Step-by-Step Guide: How to Migrate VMs from VMware ESXi to Proxmox VE 8.3

Last updated: Apr 12, 2025

In this article, I will cover how to migrate VMware ESXi virtual machines to a Proxmox hypervisor. I will show here how to transfer Debian and Windows virtual machines with BIOS or UEFI boot. For a reminder I have already discussed the Proxmox hypervisor in previous articles concerning the installation process: here and the creation of virtual machines: here.

As explained, and as you have surely seen, it seems that Broadcom's strategy is to scare off their customers (at least the smaller ones), so since they don't want our money, I suggest you follow their lead by reading this article and migrate to Proxmox. The full process is detailed below. 😉

Architecture

In this article I will work on this very simple architecture where virtual machines stored in the /vmfs/volumes/4T_RAID1 partition of a VMware ESXi will be migrated to a Proxmox server inside the /ZFS_RAID10 partition.

Network diagram showing the migration of virtual machines from a VMware ESXi server to a Proxmox VE server via SSH, with ZFS storage and Debian/Windows VMs.

Enabling SSH on ESXi

The first step consists in enabling the SSH service on the VMware ESXi host from which we want to migrate the virtual machines. We can do it from VCSA or from ESXi, I will show you both methods below. Indeed we will use the SSH protocol to transfer the VMDK files from our ESXi server to our Proxmox hypervisor.

From ESXi, go to Host > Manage > Services. From here, select the SSH service and click on Start:

ESXi client interface showing how to activate the SSH service via the Manage > Services tab.

From VCSA, select the ESXi host then go to Configure > Services and select the SSH service. Finally click on Start:

Enable SSH service on an ESXi host via the vSphere Client in the Configure > Services tab.

Once the SSH service has been activated, check that you can connect to your ESXi host:

user@debian:~$ ssh -l root 192.168.1.250

Once connected, check the partitions where the virtual machines you want to transfer are located:

[root@localhost:~] df -h
Filesystem   Size   Used Available Use% Mounted on
VMFS-6       3.6T   2.7T    953.9G  74% /vmfs/volumes/4T_RAID1
VMFSOS     119.8G   4.1G    115.7G   3% /vmfs/volumes/OSDATA-4203c1ce-de1feb81-e534-fa7e52a7d43e
vfat         4.0G 280.0M      3.7G   7% /vmfs/volumes/BOOTBANK1
vfat         4.0G 258.3M      3.7G   6% /vmfs/volumes/BOOTBANK2

List virtual machines folders:

[root@localhost:~] ls -lh /vmfs/volumes/4T_RAID1
drwxr-xr-x    1 root     root       76.0K May 26  2024 Alpine
drwxr-xr-x    1 root     root       76.0K Dec 12 16:49 Debian_10_IPsec
drwxr-xr-x    1 root     root       76.0K Dec 11 13:40 Debian_12
drwxr-xr-x    1 root     root       84.0K Feb 14 16:54 W11
drwxr-xr-x    1 root     root       96.0K Feb 20 09:09 W2K25

Copy VMDK files

Shut Down the Virtual Machine

Before transferring a virtual machine from ESXi to Proxmox, we must first shut it down. Either via the graphical web interface, the CLI, or directly from the VM.

Example here with the W2K25 virtual machine. In CLI, we first need to get the virtual machine ID (note, we can also get the path information from here):

[root@localhost:~] vim-cmd vmsvc/getallvms | grep -i W2K25
373    W2K25                                 [4T_RAID1] W2K25/W2K25.vmx                                                                        windows2019srvNext_64Guest   vmx-21

Once we retrieve the virtual machine ID, we can shut down the virtual machine:

[root@localhost:~] vim-cmd vmsvc/power.shutdown 373

Copy with SCP

Once we have identified the full path of the virtual machine we want to import we will use the SCP command to transfer the files from the ESXi to Proxmox.

Connect to the Proxmox host:

user@debian:~$ ssh -l root 192.168.1.240

Once connected, list the datastores:

root@proxmox:~# df -h
Filesystem        Size  Used Avail Use% Mounted on
udev               32G     0   32G   0% /dev
tmpfs             6.3G  8.9M  6.3G   1% /run
rpool/ROOT/pve-1  707G  2.2G  704G   1% /
tmpfs              32G   46M   32G   1% /dev/shm
tmpfs             5.0M     0  5.0M   0% /run/lock
efivarfs          304K  104K  196K  35% /sys/firmware/efi/efivars
rpool/var-lib-vz  724G   20G  704G   3% /var/lib/vz
rpool             704G  128K  704G   1% /rpool
ZFS_RAID10       8.9T  2.9T  6.0T  33% /ZFS_RAID10
rpool/data        704G  128K  704G   1% /rpool/data
rpool/ROOT        704G  128K  704G   1% /rpool/ROOT
/dev/fuse         128M   28K  128M   1% /etc/pve
tmpfs             6.3G     0  6.3G   0% /run/user/0

