In some cases, existing IP routers are deployed as spokes. As previously described, the IETF calls such a device a PE-r, to indicate routing functionality only. These routers are not capable of bridging and cannot switch packets based on MAC addresses. To offer an L2 service using the VE^ it |s possible to create VWs between the VE-r and the CO VE-rs, where all the L2 switching functions are done at the CO. This model creates more overhead, because unlike the MTU-s, where all access ports belonging to the same VVLS are mapped to a single PW, the VE-r requires that each a ccess port is mapped to its own PW. This is illustrated in Figure 4-21.
Qjgute <4-2 1 us es a PE-o as a spoke. Note that VVLS A n ow dequ^es two VWs—one foe CE1e and me for CE 12—that: belong te the same customer. For any l:r af'fic that needs to ire swi tche d notweeu th e two access ports of thq same ctstom eo t hat are conneote d to CEH and CE12, trat 0ra ffic needs to be ttansporte d to uhe CO ahd sw Itch ed at the VE-rs.
Was this article helpful?