HAOS Supervisor - Failed to Setup after updating core to 2023.6.0 and Supervisor to 2023.06.1 #4360
Comments
|
Hello, same issue on my end, but running supervised on Debian. Rolling back to 2023.05.4 backup gets it back. Also tried, without success : Observer also says Supervisor is connected Issue remains after 'ha core update' to 2023.06.1 |
|
I have to say a similar issue, For the moment, I don't see the backup I have done,maybe it is because the home assistant supervisor is down🤔 |
|
@frangigu @Kathou99 do you have this repo https://github.com/alexbelgium/hassio-addons on your installation, by the way? On a clean install, I noticed hat the issue starts to happen immediately after I add this repo... If I remove it, the issue is gone....
|
|
this repo does not belong to HACS... it's a standalone repo maintened by @alexbelgium... As soon as I got my system stable, I'll open an issue on that repo. |
Sorry I misread, I thought it was related to HACS itself... I also had that repo, but I was no longer using any add-ons from it, I just deleted the "Repo" in the repositories menu, restarted HA-core and Supervisor started correctly. Nice catch ! |
|
I just rebooted my whole system, it also seems to fix a recent bug where I had to restart core (only) to get back everything related to supervisor (addons), just like the supervisor was too slow to load. |
Yes I have it |
|
I've indeed read many issues on my repo following 2023.06.01. Guess third party add-ons are not really encouraged anymore. Well that would free me some time :) This is a pure supervisor thing apart from removing the 80 add-ons I support I don't see what I can do |
Oh man, @alexbelgium, don't take it like this! You have done a brilliant and important work for the community, being a life saver in many occasions! I'm a truly admirer of your work. I believe it is more a technical issue than an attempt to jeopardize third party add-ons. I do have other 6 third party repos, but this issue only happens to yours. As soon as I get more time, I'll try to troubleshoot a little bit more, bit it feels just like some link used by your repo is now broken... Perhaps is that statistics widget? Keep the good work and good spirit! 😉 |
What about creating some SSL tunnels with Cloud flare solution? https://github.com/w35l3y/hassio-addons/tree/main/cloudflare_tunnel This is what I use to circumvent my ISP puting me behind a CGNAT... It is the same approach taken in the epic games addon for captcha links.... Perhaps some permanent tunnels to those 80 add-ons could bypass the whole thing... |
Thanks! The statistics widget is static : images are generated daily and stored as image on the repo, statistics are generated weekly and stored as text. Perhaps it's a question of repo size - that's the only thing that I see. HA periodically downloads the whole repo. Not sure the size for the moment but it's about 12mo I would say. I have a script running to optimize images on a weekly basis to keep it as low as possible. HA perhaps choke on this size... EDIT : actually the repo is only 12Mo, with 1498 files and 82 addons. The HA community repo is 2Mo though so the difference isn't THAT great |
https://www.deadlinkchecker.com/website-dead-link-checker.asp Perhaps is a javascript function aggregating/generating the links?
|
|
Thanks I didn't knew that tool. I'll correct both now Actually I've corrected only the first one - the second is based on github templates and can't be found when checking the files manually (or using vscode) : https://github.com/search?q=repo%3Aalexbelgium%2Fhassio-addons%20refname&type=code |
|
Btw in the initial log it seems a full backup is ongoing while refreshing the addons using an (excellent) third party tool : hassio google drive backup. To avoid that, it might be worth scheduling backups at night to avoid impacting on normal working. |
|
once again, you are right @alexbelgium .... I did not noticed that... But... I was looking into the Developers docs from HA, and the Repository configuration file should be an repository.yaml on the root, but you have instead a repository.json file in your repo.... Perhaps this is related? Doc: https://developers.home-assistant.io/docs/add-ons/repository/ Even the Sample Repo has an .yaml file in the root: https://github.com/home-assistant/addons-example/blob/main/repository.yaml Perhaps it is something that have changed in the last releases? |
|
That's not an issue : older repos were based on json (repository and config files) but then HA switched to yaml. Both are supported according to documentation. BTW, the hassio Google drive backup repo uses json : https://github.com/sabeechen/hassio-google-drive-backup |
I can confirm, removing this repo made the supervisor integration load again. |
I confirm it too. |
|
I can replicate that it makes this error : 23-06-13 20:02:26 ERROR (MainThread) [aiohttp.server] Error handling request
Traceback (most recent call last):
File "/usr/local/lib/python3.11/site-packages/aiohttp/web_protocol.py", line 433, in _handle_request
resp = await request_handler(request)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/aiohttp/web_app.py", line 504, in _handle
resp = await handler(request)
^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/aiohttp/web_middlewares.py", line 117, in impl
return await handler(request)
^^^^^^^^^^^^^^^^^^^^^^
File "/usr/src/supervisor/supervisor/api/middleware/security.py", line 185, in block_bad_requests
return await handler(request)
