This issue is resolved in VMware NSX-T Data Center 3.1.0, available at
VMware Downloads.
Workaround:
To work around this issue if you do not want to upgrade:
Note: Ensure all of your Management VMKernel Interfaces are on VLAN-backed Port Groups on a Distributed Virtual Switch or vSwitch. For more details see:
https://storagehub.vmware.com/t/vmware-r-vsan-tm-network-design/nsx/.
- Enable stateful install on the ESXi host by navigating to Advanced Configuration Settings > System Image Cache Profile Settings > Enable stateful installs on the host.
- If applicable to your environment, disable the core dump profile by navigating to General System Setting > Core Dump configuration.
- Set the "Flag indicating if NSX-T should be enabled on DVS” to false by navigating to Networking configuration > vSphere distributed Switch > the DVS name used within the NSX-T transport node profile.
- Disable the lockdown profile under Security and Services > Security Setting > Security > Lockdown mode.
- If NSX-T is not configured on the VDS, use a NSX-T Virtual Distributed Switch (N-VDS) in your transport node profile. No VMkernel adapter should reside on an NSX logical port group on the VDS.
- Retry to provision the ESXi host.
Note: Even after you complete these steps, the Host Profile for the cluster might display as non-compliant since NSX-T is enabled on the host, but not on the VDS. You can ignore the warning. Alternatively, you can re-set "Flag indicating if NSX-T should be enabled on DVS” to true. However, you must set the flag to false each time you use Auto Deploy to provision an ESXi host on that cluster.