Boot hangs after rebuilding system

[davwheat]

Hi all,

I've been running Proxmox for a few years at home. I've finally got a proper server rack and chassis, so I opted to rebuild my system in that. Same CPU, motherboard, RAM, PCIe cards.

After this, I tried to turn on the system, but it appears to hang on boot where I can't get past this screen:

Found volume group "pve" using metadata type lvm2
Found volume group "ssd-new" using metadata type lvm2
2 logical volume(s) in volume group "pve" now active
15 logical volume(s) in volume group "ssd-new" now active
/dev/mapper/pve-root: recovering journal
/dev/mapper/pve-root: clean, 259468/54173696 files, 125075962/216664064 blocks

Removing quiet from the grub entry shows me the last kernel module it loads is dm_mod, followed by no further output.

The system hasn't hung as numlock still works. Ctrl+Alt+Del also makes the system reboot after a few seconds, and the power button appears to perform a clean shutdown too.

I have networking attached to the system, but my Ubiquiti switch doesn't seem to detect any device on the other end of either the motherboard GbE (interface is probably down anyway) port nor my 10GbE PCIe card (what I was using before).

I've tried disabling virtualisation thinking it could be something stealing the GPU I'm using for video out (there's no on board graphics).

Booting a live Debian usb works fine.

I have a spare SSD to attempt a new install, but I have no way of performing a backup of the data on the existing install to restore it into there. Is there any workaround for this in case my install has somehow completely borked itself? I don't really fancy setting up 20 VMs/CTs again...

 

[davwheat]

Bump? I got held up by post approval so this fell way down.

 

[davwheat]

Tried doing things by myself with a YOLO approach.

Reinstalled Proxmox on the new boot SSD I bought, then plugged in the old SATA boot SSD which was borked.

Obviously the LVM ids clashed as both disks had pve LVs. Used vgdisplay -v to get a list of the LVMs with their disks and creation dates, then renamed the older LV to pve-bak with vgrename, then rebooted so the automapper would do its job.

Mounted the old LV with mount /dev/mapper/pve--bak-root /mnt/pve-bak/root/ -o ro,user

Stopped pve-cluster on the new install, then copied /var/lib/pve-cluser/config.db from the old boot disk to the new one, then updated /etc/hostname and /etc/hosts with the old node name. (From https://pve.proxmox.com/wiki/Proxmox_Cluster_File_System_(pmxcfs)#_recovery)

Since all of my other drives were picked up by the new install, most CTs and VMs automatically worked after a reboot. Some I had actually left ISOs attached to them which didn't exist on the new install. After removing those, they also booted.

I've now just got a couple of CTs with disks that were on the original boot disk which now won't start as their drives can't be found. They don't show under /dev/mapper/ like the disks on the other drives I have did, so I assume that will take some more work.