Some of my addons won't start but other do

[X-Ryl669]

Describe the issue you are experiencing

Since today (I don't remember updating anything in HA, only the debian OS via apt update/upgrade and rebooted), some of my addons don't "start". I mean that their docker container start, but the supervisor never see them. If I try to start them by hand, the supervisor says they are already running, but they never register in the interface, and I can't access them.

The addons that do start are frigate, piper, Advanced SSH & Web Terminal, Mosquitto, Remote Backup.
Those that don't start are: esphome, file editor, whisper, openwake word.

What type of installation are you running?

Home Assistant Supervised

Which operating system are you running on?

Debian

Which add-on are you reporting an issue with?

File Editor, Whisper, OpenWakeWord, ESPHome builder

What is the version of the add-on?

5.8.0 (for FE), 3.0.1 (for Whisper)

Steps to reproduce the issue

Running:

$ sudo ha addons start core_whisper --log-level Debug
DEBU[0000] Adding homedir to searchpath                  homedir=/root
INFO[0000] No configfile found
DEBU[0000] Debug flags                                   apiToken=edited cfgFile= endpoint=supervisor logLevel=Debug noProgress=false rawJSON=false
DEBU[0000] addons start                                  args="[core_whisper]"
DEBU[0000] [GenerateURI]                                 base=supervisor command="{slug}/start" section=addons
DEBU[0000] [GenerateURI] Result                          uri="http://supervisor/addons/{slug}/start" url="http://supervisor/addons/%7Bslug%7D/start" url(string)="http://supervisor/addons/{slug}/start"
Error: Post "http://supervisor/addons/core_whisper/start": context deadline exceeded (Client.Timeout exceeded while awaiting headers)

System Health information

There's no repair (sorry translation from French)

Anything in the Supervisor logs that might be useful for us?

