Because of potential interactions when both modifications are used, the policy is always applied first before any other best path selection, which might include IGP-based modifications, takes place. Because it is an established policy, this approach can be good for mandating dedicated exit points for certain destinations. In policy-based selection, an additional routing policy is applied to the information received by RR clients (and perhaps other routers). For details, please see the ORR Internet draft. Interestingly enough, the draft does not contain any details as to how the route selection calculation is changed to complete the IGP metric adjustments needed. When considering the best path to a given destination, the RR can now use the details of the virtual location in the network inside of the “IGP location” to consider the best path. This virtual location can be specified per RR, per peer group, or per particular peer.
In IGP-based selection, the RR is assigned a “virtual” IGP location and metrics. The two modifications are referred to the draft as IGP-based best path selection and policy-based best path selection.
Exporting reachability information from robotstudio software#
The nice things about these modifications is that only the RR devices and not the client routers need to be modified for the new software features of ORR. ORR devices can implement one or both of these modifications as needed to obtain the result desired. To accomplish this, ORR adds two modifications to the RR algorithm and “rules” to find the best path.
With ORR, the router reflector will reflect the proper path from PE2 to PE1, as expected from the topology.
0 kommentar(er)
