2021.10.0: Z-Wave S2 support, Tuya, secure ESPHome and 400 new icons

[da-anda]

I just updated from 2021.9.5 to 2021.10.4 and now I can not access any of my lovelace dashboards, nor the settings dashboard. I get a JS error in the console

Uncaught (in promise) ChunkLoadError: Loading chunk 12545 failed.
(error: http://XXX:8123/frontend_latest/484f044e.js)

When I try to open the file directly, FF says “content encoding error → file is using an unknown compression”. Anybody any idea? Same issue in Chrome, so it’s not a FF related bug. I basically can’t do anything via the frontend now, even in the few pages that still load because of dependency loading issues of the same type.

Anybody any idea? I asked on discord already, but besides of “ask in a different channel” I did not get any reply so far and I am out of ideas. I restarted everything already, HAOS is up2date, nothing fixes it.

[gdreelin (Greg Dreelin)]

For some reason after Release 2021.10.3 it slowed my HA down to a crawl. I could not configure anything since it would either freeze or just take a very long time to react. If I went to Integration it would take forever to configure something. I restored back down to the version 2021.10.2 and it all started working again. I did another upgrade and same thing dog slow. I am using HA on a stand alone NUC i7 with 16Gbs mem, and a 200Gb SSD. I installed the HA OS directly on to the system to run the core HA component. I am using very little SSD space and barely any of the memory. Not sure why the release is doing that but I figured I better let someone know.

Update: After backing down to last release and then reinstalling the current one it fixed the slow response. The first install must have loaded wrong. So seems it is all good now.

[ktownsend-personal (Keith)]

I’m having similar z-wave migration issue. The migration wizard aborts saying the add-on didn’t start (but it is running when I go check it myself). When this happens, I end up with both integrations when I go to configuration>integrations. If I go into the original integration and start the wizard again, it gets further, but it shows me a long list of every entity and says they won’t be migrated. There are no other lists of what will be migrated, so it just feels like a big fail. I have tried it 5 times and every time it’s exactly the same behavior. I reverted to my backup and will have to stay on the legacy z-wave until this gets figured out. I have nearly 100 devices, so I’m really wanting to avoid re-doing everything from scratch.

[Bobb]

I had something similar and the only way to access anything was via Nabu casa. This was then fixed for me in 10.4 and a complete shutdown of my system (pi 4). Restarts didn’t really do much.

[da-anda]

I downgraded my system to 2021.10.3 via SSH now, and everything seems to be fine again. Might try to upgrade to 2021.10.4 again when I have more time, since I suppose it was just a broken upgrade process on my end, as nobody else seems to have these issues.

[inkblotadmirer]

@ktownsend-personal

If I go into the original integration and start the wizard again, it gets further, but it shows me a long list of every entity and says they won’t be migrated. There are no other lists of what will be migrated, so it just feels like a big fail. I have tried it 5 times and every time it’s exactly the same behavior. I reverted to my backup and will have to stay on the legacy z-wave until this gets figured out. I have nearly 100 devices, so I’m really wanting to avoid re-doing everything from scratch.

I am actually surprised a migration wizard was attempted. My guess is that only very simple, possibly with relatively new components, will be possible to migrate mostly automatically. Kudos to the HA team trying to attempt this but it may have been a mistake resulting in unrealistic expectations.

In my case almost all zwave devices ended up with a different number of entities, many of which reported different states (“Auto low” vs “Auto Low”, or different states entirely), different possible states, etc. There is no way these are going to be migrated automatically, and battery-powered devices are notoriously finicky.

The wizard might get you over the initial hump but peruse the conversion guides (example) for a heads-up on manual steps likely to be encountered based on what the wizard was telling you (or what the initial steps in the conversion guides indicate). Compare that to starting over from scratch, which may actually be easier.

[pxuxmxa]

Sorry if such a question has already been. My indicator does not change when the light is “on”. How to fix it?

[francisp (Francis)]

Is that the tuya integration ?

[pxuxmxa]

No. I suffer separately with Tuya:) these are Sonoff switches. Integrated via Sonoff LAN

[francisp (Francis)]

Better ask in the Sonoff Lan topic :

Control Sonoff Devices with eWeLink (original) firmware over LAN and/or Cloud

Pros: work with original Sonoff firmware work over local network (LAN) work with devices without DIY-mode work with devices in DIY-mode support single and multi-channel devices support TH and POW device attributes support Sonoff RF Bridge 433 for receive and send commands instant device state update with Multicast (optional) get devices list from eWeLink servers (with names and apikey/devicekey) and save it locally (optional) change device type (switch, light or fan) (optional) set multi-chann…

[pxuxmxa]

Thanks. It broke down when switching to 2021.10.0. But if you select the “Button” type, the current is displayed correctly

[firstof9]

Add state_color: true to your first card.

[nirkons (Nir)]

Is anyone else stuck on 2021.9?

