ERROR status could be an intermediate state during failover between transport nodes observed on DHCP server status.

Products

• VMware NSX

Issue/Introduction

• DHCP leases that rely on the DHCP server configuration may not be working.
• Edge nodes may be in an unhealthy or unresponsive state. These Edges may have had a failed redeploy attempt leading to this condition.
• DHCP Profile ( NSX UI > Networking > Networking Profiles > DHCP) is set to Auto Allocate Edges on the Edge cluster containing unhealthy Edge nodes or manually set to damaged Edge nodes.
• When navigating to Tier-1 or Tier-0 gateways and viewing their DHCP server configuration (GUI > Networking > Tier-1/Tier-0 > DHCP configuration > Servers) the server status shows the error "Error - ERROR status could be an intermediate state during failover between transport nodes. Please recheck the status a few minutes later."
  
  
• When attempting to remove the unhealthy edge node from the cluster the removal is blocked with a warning stating:
  
  edge-nodes can not be deleted as it is being referenced by entity(s): LogicalDHCPServer/<Server UUID>.
  
  
  

Environment

VMware NSX 
VMware NSX-T Data Center

Cause

This can occur when the edge nodes have faced a catastrophic failure and/or attempted to be replaced but are still not yet healthy. The edge nodes that have been allocated to be used as the DHCP servers are in an unhealthy state but are still marked as the current DHCP servers. This prevents cluster removal and the DHCP does not function on the faulty nodes.

Resolution

The long term solution is for the faulty / unhealthy edge nodes to be brought back to healthy via the following two documented methods.

1. The node itself can be redeployed while in the cluster using Admin Guide - Redeploy an Edge Node , please note this is the 4.2 documentation which includes re-deploy via the GUI. This is not present in earlier versions and the API method is required.  
   
   
2. Alternatively a new node can be deployed using different network configurations and then the edge cluster member be replaced using Admin Guide - Replacing an NSX Edge Transport Node in an NSX Edge Cluster
   
   

A short term work around is for the DHCP profile to be switched to manually allocate to a known healthy node within the cluster. This can be done by navigating to the impacted DHCP Profile (GUI > Networking > Networking Profiles > DHCP) then switching from Auto Allocated to known healthy nodes. Once the edge cluster is brought back to healthy, this can be re-set to auto-allocate.