I should put the rpi ip in the dialogue when adding a new z-wave js integration. I have tried this and the integration fails. Can’t reach host.

19

When you add a 2nd zwavejs integration in HA you give it ws://192.168.44.44:3000 for the websocket.

20

Just to be clear:
Inside your HA (which is in your main house), find the zwavejs integration. And then in that integration you should already have a link to control your 1st zwave controller on the same HA Blue.

Now, you should be able to setup a 2nd link to the zwavejs2mqtt instance in your remote house. To do so, uncheck the use add-on check box, and:
image
… in here, replace localhost with the IP of your remote rpi, which (if setup correctly) should have port 3000 open.
Similar verbiages are on the document of the integration ( Z-Wave JS - Home Assistant (home-assistant.io)), specifically this paragraph:

image
image699×98 15.4 KB

In case this task of “telling HA go find the 2nd zwavejs2mqtt server in the remote house” fails, I would first check whether the remote IP being reachable from the main house, and then check whether the port 3000 being open.
And then, as firstof9 said, maybe post your docker-compose yaml for zwavejs2mqtt on your remote raspberry pi, along with the config of your remote zwavejs2mqtt.

1 Like