Failure Recovery Scenario 3

Now suppose there is a failure on the shared port link (point 3) and on the access ring at point 1.

The recovery actions for this double-failure need to be somewhat different. In this case, even though the core link has failed, both core nodes do not receive a copy of the ring traffic. For example, the only path to the VPLS network for Dist 1 is through the controller core node. In this case, the controller node does not take down its PWs.

It is possible that the customer access network could have parallel EAPS rings that attach to Core 1 and Core 2 as shown in VPLS with Parallel Redundant EAPS Rings Configuration Example. In this example, the network connections are broken at each point X and as long as any of the parallel EAPS rings are complete, there is a path to both core VPLS nodes. Thus, the controller node must take down its PWs as long as any of the parallel EAPS rings is complete.

Click to expand in new window
VPLS with Parallel Redundant EAPS Rings Configuration Example

PW Management for Network Failures

Controller Node Responses to EAPS Ring States shows how the controller node responds to multiple failures in a network with parallel EAPS rings.

Table 1. Controller Node Responses to EAPS Ring States
Ring State Common Link Up Common Link Down
Any parallel ring Up PWs Active PWs Inactive
All parallel rings Down PWs Active PWs Active