MPLS Configuration

Table of Contents

Table of Contents
Chapter 1 MPLS Configuration............................................................................................................................................. 1
1.1 MPLS Overview..................................................................................................................................................... 1
1.2 Related Concepts of MPLS ................................................................................................................................... 1
1.2.1 Forwarding Equivalence Classes .............................................................................................................. 1
1.2.2 Label.......................................................................................................................................................... 2
1.2.3 Label Forwarding Information Database ................................................................................................... 4
1.3 Reading Before MPLS Configuration .................................................................................................................... 5
1.4 MPLS Configuration .............................................................................................................................................. 6
1.4.1 MPLS Configuration Task List ................................................................................................................... 6
1.4.2 Enabling MPLS.......................................................................................................................................... 6
1.4.3 Configuring Static LSP .............................................................................................................................. 7
1.4.4 Configures the label range. ....................................................................................................................... 8
1.4.5 Clears the MPLS counter. ......................................................................................................................... 9
1.4.6 Browsing MPLS Information...................................................................................................................... 9
1.5 LDP Configuration ............................................................................................................................................... 10
1.5.1 LDP Configuration Task List .................................................................................................................... 10
1.5.2 Configuring LSR Label ............................................................................................................................ 10
1.5.3 Enabling LDP .......................................................................................................................................... 11
1.5.4 Configuring Transmission Address.......................................................................................................... 11
1.5.5 Configuring the Address List of Label Distribution................................................................................... 12
1.5.6 Configuring the Parameter About Local Discovery Mechanism .............................................................. 12
1.5.7 Configuring the LDP Neighbor ................................................................................................................ 13
1.5.8 Configuring the Discovery Mechanism of a Designated Neighbor.......................................................... 13
1.5.9 Configuring the Holdtime of the Session ................................................................................................. 14
1.5.10 Force Explicit-Null Label........................................................................................................................ 14
1.5.11 Configuring LDP Log Record................................................................................................................. 15
1.5.12 Browsing LDP Information..................................................................................................................... 15
Chapter 2 MPLS L2VPN Configuration .............................................................................................................................. 16
2.1 Overview of MPLS L2VPN .................................................................................................................................. 16
2.2 Related Concepts of MPLS L2VPN..................................................................................................................... 16
2.2.1 VFI........................................................................................................................................................... 16
2.2.2 Pseudo-Wire............................................................................................................................................ 16
2.2.3 L2VPN Network Without Loopback......................................................................................................... 17
2.3 MPLS L2VPN Configuration................................................................................................................................ 17
2.3.1 MPLS L2VPN Configuration Task List..................................................................................................... 17
2.3.2 Enabling MPLS L2VPN ........................................................................................................................... 17
2.3.3 Creating VFI ............................................................................................................................................ 17
2.3.4 Binding VFI to VLAN Interface ................................................................................................................ 18
2.3.5 Configuring Incoming VFI Interface......................................................................................................... 19
2.3.6 Configuring a Static VC Label ................................................................................................................. 19
2.3.7 Browsing L2VPN Information .................................................................................................................. 20

-I-
 Table of Contents

2.4 MPLS L2VPN Configuration Example ................................................................................................................. 20

2.4.1 Point-to-Point L2VPN Configuration Example......................................................................................... 20
Chapter 3 MPLS L3VPN Configuration .............................................................................................................................. 24
3.1 Overview of MPLS L3VPN .................................................................................................................................. 24
3.2 MPLS L3VPN Configuration................................................................................................................................ 25
3.2.1 MPLS L3VPN Configuration Task List..................................................................................................... 25
3.2.2 Configuring MPLS ................................................................................................................................... 25
3.2.3 Configuring LDP...................................................................................................................................... 26
3.2.4 Configuring VRF...................................................................................................................................... 26
3.2.5 Configuring a VPN Route........................................................................................................................ 27
3.2.6 Configuring a Route Between PE and P ................................................................................................. 27
3.2.7 Configuring a BGP Route Between PEs ................................................................................................. 28
3.3 MPLS L3VPN Configuration Example ................................................................................................................. 29
3.3.1 Configuring S1 (CE) ................................................................................................................................ 29
3.3.2 Configuring S2 (PE) ................................................................................................................................ 30
3.3.3 Configuring S3 (P)................................................................................................................................... 32

- II -
 MPLS Configuration

Chapter 1 MPLS Configuration

1.1 MPLS Overview

Multiprotocol Label Switching (MPLS) is used to encapsulate packets from the network
layer with short and long-specific labels. MPLS separates route choosing and data
forwarding, and designates the network path of a packet with a label. The MPLS
network consists of the Label Switching Router (LSR) and the Label Edge Router
(LER). Seen as the comibination of the ATM switch and the traditional router, the LSR
consists of the control unit and the switching unit. The LER is to analyze the headers of
IP packets to decide the corresponding transmission grade and the label switching
path.

In the MPLS network, the device supporting label switching is called as LSR. The
connection of these LSRs is composed of the Label Switching Path (LSP). The label
information can be switched among LSRs through the Label Distribute Protocol (LDP),
the expanded routing protocol and other methods. The ingress LSR on LSP adds a
label to a packet according to the network-layer information (for example, the IP
header), while the egress LSR will remove the label. Other LSRs just forward the
packet according to the label.

The MPLS network adds labels to packets, which makes data forwarding more
connection-oriented though the control function of the MPLS network is based on the
current IP technology. The LSR role can be taken by many types of MPLS-supported
devices, such as the router or the layer-3 switch.

MPLS is originally a protocol that is used to improve the forwarding speed of the router;
however, thanks to its fixed advantages, MPLS is also applied on traffic engineering,
VPN, etc. Therefore, it gradually becomes a vital standard of a large-scale IP network.

1.2 Related Concepts of MPLS

1.2.1 Forwarding Equivalence Classes

Forwarding Equivalence Classes (FEC) is an important concept in MPLS. FEC

presents a group of packets with the same forwarding method, such as having the
same address prefix and the same forwarding path.

When a packet is transmitted in a network under the effect of the connectionless

network protocols, each router that the packet passes through needs to conduct path
choice independently to the packet. The procedure of selecting the next hop can be
considered as performing the following two steps: (a) Classifying packets into different
FECs; (b) Deciding the next hop for each FEC. If only the function of forwarding
packets is considered, the packets that are classified to pass through an FEC are
same because they are transmitted in the same path.

In traditional IP packet forwarding, the reasom why a router thinks two packets are in
the same FEC is that the destination addresses of the two packets can match up with

-1-
 MPLS Configuration

a certain item in the routing table at full length. On each hop of the packet, the router
needs to recheck this packet to decide the FEC that the packet belongs to.

