Using a physical disk from another Promox instance in new Proxmox instance

R

Raging_Demon

New Member
Mar 17, 2025
4
0
1
The HDD containing my old Proxmox instance died. My LVM/thins were held on seperate HDD's. I've installed a new Proxmox instance on a new HDD, and attahced the existing HDD's containing the VM's to my new Promxox instance. How can I retreive my old VM's? I attempted to create a image of hte old HD to try and get the previuos Proxmox config files, but the disk was badly corrupted, so I doubt I can get those files back. What are my options to get my VM's/data back?
 
Here is the output of a command I tried to get to my VM's:

Bash: 
root@proxmox:~/test# vgscan
  WARNING: VG name pve is used by VGs RzfTyA-ATbh-Rnku-f21c-CsRJ-hphS-WekpZH and 6Ytt43-S18S-FGK0-uumG-JoSO-CuZj-IdDsPK.
  Fix duplicate VG names with vgrename uuid, a device filter, or system IDs.
  Found volume group "pve" using metadata type lvm2
  Found volume group "EightTB_thin" using metadata type lvm2
  Found volume group "pve" using metadata type lvm2
  Found volume group "TwelveTB_thin" using metadata type lvm2

I've also attached some screenshots in case that helps anybody with troubleshooting, thanks in advance.
 

Attachments

  • proxmox lvm error.jpg
    proxmox lvm error.jpg
    58 KB · Views: 2
  • proxmox storage.jpg
    proxmox storage.jpg
    84.5 KB · Views: 2
Okay, I don't think I will be able to get the config files back from /etc/pve/local/qemu-server (unless I try some more advanced disk recovery tools) so it looks like I'll have to try and remember the VM reosurce allocations. I'm guessing it wouldn't matter if I over-provisioned each VM (apart from HDD space, as i know that's a pain to resize once allocated), and then adjusted once they are back up and running.
Is there a guide I could follow to create the new VM's, and add them back from the existing storage pool?
 
Before doing anything I'd recommend you rename the old pve VG with vgrename VGIDHERE NEWNAME. vgdisplay should tell you the ID.
Check lvs for the ID of the disk(s). Create new VMs with the same ID , then run qm rescan to add the disk(s) back.
You could also mount the old pve/root volume somewhere and then extract the configs from the database with this.
Don't you have backups? If not, why?
 
Last edited:
Impact said:
Before doing anything I'd recommend you rename the old pve VG with vgrename VGIDHERE NEWNAME. vgdisplay should tell you the ID.
Check lvs for the ID of the disk(s). Create new VMs with the same ID , then run qm rescan to add the disk(s) back.
You could also mount the old pve/root volume somewhere and then extract the configs from the database with this.
Don't you have backups? If not, why?
Click to expand...
I want to clairfy the procedure so I don't make any mistakes:

  • Use vgdisplay to get the ID of the VG(?)
  • Use lvsto get the ID of the disks
  • Then use the VG_ID and DISK_ID with vgrenameto rename the "old" pve VG's. i'm assuming these are "EightTB_thin" and "TwelveTB_thin". Question: Why do I have two old VG's named pve?
  • Once I've renamed "EightTB_thin" and "TwelveTB_thin", then create new VM's with the same ID's as in the old install (using DISK_ID from the previous step?)
  • Run qm rescanand hopefully the old VM's will apeear in my new Proxmox instance?
I didn't make any backups becuase I'm an idiot :(, and I naively assumed nothing would happen to my Proxmox setup.
 
Last edited:
Please don't quote my whole message. You only rename the old pve VG. Nothing else. You have two because your old setup used LVM too,
 
You must log in or register to reply here.
Top Bottom