Copy VMware VMDK files to the local storage:

root@proxmox:~# scp -r root@192.168.1.250:/vmfs/volumes/4T_RAID1/W2K25/*vmdk /ZFS_RAID10/W2K25/

Illustration of the transfer of VMDK files from a Windows virtual machine (W2K25) via SCP from a VMware ESXi host to ZFS storage on Proxmox VE.

Convert VMDK to RAW

The VMDK format is not directly usable with Proxmox, we need to convert them to RAW format.

Convert vmdk file to raw:

root@proxmox:~# cd /ZFS_RAID10/W2K25/

root@proxmox:~# qemu-img convert -p -f vmdk -O raw w2k25.vmdk w2k25.raw

Import Virtual Machine

Final step: create a new VM in Proxmox using the previously created RAW disks. As we'll see in the examples below, the syntax and procedure will differ according to boot type (UEFI or BIOS) and operating system.

Windows with UEFI boot

Options:
- --ide0 ZFS_RAID10:0,import-from=/ZFS_RAID10/w2k25_VM/w2k25.raw: specify OS partition destination (use ide because scsi needs the virtio drivers). ZFS_RAID10 is the destination disk. 0 means virtual disk 0.
- --tpmstate0 ZFS_RAID10:1,version=v2.0 (optional): if needed, configure a disk for storing TPM state. ZFS_RAID10 is the destination disk. 1 means virtual disk 1.
- --efidisk0 ZFS_RAID10:2,efitype=4m,pre-enrolled-keys=1,size=1M: configure a disk for storing EFI vars. ZFS_RAID10 is the destination disk. 2 means virtual disk 2.
- --net0 e1000,bridge=vmbr0: specify network device. Intel E1000 does not require additional drivers with Windows.

cputype=x86-64-v2-AES (or x86-64-v3, x86-64-v4 if your CPU is compatible) option is obliged when using Windows Server 2025, otherwise, the system will reboot in an infinite loop.

root@proxmox:~# qm create 300 --name "Windows-w2k25" --memory 4096 --machine q35 --sockets 1 --cores 4 --bios ovmf --cpu cputype=x86-64-v2-AES --efidisk0 ZFS_RAID10:1,efitype=4m,pre-enrolled-keys=1,size=1M --net0 e1000,bridge=vmbr0 --ide0 ZFS_RAID10:0,import-from=/ZFS_RAID10/w2k25_VM/w2k25.raw

After booting into Windows, you may have trouble uninstalling VMware Tools. This PowerShell script can help: https://gist.githubusercontent.com/

It is also recommended to install the VirtIO drivers. See the procedure here (2.2.3 Windows Post-Installation).

Debian

Debian with Legacy boot

Run this command to create the VM with the raw disk and start the VM:

root@proxmox:~# qm create 301 --name "Debian-12-BIOS" --memory 2048 --machine q35 --sockets 1 --cores 4 --bios seabios --net0 virtio,bridge=vmbr0 --scsihw virtio-scsi-single --scsi0 ZFS_RAID10:0,import-from=/ZFS_RAID10/Debian/Debian.raw

Once booted, as the network interface name has changed, you'll need to identify the new interface name:

root@host:~# ip addr sh
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host noprefixroute 
       valid_lft forever preferred_lft forever
2: enp6s8: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether bc:24:11:24:0b:be brd ff:ff:ff:ff:ff:ff

Update /etc/network/interfaces with the new interface name, then restart the networking service:

root@host:~# systemctl restart networking

Debian with UEFI boot

Run this command to create the VM with the raw disk and start the VM:

root@proxmox:~# qm create 302 --name "Debian-12-UEFI" --memory 2048 --machine q35 --sockets 1 --cores 4 --bios ovmf --efidisk0 ZFS_RAID10:1,efitype=4m,pre-enrolled-keys=1,size=1M --net0 virtio,bridge=vmbr0 --scsihw virtio-scsi-single --scsi0 ZFS_RAID10:0,import-from=/ZFS_RAID10/Debian/Debian.raw

If the VM fails to boot and gets stuck, press any key to enter the Boot Manager Menu:

Failure to boot a UEFI VM on Proxmox, indicating that a key has been pressed to enter the Boot Manager Menu.

