Hassio fails to start after upgrade to 106.5

Continuing the discussion from Hass.io on Synology DSM (native package):

Still struggling with hassio_audio container failing to run.

Upgraded to version 106.5. Did a reload for supervisor 207. Manually downloaded hassio_audio:8 because supervisor was generating errors when trying to load hassio_audio:5 .

Hassio_audio still causing constant restart. Last few lines of the log:

20-03-04 08:10:40 INFO (MainThread) [supervisor.core] Supervisor is up and running
20-03-04 08:10:40 INFO (MainThread) [supervisor.host.sound] Update PulseAudio information
20-03-04 08:11:01 WARNING (MainThread) [supervisor.tasks] Watchdog found a problem with PulseAudio plugin!
20-03-04 08:11:01 INFO (MainThread) [supervisor.audio] Start Audio plugin
20-03-04 08:11:01 INFO (SyncWorker_7) [supervisor.docker.interface] Clean hassio_audio application
20-03-04 08:11:19 INFO (SyncWorker_7) [supervisor.docker.audio] Start Audio homeassistant/amd64-hassio-audio with version 8 - 172.30.32.4
20-03-04 08:11:40 WARNING (MainThread) [supervisor.tasks] Watchdog found a problem with PulseAudio plugin!
20-03-04 08:11:40 INFO (MainThread) [supervisor.audio] Start Audio plugin
20-03-04 08:11:40 INFO (SyncWorker_10) [supervisor.docker.interface] Clean hassio_audio application

This message is constantly recurring. The problem started when I upgraded to supervisor 207.

Thanks for any help.

I have the same problem!

I have the same problem!
My home assistant is ko from yesterday evening (ver. 0.106.4)!

I think that the problem comes from the update of supervisor.
May be from this modification ?

Not sure. But you can make your home assistant by following these commands I think :

Same here on Synology

Perhaps try to install audiostation, that might help.

I can’t reproduce this issue on my system.

Hi Fredrik,

Thanks for the response, but audiostation was already installed.

I logged an issue for similar. For me, the supervisor needed to use /run instead of /var/run (like docker.sock) Rerunning the install should fix.

Hallo,

I did a full reinstall of the hassio package on my synology and the problem remained the same. Audiostation was running when I did the install.

I then tried uninstalling Audiostation and the error changed as follows:

20-03-04 18:50:39 INFO (MainThread) [supervisor.audio] Start Audio plugin
20-03-04 18:50:39 INFO (SyncWorker_13) [supervisor.docker.interface] Clean hassio_audio application
20-03-04 18:50:56 ERROR (SyncWorker_13) [supervisor.docker] Can't start hassio_audio: 500 Server Error: Internal Server Error ("Bind mount failed: '/dev/snd' does not exists")
20-03-04 18:50:56 ERROR (MainThread) [supervisor.audio] Can't start Audio plugin
20-03-04 18:50:56 ERROR (MainThread) [asyncio] Task exception was never retrieved
future: <Task finished coro=<Tasks._watchdog_audio_docker() done, defined at /usr/src/supervisor/supervisor/tasks.py:244> exception=AudioError()>
Traceback (most recent call last):
  File "/usr/src/supervisor/supervisor/tasks.py", line 252, in _watchdog_audio_docker
    await self.sys_audio.start()
  File "/usr/src/supervisor/supervisor/audio.py", line 168, in start
    raise AudioError() from None
supervisor.exceptions.AudioError
20-03-04 18:50:59 WARNING (MainThread) [supervisor.tasks] Watchdog found a problem with PulseAudio plugin!
20-03-04 18:50:59 INFO (MainThread) [supervisor.audio] Start Audio plugin
20-03-04 18:50:59 INFO (SyncWorker_6) [supervisor.docker.interface] Clean hassio_audio application
20-03-04 18:51:24 ERROR (SyncWorker_6) [supervisor.docker] Can't start hassio_audio: 500 Server Error: Internal Server Error ("Bind mount failed: '/dev/snd' does not exists")
20-03-04 18:51:24 WARNING (MainThread) [supervisor.tasks] Watchdog found a problem with PulseAudio plugin!
20-03-04 18:51:24 INFO (MainThread) [supervisor.audio] Start Audio plugin
20-03-04 18:51:24 ERROR (MainThread) [supervisor.utils] Can't execute run while a task is in progress
20-03-04 18:51:24 ERROR (MainThread) [supervisor.audio] Can't start Audio plugin
20-03-04 18:51:24 ERROR (MainThread) [asyncio] Task exception was never retrieved
future: <Task finished coro=<Tasks._watchdog_audio_docker() done, defined at /usr/src/supervisor/supervisor/tasks.py:244> exception=AudioError()>
Traceback (most recent call last):
  File "/usr/src/supervisor/supervisor/tasks.py", line 252, in _watchdog_audio_docker
    await self.sys_audio.start()
  File "/usr/src/supervisor/supervisor/audio.py", line 168, in start
    raise AudioError() from None
supervisor.exceptions.AudioError
20-03-04 18:51:39 WARNING (MainThread) [supervisor.tasks] Watchdog found a problem with PulseAudio plugin!
20-03-04 18:51:39 INFO (MainThread) [supervisor.audio] Start Audio plugin
20-03-04 18:51:39 INFO (SyncWorker_19) [supervisor.docker.interface] Clean hassio_audio application