However, in MPLS, the operation to assign a packet to a FCC is conducted only once
when the packet enters the network. On the ingress router of the network, the FEC that
a packet belongs to is presented as a value with fixed length, that is, a label, and the
label will be placed into the packet. At the next hop, the router does not need to check
the network-layer header of the packet any more, but directly decides the new next
hop of the packet according to the label, and then replaces the previous label with a
new label and forwards it out.

1.2.2 Label

1. Definition

The label is a short, long-specific and meaningful identifier symbol only between two
neighboring LSRs, which is used to represent as a FEC. In general, The FEC where
the packet lies is up to the destination address of the network layer.

The label is meaningful only between two neighboring LSRs. For example, both router
Ru and router Rd support label switching, and they agree on this: if a packet belongs
to FEC F, router Ru adds label L to the packet when router Ru sends the packet to
router Rd, that is, a binding relation is established between Label L and FEC F. The
binding relation is valid to those packets that are transmitted from router Ru to router
Rd. It needs to be noted that label L is just a random value used between router Ru
and router Rd.

2. Structure of the Label

MPLS is to add multiple labels to a packet to form a label stack in the packet.
RFC3032 defines the label stack’s structure which is applied in the PPP link or the
Ethernet link. The structure of each item in the label stack is shown in figure 1.1.

Figure 1.1 Structure of the item in the label stack

Label: Real value of the stack label, whose length is 20 bits.

Exp: Reserved, generally appearing as COS, whose length is 3 bits

S: If the item is the last one in the label stack, S is set to 1, meaning the stack’s bottom.
S of other items is 0, whose length is 1 bit.

TTL: It has the same meaning as the TTL of the IP header, whose length is 8 bits.

3. Label's Operation

(1) NHLFE

-2-
 MPLS Configuration

The next-hop label forwarding entry is used to forward packets with label. It contains
the following information: (a) the next hop of the packet; (b) operating the label stack of
the packet; (c) other information.

(2) Label mapping

Label mapping means mapping a packet (with or without label) to one NHLFE item or
a group of NHLFE items to decide the operation mode to the packet. Label mapping
includes the label mapping at the egress router and the mapping at LSP.

The label mapping at the ingress router is also called as FEC-to-NHLFE map. After the
LSR confirms the FEC of the packet, each FEC is mapped to a group of NHLFEs.
handling the unlabelled packets at the ingress.

A mapping in LSP is also called an Incoming Label Map (ILM), which is used to map
each incoming labeled packet to one NHLFE or a group of NHLFEs.

(3) Label distribution mode

Label distribution is a process to establish LSP for a certain FEC.

Mpls supports two label distribution modes: Downstream Unsolicited (DU) and
Downstream On Demand (DoD).

In DU mode, LSR will actively distribute the labels and the FEC binding information
without receiving the request.

In DoD mode, LSR will distribute labels only after the request information is obtained.

Note:
BDCOM hi-end switches of 4.0.1A version and its former versions support the
automatic label distribution mode.

(4) Label distribution control mode

MPLS has two label distribution control modes: Independent and Ordered.

In independent mode, if a LSR receives the label requests from other LSRs, the LSR
can give prompt response to other LSRs without downstream-distributed label binding.
The independent control mode allows LSR to distribute the label binding information to
other LSRs at any time.

In ordered control mode, LSR can learn the label binding information only from the
next stop, or LSR itself is the egress router of LSP. Hence LSR can distribute the label
binding information to other LSRs.

Note:
BDCOM hi-end switches of the 4.0.1A version and its former versions support the
orderly label distribution mode.

(5) Label retention mode

Mpls has two label retention modes: conservation and liberal.

In conservation mode, LSR only saves the label binding information that it requires, not
all received information.

-3-
 MPLS Configuration

In liberal mode, LSR saves all received label binding information.

Note:
BDCOM hi-end switches of 4.0.1A version and its former versions support the free
label maintenance mode.

4. LDP

Label Distribution Protocol (LDP) applies the hop-to-hop or dynamic label distribution
arithmetic to establish the try-the-best LSP among LSRs through distributing labels to
the routes learned by IGP.

LDP mainly runs on reliable TCP connections (unless UDP is used). LDP mainly
executes the following four operations:

Discovery: It is to release and maintain the LSRs in the network.

Session: It is to establish and keep the sessions between two peer LDPs.

Advertisement: It is to distribute and spread the labels.

Notification: It is to notify the error information.

1.2.3 Label Forwarding Information Database

BDCOM Ethernet switch series are to maintain the FTN and LIM items that are used
by MPLS packet forwarding through the label forwarding information database. The
label forwarding information database consists of three parts: software forwarding
information database, adjacent information database and hardware forwarding
information database.

1.1.1.1 MLIB

MLIB is the simplified form of MPLS label information base. MLIB saves all forwarding
information, including forwarding information that is generated by dynamic LDP and
forwarding information that is configured by users manually. All the forwarding
information varies with the states of LDP and the change of user's configuration.

The MPLS forwarding module realizes the CPU forwarding of the MPLS packets and
the IP packets through MLIB query.

In the system of BDCOM S8500 switches, the MSU and each line card have
corresponding MLIBs. All forwarding information is first generated in MLIB of the MSU
and then synchronized to all MPLS line cards by MSU. MSU is always consistent with
the MLIB of the MPLS.

1.1.1.2 Adjacent Information

The adjacent information about MPLS forwarding item includes:

-4-
 MPLS Configuration

The next-hop ARP of the normal forwarding item, or the actual information about the
next hop when the next hop of the forwarding item is indirect. (It is the MPLS
forwarding item generated by MPBGP).

In the system of S8500, MSU and MPLS line card maintain the independent adjacent
information list respectively. The switch can forward the MPLS packets only when the
adjacent information is complete.

The case that the adjacent information is incomplete is also called that the adjacent
information is unresolved. For example, the ARP of the next hop’s address does not
exist. In this case, the switch forwards the packets to CPU according to
user-configured policies and then waits for ARP, or directly drops the packets.

Note:
Bdcom hi-end switches of the 4.0.1A version and its former versions will drop the
packets before the resolution of the neighbors so that the CPU resource is avoided
being occupied.

1.1.1.3 LFIB

LFIB is the simplified form of MPLS label forwarding information base. LFIB is
generally a subnet of MLIB, representing the forwarding items that have been added tp
the switch’s hardware forwarding table.

LFIB differs from MLIB at the following points:

(1) Only the next hop's IP address exists in MLIB, while the IP address, the MAC
address and the VLAN information of the next hop exist in LFIB. All the previous
information is obtained from the adjacent information table.

(2) The forwarding item in MLIB only has the outgoing logic interface (VLAN interface),
while LFIB also includes the outgoing physical interface.

In the system of S8500, MSU and MPLS line card maintain the independent LFIB
respectively.