2025-11-13 18:55:29.894 WARNING (MainThread) [supervisor.addons.addon] Timeout while waiting for addon File editor to start, took more than 120 seconds
2025-11-13 18:55:30.783 WARNING (MainThread) [supervisor.addons.addon] Timeout while waiting for addon Piper to start, took more than 120 seconds
2025-11-13 18:55:37.353 WARNING (MainThread) [supervisor.addons.addon] Timeout while waiting for addon Advanced SSH & Web Terminal to start, took more than 120 seconds
2025-11-13 18:55:43.402 WARNING (MainThread) [supervisor.addons.addon] Timeout while waiting for addon openWakeWord to start, took more than 120 seconds
2025-11-13 18:55:44.856 WARNING (MainThread) [supervisor.addons.addon] Timeout while waiting for addon Remote Backup to start, took more than 120 seconds
2025-11-13 18:55:57.962 WARNING (MainThread) [supervisor.addons.addon] Timeout while waiting for addon ESPHome Device Builder to start, took more than 120 seconds
2025-11-13 18:56:31.246 WARNING (MainThread) [supervisor.addons.addon] Timeout while waiting for addon Frigate (Full Access) to start, took more than 120 seconds
2025-11-13 18:56:36.198 WARNING (MainThread) [supervisor.addons.addon] Timeout while waiting for addon Whisper to start, took more than 120 seconds
2025-11-13 18:56:41.398 WARNING (MainThread) [supervisor.addons.addon] Timeout while waiting for addon Mosquitto broker to start, took more than 120 seconds
2025-11-13 18:56:41.399 INFO (MainThread) [supervisor.backups.manager] Restore 6f8b5740 starting stage await_home_assistant_restart
2025-11-13 18:56:41.462 INFO (MainThread) [supervisor.homeassistant.module] Update pulse/client.config: /data/tmp/homeassistant_pulse
2025-11-13 18:56:42.887 INFO (MainThread) [supervisor.docker.homeassistant] Starting Home Assistant ghcr.io/home-assistant/qemux86-64-homeassistant with version 2025.11.1
2025-11-13 18:56:42.887 INFO (MainThread) [supervisor.homeassistant.core] Wait until Home Assistant is ready
2025-11-13 18:56:49.130 INFO (MainThread) [supervisor.resolution.evaluate] Starting system evaluation with state freeze
2025-11-13 18:56:49.131 INFO (MainThread) [supervisor.resolution.evaluate] System evaluation complete
2025-11-13 18:56:52.908 INFO (MainThread) [supervisor.homeassistant.api] Updated Home Assistant API token
2025-11-13 18:56:52.910 INFO (MainThread) [supervisor.homeassistant.core] Home Assistant Core state changed to APIState(core_state='NOT_RUNNING', offline_db_migration=False)
2025-11-13 18:57:12.938 INFO (MainThread) [supervisor.homeassistant.core] Home Assistant Core state changed to APIState(core_state='RUNNING', offline_db_migration=False)
2025-11-13 18:57:12.938 INFO (MainThread) [supervisor.homeassistant.core] Detect a running Home Assistant instance
2025-11-13 18:57:12.981 INFO (MainThread) [supervisor.backups.manager] Partial-Restore 6f8b5740 done
2025-11-13 18:57:55.911 WARNING (MainThread) [supervisor.addons.addon] 5c53de3b_esphome is already running!
2025-11-13 18:58:17.205 WARNING (MainThread) [supervisor.addons.addon] core_whisper is already running!
2025-11-13 18:59:55.918 WARNING (MainThread) [supervisor.addons.addon] Timeout while waiting for addon ESPHome Device Builder to start, took more than 120 seconds
2025-11-13 19:00:17.206 WARNING (MainThread) [supervisor.addons.addon] Timeout while waiting for addon Whisper to start, took more than 120 seconds
2025-11-13 19:00:29.810 WARNING (MainThread) [supervisor.addons.addon] core_whisper is already running!
2025-11-13 19:00:48.757 WARNING (MainThread) [supervisor.addons.addon] 5c53de3b_esphome is already running!
2025-11-13 19:02:29.818 WARNING (MainThread) [supervisor.addons.addon] Timeout while waiting for addon Whisper to start, took more than 120 seconds
2025-11-13 19:02:48.758 WARNING (MainThread) [supervisor.addons.addon] Timeout while waiting for addon ESPHome Device Builder to start, took more than 120 seconds
2025-11-13 19:03:29.462 WARNING (MainThread) [supervisor.addons.addon] core_whisper is already running!
2025-11-13 19:05:29.468 WARNING (MainThread) [supervisor.addons.addon] Timeout while waiting for addon Whisper to start, took more than 120 seconds
2025-11-13 19:05:49.960 INFO (SyncWorker_9) [supervisor.docker.manager] Stopping addon_core_whisper application
2025-11-13 19:05:54.463 INFO (SyncWorker_9) [supervisor.docker.manager] Cleaning addon_core_whisper application
2025-11-13 19:06:23.573 INFO (MainThread) [supervisor.docker.addon] Starting Docker add-on homeassistant/amd64-addon-whisper with version 3.0.1
2025-11-13 19:07:53.600 WARNING (MainThread) [supervisor.addons.addon] core_whisper is already running!

Anything in the add-on logs that might be useful for us?

