Hi!
Both Cluster-id and Originator-id can serve the same purpose of avoiding the loops. But the places where it is implemented differs.
Originator ID will be kicked in only when the route originated by the router comes back to it through a diff internal path.
Cluster id on the other hand is used so that the redundant RR in the cluster shouldnot advertise the route again which is advertised by another RR in the same cluster.
But there can exist some scenarios like a 3-4 clusters connected in a cyclic order where the cluster id wont work as the route is again recieved through a different RR out of the cluster. In this case, Originator-id would be useful to prevent the loop.
Regards,
Niranjan
