IBM 6 MPLS User manual

Specialized Models User Guide 6 MPLS Model User Guide
Modeler/Release 10.0 SPM-6-1
6 MPLS Model User Guide
Multi-ProtocolLabelSwitching(MPLS)isamulti-layerswitchingtechnologythat
uses labels to determine how packets are forwarded through a network. The
firstpartofthisdocumentdescribeskeyfeaturesoftheMPLSspecializedmodel
and the second part focuses on procedures for configuring MPLS in your
network model.
Model Features
This section contains a list of the main features available in the Multi-Protocol
Label Switching model:
• The MPLS model captures the following protocol behavior:
Table 6-1 MPLS Model Features
Feature Description
LSP (Label Switched Path) configuration • LSPs can be created manually or
automatically from traffic conversation
pairs.
• LSPs are easily reused in other scenarios
or projects by using the LSP import and
export features.
• Both dynamic and static LSPs are created
using the path object.
Differential Services (DiffServ) • DiffServ extensions, as defined in
RFC-2475, are provided.
• The model enables you to perform QoS
(quality of service) analyses by accounting
for different types of service.
Traffic Engineering Traffic engineered routes are computed
using Constrained Shortest Path First
(CSPF)withOSPForIS-ISroutingprotocols.
End of Table 6-1

Specialized Models User Guide 6 MPLS Model User Guide
SPM-6-2 Modeler/Release 10.0
• MPLS models are implemented based on information available from the
following sources.
Table 6-2 Reference Documents
Model Features Document
Traffic Engineering
MPLS TE RFC-2702—Requirements for Traffic Engineering
Over MPLS
FECs RFC-3031—Multiprotocol Label Switching
Architecture
IGP shortcuts draft-hsmit-mpls-igp-spf-00
Label Switched Paths
Dynamic LSPs
Static LSPs RFC-3031—Multiprotocol Label Switching
Architecture
LSP routing
OSPF TE
IS-IS TE RFC-2676—QoS Routing and OSPF Extensions
Label distribution
LDP RFC-3036—LDP Specification
CR-LDP RFC-3212—Constraint-based LSP Setup Using LDP
RSVP-TE RFC-3209—RSVP-TE: Extensions to RSVP for LSP
Tunnels
PP-VPNs
A framework for layer-3 PP VPNs RFC-2547—BGP/MPLS VPNs
BGP/MPLS VPNs draft-ietf-ppvpn-framework-05
Quality of Service
QOS Architecture RFC-2475—An Architecture for Differentiated
Services
MPLS Support of Differentiated
Services RFC-3270—Multi Protocol Label Switching draft-ietf-mpls-diff-ext-08
Restoration and Resiliency
Fast reroute with bypass tunnels
LSP protection with ingress backup
draft-ietf-mpls-rsvp-lsp-fastreroute-00
End of Table 6-2

Specialized Models User Guide 6 MPLS Model User Guide
Modeler/Release 10.0 SPM-6-3
Node Models
The MPLS model suite supports workstation, server, router, and link models
from the standard model library. The LER (Label Edge Router) and LSR (Label
SwitchingRouter) node models in the MPLS object paletteare preconfigured to
support MPLS. However, you can configure any of the router models in the
standard model library to model LERs and LSRs.
Figure 6-1 MPLS Object Palette
Model Attributes
Global MPLS attributes, which are used to configure network-wide MPLS
parameters, are grouped in the MPLS configuration object. Router-specific
MPLS attributes are grouped in the MPLS Parameters attribute on each router.
They are described in Router Attributes on page SPM-6-6.
MPLS Configuration Object Attributes
Some of the important MPLS configuration object attributes are described
below.
•FEC Specifications This attribute specifies the Forwarding Equivalence
Class (FEC) parameters used by MPLS in the network. FECs classify and
group packets so that all packets in a group are forwarded the same way.
FECs are based on any of the IP header fields—ToS, Protocol, Source
Address Range, Destination Address Range, Source Port, and Destination
Port can all be used to define a FEC. Figure 6-2 Specifying FEC Attributes
on page SPM-6-4 shows the attribute sequence for defining an FEC.

Specialized Models User Guide 6 MPLS Model User Guide
SPM-6-4 Modeler/Release 10.0
Figure 6-2 Specifying FEC Attributes
The FEC Details Table helps define the FEC through a set of match rules,
which are combinations of TCP, UDP, and IP header fields. FECs are
determined by taking a logical AND of the column settings in a row and by
taking a logical OR of each of the rows. In other words, for a packet to qualify
for a particular FEC, the IP header fields must satisfy every condition of at
least one row of the defined FEC. For example, a FEC that consists only of
email and ftp traffic would be specified as shown in Figure 6-3.
Figure 6-3 FEC Details for E-mail and FTP Traffic
Therefore,iftheIPheaderofapacketcontained eitheremailor FTP, it would
qualify for the FEC defined in Figure 6-3, and would be sent over the
corresponding LSP.

Specialized Models User Guide 6 MPLS Model User Guide
Modeler/Release 10.0 SPM-6-5
•LSP Specification File This attribute indicates whether the network LSPs
should be configured according to the text file specified. You can update the
text file by clickingOK in the LSP Browser. Updating the file recreates the file
based on the current network LSP settings, including LSPs that might not
have been in the original file (such as those created manually).
•Traffic Trunk Profiles This attribute specifies out-of-profile actions and
traffic classes for traffic trunks in the network. Traffic trunks capture traffic
characteristics such as peak rate,average rate, and average burst size. The
default Trunk Details setting configures a trunk with a value of
32,000 bits/sec for maximum and average bit rate and 32,000 bits for
maximum burst size.
Figure 6-4 Specifying Traffic Trunk Profiles
•EXP <--> Drop Precedence and EXP <--> PHB These attributes specify
how EXP bits in the MPLS shim header are translated into diffserv
information at each LSR. For E-LSPs, LSRs determine Per Hop Behavior
(PHB), while on L-LSPs, they determine Drop Precedence. Use the default
setting unless you are analyzing the effects of QoS.

