NSX Edge status becomes Unknown due to different system time within 3 NSX Manager nodes

Products

• VMware NSX

Issue/Introduction

• Some NSX Edge nodes status become Unknown.
  
  
• One or two NSX Manager nodes time always behind the other nodes'.
  
  
• You see messages similar to the following in /var/log/proton/nsxapi.log
  

  ####-##-##T##:##:##.###Z  INFO HeatMap-MsgHandler-Thread-2 AggTnStatusQueriesImpl 4770 MONITORING [nsx@6876 comp="nsx-manager" level="INFO" subcomp="manager"] node <UUID> heartbeat timeout, current #############, ccp #############, interval 360000 in milliseconds, isExpired:true
  ####-##-##T##:##:##.###Z  INFO HeatMap-MsgHandler-Thread-2 AggTnStatusQueriesImpl 4770 MONITORING [nsx@6876 comp="nsx-manager" level="INFO" subcomp="manager"] node <UUID> heartbeat timeout, current #############, ccp #############, interval 360000 in milliseconds, isExpired:true

  
  
  
• You see messages similar to the following in /var/log/phonehome-coordinator/phonehome-coordinator.log
  

  ####-##-##T##:##:##.###Z  INFO MessagingStreamListenerImpl-1-1 MessagingStreamListenerImpl 3916 - [nsx@6876 comp="nsx-manager" level="INFO" subcomp="monitoring"] Corfu StreamListener entry UfoTable ClientHeartbeat, (<UUID>) : payload status_change_reason: "messaging_heartbeat_client_unresponsive" status_change_timestamp: #############, operationType UPDATE.
  ####-##-##T##:##:##.###Z  INFO MessagingStreamListenerImpl-1-1 MessagingStreamListenerImpl 3916 - [nsx@6876 comp="nsx-manager" level="INFO" subcomp="monitoring"] Corfu StreamListener entry UfoTable ClientHeartbeat, (<UUID>) : payload is_active: true status_change_reason: "messaging_heartbeat_client_responsive" status_change_timestamp: #############, operationType UPDATE.

Environment

VMware NSX Datacenter 

Cause

If the time between the NSX Managers is not synchronized, a false heartbeat timeout will occur, which will result in the NSX Edge being in an unknown state.

Resolution

Check each time synchronization status on NSX Managers and correct the time drift if the time is not synchronized.

ex.)

nsxmanager> get clock
Tue Jun 24 2025 UTC 02:49:59.808
Tue Jun 24 2025 UTC 02:50:00.030
Timezone:        Etc/UTC (UTC, +0000)
NTP Sync Status: yes