You can first try resolving the issue without redeploying a new NSX Controller cluster. In this approach, you first delete the broken NSX Controller node, and then deploy a new NSX Controller node.