21-10-15 19:54:40 INFO (MainThread) [supervisor.homeassistant.core] Updating Home Assistant to version 2021.10.4
21-10-15 19:54:40 INFO (SyncWorker_1) [supervisor.docker.interface] Updating image Package intel-nuc-homeassistant · GitHub to ghcr.io/home-assistant/-homeassistant:2021.10.4
21-10-15 19:54:40 INFO (SyncWorker_1) [supervisor.docker.interface] Downloading docker image ghcr.io/home-assistant/-homeassistant with tag 2021.10.4.
21-10-15 19:54:40 ERROR (SyncWorker_1) [supervisor.docker.interface] Can’t install ghcr.io/home-assistant/-homeassistant:2021.10.4: 400 Client Error for http+docker://localhost/v1.40/images/create?tag=2021.10.4&fromImage=ghcr.io%2Fhome-assistant%2F-homeassistant: Bad Request (“invalid reference format”)
21-10-15 19:54:40 WARNING (MainThread) [supervisor.homeassistant.core] Updating Home Assistant image failed

I am running Ubuntu 20.04.03 (yes I know it is not officially supported), never had an issue until today.
Tried setting manual DNS (1.1.1.1/8.8.8.8)

HA does have internet access (its accessible from outside and addons do update)

[Ghost_Buster]

Yes it does. However, you have to manually add the correct repo to HACS. The docs clearly state: “If you have HACS, you must add this repository (“https://github.com/ollo69/ha_tuya_custom”) to your Custom Repository selecting the Configuration Tab in the HACS page.
After this you can go in the Integration Tab and search the “Tuya Custom” component to configure it.”
The repo-URL has changed, the name now contains underscores, that’s it. They still support HACS.

[chairstacker]

@Ghost_Buster

must add this repository (“https://github.com/ollo69/ha_tuya_custom”) to your Custom Repository

As I wrote, I tried that but it didn’t work.

In the meantime, @stirlo suggested to use this repository and it worked fine, so I’m good for now:

GitHub - andrey-yantsen/home-assistant-tuya-old

[KennethLavrsen]

My log is flooded with this

Both are messages I am used to see a couple of times after restart

Now they just repeat flooding the log. I have no idea where it comes from

2021-10-15 19:53:41 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!

2021-10-15 19:53:41 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!

2021-10-15 19:53:43 WARNING (MainThread) [slixmpp.basexmpp] Legacy XMPP 0.9 protocol detected.

2021-10-15 19:53:44 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!

2021-10-15 19:54:43 WARNING (MainThread) [slixmpp.basexmpp] Legacy XMPP 0.9 protocol detected.

2021-10-15 19:54:51 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!

2021-10-15 19:55:43 WARNING (MainThread) [slixmpp.basexmpp] Legacy XMPP 0.9 protocol detected.

2021-10-15 19:55:44 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!

2021-10-15 19:56:43 WARNING (MainThread) [slixmpp.basexmpp] Legacy XMPP 0.9 protocol detected.

2021-10-15 19:56:44 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!

2021-10-15 19:57:43 WARNING (MainThread) [slixmpp.basexmpp] Legacy XMPP 0.9 protocol detected.

2021-10-15 19:57:50 ERROR (Thread-4) [homeassistant] Error doing job: Task was destroyed but it is pending!

2021-10-15 19:58:43 WARNING (MainThread) [slixmpp.basexmpp] Legacy XMPP 0.9 protocol detected.

2021-10-15 19:58:44 ERROR (SyncWorker_27) [homeassistant] Error doing job: Task was destroyed but it is pending!

2021-10-15 19:59:43 WARNING (MainThread) [slixmpp.basexmpp] Legacy XMPP 0.9 protocol detected.

2021-10-15 20:00:43 WARNING (MainThread) [slixmpp.basexmpp] Legacy XMPP 0.9 protocol detected.

2021-10-15 20:00:52 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!

2021-10-15 20:01:43 WARNING (MainThread) [slixmpp.basexmpp] Legacy XMPP 0.9 protocol detected.

[stephack]

@KennethLavrsen

I have no idea where it comes from

Do you have Logitech Harmony hubs on your network? I have not updated to 2021.10.x yet but I also get that on startup because of my Harmony hubs and just ignore it. If it’s now going to flood my logs, I guess this might be another reason to hold off on the update for now.

[KennethLavrsen]

Yes. I have a harmony hub. I replaced it because it was so often unstable.

I replaced it by a Broadlink and that is rock solid and pure local.

I uninstalled the Harmony integration 6 months ago and when rediscovered I asked HA to ignore it. But I left the Harmony on so I could use the remote unrelated to HA. But the remote kept on popping up in Home Assistant and could not be ignored so I pulled the battery. Now I unplugged the Harmony hub and the flooding stopped.

That Harmony integration needs some developer attention. It behaves very badly

[bdraco (J. Nick Koston)]

@KennethLavrsen

That Harmony integration needs some developer attention. It behaves very badly

There is a fix here but it needs someone to test it

Avoid setting up harmony websocket from discovery

[Asynonyx]

@nirkons

Is anyone else stuck on 2021.9?

// Omitted logs to keep this post short //

I am running Ubuntu 20.04.03 (yes I know it is not officially supported), never had an issue until today.
Tried setting manual DNS (1.1.1.1/8.8.8.8)

HA does have internet access (its accessible from outside and addons do update)

I am not exactly sure what caused it and at some point I gave up trying.
I was running the VM-image (docker container) at version 2021.10.2 and after upgrading to 2021.10.4 this issue started occurring on my install aswell.

Fiddling with the DNS settings of the root OS (nmcli) and reverting to DHCP somehow resolved my issues. As soon as I defined a static IP the issues came back. It also appeared to occur at random and would sometimes work.

Saw this as a sign and took some time to reïnstall my Home Assistant image and moved my configs over to the new install.

I am not sure whether you’re running your Home Assistant on a static IP. If so you might aswell try configuring it to use DHCP and see if that resolves your issues. If that’s the case you could try to assign a static IP from your router to the MAC address of your Home Assistant install as that seems to have resolved the issues at my end for now.