Hi folks,
I'm scratching my head with some RR configuration, and I'd like your feeling on it. Is it correct to configure one router to be RR for different clusters? Does it reflects correctly routes from one cluster to another? Let's have an example: You have on one side a dual RR (router A and B, cluster 1) for some routers, on the other side two other RR (router C and D, cluster 2) for some other routers. As they are all part of the same AS, usually you full mesh the 4 RRs together. Now, you want to create for some reasons another cluster (cluster 3) including those 4 RRs (routers A, B, C and D) and some new clients. Does it work to have on A, B, C and D 2 different clusters configured? Is there any recommendation regarding the peering between those 4 RRs (a full mesh within that new cluster, or a full mesh in another group without cluster definition)? I guess having them in the same cluster could reduce the routes exchange between them (same cluster-list so they drop the routes learned from the cluster). Will (A,B) reflects to cluster 1 the routes learned from cluster 3, and vice-versa? I guess it works, but I have never seen such a design, nor configuration anywhere... Kind Regards, Stephane. _______________________________________________ juniper-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/juniper-nsp
