Information about Seamless MPLS
The following sections provide information about Seamless MPLS.
Overview of Seamless MPLS
Seamless MPLS provides a highly flexible and scalable architecture to integrate multiple networks into a single MPLS domain. It is based on existing and well known protocols.
A large MPLS network can have several types of platforms and services in different parts of the network. Such a network would usually be divided into areas such as a core area and aggregation areas, and each of these areas have different Interior Gateway Protocols (IGPs). The IGP prefixes from one area cannot be distributed to another area. If the IGP prefixes cannot be distributed, then end-to-end Label-Switched-Paths (LSP) cannot be established. This affects the scalability of the network.
Seamless MPLS introduces greater scalability by establishing end-to-end LSPs. Seamless MPLS uses the Border Gateway Protocol (BGP) instead of IGP to forward the loopback prefixes of the Provider Edge (PE) routers. BGP distributes the prefixes end-to-end. This eliminates the need to install IGP prefixes of one domain in another domain.
Seamless MPLS introduces separation of the service and transport planes and provides end to end service independent transport. It removes the need for service specific configurations in network transport nodes.
Architecture for Seamless MPLS
The figure shows a network with three different areas: one core and two aggregation areas on the side. Each area runs its own IGP, with no redistribution between them on the Area Border Router (ABR). Use of BGP is needed in order to provide an end-to-end MPLS LSP. BGP advertises the loopbacks of the PE routers with a label across the whole domain, and provides an end-to-end LSP. BGP is deployed between the PEs and ABRs.
Seamless MPLs uses BGP to provide an end-to-end MPLS LSP. BGP is deployed between the PEs and the ABRs. BGP sends the IPv4 prefix and label. BGP advertises the loopbacks of the PE routers with a label across the whole domain and provides an end-to-end LSP.
When using IGP in the network, the next-hop address of the prefixes is the loopback prefix of the PE routers. This prefix is not known to the IGP being used in other parts of the network. The next hop address cannot be used to recurse to an IGP prefix. To avoid this the prefixes are carried in BGP. The ABRs are configured as Route Reflectors (RR). And the RRs are configured to set the next hop to self even for the reflected iBGP prefixes.
There are two possible scenarios.
The ABR does not set the next hop to self for the prefixes advertised (reflected by BGP) by the ABR into the aggregation part of the network. The ABR needs to redistribute the loopback prefixes of the ABRs from the core IGP into the aggregation IGP. Only the ABR loopback prefixes (from the core) need to be advertised into the aggregation part, not the loopback prefixes from the PE routers from the remote aggregation parts.
The ABR sets the next hop to self for the prefixes advertised (reflected by BGP) by the ABR into the aggregation part. Because of this, the ABR does not need to redistribute the loopback prefixes of the ABRs from the core IGP into the aggregation IGP.
In both scenarios, the ABR sets the next hop to self for the prefixes advertised (reflected by BGP) by the ABR from the aggregation part of the network into the core part.