Using mpls in traffic engineering – Brocade Multi-Service IronWare Multiprotocol Label Switch (MPLS) Configuration Guide (Supporting R05.6.00) User Manual

Page 33

Advertising
background image

Multi-Service IronWare Multiprotocol Label Switch (MPLS) Configuration Guide

9

53-1003031-02

Using MPLS in traffic engineering

1

The EXP field is designated for experimental usage. By default, a device uses the EXP field to define
a Class of Service (CoS) value for prioritizing packets travelling through an LSP. Please refer to

Chapter 1, “Configuring MPLS Traffic Engineering”

, for more information. Note that software

forwarded VPLS packets do not use the EXP encode table.

S (Bottom of Stack) field (1 bit)

An MPLS packet can be assigned multiple labels. When an MPLS packet has multiple labels, they
are logically organized in a last-in, first-out label stack. An LSR performs a pop or swap operation on
the topmost label; that is, the most recently applied label in the stack. The Bottom of Stack field
indicates whether this label is the last (oldest) label in the stack. When the label is the last one in
the stack, the Bottom of Stack field is set to one. If not, the Bottom of Stack field is set to zero.

A device acting as an LSR can perform one push, swap, or pop operation on an incoming MPLS
packet. The device can accept MPLS packets that contain multiple labels, but only the topmost
label is acted upon.

TTL field (8 bits)

The TTL field indicates the Time To Live (TTL) value for the MPLS packet. At the ingress LER, an IP
packet’s TTL value is copied to its MPLS TTL field. At each transit LSR hop, the MPLS TTL value is
decremented by one. When the MPLS TTL value reaches zero, the packet is discarded. Optionally,
the user can configure the LSRs not to decrement the MPLS TTL value at each hop.

Using MPLS in traffic engineering

Traffic engineering is the task of routing network traffic to avoid points of congestion and make
efficient use of high bandwidth interfaces. When used as an application of MPLS, traffic
engineering involves creating LSPs that make the best use of available network resources; that is,
traffic-engineered LSPs. This section explains the process of creating traffic-engineered LSPs.

Creating traffic-engineered LSPs involves the following tasks:

Gathering information about the network

Using the gathered information to select optimal paths through the network

Setting up and maintaining the paths

For traffic-engineered signaled LSPs, devices can perform these tasks dynamically.

Figure

5

illustrates the process that takes place to configure, establish, and activate traffic-engineered
signaled LSPs.

NOTE

Adaptive LSPs can have primary and secondary sessions up at the same time. Brocade devices only
support 16k LSPs, and no more than a total of 32k sessions.

Advertising