How VMware vNetwork Distributed Switch work if vCenter fail

Within a distributed switch, the control and IO planes are separate:

                The control plane resides in vCenter Server. The control plane is responsible for configuring distributed switches, distributed port groups, distributed ports, uplinks, NIC teaming and so forth.

The control plane also coordinates the migration of the ports and is responsible for the switch configuration. For example, in the case of a conflict in the assignment of a distributed port (say because a virtual machine and its template are powered on), the control plane is responsible for deciding what to do.

The IO Plane is implemented as a hidden vSwitch inside the VMkernel of each ESX or ESXi host. The IO plane manages the actual IO hardware on the host and is responsible for forwarding packets.

This diagram shows a more detailed look at the components of the IO plane of a distributed switch.

On each host, an IO plane agent runs as a VMkernel process and is responsible for communication between the control and IO planes.

IO filters are attached to the IO chains connecting the vNICs to the distributed ports and the distributed ports to the uplinks. vNetwork Appliance APIs make it possible to define custom filters and apply them to the IO chains. The APIs also provide the means to preserve filtering information for the virtual machine connected to each port, even after vMotion.
Inside the IO plane, the forwarding engine decides how to forward packets to other distributed ports, either toward other virtual machines on the same distributed switch, or to an uplink, requiring it to make NIC teaming decisions. Forwarding functions can also be customized using the vNetwork Appliance APIs.

So in case vCenter server goes down still Virtual machine can contact outer network with the help of hidden vSwitch