One of the NSXT manager was responding slowly to the VIP address and also to the primary IP address assigned to the NSX manager.
The NSX manager cluster became stable after rebooting the primary node which resulted in the VIP being assigned to a different node in the NSXT cluster
Pings to other two NSX manager nodes in the cluster remained successful. However 50% of pings traffic is dropped with one of the NSX manager node when pinging its primary IP address.
Troubleshooting performed were vmotion to another ESXi host and reboot the VM but the issue was not resolved
3.2.2.0.0.20737185
Issue lies with external infrastructure where the external firewall was dropping ping packets to one of the NSX manager nodes.
The issue is resolved after a failover of external firewall. Check the external firewall configuration and if required perform a reboot or failover.