No bootable os found + Factory Reset not working

Hi,
i installed “Mint 21.3” via USB on a new NitroPad. After reboot I got the message no bootable is found on any partition. After reading in the docs that 21.3 is not supported I went through the same procedure with 20.1, same effect. Then I installed latest Ubuntu, same here.
On the console : “mounting /dev/nvme0n1 on /boot failed: Invalid argument”

Then I decided to go through the “Factory Reset”, after answering if I’d like to export my public key to an USB drive, he jumps straight back to the main menu. On the console, the next entries are:

Checking for USB Security Dongle …
mount: mounting /dev … [same msg as above - invalid argument]
unable to mount /boot

Went through this procedure multiple times.

Any ideas how to proceed from here?
Kind regards Tobias

System:
Nitropad NV41 | v2.2
FW_Ver: Heads-v2.2
Kernel: 6.1.8-Heads

I don’t know about menu jumping back / aborting, but in general an installed bootable system is required for a factory reset, because it resets the measured boot, which also keeps files on /boot.

Since Ubuntu 24.04 also appears to have a heads issue, you could start with an Ubuntu 22.04 default install to get going. Once it boots, redo the factory reset to set it up. Once that’s done, you can upgrade (see), or do another fresh reinstall attempt.

Thanks for the reply,
tried to install Ubuntu 22.04, image from the documentation ubuntu-22.04.1-nitropad_nitropc-oem-amd64.iso. But the installation breaks with the Message:

No modifications can be made to the device Encrypted Volume …p4_crypt
In use by LVM Volume Group vgubuntu.

Booted into live mode, deleted all partitions, rebooted checked that there are no partitions.Then tried installation again - without going into live mode. Same msg as above.

Downloaded 22.04 direct from Ubuntu - was 22.04.4 whereas from nitrokey server its 22.04.1.
Tried to boot but screen went black and nothing happens.

Another suggestion?

After you boot into the USB live mode, wipe the disc partition table manually, for example
dd if=/dev/zero of=/dev/nvmeX bs=1M count=100
with nvmeX being the parent device name from lsblock.
Then try the install wizard again.

Same as before, the wizard creates the partition, and then complains that there are in use.

To be sure I followed your advice correctly.

  • Booted into live mode - Terminal lsblk shows nvme0n1 with Partitions p1 - p4 bellow.
  • dd … of=/dev/nvme0n1 …
  • after that lsblk shows just nvme0n1

Then I started the wizard from the Desktop, before hitting “install now” at the end of the wizard, I switched to the Terminal and did an lsblk, again nvme0n1 with Partitions p1 - p4 where back.

Hitting install now, the error message … In use by LVM Volume Group vgubuntu … showed up.

Switched back to terminal, lsblk shows now under
nvme0n1p4

  • nvme0n1p4_crypt
    • vgubuntu-root
    • vgubuntu-swap_1

What I noticed, the wizard shows the partitioning screen after the error. In the progress bar (sic!) there is a tiny message that says that the install medium is on sda1 therefore I cannot change any partitions on this device. Don’t know if this just affects the partition screen at this stage.

Installed Qubes Os 4.2.2.
Installation and factory reset without any problem.
So … think I play around with this config.

Many thanks for your help!

Ok, that’s good then. Still odd how complicated a plain Ubuntu install got. Have a look at the /boot partition files. When you decide to try something else, best let the respective installer use that /boot partition and make a backup of the /boot/grub/grub.cfg file. Heads parses it to create its boot menu, just having a working reference helps.