Failed to read back partitions (runPartprobe)

[juan_c]

Hi,

I did a raw image capture of a dual boot (win10-linux mint) disk. When deploying in the same machine, all seems good and the image is correctly deployed. But, at the end, I get a runPartprobe error and the tasks never end. The computer restart the deployment until I stop it using the BIOS menu.

I see the “Args Passed” seems not good /dev/nvme0n instead of /dev/nvme0n1

I check in debug mode as suggested in a previous issue :

source /usr/share/fog/lib/funcs.sh
getHardDisk              # gives $hd =  /dev/nvme0n1
runPartprobe "$hd"      #  gives no error

Also I put this value in the “host primary disk” variable

Any help will be appreciated.
tks

Juan

[juan_c]

I forget to say that I’m using fog version 1.5.0

[Tom Elliott]

@juan_c if running in debug worked is it reading He disk as set in the host? Based on what I’m seeing it appears your host device input is missing the 1 at the end, so it cannot find the disk.

[juan_c]

@tom-elliott thanks for your answer. The disk name seems correct in the host properties

The image is defined like this

I did again a capture followed by a deploy and got the same problem.
Is there some truncation in the partprobe command?

[Sebastian Roth]

@juan_c said in Failed to read back partitions (runPartprobe):

Is there some truncation in the partprobe command?

There shouldn’t be as far as I know. Take a look at the script code: https://github.com/FOGProject/fos/blob/master/Buildroot/board/FOG/FOS/rootfs_overlay/usr/share/fog/lib/funcs.sh#L1557

[Tom Elliott]

@juan_c I’d recommend removing the /dev/nvme0n1 from the host in question. I’m having a feeling the 1 is being trimmed by something. FOG SHouldn’t have a hard time finding nvme disks anymore, in the past this was the case.

[zamakli]

This post is deleted!

[Sebastian Roth]

Fixed now in latest version of FOG: https://github.com/FOGProject/fos/commit/5600a5adf952d6cef50f60c4fde1a1551c2ba7ee