My disk died :(

Hi folks, I had my qubes running with a lvm pool constitued by 2 ssd’s.
When I try to decrypt my luks only 1 is showing up and the system won’t boot after the luks decrypt.
I booted tails and that disk appears with only 1 partition called “unknown”, and the one that shows up has the default partitions of luks. So it died.

Is it possible with some tweak in the uefi startup commands to boot with only 1 disk? If so, how ?
If not, is it possible to recover the qubes inside the working disk?

Thanks for the attention :slight_smile:

I cant advise on the UEFI recovery without knowing why the system wont
boot, and you havent shared that. Nor which version Qubes you run.
Hit ESC to see the boot process and enter the luks decryption - you
should see something leading up to error.

You also havent said how you had configured the pool to use the SSDs, or
which of them had failed. perhaps you dont know the last, but you should
know the first.
If the data is of any value, and you have no recent backup, I suggest
you take copy of drive before doing anything else. Use dd to copy whole
drive to another, like this: dd if=/dev/sda of=/dev/sdb, where sda is
the "working " disk, and sdb is your new backup drive.

Recovering qubes may be possible, but you’ll have to give far more
Remember, it’s not the qubes that are important - it’s the data. You
should be able to at least access the data by decrypting the luks
partitions, mounting the private volumes, and pulling off the data they

Thanks for your reply @unman

I’m using 4.2.0 version.
I don’t know much tbh the configuration of the pool. In the installation process I selected my 2 ssds in the storage field. I don’t understand when you talk bout which one of them failed? They aren’t the same, is that the point?

I’ve decrypt the working disk sucessufully at tails but couldn’t mount the the /dev/mapper/luks-*** nor the LV paths.

About the error itself. At qubes start, once I enter the passphrase for the working disk I recive an OK “crypography Setup for luks-***”, followed by 3 warnings

  • File locking is disabled
  • Couldn’t find device with uuid *****
  • VG qubes_dom0 is missing PV *** (last written to /dev/mapper/luks-****)

It then gets stucked forever.

I don’t know if it’s possible to recover, but you can search how to recover the LVM that have multiple disks with one of them failing.