Running RbPi3, Hass.io
I added an Aeotec Recessed Door Sensor Gen5. It did not add to the device list.
I rebooted homeassistant and got stuck at Starting Z-Wave Network. I have an Unknown Node 0 (Node:0 Probe). Listed. Z-Wave devices are actually working. But I can’t add or remove nodes, due to the UI thinking the network is not started.
I am aware that the Z-Wave network can take a few minutes to start up. I’ve exceeded an hour. I’ve tried several reboots.
All advice is welcomed!! Thanks
Might seem like a really silly thing to suggest, but if you are using a USB stick, try identifying it via Supervisor>System>Hardware in the GUI and put the resultant device address in your configuration.yaml under the zwave: section. I ran hassio on RPi for a couple of years and I think it is the same.
My configuration.yaml looks like this:
# z-wave component
zwave:
usb_path: /dev/serial/by-id/usb-0658_0200-if00
network_key: !secret zwave_network_key
config_path: /config/openzwave_config
Ignore the config_path: unless you have downloaded a separate copy of the open-zwave config from github.
I run HA in docker and had issues when I plugged a USB keyboard dongle into the server tin…/dev/ttyACM0 suddenly became /dev/ttyACM1 and Z-Wave wouldn’t start. No poop Sherlock! Probably not this, but worth a try.
Setup is completely hosed. It was suggested my zwcfg was corrupted and I should delete it. When I did the network did come back up, but it completely farked my device listing. I’m trying to restore a snapshot at this point, but getting no UI feedback nor logs for several minutes now. I saw the add-ons got added back but nothing else. This is a really terrible UX, that rubs salt in the wounds of stuff just taking a crap for no reason. I don’t know if I can continue with a system this flakey.
Try deleting the HA database file and reboot. It should recreate itself after the reboot.
Home-assistant_v2.db
Followed that advice. Tried restore again, still absolutely nothing happening. No feed back, no logs, no nothing. Should I expect anything to be happening here?
Are you restoring from a full snapshot (all check boxes checked when you took the snapshot in the first place)? If so, are you performing the scarily named “wipe and restore” or the more friendly named “restore selected”? I’d opt for the scary option in the GUI even though I suspect they might both result in the same outcome.
When you hit restore, from the GUI, you would expect to get the “lost connection” pop-up in the bottom left of your screen, then once the restore has gotten to the point that the front-end is up, you should regain access to the GUI (the lost access/connection message disappears) and depending on what you have added and the size of your config, things continue to load in the background for some time - one of those is your z-wave config if the config is valid. I’m hovering around the 550 entities mark at the moment across my config, and my 20 odd z-wave nodes take around a minute to two minutes to transition from cache-load to ready state.
If running from the CLI, it’s: hassio sn list (and then pick the correct snapshot) then: hassio sn restore -slug xxxxxyyyyy.
Depending on the size of your db, a snapshot and/or restore can take ages - before I moved to a MySQL instance running on FreeNAS, my db was hovering around the 4GB mark which was a nightmare to backup/restore on the RPi with an SD card or USB pen drive. Sloooooooow.
Which leads me on to…Could it be a duff USB/SD? I’ve had many go bad in RPis over the years.
Also, did you happen to install the Samba add-on? If so, you can browse from a pc or Mac to \\RPiNameOrIP\backup and get at the individual config files within the zipped backup/snapshot - like anything you may have deleted in error or haste such as any z-wave config files like you zwcfg.xml etc. Sorry if I’m teaching you to suck eggs!
If you have a spare SD or USB pen drive and can get at those snapshots/backups from another pc/Mac, perhaps copy over all your snapshots and try a fresh hassio install to a new USB/SD. If that works, install the Samba add-on, use your pc/Mac to copy over the relevant snapshot to the \\rpinameorip\backup folder, reboot and attempt a wipe and restore from the snapshot you copied over.
bikefright.
