When both active and standby VMware NSX Global Managers fail, the environment must be completely rebuilt from a valid backup. First, administrators must delete the corrupted active and standby Global Manager appliances across all clusters to prevent synchronization issues or IP conflicts. Second, deploy a brand-new Global Manager appliance using the exact same IP address and FQDN as the original active node. Third, access the newly deployed appliance to restore the active Global Manager configuration directly from the remote SFTP backup server. Finally, to re-establish high availability and Federation synchronization, deploy a new Global Manager at the secondary site and onboard it as a standby node to the newly restored active Global Manager.

VMware by Broadcom. VMware NSX-T Data Center Administration Guide. Chapter: "Backing Up and Restoring VMware NSX", Section: "Restore a Global Manager Cluster".

VMware by Broadcom. VMware NSX-T Data Center Administration Guide. Section: "Restore an Active Global Manager". (These sections explicitly detail the mandatory 4-step workflow for total Global Manager cluster failure recovery).