VMs config lost after pool importing

bogdan.solga

Member
Nov 26, 2020
13
0
6
44
Hello, everyone!

Recently I have moved my Proxmox host from one office to another one. That move implied the changing of the network.
Between the two moves, I have repeatedly tried to start the host, both by the normal boot process and using the Proxmox 'ZFS rescue boot' (by booting from a Proxmox USB installer).

After I have successfully reconnected to the running Proxmox host, I have noticed that the VMs that I had configured and used were gone. I had a look in the ZFS datasets and on the filesystem, and their content is there:

Bash:
root@ws:~# zfs list

NAME                       USED  AVAIL     REFER  MOUNTPOINT
rpool                     87.4G   318G      128K  /rpool
rpool/ROOT                16.4G   318G      128K  /rpool/ROOT
rpool/ROOT/pve-1          16.4G   318G     16.4G  /
rpool/data                71.0G   318G      149K  /rpool/data
rpool/data/vm-100-disk-0  26.2G   318G     26.2G  -
rpool/data/vm-100-disk-1   117K   318G      117K  -
...

The VM data is also displayed by lsblk:
Bash:
root@ws:~# lsblk

zd0               230:0    0    95G  0 disk
|-zd0p1           230:1    0   200M  0 part
|-zd0p2           230:2    0  94.8G  0 part
|-vm-100-disk-0p1 253:0    0   200M  0 part
`-vm-100-disk-0p2 253:1    0  94.8G  0 part
...

and
Bash:
root@ws:~# ls -lh /dev/zvol/rpool/data/

total 0
lrwxrwxrwx 1 root root 12 Jan  5 17:51 vm-100-disk-0 -> ../../../zd0
lrwxrwxrwx 1 root root 14 Jan  5 17:51 vm-100-disk-0-part1 -> ../../../zd0p1
lrwxrwxrwx 1 root root 14 Jan  5 17:51 vm-100-disk-0-part2 -> ../../../zd0p2
...

Any idea how to recover that VM?
 
Hello, everyone!

Recently I have moved my Proxmox host from one office to another one. That move implied the changing of the network.
Between the two moves, I have repeatedly tried to start the host, both by the normal boot process and using the Proxmox 'ZFS rescue boot' (by booting from a Proxmox USB installer).

After I have successfully reconnected to the running Proxmox host, I have noticed that the VMs that I had configured and used were gone. I had a look in the ZFS datasets and on the filesystem, and their content is there:

Bash:
root@ws:~# zfs list

NAME                       USED  AVAIL     REFER  MOUNTPOINT
rpool                     87.4G   318G      128K  /rpool
rpool/ROOT                16.4G   318G      128K  /rpool/ROOT
rpool/ROOT/pve-1          16.4G   318G     16.4G  /
rpool/data                71.0G   318G      149K  /rpool/data
rpool/data/vm-100-disk-0  26.2G   318G     26.2G  -
rpool/data/vm-100-disk-1   117K   318G      117K  -
...

The VM data is also displayed by lsblk:
Bash:
root@ws:~# lsblk

zd0               230:0    0    95G  0 disk
|-zd0p1           230:1    0   200M  0 part
|-zd0p2           230:2    0  94.8G  0 part
|-vm-100-disk-0p1 253:0    0   200M  0 part
`-vm-100-disk-0p2 253:1    0  94.8G  0 part
...

and
Bash:
root@ws:~# ls -lh /dev/zvol/rpool/data/

total 0
lrwxrwxrwx 1 root root 12 Jan  5 17:51 vm-100-disk-0 -> ../../../zd0
lrwxrwxrwx 1 root root 14 Jan  5 17:51 vm-100-disk-0-part1 -> ../../../zd0p1
lrwxrwxrwx 1 root root 14 Jan  5 17:51 vm-100-disk-0-part2 -> ../../../zd0p2
...

Any idea how to recover that VM?

Cool, We have exactly the same thing after the cluster mess up.
 
is pmxcfs starting normally (systemctl status pve-cluster)?
what are the exact steps you did to "reconnect"?