Looks like I was fortunate. Was running 14.2 on my ODROID N2+. Hung for long time, so flipped KVM to see HA monitor, & it was showing lots of errors & retries, but obviously not running system yet.
EDIT: It was trying (& failing) to do PXE boot.
So I did a power cycle. It came up fine, in normal amount of time. And is now at 15.0.
Same N2+ and the same thing happened I got lucky as well.
PDA: I have No idea what you are saving there. I have a dead Pi5 after trying to upgrade to v15 OS and its literally thousands of miles away from me.
Can you clarify what you mean in your instructions, or point me to where I can find out how to do things like:
Thanks
@lianyu ,
The “thousands of miles away” part is an issue.
The PI is actually not dead, just the OS is not completing properly.
You issue is that it’s failing before there’s any GUI / SSH connection available, so you’ll have to find another way to communicate with the server.
Isn’t there a local person who can do that for you (or even just plug-in a Linux USB, so you can do the dd part)?
Oh no, I was afraid you were going to say that. So even the Terminal & SSH Add-On would not have helped (didn’t have it installed anyway).
There is someone who could unplug/plug it in again, but beyond that would be like doing heart surgery over the phone, and I don’t even know how to do the ‘Linux USB’ thing if I was sitting right there.
I can’t believe HA is so fragile 
@lianyu ,
I have a workaround for you:
After 3 failed boots, HA automatically switches boot slot.
You need that someone to plug it out and then in and wait 5 minutes.
You can check if the server is up, and if not, repeat.
After 3-4 loops, it should be up again.
This will not fix your issue (you will still have a bad slot), however, the server would be working again.
I’m looking at my local HA Pi5 server in Terminal/SSH and not seeing ANY /dev/sda(s) at all.
No /dev/sda1 or 2 or 3 or 4 of 5 or anything.
The list goes from /dev/rtc0 to /dev/serial
What am I missing?
Thanks
Thank you! I will cross my fingers and give that a try.
I will also dig around on how to fix a bad slot (providing I can find those ‘slots’ !)
Once you have the server running with the workaround, install the SSH add-on.
Type the command “ha os info”.
You’ll see the slots.
That depends on your storing device (ssd / hdd / usb / sd).
You could try the command “ls -l /dev/disk/by-id/” to list all your disk devices.
Cool, they are mmcblk0p0 to p8 (on my local HA server)
Thanks!
Clicked upgrade to then while waiting opening google feed and finding this thread. Got half a heart attack…luckily my upgrade went through. Not running on a Pi, though.
Unfortunately, my HAOS cannot be accessed (running in PI5) after the attempted upgrade to 15
Tried rebooting (unplu power) a couple of times, but still no luck.
Will be following this page for slutions…
In my setup on a dell wyse I have the same issue, I selected the second option on boot screen to get back to 14.2. (I had to connect monitor and keyboard to the machine)
But on proxmox VM it went through allright without problems.
I have shut down my RPI 5 by pressing the button until the led became red and often unplugged it afterwards. I guess this counts as a hard reset. It never fell back to the other slot despite doing it many times.
After rebooting a couple of times, still nothing happened.
Then, I noticed that the LED was solid red.
I replaced the PoE splitter I have been using for over a year now with USB-C power cable, LED turned green and HAOS boots.
Does 15.0 have different power requirements all of a sudden? HAOS is on 15.0 and seems stable.
If you have a keyboard connected to the PI, you could try to switch slots even without a monitor connected.
Turn the PI on, and keep tapping many times the “arrow down” key for a minute.
This will stop the boot on the GRUB menu.
Then, tap the “up arrow” key 4 times, to get back to slot A.
Then, tap the “down arrow” key once, to get to slot B, and press “enter” to boot.
Hey everyone,
In response to the thread “Home Assistant unreachable after update to HAOS 15,” I wanted to share how I solved this with my HP T620 issue and highlight some noteworthy points from my logs afterward.
How I Solved It
Key Log Observations
STATE_CLEANING, STATE_DOCKED) that will be removed in a future release. Be prepared to update any custom code or configs relying on these constants.local_key in one of the custom integrations. If you see similar messages, verify that your keys match the devices’ actual configurations.state class 'total_increasing' together with device class 'monetary'. These settings aren’t compatible, so you’ll need to align them appropriately in your configuration.Hopefully this helps anyone dealing with similar issues after updating to HAOS 15. If you’ve encountered the same problems or have suggestions for fixes, feel free to chime in on the thread!
It would appear that the GitHub issue report has been closed with a comment of “not planned”… which I’m not going to lie is rather concerning - I fully appreciate HA is open source, and everyone working on it is a volunteer, but are users who are stuck unable to update to 15.0 going to get any form of support?
Edit - I may have been too hasty… issue #3943 has been raised as a collector for RPi issues after the original devolved into a bit of a comment thread! My apologies to the contributor team… Note to self - don’t GitHub or Forum before coffee!
Any HP T620 users here who are on HAOS 15.0 without issues?