Auto Deploy recognizes ESXi hosts as new hosts post reboot due to MAC address change

Products

• VMware vCenter Server

Issue/Introduction

• Stateless ESXi hosts (version 7.0 Update 3q) managed via VMware Auto Deploy are not reconnecting to their existing vCenter Server objects after a reboot. vCenter identifies them as new hosts, leading to inventory conflicts and VM downtime.

• ESXi hosts appear as Not Responding state in vCenter.

• Virtual Machines (VMs) on affected hosts show as Disconnected.

• vCenter Server logs (vpxd.log) shows "Error Datastore 'datastore1' conflicts with an existing datastore in the datacenter that has the same URL (ds:///vmfs/volumes/...), but is backed by different physical storage"

• Auto Deploy logs (rbd-server.log) shows: "domain has already been deleted or has not been completely created"

• The physical MAC address of the ESXI host has changed. verified via iLO/BMC and vCenter inventory

Environment

VMware vSphere ESXi 7.x
VMware vCenter Server 7.x

Cause

The primary cause is a change in the physical MAC address at the hardware level. Auto Deploy uses the MAC address as a primary identifier to map a booting host to its existing record in the vCenter inventory. When the MAC address changes (often due to motherboard replacement, NIC firmware updates, or configuration resets), Auto Deploy fails to associate the hardware with the existing host object, causing vCenter to treat it as a new, conflicting entity.

Resolution

To resolve this issue, the host identity must be aligned between the hardware and the vCenter inventory.

• Log in to the host's hardware management interface (iLO, iDRAC, etc.).

• Modify the MAC address of the primary boot interface to match the original MAC address registered in the vCenter inventory.

• Reboot the host. Auto Deploy should now correctly map the host to the existing vCenter object.