12.5. Discovery Processes and Hierarchy
12.5.1. Interdomain Discovery
The neighbor discovery process described in Chapter 6 was used by a node to automatically discover its connectivity to its neighbors. A portion of the discovery process must run in-fiber as discussed in Chapter 6.
To facilitate interdomain optical control, it is possible to incorporate the following information in the neighbor discovery process running over a link between nodes in different control domains:
The lowest level routing area that contains both end points of the link.
Additional domain capability information including information concerning types of routing or signaling protocols supported on the interdomain interface, or other domainwide properties such as SRLG information.
Knowing (1) allows two end points, which are in different lowest-level areas, to notify their respective RCs. These RCs may or may not choose to set up a control adjacency. The following example illustrates how the lowest level common routing area could be found.
Consider Figure 12-12, which depicts a network containing six RAs arranged in a hierarchy. Let us look at how the end points of links B and C can determine the appropriate lowest level common RA for each link. For link B, the two RA containment hierarchies, starting from the lowest level RA-ID to the highest level RA-ID are (2112, 113, 1313) and (505, 113, 1313). Suppose each end point is configured with the appropriate area list. By exchanging these two lists, the end points can determine that RA 113 is the lowest level common RA. For link C, the RA lists for the two ends are (505, 113, 1313) and (417, 1313). By exchanging these lists, the end points can determine that RA 1313 is the lowest level common RA.
After determining the lowest level common RA, the link end points can exchange information on the identities of the RCs between which a control adjacency should be established.