$ sudo docker ps
CONTAINER ID   IMAGE                                                       COMMAND               CREATED          STATUS                    PORTS                                                                                                                                                                                                                                       NAMES
dd1fb2bb1582   homeassistant/amd64-addon-whisper:3.0.1                     "/init"               15 minutes ago   Up 15 minutes (healthy)                                                                                                                                                                                                                                               addon_core_whisper
bffc5775a3a8   ghcr.io/home-assistant/qemux86-64-homeassistant:2025.11.1   "/init"               25 minutes ago   Up 25 minutes                                                                                                                                                                                                                                                         homeassistant
09674bfb096b   homeassistant/amd64-addon-mosquitto:6.5.2                   "/init"               27 minutes ago   Up 27 minutes             0.0.0.0:1883-1884->1883-1884/tcp, [::]:1883-1884->1883-1884/tcp, 0.0.0.0:8883-8884->8883-8884/tcp, [::]:8883-8884->8883-8884/tcp                                                                                                            addon_core_mosquitto
5abf842b9386   ghcr.io/blakeblackshear/frigate:0.16.2                      "/init"               27 minutes ago   Up 27 minutes (healthy)   0.0.0.0:1984->1984/tcp, [::]:1984->1984/tcp, 8554/tcp, 0.0.0.0:5006->5000/tcp, [::]:5006->5000/tcp, 0.0.0.0:8556->8555/tcp, 0.0.0.0:8556->8555/udp, [::]:8556->8555/tcp, [::]:8556->8555/udp, 0.0.0.0:5005->8971/tcp, [::]:5005->8971/tcp   addon_ccab4aaf_frigate-fa
8b6aa775140c   ghcr.io/esphome/esphome-hassio:2025.10.5                    "/init"               27 minutes ago   Up 27 minutes                                                                                                                                                                                                                                                         addon_5c53de3b_esphome
079fb13e8d26   homeassistant/amd64-addon-openwakeword:2.1.0                "/init"               28 minutes ago   Up 28 minutes (healthy)                                                                                                                                                                                                                                               addon_core_openwakeword
2afe015f1a2e   ghcr.io/hassio-addons/ssh/amd64:22.0.0                      "/init"               28 minutes ago   Up 28 minutes                                                                                                                                                                                                                                                         addon_a0d7b954_ssh
ad8b044459af   homeassistant/amd64-addon-piper:2.1.1                       "/init"               28 minutes ago   Up 28 minutes (healthy)                                                                                                                                                                                                                                               addon_core_piper
0d748d1e5e15   homeassistant/amd64-addon-configurator:5.8.0                "/init"               28 minutes ago   Up 28 minutes (healthy)                                                                                                                                                                                                                                               addon_core_configurator
a7c65da654d2   ghcr.io/home-assistant/amd64-hassio-multicast:2025.08.0     "/init"               2 hours ago      Up 2 hours                                                                                                                                                                                                                                                            hassio_multicast
7e03a3797e65   ghcr.io/home-assistant/amd64-hassio-audio:2025.08.0         "/init"               2 hours ago      Up 2 hours                                                                                                                                                                                                                                                            hassio_audio
9badf8325009   ghcr.io/home-assistant/amd64-hassio-dns:2025.08.0           "/init"               2 hours ago      Up 2 hours                                                                                                                                                                                                                                                            hassio_dns
0a60c946934b   ghcr.io/home-assistant/amd64-hassio-cli:2025.09.0           "/init"               2 hours ago      Up 2 hours                                                                                                                                                                                                                                                            hassio_cli
ac3d7c54201c   ghcr.io/home-assistant/amd64-hassio-supervisor:latest       "/init"               23 hours ago     Up 2 hours                                                                                                                                                                                                                                                            hassio_supervisor
93993dc27123   ghcr.io/home-assistant/amd64-hassio-observer:2025.02.0      "/usr/bin/observer"   7 months ago     Up 2 hours                0.0.0.0:4357->80/tcp, [::]:4357->80/tcp


There's absolutely no error in the core (ha core logs is all green)
The addons logs are good, for example, Whisper contains:

s6-rc: info: service s6rc-oneshot-runner: starting
s6-rc: info: service s6rc-oneshot-runner successfully started
s6-rc: info: service fix-attrs: starting
s6-rc: info: service fix-attrs successfully started
s6-rc: info: service legacy-cont-init: starting
s6-rc: info: service legacy-cont-init successfully started
s6-rc: info: service whisper: starting
s6-rc: info: service whisper successfully started
s6-rc: info: service discovery: starting
INFO:__main__:Ready
[19:06:27] INFO: Successfully send discovery information to Home Assistant.
s6-rc: info: service discovery successfully started
s6-rc: info: service legacy-services: starting
s6-rc: info: service legacy-services successfully started

Additional information

I don't understand how the esphome container's addon is supposed to speak to the core, since there's no port mapping in the docker ps command. I do see that this container has a /var/run/esphome.sock file (I guess it's a unix socket), but I don't see how HA can make use of it since it's in a different container. There's no volume mapping for /run or /var/run for this container either.

