The NCC health check duplicate_hypervisor_ip_check detects IP addresses that conflict with any of the PE/clusters hypervisor host IPs on the same network.
It does so by enumerating all available IPs for the external hypervisor interfaces in the cluster, checks responses to these IPs on the local network, and calls out any duplicate IP's which may be detected.
This check results in a PASS status if none of the Hypervisor IPs have been duplicated in the network.
If this check returns a FAIL status, that means that either the Hypervisor External IP or Hypervisor backplane network has a conflicting IP address assigned with another device on the same network.
What can be the impact?
Hypervisor host connectivity can become unstable or unavailable, leading to performance impact, redundancy concerns, and potential downtime.
Below is an overview of the steps involved in case the check reports a failure:
Run an arping command from any of the CVM and ensure a reply is received from only one MAC address for the listed hypervisor IP in the failed check.
Check if you receive two distinct MAC addresses in the output.
Check the MAC address of the host and match it with the arping command output.
Discover the device to which the other MAC address belongs and change the IP setting on that device.
Re-check with arping command for only a single MAC address for the IP.
For more information on the commands involved and to take a better look at the outputs, please refer to KB-5578.