Won't Boot - Proxmox 8 - Slot A (OK=1 TRY=1)

Hey,

My setup’s been working without issue until this morning when I did a quick check config within Home Assistant, all green so I restarted.

Home Assistant didn’t appear to come back for awhile so I checked Proxmox 8 and all I see in the console is:

Proxmox Slot A (OK=1 TRY=1)

After trying a reboot I get the same thing. I noticed a screen flashed up so by toggling the arrow keys I got this menu:

I’ve no idea if a Home Assistant OS update or some Proxmox update has caused this issue. Is anyone else getting it as of today? Also please note I’m very much a beginner with Proxmox, I know enough to get myself running but that’s about it. So please add additional details if you can around things I could look at doing.

Additional info on Hardware and Proxmox setup:

Hardware -
Device: Intel NUC 11
CPU: Intel Core i7-1165G7
RAM: 32 GB
Storage: 1TB Kingston PCIe SSD

Many Thanks

1 Like

We need to see hardware and options.

I just added additional screenshots with Hardware info contained. Do you need more than that?

Thank you.

You have provided Summary only. With VM selected on the left click on Hardware and Options on the right.

Hardware screenshot:

Ar, sorry about that. Here you go:
image

Try to change machine to q35. If that does not do the trick it is most likely a UEFI issue. Have you played around with UEFI. If not some sort of corruption. With VM shutdown make a full clone of the VM and try to start the new VM.

One more suggestion. Change memory to 4G and max mem to 8G, although both values are too high, I would suggest 2G/4G.

Before I change this, I just wanted to ask why it might be that it’s been working for a month+ and now suddenly it does not given I’ve changed nothing in the general setup.

Feels like changing a setting like this would only apply if something major had changed given it was working? Or maybe that’s not true for some reason?

Computers are logical machines. Your system stopped responding and the issue is either something you changed and are not aware of (in most instances this is the case even if people are adamant they changed nothing) or some sort of corruption.

I have suggested changes which are beneficial to any HAOS setup. Up to you to apply or not, as they are suggestions.

If you feel uncomfortable just follow my last suggestion to clone the VM and try to start the cloned VM.

K I’ve just tried a clone and the result was the same.

I then change the machine setting for the clone from i440fx to q35. The result was for the VM to hang on this screen:

K things have become stranger now. I rebooted the entire system and while I’m still stuck on my original VM at this screen:

I can now access Home Assistant. :confused:

Is this ok? Doesn’t feel ok :laughing:

Definitely this is not right. To me it seems disk or memory corruption. I see you are running ZFS and the issue might be ZFS might be starved of resources, although you seem to have enough RAM. Have you tried to get HAOS on memory 2G/4G, instead of 8G.

One other thing you could do is follow my guide to install HAOS from scratch. If it goes through OK then the VM is corrupt. If installation does not succeed then most likely ZFS issue.

I did change the RAm to 4GB, that didn’t seem to help so I put it back. I’m not sure I understand why having 8GB of RAM could cause any issues. The reason I did it was because I’m using Frigate and I wanted to ensure it had more than enough RAM to handle the 8 cameras it runs.

While your HAOS is accessible try a backup, so at least you can attempt to restore on a new HAOS install.

In terms of Home Assistant, I have a samba backup that runs every night to my NAS. So I’ve got a full back of Home Assistant to recover to if needed. Not that its ever as simple as that :laughing:

If you have corrupt RAM then not accessing those portions might point us to come closer to the source of your problems. HAOS does not really need this much RAM. You ZFS needs RAM more than HA does and you have allocated 8G to HAOS which is going to waste.

Please take a bit of time to read my guide and some of the fog might go away.

It is better to run Frigate outside HA. I have a guide almost ready which will hopefully go live next week. The secret to a reliable HA is to run everything possible in a KVM or LXC outside HAOS and then communicate using VMs. This is the beauty of Proxmox and it provides more flexibility then running services inside HA.

The issue here was the PCI device “TigerLake-LP GT2 [Iris Xe Graphics]”. I booted the VM without this device and all is back to normal.

1 Like

The PCI Device, got it.:+1: