ESXi host fails to connect to storage array and iSCSI VMkernel adapters show void state due to incorrect VDS uplink mapping

Products

• VMware vSphere ESXi

Issue/Introduction

• When adding a new ESXi host to an existing cluster, the host is unable to connect to the storage array and fails to mount datastores.
• Network tests reveal that the affected host cannot ping both the storage array IP and the peer ESXi iSCSI IPs.
• When reviewing the network configuration via CLI on the affected host, the iSCSI VMkernel adapters (e.g., vmk3 and vmk4) show an uplink state of "void".

Environment

VMware vSphere ESXi

Cause

• This issue occurs when the physical network adapters (vmnics) on the affected ESXi host are mapped to the wrong uplinks on the vSphere Distributed Switch (VDS).
• If the iSCSI port groups are strictly configured to use specific active uplinks (e.g., uplink1 and uplink2), but the host's physical adapters are mapped to different uplinks (e.g., uplink3 and uplink4), the iSCSI VMkernel adapters will have no physical path out to the network. This results in the "void" interface state and drops all iSCSI communication.

Resolution

To resolve this issue, you must remap the physical vmnics on the affected ESXi host to match the active uplink configuration required by the iSCSI port groups.

1. Log in to the vSphere Client and navigate to the vSphere Distributed Switch (VDS) topology.

2. Review the Teaming and Failover policies for the iSCSI port groups (e.g., iSCSI-A and iSCSI-B) to determine which uplinks are strictly configured as active (e.g., it is mapped to uplink1 and uplink2).

3. Open the physical network adapter management screen for that host (this can typically be done by right-clicking the VDS and selecting Add and Manage Hosts, then choosing Manage host networking).

4. Remap the physical vmnics from the incorrect uplinks (e.g., uplink3 and uplink4) to the required active uplinks (uplink1 and uplink2) and save the configuration.

5. Verify via CLI that the iSCSI VMkernel adapters no longer show a "void" state and test network communication by pinging the storage array IP address. Datastores should now be fully visible and mounted.
   
   Commands to verify the uplink mapping from CLI :
   
   netdbg vswitch instance list (DVS)
   or
   nsxdp-cli vswitch instance list (NSX prepared DVS)

Additional Information

Troubleshooting ESXi connectivity to iSCSI arrays using software initiators

Troubleshooting ESX and ESXi connectivity to iSCSI arrays using hardware initiators