Hello @Mark Andrew
If the existing Workstation VM is stable, fully configured, and already tested, I would normally migrate it to ESXi rather than rebuild it from scratch. Rebuilding introduces additional configuration differences and application-level risk without much benefit unless you specifically want a clean OS/application deployment.
For a production migration, however, I would treat this as a controlled conversion rather than simply copying the VM and powering it on.
Suggested approach
1. Prepare the Workstation VM
Before migration:
- Verify the guest OS is supported by your target ESXi version.
- Remove unnecessary snapshots.
- Confirm VMware Tools is installed and current.
- Check filesystem/disk health.
- Record the VM's CPU, RAM, disk layout, IP configuration, DNS, gateway, VLAN, and MAC address if anything depends on it.
- Back up the VM.
- Shut it down cleanly.
Do not run the Workstation VM and migrated ESXi VM simultaneously if they have the same hostname/IP.
2. Move/convert the VM
For a straightforward Workstation → ESXi migration, I would first consider exporting the powered-off VM as OVF/OVA and deploying it to ESXi/vCenter.
If OVF export/import isn't practical for the VM, VMware Converter can also be considered where the source/target versions are supported.
You generally don't need to manually copy individual VMDK files unless you have a specific reason to do so.
3. Review the VM hardware before production
Don't automatically duplicate every Workstation setting.
Review:
- CPU/RAM: Right-size them based on workload requirements.
- Virtual hardware compatibility: Set this according to the ESXi/vSphere version you're running.
- Disk controller: Prefer a controller supported by both the guest OS and your ESXi design. Be careful changing the boot disk controller because the guest may not have the required driver.
- Network adapter: VMXNET3 is normally preferred for supported production guests, but ensure VMware Tools/drivers are available before changing it.
- Disk provisioning: Choose thin or thick according to your storage and operational requirements.
- BIOS/UEFI: Keep the existing firmware type initially. Changing BIOS ↔ UEFI during migration can make the OS unbootable.
- VMware Tools: Verify/update after migration.
4. Handle networking carefully
I would initially connect the migrated VM to an isolated/test port group or leave the NIC disconnected.
Power it on and verify:
- OS boots normally
- VMware Tools/services start
- Applications/services start
- Disks are present
- Event/system logs don't show unexpected errors
- Network adapters are detected correctly
If you're assigning the original static IP to a new virtual NIC, check for hidden/old NICs inside the guest OS. Windows in particular can retain the previous adapter and its static IP configuration.
5. Cut over to production
Once validation is complete:
- Confirm the original Workstation VM is powered off.
- Connect the ESXi VM to the correct production port group/VLAN.
- Configure/verify the production IP, DNS, gateway, and firewall settings.
- Test network connectivity.
- Test application connectivity from actual clients.
- Verify monitoring, backup, security/EDR, time synchronization, and other infrastructure services.
I would also check that no licensing or application configuration is tied to the VM's previous virtual hardware/MAC address.
One important point about snapshots
I wouldn't use an ESXi snapshot as the primary rollback strategy for this migration.
Keep the original powered-off Workstation VM and a proper backup until the ESXi VM has passed production validation. A VMware snapshot is useful for short-term operational rollback, but it isn't a backup.
Migrate or rebuild?
For your scenario:
Healthy, tested Workstation VM → Migrate
Old/unsupported OS, questionable configuration, accumulated problems, or an opportunity to modernize the server → Rebuild
Since you've stated that this VM is fully configured, working correctly, and already tested, I would migrate it, validate it on an isolated network, and then perform a controlled production cutover.
That gives you the lowest-risk path while preserving the original VM as your rollback option.