A consultant is unable to ping the CVM from the hypervisor using the external interface. The customer's network is configured with the 10.3.12.0/24 subnet without VLAN tagging. The CVM and hypervisor are using the same address space. What is causing this issue?

The scenario describes a layer 2 connectivity issue on a single subnet. The key information is that the network is configured "without VLAN tagging." This means the hypervisor and CVM are sending untagged Ethernet frames. If the physical switch port they connect to is configured as a trunk port (which expects tagged frames for specific VLANs), it will handle untagged traffic according to its "native VLAN" setting. If the native VLAN is not the VLAN associated with the 10.3.12.0/24 subnet, the untagged frames from the host will be placed in the wrong network segment, breaking communication.

A. The 1GbE and 10GbE NICs have not been separated: Mixing NIC speeds is a performance and best practice issue; it does not typically prevent basic IP ping connectivity on a correctly configured network.
B. The default gateway is incorrect: A default gateway is used for routing traffic between different subnets. It is not involved in communication between two devices on the same subnet.
D. The CVM/hypervisor only communicates over 192.168.5.0/24: This is false. The 192.168.5.0/24 network is a private backplane for internal services, but all external communication uses the configured external IP addresses.

1. Nutanix, AHV Networking Best Practices Guide, "VLAN Configuration" section. This guide details the proper configuration of switch ports for AHV hosts, explaining the difference between access ports (for untagged traffic) and trunk ports (for tagged traffic) and the concept of a native VLAN.

2. Nutanix University, Nutanix Hybrid Cloud Fundamentals Course, "Module 4: Networking," Section: Acropolis Networking Overview. This section covers the interaction between the AHV host's virtual switch and the physical network infrastructure, including VLANs.