HELP - GUI won't load after failed update and I'm stuck!

1

Hello dear community, I’m really hopeful someone can help save my config…

I’m running HA in a VirtualBox VM on an old Mac Mini. Everything’s been working smoothly until last week when I (foolishly) clicked ‘Update’ in the UI and walked away. I returned to the room some minutes later and the entire MacOS had rebooted! After logging into MacOS I saw the dreaded prompt “Your computer restarted because of a problem…” . This made me immediately think "Oh, **** ", and sure enough, HA hasn’t been accessible since.

The VM boots up and looks normal, I get the big banner and no obvious errors along the way, but the UI both via browser and mobile app do not load.

In the VM I see:
OS Version: Home Assistant OS 9.4
Home Assistant Core: 2023.2.4

I started with the most basic troubleshooting step of a graceful reboot of both the VM and the MacOS host. No change. I tried running some basic CLI commands in the HA, telling it to look for updates, to inspect its config, and no luck.

If I list the backups I see the ‘partial’ backup the system made before the failed update, but I cannot to restore that. I get an immediate response ‘Processing… Done. Error: Unknown error, see supervisor’, and if I use ‘supervisor logs’ I see the last entry that says ‘ERROR (MainThread) [supervisor.backups.manager] us is only a partial backup!’ . Unfortunately I don’t know how to view the logs completely, only the glimpse that I can see on the screen.

I do have a backup listed from 2021 as ‘full’, but I’m nervous to restore that and loose a lot of my config I’ve made since then.

This is where I’m stuck. I’ve done a bunch of Googling and unfortunately the few leads that I find are beyond my technical ability. Can anybody provide a real “hand holding” response with some ideas? In the end, I think I’m technical enough to get myself into trouble, but not get out again :slight_smile:

My only goal is really to save my config (integrations, automations, dashboards, etc.) by any means. I would gladly create a new installation (spin up a new VM?) if I could copy out those files from the old one…I just don’t know how :frowning:

Thank you all in advance for any ideas!!

2

So, I will start with the very basic questions…

  1. Is the HA system responding to a ping from the machine where you are running the browser?
  2. What is the error message on your browser when you try to access that IP address on port 8234?
  1. Have you tried to restore this backup in a fresh new HA installation or just on the old installation where you are troubleshooting?
3

Thanks a lot for the fast response.

  1. Good question, yes, I just pinged the IP I see on the HA banner and it’s responding like normal
  2. Browser error is: “This site can’t be reached refused to connect”
  3. Only the old installation, I don’t know how to get at the backup files themselves to copy them into a new installation…
4

Minor updates:

  1. Thanks to this super helpful thread, I learned to pagefy the logs page. Now I see a whole series of Warnings relating to Zigbee2MQTT addon, all fitting this pattern:
    WARNING (SyncWorker_2) [supervisor.addons.options] Option <option_here> does not exist in the schema for Zigbee2MQTT (45df7312_zigbee2mqtt)
    For fun I unplugged my Conbee stick from the Mac and restarted the VM to see if that did anything, and no change.

  2. I noticed a new screen flash by quickly when booting up the VM and managed to grab a screenshot to read what it said. It’s a menu list that reads “GNU GRUB version 2.04” with four options listed:

    • Slot A (OK=1 TRY=0)
    • Slot B (OK=1 TRY=2) ← this one is highlighted by default so I assume it’s selected
    • Slot A (rescue shell)
    • Slot B (rescue shell)

    ^^I have no idea what these mean, but just discovered this and wanted to share in case it’s relevant at all!

Thanks again for any ideas.

5

Any further ideas i could try here? I haven’t pushed any further into troubleshooting, but the two paths i see are potentially:

  1. Try to install an SSH Client on the existing server, get the config files copied out, and spin up a brand new VM to copy those files into

  2. Look further into those Zigbee2MQTT errors i was seeing. I use that addon to integrate all my ZigBee devices, so it is eventually important to my system, but maybe i could remove it from the CLI to see if it boots and reinstall it later if that works?

Any other ideas? Option 1 above I’d appreciate any advice on how to do this, I wouldn’t know how to install the ssh client onto the existing server…

Thank you all!