1.3 Reading Before MPLS Configuration

Before configuring the MPLS function on a switch, please read the following content
first:

 The Ethernet switch depends on the hardware to realize the wire-speed

forwarding of the MPLS packets or IP packets. The MPLS function does not
collide with the ip exf function. It is recommended to enable the ip exf function
when the MPLS function is running.

 The Ethernet switch supports the MPLS configuration only on the VLAN
interface, while does not support the MPLS function on the SuperVlan interface.

 The SuperVlan interface does not support the VFI binding of MPLS L2VPN and
the VRF binding of L3VPN at the same time. Hence, please do not perform
related configuration on the SuperVlan interface.

-5-
 MPLS Configuration

 BDCOM hi-end switches of the 4.0.1A version and its former versions supports
up to 64 L2VPN cases. Among L2VPN, up to 64 VFIs can be created.

 At the mixture of LDP and LSP, BDCOM hi-end switches of the 4.0.1A version
and its former versions have the following limits: as to an address prefix (route),
if the LDP generated dynamic label binding coexists with the command
configured static label binding (through the mpls static command in global
mode), they will collide with each other.In actual operation, please avoid as
much as possible enabling LDP during configuration of the static label binding.If
static binding is necessary, please shut down LDP before configuration.

 The LS12GE-MPLS card of the BDCOM hi-end switch only supports the
point-to-point L2VPN.

 The Ethernet switch does not support L2VPN and L3VPN at the same time.
Please avoid to create VFI instances and VRF instances at the same time.

The following cautions can be applied to S8500 switches and S6800 switches:

 In software version 4.0.1A and earlier software versions of BDCOM hi-end

switch, MPLS only run on the MPLS wire-card. Please make sure that the VLAN
interface which is used to create LSP contains the physical interface of the
MPLS wire card, and that the physical interface CE lies at the MPLS wire card in
the L2VPN and L2VPN configurations.

 As to BDCOM LS12GEMP wire card, in L2VPN, if a VFI-binding VLAN interface

contains multiple physical interfaces, these physical interfaces cannot
interconnect with each other. When using the LS12GEMP wire card to
establish L2VPN, make sure each VFI VLAN contains only one physical
interface.

1.4 MPLS Configuration

1.4.1 MPLS Configuration Task List

 Enabling MPLS

 Configuring Static LSP

 Configuring the Label Range

 Clearing the MPLS counter

 Browsing MPLS Information

1.4.2 Enabling MPLS

To allow a switch to forward MPLS packets, you need to enable MPLS on the switch.
MPLS startup contains two steps: global startup and interface startup. Only after the
two steps are finished, the interface of the switch can handle the MPLS packets.

-6-
 MPLS Configuration

Run the following commands to enable MPLS in global mode:

Command Purpose
mpls ip Enables MPLS in global mode.

no mpls ip Disables MPLS in global mode.

Run the following commands to enable MPLS in interface configuration mode:

Command Purpose
mpls ip Enables MPLS on the VLAN interface.

no mpls ip Disables MPLS on the VLAN interface.

Remark:
1. MPLS can be enabled only on the VLAN interface.
2. The MPLS in interface configuration mode validates only after the MPLS is
enabled in global mode.

1.4.3 Configuring Static LSP

Note:

At the mixture of LDP and LSP, BDCOM hi-end switches of the 4.0.1A version and
its former versions have the following limits: as to an address prefix (route), if the
LDP generated dynamic label binding coexists with the command configured static
label binding (through the mpls static command in global mode), they will collide
with each other.In actual operation, please avoid as much as possible enabling LDP
during configuration of the static label binding.If static binding is necessary, please
shut down LDP before configuration.

You can forbid LDP on the VLAN port to realize the dynamic and static bindings'
coexisting. You can shut down LDP on the VLAN port by running the no mpls ldp
enable command. If there is any question, please contact our technical service staff.

MPLS supports the static forwarding information configuration. The information

includes the static FTN and the static ILM. If the network scale is small, the static LSP
can be used to replace the complicated dynamic label distribution protocols.

Run the following commands in global mode to configure static MPLS forwarding information:

Command Purpose
mpls static binding ipv4 dest mask Configures static FTN items.

output nexthop outgoing-label Dest: stands for the address prefix of the
FTN item.

Mask: stands for the mask of the address

prefix.

Nexthop: stands for the IP address of the

next hop.

Outgoing-label: stands for the outgoing

-7-
 MPLS Configuration

label, whose range is from 16 to 1048575.

mpls static crossconnect incoming-label Configures static ILM items.

outgoing-intf nexthop outgoing-label incoming-label: stands for the incoming

label whose range is the static label
range.

outgoing-intf: stands for the outgoing

interface of the next hop.

Nexthop: stands for the IP address of the

next hop.

Outgoing-label: stands for the outgoing

label, whose range is from 16 to 1048575.

The negative forms of the previous commands can be used to cancel the
corresponding static LSP configuration.

The static FTN items are generally configured on the edge router of LSP. The next
hop’s IP address of the item can be a direct address or an indirect address. The
indirect next hop’s FTN is normally used in MPLS L3VPN. In this case, the switch
searches for an effective next-hop FTN for the indirect address automatically, and then
binds the detected FTN to the indirect FTN to form a multi-layer label stack.

It is noted that, after the static FTN item is configured, you need to add a static route
for a static FTN item or make sure that the FEC route corresponding to FTN is already
learned by the dynamic routing protocol. As to a destination subnet, if you only
configure the FTN item but the corresponding route does not exist, the LER switch
cannot encapsulate packets into a label to forward them to the expected LSP.

1.4.4 Configures the label range.

MPLS allows to configure the dynamic or static label range.

Run the following commands in global mode to configure the label range:

Command Purpose
mpls label range min max Configures the label range.

[ static min max ] min: Means the minimum value of the

label range, which is 1024 by default.

max: Means the maximum value of the

label range, which is 1048575 by default.

static: Configures the range of static

labels.

By default, the static label range is between 16 and 1023, while the dynamic label
range is between 1024 and 1048575.

At present, our label range is automatically adjusted to the multiples of 32 for speeding
up the research speed. In fact, labels of a 64K volume are supported. Labels 0-15 are
values reserved by IETF, so you cannot use them during configuration.

If the label range is configured and the labels are not distributed to other modules,
when the distribution range of the labels is changed, the new label range validates
immediately; otherwise, the new label range will validate at the next startup.

-8-
 MPLS Configuration

If the value range 16-1023 is the static label range and this static label range need be
modified, the modified static label range cannot exceed the range 16-1023.

1.4.5 Clears the MPLS counter.

MPLS will automatically count the forwarded packets. In privileged mode, you can run
the following command to clear the MPLS counter.

Command Purpose
clear mpls counters Clears the MPLS counter.

1.4.6 Browsing MPLS Information

In global or interface configuration mode, run the following commands to browse the
information about MPLS.