^^^^^^^^^^^^^^^^^^^^^^
File "/usr/src/supervisor/supervisor/api/middleware/security.py", line 201, in system_validation
return await handler(request)
^^^^^^^^^^^^^^^^^^^^^^
File "/usr/src/supervisor/supervisor/api/middleware/security.py", line 268, in token_validation
return await handler(request)
^^^^^^^^^^^^^^^^^^^^^^
File "/usr/src/supervisor/supervisor/api/middleware/security.py", line 280, in core_proxy
return await handler(request)
^^^^^^^^^^^^^^^^^^^^^^
File "/usr/src/supervisor/supervisor/api/utils.py", line 62, in wrap_api
answer = await method(api, *args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/src/supervisor/supervisor/api/store.py", line 178, in store_info
ATTR_ADDONS: [
^
File "/usr/src/supervisor/supervisor/api/store.py", line 179, in <listcomp>
self._generate_addon_information(self.sys_addons.store[addon])
File "/usr/src/supervisor/supervisor/api/store.py", line 114, in _generate_addon_information
ATTR_ADVANCED: addon.advanced,
^^^^^^^^^^^^^^
File "/usr/src/supervisor/supervisor/addons/model.py", line 227, in advanced
return self.data[ATTR_ADVANCED]
^^^^^^^^^
File "/usr/src/supervisor/supervisor/store/addon.py", line 19, in data
return self.sys_store.data.addons[self.slug]
~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^
KeyError: '243ffc37_rsync'Can't understand what is |
|
Ok, I've removed all translations files and replaced all big-sized png icons. Now the repo size is reduced by 46% to 6.5Mo. Let's see if it changes something Edit : on my system the error message is now removed. I've performed 3 quick successions repo updates without any issues although previously it was seen at each time see log
|
|
@alexbelgium I just added the repo on my clean installation, and indeed it was way faster to add than previously. I've restarted my HA instance and supervisor integration is just fine up to now.. I'll leave the repo there for some more time and report if anything is found... Thank you for all the efforts!
|
|
Thanks - if indeed size is the issue then I'll see how to further optimize (like running a github action to identify any file of more than 100ko) ; looks like the real solution would be that they implement a slightly longer timeout |
|
almost 3 hours stable and I decided to reboot the entire host to see what would be the behavior, and no issues at all :D 🥇 Let's reinstall some addons :D |
|
As a simple user, I monitored this topic, and I can confirm that this morning, everything works after rebooting my HA dedicated MinowBoard! :) |
|
I did some more tests - the culprit was actually the presence of translations and not the repo size. Re-adding them crashes things. |
|
Having suffered from this on my Pi3B+ (supervised) and not my Pi4 (HAOS), I finally got the Pi3 to update to 2023.6.1 this morning after removing the repo before the update. Are we safe to add it back now or is there still some work to be done still to make it stable ? |
|
Hi, well I can mimic that adding the translation files prevents supervisor from loading the repo but removing them removes all error messages which seems to indicate the root cause if identified and removed. I've also further reduced the repo size vs the tests above by continuing to optimize images. So, based on that and the comments in the posts above, in theory it should be safe :) To be on the safe side I've also removed all third party repos I wasn't using as I'm not sure if the timeout is per repo or global |
|
I've been running for 24h+ after adding the repo back with no issues. I think it's safe now. Question: what's the best way to close this issue? Thinking about further reference for others, should we do anything besides jus closing it? |
|
We have found the cause, and a fix around it. I see no reason to keep this issue open. It will always be visible for anyone else that has the same or a similar issue. |
Describe the issue you are experiencing
Hello,
Since I've update the HA Supervisor to 2023.06.1 and HA Core to 2023.06.01, i'm getting the "Failed to setup" message for Supervisor component on the new Integrations dashboard.
The issue is similar to #91461 and to #90558, but as they are for a different Core/Supervisor versions, I've decided to open this new issue as it could be related to the new one.
Here is what I've already tried:
Trough SSH:
ha core rebuild
ha supervisor reload
ha supervisor update
ha supervisor repair
ha supervisor restart (several times)
banner
On host:
Restart HA Core and the whole host several times.
No Repair available from the web interface -> Config -> System -> Repairs
The Observer component report that everything's fine.
Sys Info:
Home Assistant 2023.6.0
Supervisor 2023.06.1
Operating System 10.2
Frontend 20230607.0 - latest
Logs:
I'm running the AdGuard Home addon on this instance as the main DNS server for my entire network, and I'm not able to see any blocked requests to the host neither the containers.
Any ideas how to troubleshoot this further?
What type of installation are you running?
Home Assistant OS
Which operating system are you running on?
Home Assistant Operating System
Steps to reproduce the issue
...
Anything in the Supervisor logs that might be useful for us?
System Health information
System Information
Home Assistant Community Store
AccuWeather
Home Assistant Cloud
Home Assistant Supervisor
Dashboards
Recorder
Spotify
Supervisor diagnostics
the function is not available in my case.
Additional information
No response
The text was updated successfully, but these errors were encountered: