[SOLVED] mounting 'drive-scsiXX.img.fidx/part/1' failed: all mounts failed or no supported file system (400)

[MisterDeeds]

Dear all

I have a PVE system in use together with a Proxmox backup server. Now when I want to perform the "file restore" from the PVE, the following message appears:

mounting 'drive-scsiXX.img.fidx/part/1' failed: all mounts failed or no supported file system (400)

I use a normal GPT partition. However, I am not sure if it is because of the 128MB Reserved partition:



Does anyone know the problem? Thank you and best regards

 

[MisterDeeds]

I have now changed the partition to have only one and run a new backup. Unfortunately, the same message still appears. Can anyone help me?

 

[MisterDeeds]

Dear all

In consultation with Proxmox support, it turns out that the restore via web gui mounts the disk into a temporary virtual machine with the "ntfs-3g" command. This shows in the logfile of the PVE host under:

/var/log/proxmox-backup/file-restore/

an error.

There it was apparent that the file system had an error. Windows seems to have no problem with it, but the ntfs-3g command does.
A simple

CHKDSK D: /F /R /X

fixed the problem.

Thanks again to the great proxmox support!

Best regards

 

[JL17]

Hello, i have the same problem with an NTFS disk backup.
If I try to do a "file restore" from GUI, i get the message:

mounting 'drive-scsiXX.img.fidx/part/1' failed: all mounts failed or no supported file system (400)

The file-restore log (/var/log/proxmox-backup/file-restore/qemu.log) show this information about ntfs:

ntfs: (device vda1): read_ntfs_boot_sector(): Primary boot sector is invalid. ntfs: (device vda1): read_ntfs_boot_sector(): Mount option errors=recover not used. Aborting without trying to recover. ntfs: (device vda1): ntfs_fill_super(): Not an NTFS volume.

If i map the backup with "proxmox-backup-client map" and then mount the mapping i can access all the files without problem. Seems like "file-restore" and "map" works different. Is there a way to configure "file-restore" so i can force the mount an NTFS dirty-flagged partition?

Thanx in advance

 

[fabian]

how are you mounting the mapped snapshot? please give the exact command line/options used!

 

[JL17]

how are you mounting the mapped snapshot? please give the exact command line/options used!

Using this commands:

proxmox-backup-client map vm/101/2024-01-10T07:15:12Z drive-virtio1.img --repository root@pam@192.168.1.2:8007:bk1 --ns jls ls -lah /dev/loop0* mkdir /mnt/test mount -r /dev/loop0p1 /mnt/test ls /mnt/test/

 

[fabian]

what does "mount" say about the /mnt/test mountpoint afterwards?

 

[JL17]

what does "mount" say about the /mnt/test mountpoint afterwards?

/dev/loop0p1 on /mnt/test type fuseblk (ro,relatime,user_id=0,group_id=0,allow_other,blksize=4096)

Mounts ok using that commands, I can acces all the files.
The problem is when I try to do a "file restore" from GUI

 

[fabian]

yeah, that seems to use a very different driver (ntfs-3g?) than the file-restore VM (which uses the in-tree kernel "ntfs" one)..

 

[JL17]

can I change the driver in the file-restore?

 

[fabian]

no, that's hard-coded at the moment..

 

[JL17]

Ok thanks for your help!