Also, if I do:

$ sudo ha addons info 5c53de3b_esphome

I get this in the output:

machine: []
name: ESPHome Device Builder
network:
  6052/tcp: null
network_description: null
options: {}
privileged: []
protected: true

I don't know if it's expected, I couldn't find anything with google with the expected output. I'm dubious about the network part above. The port 6052 isn't mapped in the docker container.

Same for core_whisper:

machine: []
name: Whisper
network:
  10300/tcp: null
network_description: null

Addons that have started, like SSH, returns this instead:

name: Advanced SSH & Web Terminal
network:
  22/tcp: 2222
network_description:
  22/tcp: SSH server port (make empty to disable)

and Frigate:

machine: []
name: Frigate (Full Access)
network:
  1935/tcp: null
  1984/tcp: 1984
  5000/tcp: 5006
  8554/tcp: null
  8555/tcp: 8556
  8555/udp: 8556
  8971/tcp: 5005
network_description:
  1984/tcp: go2rtc API
  5000/tcp: Unauthenticated Web interface (not required for Home Assistant Ingress)
  8554/tcp: RTSP Restream
  8555/tcp: WebRTC over tcp
  8555/udp: WebRTC over udp
  8971/tcp: Authenticated Web interface

Also, restoring a backup from yesterday didn't solve the issue either.

[sheminasalam]

same issue for me. any solution?

[X-Ryl669]

I've done this and it resolved the situation although I'm 90% sure it's bullshit. Let me know:

1. Stopped the machine via the HA's interface (not via SSH), click on the Setting page, top right button, "Restart HA", "Advanced", "Power off machine"
2. Restarted the machine (pressing its power button)
3. HA core didn't start on boot (don't know why, it usually does)
4. Addons' containers started on boot (don't know why they did, since the core didn't)

Typed this:

1. sudo ha core start
2. sudo ha addons restart core_mosquitto
3. sudo ha addons restart a0d7b954_ssh --log-level=debug

At that point, all the addons container have restarted as seen in the logs, but HA doesn't still don't see them

1. sudo ha supervisor info

(it doesn't report the addons as started, state:unknown, but it reports some as state:error). Yet sudo docker ps show the containers as healthy

1. sudo ha supervisor --log-level=debug restart
2. sudo ha supervisor --log-level=debug reload
3. sudo ha supervisor --log-level=debug logs
4. sudo ha supervisor --log-level=debug repair

Notice that I did try to repair previously and it always said that it succeeded but it didn't change anything.
Yet previously, it took a lot of time, this time it was very fast. The logs didn't contain anything special

Then I tried to manually stop the addons that were still running (and start them afterward), instead of issueing a restart I did a stop, waited until it said it was successful, then a start:

1. sudo ha addons stop a0d7b954_ssh
2. sudo ha addons start a0d7b954_ssh

This magically "reconnected" everything. All addons (not only SSH) were seen by the core and the supervisor. I clearly don't understand why it worked.

[Egorjiv]

https://github.com/home-assistant/addons/issues/4228#issuecomment-3533533489 Thank you, this method helped me. But it only works until the next restart of Home Assistant. Why do add-ons fail to launch during startup and when Home Assistant is restarted? This method seems to work. It works not only when the system is shut down, but also when rebooting the system from Home Assistant itself (not via SSH). I would like a permanent solution to this problem. I would be very grateful if you could resolve this issue with Home Assistant restarting and if you could share your solution.

[erdosip]

With Debian 13 (trixie) and supervised - I have the same issue, after updating the docker-ce and containterd.io to the recent v29+. Now they're in hold...

[fazy(MoPeHOME:0)] <> dpkg --list |grep conta
hi containerd.io 1.7.29-1-debian.13-trixie amd64 An open and reliable container runtime
hi docker-ce 5:28.5.2-1-debian.13-trixie amd64 Docker: the open-source application container engine
hi docker-ce-cli 5:28.5.2-1-debian.13-trixie amd64 Docker CLI: the open-source application container engine
[fazy(MoPeHOME:0)] <>

[alanmilinovic]

In my case all addons stopped working and this is what helped me.

