The recovery scenario is more complicated when the common link fails between Core 1 and Core 2.
The common link is configured to be an EAPS common link, and this configuration requires that one node be designated as the EAPS controller node and the other node be designated as the EAPS partner node. As described below, the selection of which node is assigned which role must consider the overall customer topology.
If the shared port link fails, the EAPS master node unblocks its port, and VPLS on the EAPS controller node takes the additional action to remove the PWs (point 4 in the following figure) associated with the VPLS from hardware.
In this recovery mode, all traffic to and from the access ring and the rest of the VPLS instance passes through Core 2. When the EAPS controller node detects that the common link is repaired, it enters the preforwarding state as normal. When the controller node exits the preforwarding state, EAPS informs VPLS so that VPLS can reestablish the PWs.