Configuring an L2VPN Service on a QinQ Stacking Sub-interface

Usage Scenario

In early stages, QinQ was primarily deployed on CEs on Layer 2 networks. VLAN tags are added to packets using VLAN stacking and services are forwarded on Layer 2 networks based on the outer VLAN tags. QinQ stacking sub-interfaces are configured on PEs to identify user VLANs and add outer VLAN tags to Layer 2 frames.

This implementation, however, faces a problem that one physical interface cannot provide L2VPN access to multiple users. To address this problem, you can configure a QinQ stacking sub-interface and bind it to a VSI or L2VC to provide L2VPN access to multiple users.

QinQ stacking sub-interfaces cannot forward packets at Layer 2 and must be deployed with the L2VPN.

• VPWS

  VPWS is a point-to-point virtual leased line technology and supports almost all link layer protocols. VPWS simulates the traditional leased line services on IP networks and provides asymmetric and low-cost digital data network (DDN) services. For users on both ends of the leased line, VPWS is similar to the traditional leased line services.

• VPLS

  VPLS makes a multipoint-to-multipoint VPN networking possible. With VPLS, the carrier can transmit Ethernet-based multipoint-to-multipoint services for users over an MPLS backbone network.

After receiving a user-destined Layer 2 multicast packet, a QinQ stacking or QinQ mapping sub-interface that connects to a VPLS network removes the outer tag from the packet, adds the learned inner and outer tags to the packet, and then forwards the packet to a downstream device.

Pre-configuration Tasks

Before you configure the QinQ stacking sub-interface provide L2VPN access, plan user VLANs properly so that packets received by QinQ stacking sub-interfaces carry one VLAN tag.