Hello Guys,
I am working with NSX-T 2.5, with two vSphere clusters under one vCenter 6.7 U3 and Hosts version 6.7 U1.
We prepared the first cluster with NSX-T after Creating/Preparing the Transport Node Profile.
We got a compatibility error between 2.5 and 6.7 U1. After checking the compatibility matrix we found that we have to patch our hosts to P06.
We patched all hosts, then try to prepare clusters. Successfully prepare the second one. Regarding the first one we un-configure it then try to prepare with no luck.
Below are the errors.
Please note that we followed these steps to solve this problem:
What is happening is that the host is flagged as a transport node in NSX-T manager (since the NSX-T manager attempted to prepare it but failed before due to the missing patches before the upgrade) while the agents are not running on the hosts and the NSX-T can't communicate with the hosts via the MPA agent.
So , we need to use Postman REST API client then send the below request to delete all the transport nodes from the cloud cluster.
DELETE https://<nsx-mgr>/api/v1/transport-nodes/<Transport Node ID>?force=true
You can get the ID from the NSX-T manager GUI.
After deleting all the transport nodes, you will see that the status for the transport nodes is back to "Unconfigured" and you will be able to prepare the cluster again.
Please note that we followed these steps to solve this problem:
What is happening is that the host is flagged as a transport node in NSX-T manager (since the NSX-T manager attempted to prepare it but failed before due to the missing patches before the upgrade) while the agents are not running on the hosts and the NSX-T can't communicate with the hosts via the MPA agent.
So , we need to use Postman REST API client then send the below request to delete all the transport nodes from the cloud cluster.
DELETE https://<nsx-mgr>/api/v1/transport-nodes/<Transport Node ID>?force=true
You can get the ID from the NSX-T manager GUI.
After deleting all the transport nodes, you will see that the status for the transport nodes is back to "Unconfigured" and you will be able to prepare the cluster again.