Command Purpose
show mpls forwarding-table Browses the content in the MPLS label
forwarding base.
[ { network mask } | interface name |
Network: ID of the destination network
labels value | nexthop address ]
mask: Mask of the destination mask

interface name: Displays the items with

specified outgoing port in the label
forwarding information base.

labels value : Displays the locally

distributed items with specified label in the
label forwarding information base.

nexthop address: Displays the items that

take designated neighbors as next hops in
the label forwarding information base.

show mpls ftn-table Displays the content of the

[ { network mask } | neighbor address | FEC-to-NHLFE table.

remote-label value ] Network: forwards the equivalent prefixes.

Mask: Stands for the destination mask.

Neighbor address: Stands for the next

hop's address of forwarding equivalence.

remote-label value: configured outgoing

labels

show mpls lfib [ ftn | ilm | summary ] Displays the content of MPLS LFIB.

show mpls interface Displays the information about the MPLS

[ interface-name [ all [ detail ] ] ] port.

interface-name: type and name of the

interface

All: Displays all interfaces, including the

-9-
 MPLS Configuration

unconfigured MPLS IP interface.

detail: Displays MTU of the port and

export and import the statisitics value of
the MPLS packets.

show mpls label range Displays the available label distribution

range.

show mpls static binding ipv4 | crossconnect Displays FTN or LIM items which are
statically configured.

show mpls traffic Displays the statistics information about

the MPLS packets.

1.5 LDP Configuration

1.5.1 LDP Configuration Task List

 Configuring LSR Label

 Enabling LDP

 Configuring Transmission Address

 Configuring the Address List of Label Distribution

 Configuring the Parameter About Local Discovery Mechanism

 Configuring the LDP Neighbor

 Configuring the Discovery Mechanism of the Designated Neighbor

 Configuring the Holdtime of the Session

 Force Explicit-Null Label

 Configuring LDP Log Record

 Browsing LDP Information

1.5.2 Configuring LSR Label

Before the label is started and the protocol is distributed, the router needs a
designated router ID which will mark LSR on the label switching path.

Run the following commands in global mode to configure the LSR label of the switch:

Command Purpose
mpls ldp router-id A.B.C.D Configures the LSR label.

A.B.C.D: Designates the IP address as

the label.

- 10 -
 MPLS Configuration

Only the router ID is configured can LDP be enabled on an interface.

If the mpls ldp router-id command is configured, new router ID takes effect
immediately and the LDP neighborhood is re-established.

It is recommended to configure the router ID as the address of the loopback interface

for the TCP connection triggered by LDP Targeted Hello takes the router ID of the LDP
neighbor as the addresses of two terminals of the socket, without which the neighbor
will fail to be established.

1.5.3 Enabling LDP

You can configure the mpls ip port start LDP Label Distribution Protocol.After LDP is
enabled on the interface, the LDP Hello packets will be automatically transmitted to the
network segment that directly connects the interface, and the LDP session will be
established to distribute the information about the label.

Run the following commands to enable LDP in interface configuration mode:

Command Purpose
mpls ldp enable Enables LDP.

no mpls ldp enable Disables LDP.

To enable LDP on an interface, you need to enable MPLS and the MPLS interface in
global mode first.

After LDP is enabled on an interface, the switch automatically establishes a session

with LSR that supports LDP, and distributes the information about labels. The LDP
neighbors and the state of the session can be browsed through the show mpls ldp
neighbor command. In initial conditions, the interface where LDP is enabled cannot
respond to the request of the Targeted Hello packets.

1.5.4 Configuring Transmission Address

By default, the switch uses the IP address of an interface to establish an connection

with other LSRs. The switch can be forced to use the designated IP to establish the
TCP session with LSR by configuring the transmission address.

Run the following commands in interface configuration mode to configure the

transmission address of LDP.

Command Purpose
mpls ldp discovery transport-address Configures the transmission address that
is used by the LDP session on the
ip-addr | interface
interface.

ip-addr: Designates the IP address;

interface: Designates the IP address of

the interface as the transmission address.

Before the LDP session is established, the TCP connection must be first established
between two routers to notify labels. To create the TCP connection, the router must
know the transmission address of the peer.

- 11 -
 MPLS Configuration

After the TCP connection is successfully established, the connection will not be
reestablished any more even if the IP address is changed.

1.5.5 Configuring the Address List of Label Distribution

By default, the switch where LDP is enabled distributes all established binding
relations of labels and FEC. By configuring the address list of label distribution, you
can control how to distribute the labels that interconnect the local address and the
designated address's prefix.

Run the following commands in global mode to configure the address list of label
distribution:

Command Purpose
mpls ldp advertise-labels for Complies with the regulations in the
prefix-access-list prefix-access list when the label
distribution of LDP is configured.

prefix-access-list: It is an access control

list containing the address prefix, where a
series of deny items and permit items are
used to control the policy of label
distribution.

For how to configure the access control list, refer to related user manuals.

1.5.6 Configuring the Parameter About Local Discovery Mechanism

After LDP is enabled on an interface, the switch establishes a session with the
directly-connected LSR automatically and exchanges the label information. The
establishment of the local session is controlled by the parameters related with the local
discovery mechanism. These parameters include holdtime and interval.

The holttime parameter represents the time for LSR to maintain a discovered neighbor,
while the interval parameter represents the interval for LSR to transmit the local Hello
packets. In general, the value of the interval parameter cannot be smaller than that of
the holdtime parameter. Before the holdtime exceeds, the local LSR still thinks that
the neighbor exists even if the Hello packets from a neighbor are not received. If the
holdtime exceeds, the local LSR will delete the neighbor and will terminate the
established session with the neighbor.

If the holdtime is too long, the LDP discovery will be too slow and the connection will
time out; if the holdtime is too short, abrupt flux will occur on the connection, and then
the Hello packets will be lost and the LDP will terminate the session finally.

Run the following commands in global mode to configure the parameters of local
discovery mechanism:

Command Purpose
mpls ldp discovery hello holdtime value Configures the time of maintaining a
neighbor.

Value: 1-65535 seconds, which is 15

seconds by default

mpls ldp discovery hello interval value Configures the interval of transmitting the

- 12 -
 MPLS Configuration

Hello packets.

Value: 1-65535 seconds, which is 3

seconds by default

The interval is one fifth of the holdtime by default, a value that can maintain the stable
state of the local neighbor. In case that the session with the neighbor is already
established, if you want to modify the value of the holdtime parameter of the local
switch, the switch will negotiate with the neighbor LSR to decide a comparatively small
value of the holdtime parameter as a valid value. Meanwhile, the modification of the
holdtime parameter will lead to establishing of the local session again.

1.5.7 Configuring the LDP Neighbor

For a local directly-connected neighbor, LDP can automatically establish a connection

with the neighbor through the local discovery mechanism.LDP can also establish a
connection with the indirectly-connected LSR on the remote terminal through
specifying the LDP neighbor manually.

