Petro, now youre here. I am planning on running a 2nd zwave network zwavjs (or js2mqtt) on a separate stick next to ozw. That will not bite will it?
Currently testing on my test system and switching between js and j2mqtt is painless. Although I do not know why I would ever switch back to js as I cannot configure any parameter there…
I currently have two operating zwave and zigbee sticks less than 6 inches apart and have never had any obvious interference between the two of either protocol.
I think you missed that I have two of each protocol sticks (three sticks total - one zwave, one zigbee, and one zwave/zigbee combo) and all running within 6 inches of each other with no obvious interference.
After some recent ZWave issues, I moved from OZW to ZwaveJS and then to ZwaveJSMQTT pretty quick after that. I was please to see that ZwaveJSMQTT found many of my items that regular JS was having trouble with. I’m having a little trouble with a couple of my battery Zwave items retrieving all the sensors and payload data right. When on OZW for these items [Everspring ST812 flood sensor, Kwikset 910 door locks & Nortek NGD00Z-4 garage door openers] to get them fully populating in my old smartthings and for the past couple years in HA I would put Zwave in inclusion and wake up the device. Often this would be repeatedly until finally all the features appeared.
With ZwaveJSMQTT I read a few times that I just need to wake it up and it will poll the device or I could re-interview through the control panel. I’ve tried both a couple times and they just don’t seem to be getting any new information. For the flood sensors for example I can refresh the battery levels fine every time the device wakes up but the most important sensor the “leak_state” isn’t appearing. I tried with one of these flood sensors to update it by putting the network into inclusion mode but it just refresh the same sensors again.
Should I need to do any of these extra steps or is repeatedly waking the device and letting ZwaveJSMQTT manage it from there the best route? All in all though I really like this version as it seems to offer a good mix of options not yet seen on the regular JS version.
when referring to some componant management that appeared to give some good information under the Home Assistant tab. But my control panel lacks in this department as I don’t even have that tab. Is this from a different version or did I set it up differently which caused this. I’ve been going into the integration to get a lot of this detail.
If you’re not running hass.io and don’t have the community add ons, is the only other way to go to run the docker container? If so, is there anything I need to install in my HA instance? Just the MQTT integration and point it at the MQTT broker running in ZwaveJS2MQTT?
or you can go the hard way and install the zwavejs server directly on the host. But who wants to do that when the container offers you everything you need?
it depends.
you cab run the zwavejs2mqtt in two ways - via mqtt or via websockets.
if you run via mqtt you need an mqtt broker and you need to configure HA and zwavejs2mqtt to talk to that broker. and setting up mqtt discovery is a benefit as well.
if you run it via websockets then you will need to enable that in the zwavejs2mqtt control panel. THen in HA you need to configure the zwavejs integration to communicate to the zwavejs2mqtt websocket server.
Or you could do both if you really wanted to but it’s unnecessary and could get confusing since zwavejs2mqtt would create two of everything in HA - one for MQTT and one for websockets.
I followed the guide and now have a nicely functional ZWaveJS2MQTT based setup. Thanks @petro for the guidance.
But I noticed a weird behaviour this morning that I’m wondering if anyone else is seeing. If I stop the ZWave2MQTT Addon for any reason (for example, to upgrade to the latest version like I did this morning when 0.15.0 was released), then for some reason the “official” ZWave JS Addon gets automatically installed, starts, and takes over the ZWave USB dongle. That obviously means that the ZWaveJS2MQTT Addon is blocked from using it when it starts up a few seconds later (e.g. after the update).
Fixing it is as simple as stopping the official ZWave JS Addon, uninstalling it again, and then restarting the ZWaveJS2MQTT Addon, but it’s really annoying, and it’ll just pop up again the next time I stop/restart the ZWaveJS2MQTT Addon.
Is anyone else seeing this? I checked, and the ZWave JS integration seems to be configured properly to use the ws://a0d7b954-zwavejs2mqtt:3000 server. So I’m not sure what keeps trying to default to the ZWave JS addon.
Nevermind. I figured it out. For anyone out there who’s also seeing this, here’s how I fixed it.
(Be sure to take a snapshot before doing anything)
I uninstalled the ZWave JS integration and reinstalled just as a test, making sure to uncheck the “use official addon” box like it says in big bold letters in the documentation. I was pretty sure I did that the first time around as well but I guess I fat-fingered it because this install didn’t see the same issue I described above.
But, I had already spent hours configuring device and entity names and associated automations for all my ZWave devices, and reinstalling the Integration would wipe all that out. So I reverted to my most recent snapshot, and then opened up the /config/.storage/core.config_entries file using the File Editor plugin. Look for the section with the domain “zwave_js”, and change the “use_addon” and “integration_created_addon” lines to both be “false”. Then save and restart HA. Problem solved.