MPLS L2VPN network models

As shown in Figure 67, this model connects two CEs through a PW on an MPLS or IP backbone.

Figure 67: MPLS L2VPN network model

To set up an MPLS L2VPN connection:

  1. Set up a public tunnel to carry one or more PWs between PEs:

    The public tunnel can be an LSP or MPLS TE tunnel.

    If multiple public tunnels exist between two PEs, you can configure a tunnel policy to control tunnel selection. For more information about tunnel policies, see "Configuring tunnel policies."

    If a PW is established over an LSP or MPLS TE tunnel, packets on the PW have two labels. The outer label is the public LSP or MPLS TE tunnel label that MPLS uses to forward the packet to the peer PE. The inner label is the PW label that the peer PE uses to forward the packet to the destination CE.

  2. Set up a PW to connect customer networks:

    PWs include static PWs, LDP PWs, BGP PWs, and Circuit Cross Connect (CCC) PWs.

    To establish a static PW, configure the peer PE address, and the incoming and outgoing PW labels for the PW on the two PEs. Static PWs consume a small amount of resources but have complex configurations.

    To establish an LDP PW, configure LDP and specify the peer PE address on the two PEs. LDP defines a new FEC type named PW ID FEC for PEs to exchange PW-label bindings. The new FEC type uses a PW ID and a PW type to identify a PW. The PW ID is the ID of the PW between PEs. The PW type specifies the encapsulation type for data transmitted over the PW, such as ATM, FR, Ethernet, or VLAN. PEs advertise the PW label and PW ID FEC in label mapping messages to create a PW. Dynamic PWs have simple configurations but consume more resources than static PWs.

    To establish BGP PWs, BGP advertises label block information in an extended BGP update to PEs in the same VPN. Each PE uses the received label block information to calculate outgoing labels and uses its own label block to calculate incoming labels. After two PEs complete label calculation, a BGP PW is established between them. BGP PWs have the following features:

    • Simplified configuration—There is no need to manually specify peer PEs. A PE automatically find peer PEs after receiving label block information from the peer PEs.

    • Reduced workload—Label block advertisement enables assigning labels for multiple PWs at one time.

    To establish a CCC PW, manually specify the incoming and outgoing labels for the CCC PW on the PEs, and create two static LSPs in opposite directions on P devices between PEs. CCC employs only one level of label to transfer packets. The static LSPs on the P devices transfer data only for the CCC PW. They cannot be used by other connections or MPLS L3VPN.

  3. Set up an AC between a PE and a CE:

    Set up an AC by configuring a link layer connection between a PE and a CE.

    An AC is a service instance on a Layer 2 Ethernet interface or Layer 2 aggregate interface. The AC forwards packets that are received on the interface and meet the match criteria of the service instance to the bound PW.

  4. Bind the AC to the PW:

    Bind the service instance to the PW, so the PE forwards packets between the AC and the PW.