Comprehensive and Detailed 250 to 350 words of Explanation From VMware Cloud Foundation (VCF) documents:
The architecture ofNSX Federationwithin a VCF Multi-Site design is built upon a separation of theControl Planeand theData Plane. This "decoupled" architecture ensures high availability and resiliency even when management components become unavailable.
In NSX Federation, theGlobal Manager (GM)handles the configuration of objects that span multiple locations, while theLocal Manager (LM)is responsible for pushing those configurations down to the local Transport Nodes (ESXi hosts and Edges) within its specific site. When a configuration is pushed, the Local Manager communicates with theCentral Control Plane (CCP)and subsequently theLocal Control Plane (LCP)on the hosts.
If an NSX Local Manager goes offline, the "Management Plane" for that site is lost. This means no new segments, routers, or firewall rules can be created or modified at that site. However, the existing configuration is already programmed into theData Plane(the kernels of the ESXi hosts and the DPDK process of the Edge nodes).
According to VMware's "NSX Multi-Location Design Guide," the data plane remains fully operational during a Management Plane outage. Existing VMs will continue to communicate, BGP sessions on the Edges will remain established, and Distributed Firewall (DFW) rules will continue to be enforced based on the last known good configuration state cached on the hosts. The data plane does not require constant heartbeats from the Local Manager to forward traffic. Therefore, operations continue normally "headless" until the LM is restored and can resume synchronization with the Global Manager and local hosts. Failover to a primary site (Option D) is only necessary if the actual data plane (hosts/storage) fails, not just the management components.
===========
