HassOS 3.13 installation becomes corrupted on NUC after Supervisor Reboot

I recently bought an Intel NUC7i3BNH with a 256gb 2.5 inch SATA SSD to run HA.

I was able to install ProxMox with no issues. I also had previously installed Ubuntu on this machine without issues. I have disabled secure boot and upgraded the BIOS to the most recent version.

I am able to successfully create (at least at first) a HassOS VM on Proxmox via the Whiskerz script (https://github.com/whiskerz007/proxmox_hassos_install).

If I reboot the machine by using the web interface->Supervisor->System->Host System-> Reboot, the system becomes more and more corrupted with each successive reboot, to the point where it will no longer boot after roughly 3 reboots. This has happened about five times in a row (removing the VM and starting from scratch each time), and even across reinstalls of Proxmox. In the following description, each time I say “reboot”, I mean the process described above via the web interface.

The most recent time I created a VM, I noticed after the first reboot several SQUASHFS errors:
SQUASHFS error: Unable to read data cache entry [bc8bac]
SQUASHFS error: Unable to read page, block bc8bac, size 10060.

It did boot though, and the next time I rebooted, I noticed EXT4-fs errors:
EXT4-fs error (device sda8): ext4_lookup:1591: inode #148747: comm python3

Again it booted despite the errors (though it usually takes longer with the errors). The next time I tried to reboot, I ended up at the four-option boot screen. Autoboot does not work, but I was able to progress the boot attempt by selecting Boot System 0. Then a bunch of python segfault errors occur. I am able to login as root but the web interface does not go up, even after an extended period of time.

I have also tried this on the bare metal, and I think something similar is happening, though I am less able to monitor the errors messages as I have not had success loading the CONFIG/network/my-network file and I do not have a monitor near where I can plug into my router via ethernet. But after a few boots which take longer and longer, I end up just seeing the four-option boot screen and am not able to reboot without selecting a boot option after a few reboots.

I have tried this with a few different cache settings on Proxmox (no cache, write through, and direct sync) with no clear improvement. I fiddled a bit with other virtual hardware settings, but don’t really have a good idea of what settings are likely to be most fruitful, and the possible combinations are endless. I have scoured the internet for others with similar issues, and saw a few similar reports but none that had a clear resolution.
I have also tried installing on the bare metal as I mentioned above, as well as using the qcow2 image on Proxmox (Supervised Install on Proxmox)

I am about ready to give up on HassOS and just switch to HA core on a linux desktop, but would prefer to get HassOS working if possible. I can’t commit to using HassOS unless I can reboot it stably a few times, otherwise I feel like it is a ticking time bomb that is sure to die after a few power outages. Improved stability is the whole reason I am trying to upgrade from a Pi4 to a NUC.

Thank you in advance for any advice you might be able to provide.

Maybe you should try the latest HassOS beta?
Release 4.6 RC1 is available here: https://github.com/home-assistant/operating-system/releases/tag/4.6

Thank you for the suggestion. I tried, and it booted the first time, but on subsequent tries, it no longer boots. It flashes the boot menu for a split second when I press enter, but otherwise the screen is black and it does not get assigned an IP address on my network (plugged in via ethernet).

Then you will have to open an issue on GitHub

It turns out I had a bad stick of RAM. I haven’t confirmed resolution after removing the bad DIMM but I think it was likely at fault. Marking as resolved.