ZHA not working after update to 0.115

My conbee Ii ZHA integration just blew itself up after going from 115.2 to 115.3. Even rolling back to 114.4 did not solve the problem. The migration doc is unclear: adding the text as is to my config does not fix the issue, and i can’t find any info on the configuration variables to know how to tweak for my system.

This sucks :frowning:

All device are offline and i see this in the logs:

Logger: homeassistant.components.zha.core.channels.base
Source: components/zha/core/channels/base.py:280
Integration: Zigbee Home Automation (documentation, issues)
First occurred: 3:55:53 PM (1 occurrences)
Last logged: 3:55:53 PM

[0x0C22:1:0x0201]: async_initialize: all attempts have failed: [DeliveryError('[0x0c22:1:0x0201]: Message send failure'), DeliveryError('[0x0c22:1:0x0201]: Message send failure'), DeliveryError('[0x0c22:1:0x0201]: Message send failure'), DeliveryError('[0x0c22:1:0x0201]: Message send failure')]

This too:

Logger: homeassistant.components.websocket_api.http.connection.548042838080
Source: components/websocket_api/connection.py:84
Integration: Home Assistant WebSocket API (documentation, issues)
First occurred: 3:54:35 PM (6 occurrences)
Last logged: 3:55:12 PM

Received invalid command: zha/device

Triple edit: like others, it appears I can repair each device by unpairing it and repairing it…but that’s a lot of devices.

OK it wasn’t that bad. Everything came back one by one as i unpaired it and the repaired it. No renaming devices or entities as it was all preserved. All node-RED flows back 100% too.

I need to not be such a baby.

Does anyone have a real solution?

Re-pairing 60 Zigbee items is not something that I want to do.

Updated: Tried unplugging and replugging the HUSBZB-1, my bulbs are now ALL flashing constantly.
This is getting ridiculous.

Update 1: Re-paired every single sensor. Everything worked for an hour before having all the lights (HUEs) flash again. Can anyone let us know what is going on?

I’m starting to get really fed up as well.

I have approx 150 devices, two weeks ago I wiped my NUC clean and installed HASSOS from my previous Ubuntu and docker setup as it was “not supported”

Spent about five days getting back to where I was previously.

ZHA is worse than ever.

Things dropping off the network, its not interference ( yes I have a spectrum analyser) things wont join to the network after a period of time and I have to get a cold boot of the NUC to get it operating again.

Tearing my hair out. I am sick of spending every spare hour working on HA

Did you try this:

works for me too. But i have only 6 devices… for those who has 10+ it is not a solution :frowning:

Is this a confirmed solution to not having to re-pair all my zha devices? Does this work for the HUSBZB-1 stick? This ZHA issue of not working after 0.115 upgrade is the one thing holding me back from upgrading.

You should be able to upgrade without issues, this is only for Texas Instruments based adapters.

So apparently this is not true and it did impact me. So my question remains is this a confirmed solution to avoid having to repair all my zha devices or am I screwed now and have to bite the bullet and repair everything?

**** Edit - I apologize and stand corrected, initially everything was unavailable and it took an unusually long time to come back but it appears just about everything is back. Never seen it take that long after a restart/upgrade.

1 Like

I’ve given up - moved to zigbee2mqtt to give some separation between HA and ZigBee

2 Likes

I hear you, I have gone back to Deconz

Deleting ZigBee.db and repairing devices worked for me, although it’s a pain to pair all the devices again.

It’s funny that for zwave the separation of HA and OZW is propagated, but for zigbee the reverse is True. Frenck said in one of his streams that there is no point in using z2m in a HA szenario. The EAI guy in me says decouple EVERYTHING via messaging if possible.

I just updated and experienced the same issue. Having to repair everything is a nightmare. Especially the xiaomi stuff where pairing is a pain anyway just for 1 device.

Whats the chances this will happen again. I feel now that zha will be unreliable.

Ugh for the 2nd time in about two months following an update half my devices have dropped off the network which I’m now going to have to go and repair, as others have said, pain with the Xiomi
devices which take forever. I’m using the elelabs HUSBZB-1, am I stuck with ZHA? Anybody have any suggestions?

I just setup my entire zigbee setup two weeks ago on ZHA with 40 odd devices. This is a massive pain.

After every core update my zigbee devices become unavailable. it’s a pain in the ass…

Hi, happened again when upgraded to 2021.1.5
Has anybody found a workaround for this or at least the reason why is this happening (except the obvious core upgrade)?
Any help is appreciated.
BTW my zigbee devices are unavailable but I can still see the connected nodes in the ZHA map.

This did not happen to me, even though I also have the latest update. Maybe your hub/usb is at fault here? Does it show up as connected to one of the ports in supervisor->host->…->hardware?

Hi, thanks for the fast reply but I have a Synology docker running with the core installed.
I can see that the USB device is fine and have even made it work on Windows platform so I think that the device is working properly.