Configuring PBB VPLS

Fully meshed PBB VPLS

On the HVPLS network shown in Figure 1, PBB VPLS uses the MAC-in-MAC technique instead of the QinQ technique to decrease the number of MAC address entries that SPEs must learn, improving network expansibility.

Figure 1 Fully meshed PBB VPLS

PBB VPLS with an E-Trunk determining the master/backup NPE status

On the network shown in Figure 2, UPE1 is dual-homed to two SPEs, and each SPE is dual-homed to two NPEs. An E-Trunk determines the master/backup NPE status. If SPE1 provides access for CE3, PBB VPLS can then enable CE1 to communicate with CE2 and CE3.

Figure 2 PBB VPLS residential service with an E-Trunk determining the master/backup NPE status

Figure 3 PBB VPLS enterprise service with an E-Trunk determining the master/backup NPE status

Local PBB VPLS

On the network shown in Figure 4, CE1 is dual-homed to two UPEs, and CE1 connects to CE2 through UPE2. An E-Trunk is configured between UPE1 and UPE2 to determine the master/backup UPE status. Normally, UPE1 is the master device and UPE2 is the backup device, and CE1 communicates with CE2 through UPE1 over a remote PBB VPLS connection. If the link between CE1 and UPE1 fails or UPE1 fails, CE1 communicates with CE2 through UPE2 over a local PBB VPLS connection.

In local PBB VPLS, two I-VSIs are bound to the same B-VSI and assigned the same I-tag. One I-VSI's B-DMAC address is the same as the other I-VSI's B-SMAC address.

Figure 4 Local PBB VPLS