NSX-T roubleshooting Guide, checking the nsxa or nsx-mpa status and logs could be helpful, is there specific errors that provide more detail:
https://docs.vmware.com/en/VMware-NSX-T/2.1/nsxt_21_troubleshoot.pdf
Check the Transport Node Status
Procedure
Make the following API call to get the state of the transport node.
GET https://<nsx-mgr>/api/v1/transport-nodes/<transport-node-ID>/state
If the call returns the error RPC timeout, perform the following troubleshooting steps:
- Run /etc/init.d/nsxa status to see if nsxa is running.
- Run /etc/init.d/nsx-mpa status to see if nsx-mpa is running.
- To see if nsx-mpa is connected to the NSX Manager, check the nsx-mpa heartbeat logs.
- To see if nsxa is connected to the NSX Manager, check the nsxa log. You will see the following message if nsxa is connected to the NSX Manager.NSXA_LOG(LVL_INFO, "[%s] Connected to mpa, cookie:[%d]\n", __func__ , _mpaCookieId );https://opengrok.eng.vmware.com/source/xref/nsx.git/mpa/clients/nsxa/src/core/mpaClient.cpp#419
- class="p" id="GUID-54724E5E-269A-4F0E-8665-AB40E3B67537__d5t99" style="margin-top:1em;margin-bottom:10px"
- Check to see if nsxa crashed while executing HostConfigMsg.
- To see if the RMQ messages are taking a long time to be delivered to the host, compare the timestamps of log messages on the NSX Manager and the host.
If the call returns the error partial_success, there are many possible causes. Start by looking at the nsxa logs. On the ESXi host, check hostd.log and vmkernel.log. On KVM, syslogholds all the logs.
