Hard disk format issue when importing ESXI to PVE 8.2

jordnmao

New Member
Aug 14, 2026
2
1
3
When importing from ESXI to PVE 8.2, the original virtual machine centost7 had six hard disks, all of which were in scsi format. After the import, the hard disk format was changed to ide, but PVE only supports four ide devices. When booting, it gets stuck at the progress bar. Is there any solution to this?
 
The is no ide/scsi "disk format". You define a Storage Controller in VM settings, and need the driver/os support for the device. And yes, old IDE was limited to 4 devices, but OSes have native driver support.

Next steps: install virtio (scsi) drivers and quemu geust agent, then change HW disk controller type to scsi.
 
I suspect that while the VirtIO drivers are installed on CentOS, they are not loading at boot time, preventing the use of SCSI.
I recommend fixing this so that you can use SCSI instead of IDE.

Below are notes from a procedure I performed previously.
If you boot CentOS using IDE and run the following commands, you should be able to use SCSI.

Code:
sudo tee /etc/dracut.conf.d/90-proxmox-virtio.conf >/dev/null <<'EOF'
# VMware PVSCSI -> Proxmox VirtIO SCSI migration support
# Include VirtIO drivers in initramfs and try loading them early during boot.
force_drivers+=" virtio virtio_ring virtio_pci virtio_scsi virtio_blk "
add_drivers+=" virtio_net "
EOF

sudo mkdir -p /root/initramfs-backup-$(date +%Y%m%d-%H%M%S)
sudo cp -a /boot/initramfs-*.img /root/initramfs-backup-$(date +%Y%m%d-%H%M%S)/ 2>/dev/null || true

sudo dracut -f --regenerate-all

KVER="$(uname -r)"
sudo lsinitrd "/boot/initramfs-${KVER}.img" | egrep 'virtio(_scsi|_pci|_ring|_blk|_net)?'
 
  • Like
Reactions: Onslow