after the migration for zwave I’m still using HUSBZB
after the migration for zigbee I’m using the one comes with yellow
And after the migration everything seems ok, until a restart.
Today I restarted yellow for some non-zigbee/zwave related changes, and after that, yellow is trying to reach out to my HUSBZB stick for zigbee rather than the built-in one.
Error setting up entry HubZ Smart Home Controller - /dev/serial/by-id/usb-Silicon_Labs_HubZ_Smart_Home_Controller_xxxxxx-port0, s/n: xxxxx - Silicon Labs - xxx
2:09:17 PM – (ERROR) components/zha/core/gateway.py
Couldn't start EZSP = Silicon Labs EmberZNet protocol: Elelabs, HUSBZB-1, Telegesis coordinator (attempt 1 of 3)
2:09:17 PM – (WARNING) Zigbee Home Automation - message first occurred at 2:09:01 PM and shows up 3 times
And obviously, zigbee automation was completely dead(since during the migration the HUSBZB was reset for zigbee part).
I deleted the zha integration and re-added it back by choosing the correct radio, and now it seems ok again.
But just wondering if anyone has seen this recently? It will be so annoying if I need to do this every time when need to restart home assistant. Thanks
Hi,
Glad the “stuff I got wrong” post helped you!
I’m running vanilla:
Zigbee Home Automation + Yellow Sil Labs (EZSP = Silicon Labs EmberZNet protocol: Elelabs, HUSBZB-1)
Z-Wave JS Supervisor add-on + Z‐Stick Gen5 USB Controller
The HASSOS docker virtualisation makes it surprisingly difficult to workout the /dev/tty* device mapped through to the integration - I thought it was in the Download Diagnostics, or in the binding config, but couldn’t find it today.
From a web terminal, ls -la /dev/serial/by-id/usb-* only came back with one file - a symbolic link to ../../ttyACM0 which sounds like the Zigbee SilLabs device on my Yellow (there are several other /dev/tty* serial device files likely for Z-Wave, serial console, USB console, etc).
Is you ls -l the path mentioned in the first line of the HubZ Smart Home Controller error, what comes back?
That it apparently shows the s/n: REDACTED makes me wonder if the integration is connected to the hardware (e.g. NOT wrong port/device/moon phase), but there’s something deeper wrong.
My best suggestion is take a full backup, backup the adapter config from the integration, then at least you have a baseline configuration whatever you try next.
Given a chunk of Zigbee config is in adapter flash, I don’t know if deleting and reinstalling the integration will work? I guess it comes down to how many devices you might need to re-pair if a re-install fails.
I think there is something else going on, as none of those two path matches the log error message… anyway since re-adding integration currently works for me I’ll take it, and keep monitoring, if this is not an uncommon scenario, I believe there will be more topics in forum like this.