- Overview
- Carrier Ethernet
- Coarse Wave Division Multiplexing Solution
- Commercial Services Solution
- IP Video Surveillance
- Layer 2 Virtual
Private Networks - Network Resiliency
- OAM
- Provider Backbone Bridging — Traffic Engineering
- Service Assurance
Hard QoS - Switched Ethernet vs. TDM-PON
- Wireless Backhaul Infrastructure
Layer 2 Virtual Private Networks (L2VPN)
MPLS-based L2VPNs
With additional key concepts defined, we will now introduce Multi-Protocol Label Switching (MPLS) based L2VPNs, which encompass encapsulation mechanisms useful in many metro networks. MPLS may be used to scale and simplify network provisioning. Where Ethernet-based L2VPNs simply add S-VLANs, MPLS encapsulation, as defined in draft Martini, adds a L2 header containing PE destination MAC address, source MAC address, a VC label, and, optionally, an MPLS Tunnel label.
MPLS Tunnel
As larger access and metro edge networks are deployed, logically grouped VCs may be transported in an MPLS tunnel for more efficient transport over existing MPLS core networks. An MPLS tunnel defines an optional transport connection that carries one or more VCs between two provider endpoints.
Utilizing an optional MPLS tunnel results in three service provider benefits:
- Support more subscribers for the same capital investment by encapsulating all subscriber MAC addresses using tunnel MAC addresses defined by the service provider.
- Support more services for the same capital investment by aggregating several VCs using tunnel identifiers defined by the service provider.
- Improve resiliency and service availability by reducing the number of routes that needs to be updated in the event of a network failure.
VPLS/H-VPLS
MPLS-based L2VPNs create a full-mesh of VCs/PWs between devices. Hierarchical VPLS (H-VPLS) was defined by the IETF as a means of enabling larger-scale MPLS access and metro-edge networks. World Wide Packets’ solution enables MPLS-based L2VPNs by creating an MPLS Tunnel and VC/PW between two CE devices, or between a CE and a PE device for H-VPLS.
MPLS signaling
The World Wide Packets solution supports static or dynamic configuration of MPLS-based L2VPNs. Dynamic configuration of VCs is accomplished by using Label Distribution Protocol (LDP). Tunnels are dynamically configured using Resource ReSerVation Protocol Traffic Extensions (RSVP-TE). Provisioning consists of the following four steps:
- Network provisioning – Prior to provisioning VCs and MPLS Tunnels, the network must be provisioned. This step involves port, interface, and platform provisioning. In addition, this step associates the IP addresses with specific physical port/VLAN interfaces and the loopback interface.
- OSPF and LDP provisioning – Once the network is provisioned, Open Shortest Path First (OSPF) is used to set up and identify peer MPLS signaling capable devices and create neighbor adjacencies. Next, LDP is used to create targeted sessions with specific remote PE or CE devices.
- MPLS Tunnel provisioning – RSVP-TE tunnels are then established along a traffic engineered path identified using Constrained Shortest Path First (CSPF) or may be established by following a user configured explicit path to the destination device.
Specific parameters are associated with each MPLS tunnel, including Class of Service (CoS), Time-to-Live (TTL), and bandwidth information. - VC provisioning – Once the MPLS tunnels are provisioned, VCs may be provisioned through LDP.
MPLS Redundancy
Each VS is associated with a primary and optionally a backup VC. VCs, in turn, are associated with a primary and optional backup MPLS Tunnel. This is known as MPLS Tunnel redundancy. The VC/PW redundancy allows you to dual home the service to two different PEs in the same PE mesh.
World Wide Packets solution enables L2VPNs to be carried by redundant MPLS Tunnels. A primary tunnel and a secondary tunnel may be configured for a given VC. If the primary tunnel is available, the VCs are carried by it. If the primary tunnel becomes unavailable, the VCs will switch over to the secondary tunnel.