In global mode, run the following commands to designate the indirectly-connected LDP
neighbor.

Command Purpose
mpls ldp neighbor ip-addr targeted Designates the LDP neighbor.

ip-addr: Designates the IP address of the

neighbor;

no mpls ldp neighbor ip-addr targeted Deletes the LDP neighbor.

It is to be noted that you should make configuration to enable the switch to accept the
Targeted Hello packets after the configuration of an indirectly connected neighbor so
that the indirectly connected neighbor can establish a connection with a remote
neighbor.

You can run show mpls ldp neighbor to browse the state of the session between the
switch and the indirectly connected LDP neighbor.

Remark:
In case that MPLS L2VPN is not enabled globally, the switch can establish and
maintain a session with a designated target on the remote LSR, but cannot
interchange the label information with the designated target.

1.5.8 Configuring the Discovery Mechanism of a Designated Neighbor

Similar to the local neighbor, LDP also support a set of parameters conduct settings to
the parameters related to the indirectly-connected session.

Run the following commands in global mode to do relevant configurations:

Command Purpose
mpls ldp discovery targeted-hello accept Allows the local LSR to process the
Targeted-Hello packets.

mpls ldp discovery targeted-hello holdtime Configures the holdtime of a remote

- 13 -
 MPLS Configuration

value session.

Value: 1-65535 seconds, which is 45

seconds by default

mpls ldp discovery targeted-hello interval Configures the interval of sending

value Targeted-Hello packets of the remote
session.

Value: 1-65535 seconds, which is 15

seconds by default

After a neighbor is detected on the port before the hold time times out, the target hello
information from the neighbor is not received, the neighbor will be deleted and the
target LDP session with the neighbor will be terminated.

If the target holdtime is too long, the target LDP discovery will be too slow and the
connection will time out; if the target holdtime is too short, abrupt flux will occur on the
connection, and then the target Hello packets will be lost and the target LDP will
terminate the session finally.

1.5.9 Configuring the Holdtime of the Session

After a session between the switch and other LSRs is established, the keepalive
packets need be sent regularly to maintain the session.In the case that the holdtime
does not exceed, LSR still thinks the session valid even if no Keepalive packets is not
received.

Run the following commands in global mode to configure the holdtime of a session:

Command Purpose
mpls ldp holdtime value Configures the Holdtime of the LDP
session.

Value: 1-65535 (the default value is 60

seconds)

During the establishment of a session, LSRs will negotiate the comparatively small
value of the holdtime as an effective value.

1.5.10 Force Explicit-Null Label

By default, as to the local direct-through route, LDP notifies the upstream neighbor of
the implicit null label so that the upstream neighbor adopts the next to the last hop to
pop up the outside label when the upstream neighbor is forwarding the MPLS packets.
In order to prevent the next to the last hop from popping out the outside label, you
need to take the strategy that uses the explicit null label to replace the outside label
and to configure this command.

Note:
To enable MPLS L2VPN or MPLS L2VPN on the switch, please make sure that the
force explicit-null label is not configured.

You can run the following command in global mode to force the switch to notify the
Explicit-Null label.

- 14 -
 MPLS Configuration

Command Purpose
mpls ldp explicit-null prefix-acl: Means that the ACL list which

[ for prefix-acl [ to peer-acl ] ] complies with the designated prefix will

replace the implicit null label with the
explicit null label and then notify the
upstream LDP neighbor.

peer-acl: Means replacing the implicit null

label with the explicit null label and then
notifying the designated upstream LDP
neighbor.

If mpls ldp explicit-null is configured, the direct-through routes that meet prefix ACL
will replace the Implicit-Null label with the Explicit-Null label and notify the upstream
neighbors that meet prefix ACL; if prefix-acl is not configured, all direct-through routes
will replace the Implicit-Null label with the Explicit-Null label.

1.5.11 Configuring LDP Log Record

LDP supports recording the change of neighbors through the log.

Run the following commands in global mode to enable the LDP log:

Command Purpose
logging ip-addr Enables the systematic log.

ip-addr：Means the address of the log

server.

mpls ldp logging neighbor-changes Enables the LDP log to record

information.

1.5.12 Browsing LDP Information

In global or interface configuration mode, run the following commands to browse the
LDP information.

Command Purpose
show mpls ldp bindings Displays the content of the label
database.

show mpls ldp discovery Displays the port list in which all the ports
have the LDP discovery mechanism.

show mpls ldp neighbor Displays the state of the session between
LDP neighbors.

show mpls ldp parameters Displays the current parameters of LDP.

- 15 -
 MPLS Configuration

Chapter 2 MPLS L2VPN Configuration

2.1 Overview of MPLS L2VPN

MPLS L2VPN is also called as Virtual Private LAN Service (VPLS) that provides
layer-2 virtual private network services on the basis of the MPLS network

BDCOM Ethernet switch series provide L2VPN function that takes the Label
Distribution Protocol (LDP) as the signaling protocol.

The switch that gets L2VPN service running is always served as the Provider Edge
(PE) equipment of which one end is connected to the Client Edge (CE) equipment and
the other end to the MPLS core of the provider to establish the label distribution path
with other providers' router (P).The establishment of the label distribution path in the
switch is finished by the software control layer while the forwarding of packets is
completed by the data layer.As to the packets from the client equipment, the switch
conducts the local layer-2 switching, if packets need be forwarded to the MPLS
network, adds the MPLS label stack to packets automatically and encapsulates them
to the header of the new Ethernet network and then forwards them to the next hop of
LSP; as to packets from the MPLS network, the switch first remove their headers and
then forwards them according to the Ethernet information carried by original packets.

2.2 Related Concepts of MPLS L2VPN

2.2.1 VFI

Virtual Forwarding Instance (VFI) acts as the bridge to forward layer-2 packets in the
switches where L2VPN is enabled. Different from general switches, VFI connects not
only common Ethernet interfaces but also PW of the remote VFI.

VFI can plot the broadcast domains in the switch and thus a broadcast packet can be
forwarded to all interfaces of VFI.That's the exact reason why VFI corresponds to
VLAN of a switch.

2.2.2 Pseudo-Wire

The Pseudo Wire (PW) means the connection between remote VFIs. PW consists of a
pair of unidirectional MPLS label distribution paths. Hence, a PW always corresponds
to two virtual circuit labels: the VC label generated by local PE and the VC label
generated by peer PE. The local VC label that is always carried by the received MPLS
packet can be used to represent the corresponding ILM item, while the peer VC label
that is always used by the locally transmitted MPLS packet can be used to represent
the corresponding FTN item.

The establishment of a PW also requires that the next hop of LSP must be definite.The
label that corresponds to the next hop is called as the tunnel label. When an MPLS
packet is being transmitted, the VC label and then the tunnel label will be added to the

- 16 -
 MPLS Configuration

label stack of the MPLS packet and a decision as to where the packet to go will be
made before the reciprocal twice hop.

