In an ESXi host environment where NSX has been recently removed, the Configure Management Network option in the Direct Console User Interface (DCUI) appears greyed out or inaccessible.
Symptoms:
esxcli network ip interface list or similar commands return: Error performing operation: Unable to get node: Not found.
/var/log/vmkernel.log contains: Load virtual NICs failed: Error retrieving VMkernel NIC list: Unable to get node: Not found.
The host may still contain stale NSX Virtual Distributed Switch (N-VDS) configuration entries despite the VIBs being removed.
VMware vSphere ESXi.
VMware NSX.
The issue is caused by a stale or orphaned N-VDS switch configuration remaining in the ESXi ConfigStore database. When the ESXi networking stack attempts to initialize, it encounters a reference to a switch that no longer exists or has no associated VIBs, leading to a failure in populating the VMkernel NIC list.
To resolve this issue, the stale switch configuration must be manually removed from the ConfigStore.
Log in to the ESXi host via shell console as root.
Identify the stale switch name by running the following command: configstorecli config current get -c esx -g network_vds -k switches
Note the <vDS_Name> of the orphaned NSX switch from the output.
Delete the stale switch entry: configstorecli config current delete -c esx -g network_vds -k switches -i <vDS_Name>
Restart the host management agents to refresh the state: /etc/init.d/hostd restart /etc/init.d/vpxa restart
Verify that "Configure Management Network" is now available in the DCUI.
This behavior is typically seen in VCF or standalone NSX environments where a host was not gracefully decommissioned or the Transport Node unconfiguration failed to clean up the local database.