My Home Assistant installation has become completely unstable and nonfunctional lately. I had issues with my Orbi wifi (satellites have both died) so I had to move the HA Yellow to plug into the base unit. Now the system is not functioning at all, it seems to be running the CM4 full blast constantly and restarting itself–that is when I can get logged in at all.
The leds are all on. Red is solid, power is good. Yellow is constantly blinking (not the heartbeat), and green is constantly flickering. I’m not sure what that’s trying to indicate but it’s clearly not good.
I can ping the device with little latency. I can sometimes get on SSH via MobaXterm, but that keeps getting killed and is often very slow to respond. When I can get the web interface to respond at all, I can get to the base Settings menu page, but any sub pages usually get an “Error while loading page ” message. Trying to load a Lovelace page gets nothing but a (!) icon with a Reload UI button, which does nothing; or I get the underlying YAML with errors because none of the custom controls are getting loaded.
Is there anything to be done with this short of a full factory reset and start from scratch? A lot of this seems to have started with the 2024.5 updates for me ( I have Core 2024.5.5 and Frontend 2025001.1 installed-it will show me that much right now).
I would try a factory reset if I have a good backup.
https://yellow.home-assistant.io/guides/factory-reset/
If you are still getting the same behavior after a reset / restore, there might be an issue with one of your integrations. See this to troubleshoot a misbehaving integration:
I do keep regular backups. OK, factory reset it is. I was hoping that wasn’t needed but oh well.
I was able to get it to where I could restart in safe mode. After a couple of minutes it restarted itself again. Seems completely borked at this point.
Restarting while in safe mode might be hardware issues. I’m not familiar with the Yellow hardware but you might want to see some of these suggestions
We will see. I’ve done the factory reset and am running a restore from a full backup from a couple of days ago. The hardware has been behaving so far.
The hardware is the Yellow kit, with the spec power supply included so I know that’s not the issue–it’s been running well for a while now.
Corrupt files from a crash is a distinct possibility, especially since it’s had a few power off reset events.
Power supplies do fail / go out-of-spec.
True. I’d expect it to last longer than I’ve had it. Restore just finished, and things are coming back online.
Interesting. Just poking around Settings pages, About is showing more info. It’s now displaying Core, Supervisor, OS, and Frontend versions. It was missing half that before resetting.
It starts out ok, then goes right back to the previous behavior. I suspect something seriously corrupt in this last backup. I have one from about a month ago I’ll try. Failing that, I’ll have to try exporting automations and dashboards and start from scratch.
I see the correct yellow heartbeat LED, then when the system is freaking out it’s back to just the continuous blinking. When it’s doing that, the hardware isn’t being recognized any more.
This will be about the third or fourth full rebuild I’ve had to do.
It seems there is something seriously corrupted in my setup for a while. I can restore the older backup, but as soon as I attempt to update anything it goes back into this borked state.
Looks like I’m going to have to take notes & export what I can, then rebuild from a fresh factory reset. Oh well, I guess it’s another opportunity to do some deep housecleaning of the installs.