Failure Recovery Scenario with VPLS STP Redundancy
If VPLS STP redundancy is configured on switches S1 and S2 in the following figure and the S2-R1 link fails, S1 unblocks S1-R1 and S2 flushes its FDB on the S2-R1 port.
When S2 flushes its FDB, it also sends a flush message to its core peer C2. Upon receipt of this flush message, C2 flushes its FDB entries learned over the PW to S2. C2 also sends flush messages to its core peers C1, C3, and C4. These core nodes then flush their FDB entries learned over their PWs to C2. Any traffic from the VPLS core destined for R1 is flooded until such time as traffic from R1 is forwarded into the VPLS core.
NoteIn the example shown in Redundant Edge H-VPLS Network with STP Example, the core nodes are unaware that STP redundancy is configured on S1 and S2. There is no STP redundancy configuration on C1 and C2.
The following figure shows an H-VPLS configuration where the STP network directly connects to redundant spoke nodes. A similar configuration is possible where the STP network directly connects to redundant VPLS core nodes. In this case, the core nodes participating in STP are configured for STP redundancy and originate flush messages to their core peers whenever STP causes a flush on an STP port.