Virtual Private LAN Services (VPLS) is a solution that can provide layer 2 Virtual Private Network (VPN) services over Ethernet networks. It uses a combination of Ethernet and MPLS to meet the needs of carriers and customers alike. VPLS allows customer networks at geographically diverse locations to communicate with each other as if they were in the same LAN. The WAN and MAN becomes transparent to all customer locations. Ethernet VPN based on VLPS and MPLS provides more benefits than other alternative layer 2 or 3 VPN technologies:

• Lower capital expenditure required for deploying Ethernet infrastructure by the Service Providers and customers
• Better scalability due to unlimited scalability o MPLS
• Better reliability because MPLS provides many advantageous reliability features
• Better QoS management because the traffic engineering capabilities in MPLS allow providers to support service level guarantees across the entire network.
• Improved OAM: MPLS' dynamic signaling is instrumental in providing quicker changes and reconfigurations of service.
• Protection of investment on existing technologies because VPLS can be used to offer not only Metro Ethernet services but can also interconnect with existing ATM and Frame Relay access networks and IP-VPN core networks running over various core technologies such as Next Generation SONET/SDH and Dense Wave Division Multiplexing (DWDM).

VPLS Standards

VPLS is currently being defined in the Internet Engineering Task Force (IETF) with the broad support of carriers and vendors. Most of VPLS related standards are still in the drafting stage by the l2vpn and pwe3 working groups of IETF. There are primaried two groups of technologies to address point-to-point communication and point-to-multi-point commnucation.

Pseudowires are point-to-point connections setup between pairs of Provider Edge routers. Their primary function is to emulate services like ATM, Frame Relay, Ethernet and TDM over an underlying common MPLS network. To achieve this, each of these technologies is encapsulated into a common MPLS format. These encapsulation standards were previously known as the martini drafts. By encapsulating services into a common MPLS format, pseudowires allow carriers to converge their services to an MPLS network. Ethernet encapsulating pseudowires are the building blocks of VPLS. The pseudowire encapsulation standards are being defined in the IETF's pwe3 working group.

For the point-to-multipoint communication network, the customer sites are connected through a service provider network, which appears as a Layer 2 switch capable of learning and aging. Customer sites are connected to the service provider network at the Provider Edge (PE). All PEs in the network are connected together in a full mesh of tunnels with each tunnel carrying multiple pseudowires. Depending on the location and the number of customer sites, the number of pseduowires setup for a customer/service may range from one (for a customer with only two locations) to a full mesh (for a customer who has locations connected to every PE). All unknown unicast, multicast and broadcast packets are flooded to all the PEs participating in a customer VPN. This network model assumes that all PEs in a service (or VPLS instance) are connected in a full mesh of pseudowires which obviates the need to keep the network loop free. The VPLS network model is being standardized as part of the VPLS drafts in the IETF's l2vpn working group.

To improve its scalability, Hierarchical VPLS (HVPLS) is introduced. The HVPLS standards allow the creation of hierarchies with a hub-and-spoke arrangement. The full mesh of tunnels is maintained between the hub sites (designated as PEs). The CE equipment is connected to an MTU-s router, which is connected to a PE router, thus providing the hierarchy.

Layer 2 Ethernet VPN and Virtual Private LAN Services (VPLS)

Related Terms: VPLS, HVPLS, Ethernet VPN, Layer 2 VPN, pseudowire, MPLS