Three layer virtual private network and its construction method

Abstract

A virtual private network of three layers incldues P and PE equipments of a backbone network, a number of sites and CE equipments of virtual private network (VPN) for the client, layering PE equipment which is used as router as the edge of the backbone network and being connected with P equipment of the backbone network and the sites, CE equipment of VPN as well as includes lower layer PE, (VPE) zero or a number of middle layer PE (MPE) and upper laeyr PE (SPE) to undertable different task of the functions in one certralized type PE. It has high requirements for the performance of PE route and retransmission at lower layer when BGP / MPLS VPN Is arranged in layering disposition for solving the problem of expansibility.

Description

technical field [0001] The invention relates to a network structure of a three-layer virtual private network (PP VPN, Provider Provide VPN) provided by an operator and a construction method of the network. Background technique [0002] Existing PP VPNs have many different forms in specific applications, such as BGP / MPLS VPN, VR VPN, and MPLS 12 VPN. Taking Border Gateway Protocol-based Multi-Protocol Label Switching VPN (BGP / MPLS VPN) as an example, PP VPN is a network that uses MPLS technology and Border Gateway Protocol (BGP protocol) in public networks to provide IP VPN services. It mainly includes A backbone network composed of P equipment and PE equipment provided by an Internet access service provider (ISP), and a user's virtual private network (VPN), and the VPN mainly includes multiple sites (Sites) and CE equipment. [0003] Among the above devices, the P (Provider Router, backbone network core router) device is mainly responsible for MPLS forwarding. PE (Provider...

AI Technical Summary

Problems solved by technology

[0005] 1) A large number of interfaces and sub-interfaces are required between PE and VCE, so they consume limited interface and IP address resources;

[0006] 2) Multiple VRFs need to be configured on PE and CE, and the configuration workload is heavy and repetitive;

[0007] 3) When exchanging routes between PE and VCE, if a dynamic routing protocol is used, both PE and VCE need to run multiple instances, and if static routing is used, the configuration workload will be heavy;

[0008] 4) If the PE and VCE are not directly connected, but using a tunnel interface, each VRF needs a tunnel, which consumes a lot of resources;

Images