Problems with file restore

[supermicro_server]

Good morning,

we are encountering an issue with the "file restore" function.
The infrastructure is composed by a ceph cluster and a PBS:
- PVE version 9.2.3
- Ceph version 19.2.3
- PBS version 4.2.1

The issue affect a Windows VM backed up to PBS.

The VM contains 4 GPT/NTFS disks:
scsi0 - 60 GB
scsi1 - 3.54 TB
scsi2 - 3.65 TB
scsi3 - 570 GB

When browsing the backup through PBS File Restore, the following error is returned for both large data disks:

mounting 'drive-scsi2.img.fidx/part/2' failed: all mounts failed or no supported file system (400)

watching the qemu.log file it seems to be related with:
[2026-07-20T11:10:48.943Z WARN proxmox_restore_daemon:roxmox_restore_daemon::disk] mount error on '/dev/vdd2' (ntfs3) - ENOENT: No such file or directory

The issue is affecting only the largest disks (scsi1, scsi2), on the smallest the file restore works properly.

An entire restore of the VM works properly, all the disks are accessible from Windows.

Additional checks:

- Both affected volumes are standard NTFS 3.1 filesystems.
- No ReFS is involved.
- No Windows Dynamic Disks are used.
- No Storage Spaces are configured.
- Volumes use standard 4 KB clusters.
- Repair-Volume -Scan reports no filesystem errors.

We can provide the complete qemu.log and any additional debugging information if required.

Thanks in advance.