問題文
Two switches at a small site have to carry one VLAN across a routed link between them, and the design keeps to the arrangement that the documentation recommends for a network of that size. What does that arrangement learn addresses by, and what has to be written on each of the two switches?
選択肢
- It learns by flooding in the data plane, but traffic with more than one destination is not carried at all, so a second path outside the tunnel has to be built for it before the VLAN can be stretched across the routed link.
- It learns by flooding in the data plane, and a list of the remote endpoints, replication at the ingress node and an anycast address on lo0 have to be written.
- It learns by flooding in the data plane and needs no list, since the endpoints find one another from the identifier of the VLAN, and the only thing to write is the address that the tunnel is sourced from.
- It learns from the routes that the other endpoint sends, which is what removes the need for a list of endpoints, and the address of lo0, the identifier of the VLAN, the address of the far switch and a route toward it have to be written on each of the two switches.