Specialized Models User Guide 6 MPLS Model User Guide
SPM-6-6 Modeler/Release 10.0
Figure 6-5 Mapping EXP Bits to Drop Precedence and PHB
Router Attributes
Someof theimportantMPLSParametersattributessetonroutersaredescribed
below.
Traffic Mapping Configuration
This attribute specifies bindings between FECs and LSPs. Each row of the
Traffic Mapping Configuration table specifies a distinct traffic engineering (TE)
binding.EachTE bindingspecifiestheFEC,traffictrunk,andLSPthatisapplied
to the label of the incoming packet.
Only previously defined values appear in the attribute pull-down lists. If no
values appear in the attribute pull-down lists, verify that you have defined the
FECs and traffic trunks in the MPLS Configuration object, and that the LSPs
appear in the network path browser.

Specialized Models User Guide 6 MPLS Model User Guide
Modeler/Release 10.0 SPM-6-7
When an unlabeled packet arrives at an ingress LER, the following sequence
occurs to determine the correct label for the packet:
1) The TE binding is selected based on the packet FEC and the incoming
interface.
2) The packet is checked to make sure that its traffic characteristics conform
to those specified for the TE binding’s traffic trunk.
3) Thepacketis labeled forand sentthroughthe primaryLSPspecifiedforthe
TE binding.
Figure 6-6 Configuring TE Bindings
\
•EXP <--> Drop Precedence and EXP <--> PHB These attributes specify
which mappings, defined in the MPLS configuration object, are used by the
router.
•LDP Parameters—specifies Label Distribution Protocol parameters used by
the LSR. LDP Parameters is a compound attribute, composed of the
following sub-attributes:
—Discovery Configuration—specifies Hello message parameters
needed to learn of neighboring routers
—Session Configuration—specifies Keep-alive message parameters
used to establish LDP sessions
—Recovery Configuration—specifies how node and link failures are
detected
This weight attribute
configuration uses LER2-LER5
75%ofthetimeandLER2-site9
25% of the time.

Specialized Models User Guide 6 MPLS Model User Guide
SPM-6-8 Modeler/Release 10.0
Figure 6-7 Configuring LDP Parameters
Simulation Attributes
The following simulation attributes are available (Configure/Run Discrete Event
Simulation dialog box) when using the MPLS model suite.
•CR-LDP Routing—specifies if CR-LDP routing uses CSPF or conventional
IGP to determine routes in loosely defined LSPs. The default value is IGP.
•CSPF Retry Timer—specifieshowlongan ingressLER waitsafterdetecting
a node or link failure before rerouting an LSP that traverses the failed node
or link. The default value for this attribute is 45 seconds.
•LDP Discovery End Time—specifies when LDP discovery ends. After this
time, no more LDP discovery packets are sent through the network. This
value should occur after the network reaches a final, constant state in the
simulation since no network topology or device status changes are reflected
in the LDP routing tables after LDP Discovery End Time.
•LDP Discovery Start Time—specifies when LDP starts sending discovery
packets through the network. Set this attribute to a value other than Do Not
Start to enable LDP.
•LSP Signaling Protocol—specifies whether dynamic LSPs are signaled
using CR-LDP (constraint-based routed LDP) or RSVP. The default value is
CR-LDP.

Specialized Models User Guide 6 MPLS Model User Guide
Modeler/Release 10.0 SPM-6-9
LSP Attributes
Some of the important LSP attributes are described below. Most of these
attributes can also be configured in the LSP browser, which is described in the
next section.
Figure 6-8 Configuring an LSP’s Attributes
•Directionality—specifiesifanLSPisunidirectionalorbidirectional.Dynamic
LSPs are always unidirectional.
•LSP Type—specifies whether the LSP is of type E-LSP or L-LSP. For
E-LSP, three experimental bits in the shim header carry the Diff-Serv
information. Thisprovides eight different types ofservice (TOS)per LSP. For
L-LSP, TOS information is contained in the MPLS label and all packets
traversing the link are treated equally.
•Path Details—specifies which packets use the LSP and defines how
packets are forwarded through the LSP. This attribute is automatically
configured for dynamic LSPs. To configure this attribute for static LSPs,
select Update LSP Details from the Protocols > MPLS menu.
Figure 6-9 Path Details for a Static LSP
•Recovery Parameters—specifies recovery parameters that are used to
reroute traffic on this LSP if there is a link or node failure along the LSP.

Specialized Models User Guide 6 MPLS Model User Guide
SPM-6-10 Modeler/Release 10.0
Figure 6-10 Recovery Parameters Configuration
•Setup Parameters—specifies the duration of the LSP.
Figure 6-11 Setup Parameters Configuration
•TE Parameters—specifies the traffic engineering constraints used by
CR-LDP to find a route through the network. CR-LDP uses Constrained
Routing to find the route that is the best fit for the specified constraints. This
attribute applies to dynamic LSPs only. Make sure you account for network
bandwidth availability when configuring static LSPs.
Figure 6-12 TE Parameters Configuration