2.2.3 L2VPN Network Without Loopback

In point-to-multipoint mode, VPLS requires that full-mesh connections should be

established among all PEs. A PW thus exists between any two PEs.When packets are
forwarded, the split-horizon strategy is carried out and the broadcast packets that PE
receives from PW will not be forwarded to PW.In this case, PE need not start STP to
avoid loopback.

Note:
The LS12GE-MPLS card of the BDCOM hi-end switch only supports the
point-to-point MPLS L2VPN.

2.3 MPLS L2VPN Configuration

2.3.1 MPLS L2VPN Configuration Task List

 Enabling MPLS L2VPN

 Creating VFI

 Binding VFI to VLAN Interface

 Configuring the Incoming VFI Interface

 Configuring a Static VC Label

 Browsing L2VPN Information

2.3.2 Enabling MPLS L2VPN

Run the following commands in global mode to enable L2VPN:

Command Purpose
mpls l2vpn Start L2VPNEnables L2VPN.

no mpls l2vpn Shuts down L2VPN.

2.3.3 Creating VFI

After L2VPN is enabled, VFI need be created and bound to a VLAN interface, which
means that the interface of this VLAN will be connected to a client device.

Run the following command in global mode to create a VFI:

Command Purpose
mpls vfi vfi-name Creates VFI and enters the VFI

- 17 -
 MPLS Configuration

[ ptop | vpls ] configuration mode.

Vfi-name: Means the name of VFI.

ptop: Means that VFI is in point-to-point

mode and there is only one remote
neighbor.

vpls: means that VFI is in

point-to-multipoint mode.

In VFI configuration mode, you can run the following commands to configure relevant
parameters:

Command Purpose
pwid value Configures the PW ID used by VFI.

neighbor peer-addr encapsulation mpls Designates the VFI neighbor.

peer-addr: IP address of a neighbor

PW ID is the ID that is used by VFI when the fake line is established. Different VFIs
have different PW IDs. The same PW ID must be used to identify the same VFI among
VFI neighbors; otherwise, the PW cannot be established. After a PW ID is designated
for a VFI, the ID cannot be modified. If you want to modify the PW ID of a VFI, you
have to delete the VFI and then reestablish it.

It is noted that you cannot establish the designated target session between two LSRs
only by designating the VFI neighbor. You also need to designate the VFI neighbor as
the LDP neighbor through the mpls ldp neighbor command.

Note:
1. Only one neighbor can be designated for the point-to-point VFI.
2. In multipoint mode, each PE device requires a complete VFI neighbor list to be
designated so that a full-mesh connection can be established.

2.3.4 Binding VFI to VLAN Interface

Note:
As to BDCOM hi-end switches of version 4.0.1A and previous version, do not
configure the IP address on the VLAN interface where VFI is bound.

After VFI is created, you need to bind it to the VLAN interface to show that the physical
interface in VLAN is used to connect the client device.

Run the following command in VLAN interface configuration mode to bind a VFI:

Command Purpose
mpls l2vpn vfi vfi-name Binds VFI to an interface.

vfi-name: Means the name of bound VFI.

Note:

- 18 -
 MPLS Configuration

VFI corresponds to VLAN interface one by one, that is to say, one VFI can be bound
to only one VLAN interface.

2.3.5 Configuring Incoming VFI Interface

When using BDCOM LS12GE-MPLS card, except that VFI need be bound to the
VLAN interface, the settings of the physical interfaces in this VLAN to be access
interfaces of VFI also need be conducted.

Run the following command in physical interface configuration mode to add interfaces
to a VFI:

Command Purpose
switchport mpls l2vpn vfi vfi-name Sets an interface to be the access
interface of the designated VFI.

vfi-name：Means the name of the bound

VFI.

BDCOM LS12GE-MPLS card supports configuring multiple physical interfaces in a VFI

as the access interfaces simultaneously. You can run show vlan id to browse the
physical interface set in a VLAN which correspond to VFI and select the access
interface to conduct configuration.One physical interface can be configured as the
access interface of multiple VFIs at the same time after you run switchport mpls
l2vpn vfi on each VFI.

Note:
The configuration of the VFI access interface cannot change the VLAN attributes of
a physical interface, but you have to make sure that the physical interface has been
added to the VLAN corresponding to VFI.

To cancel the configuration of the VFI access interface, please run no switchport
mpls l2vpn vfi in physical interface configuration mode.It should be paid attention that
the physical interface will be removed from all VFIs if the VFI name is not entered.

2.3.6 Configuring a Static VC Label

Like MPLS, L2VPN also supports static information forwarding.

Run the following commands in global mode to configure static information forwarding
for VFI:

Command Purpose
mpls static vfi vfi-name incoming-label Configures static VFI ILM items.

output peer-addr outgoing-label Vfi-name: Means the name of VFI.

incoming-label: Incoming label whose

range is the static label range

peer-addr: means the address of the VFI

neighbor

Outgoing-label: stands for the outgoing

label, whose range is from 16 to 1048575.

- 19 -
 MPLS Configuration

Remark:
A static VFI item is used to stand for a PW connection, while incoming-label and
outgoing-label stand for the local VC label and the peer VC label respectively.The
static VFI item takes effect only when PW is effective at the next hop.

2.3.7 Browsing L2VPN Information

In global or interface configuration mode, run the following commands to browse the
information about MPLS L2VPN.

Command Purpose
show mpls forwarding-table pwid value Displays the ILM information that
corresponds to the designated PW ID in
the label forwarding base. The value
parameter ranges from 1 to 4294967295.

show mpls ftn-table vfi vfi-name Browses the forwarding items in the
designated VFI.

vfi-name: Means the name of VFI.

show mpls ldp bindings vfi vfi-name Browses the information about VFI-related
label binding.

show mpls l2vpn vfi Browses the states of all VFIs.

show mpls l2vpn pw Browses PW information.

[ pwid | neighbor peer-addr | pwid: Designates PW ID.

interface interface-name | detail ] neighbor peer-addr: Designates a
neighbor.

interface interface-name: Stands for the

type and name of the interface.

detail: Stands for detailed PW information.