Inside the Boot Manager Menu, enter inside the Boot Maintenance Manager and select Boot From File:

UEFI navigation sequence from Boot Manager to boot file selection in a Proxmox environment.

Navigate to EFI >debian, then select shimx64.efi:

UEFI file explorer showing selection of the shimx64.efi file in the EFI/debian folder for booting Debian on Proxmox.

Now the virtual machine should now boot successfully:

GRUB boot menu displaying Debian GNU/Linux after manual addition of the EFI entry via efibootmgr on Proxmox.

The first thing to do after booting is to add the EFI entry. After that, Debian will boot automatically on subsequent reboots:

root@host:~# efibootmgr --create --disk /dev/sda --part 1 --label "debian" --loader "\EFI\debian\shimx64.efi"

As we did with the BIOS, identify the new network interface name:

root@host:~# ip addr sh
1: lo: <.gLOOPBACK,UP,LOWER_UP>.g mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host noprefixroute 
       valid_lft forever preferred_lft forever
2: enp6s8: <.gBROADCAST,MULTICAST>.g mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether bc:24:11:24:0b:be brd ff:ff:ff:ff:ff:ff

And modify /etc/network/interfaces with the new interface name and restart networking service:

root@host:~# systemctl restart networking

OpenBSD

OpenBSD with UEFI boot

For this, I used the following source: https://forum.proxmox.com/.

Run this command to create the VM and start it using the RAW disk:

root@proxmox:~# qm create 303 --name "OpenBSD-UEFI" --memory 2048 --machine q35 --sockets 1 --cores 4 --agent 1,type=isa --cpu kvm64 --net0 virtio,bridge=vmbr0 --scsihw virtio-scsi-single --boot order=scsi0 --scsi0 ZFS_RAID10:0,import-from=/ZFS_RAID10/OpenBSD/OpenBSD.raw --bios ovmf --efidisk0 ZFS_RAID10:1,efitype=4m,pre-enrolled-keys=1,size=1M

Optional: use pre-enrolled-keys=0 option when creating VM to disable Secure Boot directly.

The VM may fail to boot and get stuck on this screen. As requested, press any key to enter the Boot Manager Menu:

UEFI boot failure of a Proxmox virtual machine with ‘Access Denied’ message and option to enter Boot Manager.

Inside the Boot Manager Menu, enter the Device Manager, select Secure Boot Configuration and disable Secure Boot:

Disable the Secure Boot option in the UEFI Device Manager of a Proxmox VM via Secure Boot Configuration.

Once booted, since the network interface name has changed, you need to identify the new name:

host# ifconfig
lo0: flags=2008049<UP,LOOPBACK,RUNNING,MULTICAST,LRO> mtu 32768
	index 5 priority 0 llprio 3
	groups: lo
	inet6 ::1 prefixlen 128
	inet6 fe80::1%lo0 prefixlen 64 scopeid 0x5
	inet 127.0.0.1 netmask 0xff000000
vio0: flags=2008843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST,LRO> mtu 1500
	lladdr bc:24:11:53:f1:5a
	index 1 priority 0 llprio 3
	groups: egress
	media: Ethernet autoselect
	status: active

Rename your hostname.interface file using the new interface name:

root# mv /etc/hostname.OLD /etc/hostname.vio0

Apply network configuration changes:

root# sh /etc/netstart

OpenBSD with Legacy boot

Run this command to create the VM with the raw disk and start the VM:

root@proxmox:~# qm create 304 --agent 1,type=isa --memory 4096 --bios seabios --name "OpenBSD-BIOS" --sockets 1 --cores 2 --cpu kvm64 --net0 virtio,bridge=vmbr0 --scsihw virtio-scsi-single --boot order='scsi0' --scsi0 ZFS_RAID10:0,import-from=/ZFS_RAID10/OpenBSD/OpenBSD.raw

Alpine Linux with UEFI boot

Run this command to create the VM with the raw disk and start the VM:

root@proxmox:~# qm create 305 --sockets 1 --cores 2 --memory 2048 --name "Alpine-UEFI" --bios ovmf --efidisk0 ZFS_RAID10:1,efitype=4m,pre-enrolled-keys=1,size=1M --net0 virtio,bridge=vmbr0 --ide0 ZFS_RAID10:0,import-from=/ZFS_RAID10/Alpine/Alpine.raw

Optional: use the pre-enrolled-keys=0 option when creating the VM to disable Secure Boot directly.

The VM may fail to boot and get stuck on this screen. As requested, press any key to enter the Boot Manager Menu:

Inside the Boot Manager Menu, enter the Device Manager, select Secure Boot Configuration and disable Secure Boot:

Computing

Music

Misc