Have the same issue today. After updating supervisor to the latest version.
Same issue today, getting connection error when trying to do ha core rebuild via hass-cli
ssl.SSLError: [SSL: WRONG_VERSION_NUMBER] wrong version number (_ssl.c:1108)
tried with --insecure arg, didn’t help too… still digging.
Okay I might have fixed it, at least this was the case for me. I used Portainer for managing docker containers via GUI.
Digging through the logs of homeassistant container I noticed that it could not connect to supervisor at ip 172.30.32.2. Someone had also mentioned network discrepancies in this thread. The homeassistant container had ENV variables HASSIO and SUPERVISOR set to the same ip
In hassio_supervisor container I had noticed the ip of hassio network being different from 172.30.32.2. Bizarrely it was set to 172.30.33.2 and a gw set to 172.30.32.1. At this point it was obvious there’s something wrong with the network settings.
So I stopped hassio_supervisor container, removed it from hassio network, recreated the container and then added hassio network back. Restarted the container and it got the correct ip 172.30.32.2.
After that I recreated a couple of add-on containers which wouldn’t start, and finally after many hours I had everything back up and running!
Hope that will help someone.
6 Likes
How did you change it?
I use UNRAID as the OS. There you can specify the settings of your docker easily.
As wrote @mihma…
- stop supervisor container
- leave hassio network
- recreate container
- join to hassio network
…than container should have assigned the same network as definned in homeassistant container (172.30.32.2 in our case)
2 Likes
227 version seems a bit broken, auto updated and then stopped without the ability to start correctly.
Had to recreate the container few times before started to work again.
I really don’t like that auto update feature of supervisor: isn’t there a way to do it manually like home assistant core?
1 Like
Have this issue now. I wish i understood how to do all that…
Edit: Nevermind, figured it out with google and working again following these steps
i cannot ssh into the homeassistant container. i just have ssh access to the buster host system.
can you explain how i am supposed to run the ‘ha core rebuild’ command?
EDIT: nvm, i was able to access the terminal via visual studio addon and run it from there.
Same message and problems here. Setup Home Assistant 0.110.3 on RPi3+.
Did a few warm resets, a few cold resets but my HA op RPi3+ was really dead now and image looked as in a loop as led’s flashed but I couldn’t ssh and ping the RPi…
Did a complete rebuild from scratch with a new 32GB Samsung SDHC card and a restore from a backup (have “Hass.io Google Drive Backup” running daily backups luckily). Had to do a few more restore things as my zigbee2mqtt didn’t startup. Now everything runs again on Home Assistant 0.110.3
I’m still not sure what has happened (I didn’t do any installation things last days) but perhaps my memory card wasn’t stable anymore.
Thankyou for this advice, unfortunately mine starts up again on the old incorrect network. Not sure why.
Add me to the list of RPi users getting the
Unable to load the panel source: /api/hassio/app/entrypoint.js.
out of the blue when trying to access Supervisor, File Editor, etc
Really don’t want to be pulling the plug on the RPi on a regular basis.
Same here with issues using Rpi though I didn’t realise auto update was a thing now (would love to stop that). Despite Home Assistant breaking during updates so, so, so many times, I’ve still not learnt my lesson to just leave it alone and stop trying to update it 
randomusername said:
Despite Home Assistant breaking during updates so, so, so many times, I’ve still not learnt my lesson to just leave it alone and stop trying to update it
In my experience so far, this issue is not related to updates. It just happens every 2-4 weeks… then it’s time to re-install everything again. Fortunately, the restore-from-backup seems to be working.
I’m undecided about the whole system. The UI is nice, but a rPi with MQTT, Node-Red, and motioneye Might improve reliability tremendously and provide all the functionality minus the UI and Nabu Casa’s remote access.
-Mike
Just for data - I’ve had this happen to me on two different installations in two different locations, when the Supervisor updated recently. Both are NUCs running HA Supervised on Ubuntu. Both needed a full shutdown and cold reboot to get back to normal. This was a major issue for the remote installation, as it had to wait until someone was able to go there and deal with the issue on location.
I’ve just changed one of the installs to HA (with HassOS) on Proxmox, so I’ll be interested to see if it happens again.
This happend to me the second time, the first time on my RPI 3B and the second time today running on my nas.
Fortunately I found a solution, which was working:
I’m using portainer, stopped the homeassistant container, checked the access_token of the home-assistant.json and deployed the container again. 
1 Like
Same song for RPi and v111.4 here. I am experiencing the /api/hassio/app/entrypoint.js. errors and have been reading to try and find a solution. I decided to completely reload from scratch a couple of days ago and been working through SSL problems and error messages, to the point that I’m trying to stay motivated with HA. I love the vision, but the stability and integrity with so many external variables, just makes this difficult for us that have day jobs. Very sad, as I want to learn all the things HA will be able to do, but I’m stuck in managing it’s basic interaction.
If anyone has advice on “best practices” to avoid so much heartache, with very little functional results, I’m all ears.
1 Like
This started to happen to me this morning…power cycle the Raspberry pi4 and now it won’t come up at all…
This is started to happening to me also since v109. Now I’m using v111.4 and it’s happening every 2 days. It doesn’t come up with a power cycle, I have to restore it from snapshot everytime. It is practically impossible to use the HomeAssistant anymore. Doing more recoveries than I use the device for its own purposes. PLEASE FIX THIS PROBLEM ASAP! For now I wont use it anymore…
1 Like
Pulling the power and letting it reboot fixed the issue for me.