2.4 MPLS L2VPN Configuration Example

2.4.1 Point-to-Point L2VPN Configuration Example

Figure 2.1 shows a typical point-to-point MPLS L2VPN network frame:

- 20 -
 MPLS Configuration

VPN 1
Site 2
PE 1
CE 2
G0/1
VLAN 100 P
VLAN 100

G1/1 P VPN 1
PE 2 Site 1
CE 1

Figure 2.1 Point-to-Point MPLS L2VPN

The configuration of PE1 is shown below:

(1) VLAN configuration on an interface

switch_config# interface g0/1
switch_config_g0/1# switchport pvid 200（Here, VLAN200 connects the MPLS nwtwork of the
service provider.）
switch_config_g0/1# switchport mode trunk
switch_config_g0/1# exit
switch_config# interface vlan 200
switch_config_v200# ip address 91.0.0.1 255.0.0.0
switch_config_v200# exit

(2) Creating the loopback interface that is used as LSR ID

switch_config# interface loopback 1
switch_config_l1# ip addr 101.0.0.1 255.0.0.0

(3) MPLS configuration

switch_config# mpls ip
switch_config# interface vlan 200
switch_config_v200# mpls ip

(4) LDP configuration

switch_config# mpls ldp router-id 101.0.0.1
switch_config# mpls ldp discovery targeted-hello accept
switch_config# mpls ldp neighbor 102.0.0.1 targeted
switch_config# interface vlan 200
switch_config_v200# mpls ldp enable

(5) MPLS L2VPN configuration

switch_config# mpls l2vpn
switch_config# mpls vfi vpn-red ptop
switch_config_vfi_vpn-red# pwid 200
switch_config_vfi_vpn-red# neighbor 102.0.0.1 encapsulation mpls

- 21 -
 MPLS Configuration

switch_config_vfi_vpn-red# exit
switch_config# interface vlan 100（here VLAN100 means the client’s VLAN）
switch_config_v100# mpls l2vpn vfi vpn-red
switch_config_v100#exit
switch_config#interface g0/5（Configuring an access interface）
switch_config_g0/5# switchport pvid 100
switch_config_g0/5# switchport mpls l2vpn vfi vpn-red
switch_config_g0/5# exit

(6) Routing protocol configuration

switch_config# router rip
switch_config_rip# version 2
switch_config_rip# network 101.0.0.0
switch_config_rip# network 91.0.0.0
switch_config_rip# exit

The information about PE2 configuration is displayed below after the PE2 show
running-config command is run:
mpls ip
mpls ldp discovery targeted-hello accept
mpls ldp neighbor 101.0.0.1 targeted
mpls ldp router-id 102.0.0.1
！
mpls l2vpn
mpls vfi vpn-red ptop
pwid 200
neighbor 101.0.0.1 encapsulation mpls
！
interface loopback 1
ip address 102.0.0.1 255.0.0.0
！
interface gigaethernet 1/1
switchport mode trunk
switchport pvid 200
!
interface gigaethernet 1/5
switchport pvid 100
switchport mpls l2vpn vfi vpn-red
！
interface vlan 200
ip address 92.0.0.1 255.0.0.0
mpls ip
mpls ldp enable
！
interface vlan 100
no ip address
mpls l2vpn vfi vpn-red

- 22 -
 MPLS Configuration

！
router rip
version 2
network 102.0.0.0
network 92.0.0.0

- 23 -
 MPLS Configuration

Chapter 3 MPLS L3VPN Configuration

3.1 Overview of MPLS L3VPN

MPLS L3VPN can also be called as MPLS VRF VPN or MPLS BGP VPN, a
combination between MPLS and VRF.

In general, one VPN comprises a team of client networks that share a public routing
table on the ISP's routers.Each client network is connected to the interface of the
network devices of ISP, while ISP's device will relate each interface to a VPN routing
table.One VPN routing table is also called as a VRF (VPN Routing /Forwarding table).

MPLS L3VPN network mainly consists of three kinds of devices: client's edge router
(CE) and provider's edge router (PE) and provider's router (P).

CE is connected to PE at the edge of the client network, and the routes between CE
and PE are established through dynamic protocols or static protocols.

PE can connect different CEs of different client networks simultaneously and allocate
routes of different client networks to different VPN routing tables. The VRF routes can
be learned through MP-BGP protocol among PEs and at the same time the MP-BGP
protocol can allocate a VPN label for each VRF route.

VRF need not be deployed at the interior of the provider's network and therefore P
does not know the routing information about the client network.P and PE can establish
dynamic LSP through running LDP and the packets from the client network are
encapsulated as MPLS in the ISP's network and then forwarded.

Figure 3.1 MPLS VRF VPN

- 24 -
 MPLS Configuration

The packets from the client network are first forwarded from CE to an interface of PE,
PE then search the relevant VRF routing table according to the VRF configuration of
the interface and find the route according to the destination IP address of these
packets.In case that an MP-BGP session among PEs is already created, PE finds that
the packets from the client network need be forwarded to the remote PE. You need first
add the VPN label generated by MP-BGP to the front of the IP header of the client
packet, then PE searches the real next-hop P of the peer PE on LSP and adds the
next-hop label to the outside of the VPN label.In the process from PE to the
next-to-the-last P, the VPN label and the tunnel label always exist before the IP header
of the client packet.When the packets arrive at the next-to-the-last-hop P, P removes
the outside labels from the packets and then transmits the packets with the VPN label
to the destination PE, while the destination PE, according to the VPN label, forwards
the client packets to CE.

Remark:
The Ethernet switches that support VRF can function as MCE devices to be
deployed between PE and client network so as to save the port resources of
PE.Please see Multi-VRF CE Configuration of BDCOM Ethernet Switch.

Note:
On a PE device, you can configure MPLS L2VPN and MPLS L3VPN at the same
time.

3.2 MPLS L3VPN Configuration

3.2.1 MPLS L3VPN Configuration Task List

 Configuring MPLS

 Configuring LDP

 Configuring VRF

 Configuring a VPN Route

 Configuring a Route Between PE and P

 Configuring a Route Between PEs

3.2.2 Configuring MPLS

Run the following commands one by one to enable the MPLS function of a switch:

Command Purpose
Switch# configure Enters the global configuration mode of the
switch.

Switch_config# mpls ip Enables MPLS in global mode.

Switch_config# interface vlan id Creates a VLAN interface and enters the VLAN

- 25 -
 MPLS Configuration

configuration mode.

Switch_config_intf# mpls ip Enables the MPLS function of a VLAN

interface.

Switch_config_intf# exit Exits from Vlan interface configuration mode.

Switch_config# show mpls interface Browses the state of the MPLS interface.

3.2.3 Configuring LDP

Run the following commands one by one to enable the LDP function of a switch:

Command Purpose
Switch# configure Enters the global configuration mode of the
switch.

Switch_config# mpls ldp router-id A.B.C.D Configures the ID of the LDP router.

Switch_config# interface vlan id Enters the VLAN configuration mode.

Switch_config_intf# mpls ldp enable Enables LDP of the VLAN interface.

Switch_config_intf# exit Exits from the interface configuration mode.

3.2.4 Configuring VRF

To configure one or multiple VRFs and bind the L3 interface to VRF, run the following
commands:

Command Purpose

Switch# configure Enters the switch configuration mode.

Switch_config# ip vrf vrf-name Creates VRF and enters the VRF

configuration mode.

