для Cisco Cisco IOS Software Release 12.0(22)S

MPLS traffic engineering provides an integrated approach to traffic engineering. With MPLS, traffic 
engineering capabilities are integrated into Layer 3, which optimizes the routing of IP traffic, given the 
constraints imposed by backbone capacity and topology.

MPLS traffic engineering enhances standard Interior Gateway Protocols (IGPs), such as IS-IS or OSPF, 
to automatically map packets onto the appropriate traffic flows.

Transports traffic flows across a network using MPLS forwarding.

Determines the routes for traffic flows across a network based on the resources the traffic flow 
requires and the resources available in the network.

Employs “constraint-based routing,” in which the path for a traffic flow is the shortest path that 
meets the resource requirements (constraints) of the traffic flow. In MPLS traffic engineering, the 
traffic flow has bandwidth requirements, media requirements, a priority that is compared to the 
priority of other flows, and so forth.

Recovers from link or node failures by adapting to the new constraints presented by the changed 
topology.

Transports packets using MPLS forwarding crossing a multihop label-switched path (LSP).

Uses the routing and signaling capability of LSPs across a backbone topology that

Understands the backbone topology and available resources

Accounts for link bandwidth and for the size of the traffic flow when determining routes for 
LSPs across the backbone

Has a dynamic adaptation mechanism that enables the backbone to be resilient to failures, even 
if several primary paths are precalculated off-line

Includes enhancements to the IGP (IS-IS or OSPF) shortest path first (SPF) calculations to 
automatically calculate what traffic should be sent over what LSPs.

WAN connections are an expensive item in an ISP budget. Traffic engineering enables ISPs to route 
network traffic to offer the best service to their users in terms of throughput and delay. By making the 
service provider more efficient, traffic engineering reduces the cost of the network.

Currently, some ISPs base their services on an overlay model. In the overlay model, transmission 
facilities are managed by Layer 2 switching. The routers see only a fully meshed virtual topology, 
making most destinations appear one hop away. If you use the explicit Layer 2 transit layer, you can 
precisely control how traffic uses available bandwidth. However, the overlay model has numerous 
disadvantages. MPLS traffic engineering achieves the traffic engineering benefits of the overlay model 
without running a separate network, and without needing a nonscalable, full mesh of router 
interconnects.

MPLS traffic engineering automatically establishes and maintains LSPs across the backbone by using 
RSVP. The path that an LSP uses is determined by the LSP resource requirements and network resources, 
such as bandwidth.

Available resources are flooded by means of extensions to a link-state based IGP.