Management Channel To Transport Node Down or Management Channel To Transport Node Down Long alarm is reported in NSX Manager

Products

• VMware NSX

Issue/Introduction

• NSX Manager UI reports Alarms as "Management Channel To Transport Node Down" or "Management Channel To Transport Node Down Long".
• Host Transport nodes indicate NSX configuration of "Host Disconnected".
• Attempts to resolve Host Transport node Alarm "Host Disconnected" fail and remain in a disconnected state.
• Ping attempts from Host transport node to NSX Manager IP are successful.
• Alarm details:

  Management Channel To Transport Node Down
  This alarm is triggered when the messaging channel between the NSX Manager node and the Transport node remains down for a continuous period of 5 minutes

  Management channel to Transport node is down for too long.
  This alarm is triggered when the messaging channel between the NSX Manager node and the Transport node remains down for a continuous period of 15 minutes

Review this https://knowledge.broadcom.com/external/article/410581 which is a comprehensive guide to troubleshoot connectivity issues between the Control Channel and the Transport Node.

Environment

VMware NSX

Cause

• The MPA connectivity (TCP port 1234) is down between the transport nodes and the manager.
• Heartbeating between NSX management node and host is down.

Resolution

To resolve this issue, please verify the following: 

• Verify Network Connectivity:
  Ensure that MPA connectivity is permitted on TCP port 1234 between the Transport Nodes and the NSX Manager.

• Duplicate IP address:
  Ensure the IP address assigned to the Edge management interface is not in use elsewhere on the network.

• Test Port Communication:
  From the impacted Transport Node, run the following command to validate connectivity to the manager:
  # nc -zv <NSX_Manager_IP/FQDN> 1234

• Check Service Status:
  On the NSX Manager nodes, verify that the messaging-manager service is active by executing the command below while logged in as root user:
  # /etc/init.d/messaging-manager status

• Restart Service (If Required):
  If the service is reported as stopped or not running, restart it using the following command:
  # /etc/init.d/messaging-manager restart

Additional Information

Reference documents:

• Transport Node Connectivity to Controller is Down
• "Control Channel To Transport Node Down Long" or "Management Channel to Transport Node Down Long" alarms for removed Transport Nodes
• NSX Event Catalog
• VMware NSX-T Data Center 3.2 Release Notes