Shutting down the VMware NSX for vSphere 6.x controllers flushes out the VTEP tables on the ESXi host

Products

• VMware NSX

Issue/Introduction

Symptoms:
When using replication type Hybrid mode in a VMware NSX for vSphere 6.x environment, you experience these symptoms:

• After failure of all NSX Controllers or manually shutting down the NSX Controllers, the VTEP tables on the ESXi host are flushed
• Network connectivity between the NSX Edge Services Gateway (ESG) to the Distributed Logical Router (LDR) is lost.
• In the /var/log/vmkernel.log file of the ESXi host, you see entries similar to:
  
  WARNING: vxlan: VDL2CPCheckConnUpCB:xxx: Control plane connection of VXLAN network[xxxx] is down
  WARNING: vxlan: VDL2CPCheckConnUpCB:xxx: Control plane connection of VXLAN network[xxxx] is down

Environment

VMware NSX for vSphere 6.2.x
VMware NSX for vSphere 6.1.x

Cause

This issue occurs due to the current design of VMware NSX for vSphere 6.x. VTEP tables are flushed after a grace period when the NSX Controller is no longer available and the Multicast Tunnel Endpoint (MTEP) and VXLAN Tunnel Endpoint (VTEP) counts are re-initialized.

Resolution

This issue is resolved in VMware NSX for vSphere 6.2.2.

Note: This is a known issue affecting VMware NSX for vSphere 6.1.x.

Currently, there is no resolution or workaround in Hybrid mode on VMware NSX for vSphere 6.1.x.

Additional Information