NSX managed hosts HBA adapters offline after host prepared for NSX

Products

• VMware NSX

Issue/Introduction

When an ESXi server is prepared in NSX managed cluster, the HBA (CNA) adapters on the host may become offline and all vSAN storage are gone from the host. 

The N-VDS's LLDP profile is set to "Listen" on NSX-T 3.x or "Discovery Protocol" on the DVS in vCenter is set to use LLDP. 

 

On the top of rack switch's interface, there may have a similar output as:

From the ESXi host log

/var/log/tdlog/vobd.all

There may have similar entries as:

2024-07-12T06:10:35.417Z: [netCorrelator] 82260298us: [vob.net.dvport.uplink.transition.down] Uplink: vmnic2 is down. Affected dvPort: <UUID>/<port ID> 1 uplinks up. Failed criteria: 128

Environment

VMware NSX-T 3.x

VMware NSX 4.x

Cause

This is caused by top-pf-rack switch incorrectly counting the number of LLDP neighbor when NSX onboards the DVS on the host. 

Resolution

For NSX-T 3.x:

• Disabling the LLDP on the Transport Node Profile or from the vCenter change the type of Discovery Protocol of DVS switch to "Cisco Discovery Protocol" or "disabled". 

For NSX 4.x:

• From the vCenter change the type of Discovery Protocol of DVS switch to "Cisco Discovery Protocol" or "disabled".