Thanks for chiming in. I’ve given up on the restore. I started with a fresh install and then went to restore. All it would ever do is restore the add-ons. Yes, it’s a full snapshot, etc. I opened the gz and found all the config files, it should’ve worked. When I deselected the add-ons, the popup message for restore complete was within seconds.
I’m moving off the RbPi, to a NUC, so I’m done fiddling with this until the hardware arrives. I’m tied up all week anyway.
I find that selecting a node while it still says Zwave Network Starting going to Services > zwave.add_node and pressing Call Service gets the GUI back in order, making sure that the entities have finished loading from the Nodes drop down first in the Zwave panel of course.
I’m in this situation right now, and it doesn’t seem like zwave.add_node is available as an option in Services while “Starting Z-Wave Network” is in progress. Looks like my nodes are available in the dropdown though.
I’ve gotten pretty familiar with the ‘nuances’ of z-wave on HA. First and foremost, never reboot while z-wave is still starting up. Check the logs, don’t trust the UI. You might have a node causing pollution, it’s pretty difficult to determine that as far as I’ve found. Your zwcfg could be corrupted.
Thanks. I was pretty early in my move from Vera to HA Z Wave, so I just started over and am being extra cautious, making sure everything is as it should be, restarting after healing to make sure the node is still present on restart, modifying entity ids, etc. So far so good.
Is there any proper solution for this? I have a device that dropped off my network, I tried to reset it and “add devices” there were issues… ever since then I get the dreaded “starting z-wave network” message…
I did delete the database, didn’t help. I added the lines:
zwave:
usb_path: /dev/serial/by-id/usb-0658_0200-if00
to configuration.yaml… the zwave logs show that stuff is happening…
2021-10-07 10:43:27.123 Detail,
2021-10-07 10:43:27.124 Info, Node009, Received Alarm report: type=0, level=0, sensorSrcID=0, type:Burglar event:8, status=255
2021-10-07 10:43:27.125 Detail, Node009, Refreshed Value: old value=0, new value=0, type=byte
2021-10-07 10:43:27.125 Detail, Node009, Changes to this value are not verified
2021-10-07 10:43:27.126 Detail, Node009, Refreshed Value: old value=0, new value=0, type=byte
2021-10-07 10:43:27.126 Detail, Node009, Changes to this value are not verified
2021-10-07 10:43:27.127 Detail, Node009, Refreshed Value: old value=0, new value=0, type=byte
2021-10-07 10:43:27.127 Detail, Node009, Changes to this value are not verified
2021-10-07 10:43:27.127 Detail, Node009, Refreshed Value: old value=8, new value=8, type=byte
2021-10-07 10:43:27.128 Detail, Node009, Changes to this value are not verified
2021-10-07 10:43:27.128 Detail, Node009, Notification: ValueChanged
2021-10-07 10:43:27.132 Detail, Node009, Notification: ValueChanged
2021-10-07 10:43:27.137 Detail, Node009, Notification: ValueChanged
2021-10-07 10:43:27.142 Detail, Node009, Notification: ValueChanged
2021-10-07 10:44:36.138 Detail, Node009, Received: 0x01, 0x11, 0x00, 0x04, 0x00, 0x09, 0x09, 0x71, 0x05, 0x00, 0x00, 0x00, 0xff, 0x07, 0x08, 0x00, 0xd5, 0x00, 0xbb
2021-10-07 10:44:36.138 Detail,
2021-10-07 10:44:36.140 Info, Node009, Received Alarm report: type=0, level=0, sensorSrcID=0, type:Burglar event:8, status=255
2021-10-07 10:44:36.141 Detail, Node009, Refreshed Value: old value=0, new value=0, type=byte
2021-10-07 10:44:36.141 Detail, Node009, Changes to this value are not verified
2021-10-07 10:44:36.142 Detail, Node009, Refreshed Value: old value=0, new value=0, type=byte
2021-10-07 10:44:36.143 Detail, Node009, Changes to this value are not verified
2021-10-07 10:44:36.144 Detail, Node009, Refreshed Value: old value=0, new value=0, type=byte
but no such luck in the GUI.
Switch to zwavejs.