Kompella Mpls L2Vpn; Configuring Mpls L2Vpn - 3Com MSR 50 Series Configuration Manual

Kompella MPLS L2VPN

Configuring MPLS
L2VPN
Once LDP establishes an LSP between the two PEs and the label exchange and the
binding to CE are finished, a VC is set up and ready to transfer Layer 2 data.
To allow the exchange of VC labels between PEs, the Martini method extended
LDP by adding the forwarding equivalence class (FEC) type of VC FEC. Moreover,
as the two PEs exchanging VC labels may not be connected directly, a remote LDP
session must be set up to transfer the VC FEC and VC labels.
With Martini MPLS L2VPN, only PEs need to maintain a small amount of VC labels
and LSP mappings and no P device contains Layer 2 VPN information. Therefore, it
has high scalability. In addition, to add a new VC, you only need to configure a
one-way VC for each of the PEs. Your configuration will not affect the operation
of the network.
The Martini method applies to scenarios with sparse Layer 2 connections, such as
a scenario with a start topology.
Kompella MPLS L2VPN is different from Martini MPLS L2VPN in that it does not
operate on the connections between CEs directly. It organizes different VPNs in
the whole service provider network and encodes each CE in a VPN. For a
connection to be established between two CEs, you only need to perform these
tasks on the PEs:
Configuring CE IDs of the local and remote CEs respectively
■
Specifying the circuit ID that the local CE assigns to the connection, such as the
■
VPI/VCI with ATM.
Kompella MPLS L2VPN uses extended BGP as the signaling protocol to distribute
VC labels. Its label block mode allows it to assign labels to multiple connections at
a time.
With Kompella MPLS L2VPN, you can specify a local CE range to indicate how
many CEs can be connected with a CE. Then, the system assigns a label block of a
size equal to the CE range for the CE. In this way, you can reserve some labels for
the VPN for future use. This wastes some label resources in a short term, but can
reduce the VPN deployment and configuration workload in the case of expansion.
Imagine that an enterprise VPN contains 10 CEs and the number may increase to
20 in future service expansion. In this case, you can set the CE range of each CE to
20. Thus, when you need to add a CE to the VPN later, you only need to modify
the configurations of the PE to which the new CE is connected. No change is
required for the other PEs. This makes VPN expansion extremely simple.
Similar to MPLS L3VPN, Kompella MPLS L2VPN also uses VPN targets to identify
VPNs. This brings excellent VPN networking flexibility.
In addition, Kompella MPLS L2VPN supports local connections and inter-provider
VPN solutions.
You can select any of the implementation methods for MPLS L2VPN as needed.
However, no matter what method you select, you must complete these two tasks:
Configuring MPLS basic capability
■

Configuring MPLS L2VPN

1429