Error "Node with same ip already exists" when preparing VxRail ESXi for NSX

Products

• VMware NSX

Issue/Introduction

• VxRail ESXi host is being prepared for NSX.
• The ESXi host is using two vmknics for management, vmk0 is for VxRail and vmk1 is for vSphere.
• vmk0 is configured with a static IPv4 IP address and static IPv6 address, as can be seen with ESXi command:

host1#esxcfg-vmk -l | grep vmk0

vmk0       29                                      IPv4      169.254.0.2                             255.255.255.0   169.254.0.255   ##:##:##:##:##:8d 1500    65535     true    STATIC              defaultTcpipStack
vmk0       29                                      IPv6      ####::###:####:####:1234                64                              ##:##:##:##:##:8d 1500    65535     true    STATIC, PREFERRED   defaultTcpipStack

host2#esxcfg-vmk -l | grep vmk0

vmk0       83                                      IPv4      169.254.0.2                             255.255.255.0   169.254.0.255   ##:##:##:##:##:91 1500    65535     true    STATIC              defaultTcpipStack
vmk0       83                                      IPv6      ####::###:####:####:1235                64                              ##:##:##:##:##:91 1500    65535     true    STATIC, PREFERRED   defaultTcpipStack

• The IPv4 address configured on vmk0, is also configured on another ESXi host vmk0 and prepared for NSX. 
• When adding a new host, the following error is observed:
  Error: Node host1.example.com with same ip <IP address of host host1.example.com>  already exists. Check the state of existing transport node. If it is a failure in deletion then use force option to delete previous transport node and retry the operation. If the host is part of a cluster having transport node profile applied then then move the host into datacenter before deletion. Refer KB article number 319975 for more details. (Error code: 26210)

Environment

• VMware NSX
• VMware NSX-T Data Center

Cause

• The same static IPv4 IP address was configured on different ESXi host's on their vmk0 interfaces.
• This can be observed in VxRail environments, as VxRail will use two management vmknics:
  • The first, usually vmk0, will be using IPv6, IPv4 should not be manually set on the ESXi host. This vmknic is used for internal management within VxRail. 
  • The second, usually vmk1, is used for vSphere management of the ESXi host (e.g. connection of the host to the vCenter, SSH connectivity to the host, etc.).

Resolution

The IPv4 APIPA (169.254.0.2) address was manually configured on the ESXi host.

If vmk0 is using IPv6, as seen above, managed by VxRail, clear the manually configured IPv4 address:

#esxcli network ip interface ipv4 set -t none

If there is no duplicate IP address found, please review this KB mentioned in the error, which deals with stale host which can also lead to the same error message:

Installing or Upgrading NSX on an ESXi host fails reporting the node already exists

If in doubt, please consult VxRail support to confirm correct vmk0 IP assignment.

Additional Information

Installing or Upgrading NSX on an ESXi host fails reporting the node already exists