After I reinstalled Audiostation the errors in the supervisor log changed to this:

20-03-04 18:56:20 WARNING (MainThread) [supervisor.tasks] Watchdog found a problem with PulseAudio plugin!
20-03-04 18:56:20 INFO (MainThread) [supervisor.audio] Start Audio plugin
20-03-04 18:56:20 INFO (SyncWorker_4) [supervisor.docker.interface] Clean hassio_audio application
20-03-04 18:56:58 INFO (SyncWorker_4) [supervisor.docker.audio] Start Audio homeassistant/amd64-hassio-audio with version 8 - 172.30.32.4
20-03-04 18:57:19 WARNING (MainThread) [supervisor.tasks] Watchdog found a problem with PulseAudio plugin!
20-03-04 18:57:19 INFO (MainThread) [supervisor.audio] Start Audio plugin
20-03-04 18:57:19 INFO (SyncWorker_15) [supervisor.docker.interface] Clean hassio_audio application
20-03-04 18:57:47 INFO (SyncWorker_15) [supervisor.docker.audio] Start Audio homeassistant/amd64-hassio-audio with version 8 - 172.30.32.4

The restart loop continues. I am still at a loss as to what to do.

Thanks for your help.

Same here on Synology. I see the new hassio_audio container in a startup loop. The log shows that adjust-audio-group.sh fails with exit code 1. After that, the container shuts down clean. I will now look at the container and what that script does.

[s6-init] making user provided files available at /var/run/s6/etc...exited 0.
[s6-init] ensuring user provided files have correct perms...exited 0.
[fix-attrs.d] applying ownership & permissions fixes...
[fix-attrs.d] persistent-data: applying... 
[fix-attrs.d] persistent-data: exited 0.
[fix-attrs.d] done.
[cont-init.d] executing container initialization scripts...
[cont-init.d] adjust-audio-group.sh: executing... 
[cont-init.d] adjust-audio-group.sh: exited 1.
[cont-finish.d] executing container finish scripts...
[cont-finish.d] done.
[s6-finish] waiting for services.
[s6-finish] sending all processes the TERM signal.
[s6-finish] sending all processes the KILL signal and exiting.

Ok, after looking at that script and see what it is doing, I got a solution: the script tries to sync the audio group between the host and the container. The issue is that if the host does not have a group audio, this fails. This will work with all “standard” Linux installations as every of those will usually have that group by default. But it fails on devices like NAS systems who don’t have (or need) the audio group.

The solution is to create a group audio on your device. For Synology NAS systems, you can just create a group, name it audio and that’s it. No permissions necessary. For other systems, you may need to edit /etc/group to add the group.

8 Likes

Filed an issue here:

2 Likes

I was having a similar issue on RPi3 with an error like this.

ERROR (MainThread) [supervisor.homeassistant] Home Assistant can't write pulse/client.config: [Errno 21] Is a directory: '/data/tmp/homeassistant_pulse'

At some point during my attempts to track this down (more specifically I ran su repair) Supervisor was updated to version 208.

Now all is well in the State of Denmark…

Added the audio group on my synology and all is well now. the container stays up. No more looping.

1 Like

Hello, I am also desperate when trying to install HA version 0.106.x. But the update was not the error, but the operator :raising_hand_man:of the hardware and software. For me HA runs on a RPi3 in the 32 bit version (because of the RaspBee and Deconz). So far I had opened the Web Ui via a shortcut on the smartphone in the browser, so everything was fine. So update to 0.106.3 or 0.106.5, the update went through and the RPi was restarted. So I press my link on the smartphone and then it says: Page is not accessible! After a long time of trying, trying, reinstalling and despair, the enlightenment was there, the problem was the connection. The link was (example): “123.456.789.12:8123/states”, or something. I then entered this address in the browser and left out the “/ states” so “123.456.789.12:8123” and the page opened. Of course, the States UI is out of date. Perhaps this will solve his problem for one or the other, if not continued success and have fun with HA

@Fizzblizz, can you please give me a hint.
I’ve added the audio group as shown below but the issue still remain.
Did I do something wrong or overlook something like a reset?


Screenshot of the group file
image

This looks good. Can you try to capture the log of the audio container? It may not work every time as DSM kills the log when the container terminates, but I managed to get the log eventually with multiple trys. Look for some script that terminates with an exit code that is not zero. Is there something?

The following is an extract of the log file. Hopefully all data is available.

For whatever reason, now it works.
8 hours after I created the audio group, it’s up and running.
Strange but I’m glad it works now.
Thanks again to @Fizzblizz

1 Like

Yeah, this looks like something that might get fixed with a complete restart. I guess there was a directory missing on the host.