#4229 (comment)

[erdosip]

It looks like it will remove not the cache, but the addon specific configurations (mqtt password, nodered config, etc... I may don't want to try this :D

[alanmilinovic]

Well you can do a backup and restore only addons configuration.

[erdosip]

But that's it: this folder only holds the configuration for most of the plugins. The ssh have an extra ssh key, the mariadb have the database files, the piper some language files, and whisper + vscode some extra stuff and log. This folder is NOT the cache folder, but the configuration one. If I backup and copy back almost everything, what is the point? The /usr/share/hassio/addons/core have the common component root folders (etc usr dirs) and the /usr/share/hassio/addons/git folder has the extra stuff (z2m, esphome, go2rtc, whisper repos)

[X-Ryl669]

I think the issue is the order the containers are starting. Like @erdosip, I did an system update before the issue appeared. The hassio-agent service seems to start all containers that were running before stopping. And the addons starts before the core (and/or) the supervisor. Thus the supervisor can't start them up and doesn't stop them either. So it's struck. Doing a restart fails too (because, I guess the supervisor thinks the addons aren't started). So you have to manually stop them before it work. Then it's able to start from scratch and it works.

This is a wild guess about what's happening.

[alanmilinovic]

I didn't understood what you mean with "stop them before it works".

Anyways in my case I uninstalled all of them, rebooted, installed just one (File editor) and tried to start, so not sure if the order could be the issue, also with uninstall a new container is created.

[erdosip]

I've started to read the docker changelog btw. The statement: "This release includes several breaking changes and deprecations. Review the release notes carefully before upgrading." is let's say: forshadowing at least? :D

So I'm trying to figure out, which change broke what exactly, and since I'm a ~network guy, and I know a few about iptables and kernel package processing, I started my investigation there. Since the containers (stick with esphome) starts successfully, and the supervisor logs says that it's already running, but after 120 sec, it goes to timeout. Is there any possibility, that some nftables/docker change broke something, and the containers cannot communicate with each other? Because of this, I've tried to enable ip.forward from sysctl, but no luck so far: "nftables: Docker will not enable IP forwarding on the host. If forwarding is needed by a bridge network, but not enabled, daemon startup or network creation will fail with an error. You must either enable forwarding and ensure firewall rules are in place to prevent unwanted forwarding between non-Docker interfaces. "

PS: Maybe this whole idea is wrong, than sorry for that :)

[X-Ryl669]

@alanmilinovic : When you try to start the addons in the "broken" state, the supervisor log states: This addon is already started (see the first post). Yet, the supervisor still consider its not, in the UI and in the CLI, thus brokage happens between the actual state and the state managed in the UI.

If you try to restart an addon, via ha addons restart, it doesn't work either. Here, I guess that since the addon isn't considered started, it's not actually stopped. However, if you stop the addon via ha addons stop it does seem to stop it (the container is stopped). Then starting it again seems to work, the supervisor finally register it as started and it's recognized in the UI.

When you reinstall the addon, you're actually stopping the container for removing it. That's why it work, but it's painful. And it won't survive a machine restart either since on reboot, the container will be started earlier than the supervisor and/or the core and it'll fail again. If you're running frigate, you don't want to uninstall the addon, since it's a lot of data to backup/restore so it's not a solution.

[X-Ryl669]

@erdosip : In my first post, I've written what the network port where mapped in the failed case. Today the addon is working for me (following the process above) and redone the info command for esphome:

network:
  6052/tcp: null

So, actually, no the port isn't required to be mapped on the host for it to work. I guess those containers are running in the same network namespace in docker so they don't need any mapping to communicate between each other. A HA dev could confirm this?

[alanmilinovic]

@X-Ryl669 I tried manual stopping and starting also reinstall of the addon and in any case it didn't worked for me. Maybe I had a different issue.

[X-Ryl669]

Did you do the supervisor restart/reload/repair dance before?