Vrf-name: VRF name with up to 16

characters

Switch_config_vrf# rd route-distinguisher Sets the route distinguisher of VRF.

route-distinguisher: Stands for the

distinguisher of the route. It consists of
autonomous domain ID and random
numbers, or IP and random numbers.

Switch_config_vrf# route-target Creates the expanded VPN attributes of VRF

input and output objects.
{ export | import | both }
route-target-extended-community: Stands
route-target-extened-community
for the distinguisher of the route. It consists
of autonomous domain ID and random
numbers, or IP and random numbers.

Switch_config_vrf# interface intf-name Enters the interface configuration mode.

intf-name: Stands for the name of an

interface.

Switch_config_intf# ip vrf forwarding vrf-name Relates the L3 interface with VRF.

- 26 -
 MPLS Configuration

vfi-name: Means the name of VRF.

Switch_config_intf# exit Exits from interface configuration mode.

Switch_config# ip exf Enables the routing of ip hardware.

Switch_config# show ip vrf Browses the VRF information.

[ brief | detail | interface ] [ vrf-name ]

Switch_config# no ip vrf vrf-name Deletes the configured VRF and the relation
between VRF and the L3 interface.

vfi-name: Means the name of VRF.

Switch_config_intf# no ip vrf forwarding Deletes the relation between the L3

interface and VRF.
[ vrf-name ]

3.2.5 Configuring a VPN Route

The VPN route refers to the route between PE and CE. The following configurations
take the OSPF protocol as an example. These configurations based on OSPF are
similar to these configurations based on RIP or BEIGRP.

Remark:
The VPN route can be established between PE and MCE through the EBGP
protocol.Please see Multi-VRF CE Configuration of BDCOM Ethernet Switch.

Refer to the following steps to create a VPN route between PE and CE.

Command Purpose
Switch# configure Enters the switch configuration mode.

Switch_config# router ospf Starts the OSPF-VRF route and enters the

process-id vrf vrf-name configuration mode.

Switch_config_ospf# network network-number Defines the OSPF network, mask and area ID.

network-mask area area-id

Switch_config_ospf# redistribute bgp ASN Forwards the designated BGP network to the
OSPF network.

Switch_config_ospf# exit Exits from the OSPF configuration mode.

Switch_config# show ip ospf Browses the information about the OSPF

protocol.

Switch_config# no router ospf process-id Deletes the OSPF-VRF routing configuration.

3.2.6 Configuring a Route Between PE and P

A route can be established between PE and P through dynamic routing protocol such
as OSPF, RIP and BEIGRP, so can a static route.The following description takes the
OSPF protocol as an example.

Command Purpose

- 27 -
 MPLS Configuration

Switch# configure Enters the switch configuration mode.

Switch_config# router ospf Starts the OSPF-VRF route and enters the

process-id configuration mode.

Switch_config_ospf# network network-number Defines the OSPF network, mask and area ID.

network-mask area area-id

Switch_config_ospf# exit Exits from the OSPF configuration mode.

Switch_config# show ip ospf Browses the information about the OSPF

protocol.

Switch_config# no router ospf process-id Deletes the OSPF-VRF routing configuration.

3.2.7 Configuring a BGP Route Between PEs

Refer to the following steps to create a MP-BGP route between PEs.

Command Purpose
Switch# configure Enters the switch configuration mode.

Switch_config# router bgp Starts the BGP protocol by designating

autonomous-system-number autonomous system number and enters the
BGP configuration mode.

Switch_config_bgp# bgp log-neighbor-changes Starts the record about BGP neighbor change.

Switch_config_bgp# neighbor address Configures the BGP neighbor and the

remote-as ASN autonomous system number of a neighbor.

Switch_config_bgp# neighbor address Configures the interface that will be used to

update-source intf-name connect a BGP neighbor.

Switch_config_bgp# address-family vpnv4 Enters the configuration mode of BGP VPN.

Switch_config_bgp_vpnv4# neighbor address Activates a neighbor in VPNV4.

activate

Switch_config_bgp_vpnv4# bgp Redistributes the VPNV4 route to the routing

redistribute-internal protocol of internal gateway.

Switch_config_bgp_vpnv4# exit-address-family Exits from the VPNV4 configuration mode.

Switch_config_bgp# address-family ipv4 vrf Enters the configuration mode of VRF

vrf-name address-family.

Switch_config_bgp_af# bgp redistribute-internal Redistributes the VRF route to the routing

protocol of internal gateway.

Switch_config_bgp_af# redistribute ospf Redistributes the OSPF routing information to

ospf-process-id the BGP network.

Switch_config_bgp_af# network network-number Configures the network number and the mask
/ prefix-len ‘s length that are distributed by BGP.

Switch_config_bgp_af# exit-address-family Exits from the configuration mode of VRF

address-family.

- 28 -
 MPLS Configuration

Switch_config_bgp# exit Exits from the BGP configuration mode.

Switch_config# show ip bgp vpnv4 [ * | all | rd | Browses the BGP-VRF routing information and
vrf ] the label’s state.

3.3 MPLS L3VPN Configuration Example

The following figure is a simple MPLS L3VPN network and the configuration of all its
devices is also listed below:

S2 S3 S4

VPN1
PE P PE
S1- CE 102.0.0.1/32 103.0.0.1/32 104.0.0.1/32
101.0.0.1/32

VLAN 2 VLAN 3

F0/31 G3/5 G3/9 F0/31

Figure 3.2 MPLS L3VPN showcase network

3.3.1 Configuring S1 (CE)

1. Configuring loopback interface

S1_config# interface Loopback0

S1_config_l0# ip address 101.0.0.1 255.255.255.255
S1_config_l0# exit

2. Configuring the physical and logical interfaces

S1_config# interface FastEthernet0/31

- 29 -
 MPLS Configuration

S1_config_f0/31# switchport pvid 2

S1_config_f0/31# exit
S1_config# interface VLAN2
S1_config_v2# ip address 92.0.0.1 255.0.0.0
S1_config_v2# exit
S1_config# vlan 1-3

3. Configuring the OSPF routing protocol

S1_config# router ospf 1

S1_config_ospf_1# network 92.0.0.0 255.0.0.0 area 0
S1_config_ospf_1# network 101.0.0.1 255.255.255.255 area 0
S1_config_ospf_1# exit

3.3.2 Configuring S2 (PE)

1. Configuring global MPLS

S2_config# mpls ip

2. Configuring the global attributes of LDP

S2_config# mpls ldp router-id 102.0.0.1

3. Creating VRF

S2_config# ip vrf vpn1

S2_config_vrf_vpn1# rd 100:1
S2_config_vrf_vpn1# route-target export 100:1
S2_config_vrf_vpn1# route-target import 100:1
S2_config_vrf_vpn1# exit

4. Creating the loopback interface

S2_config# interface Loopback0

S2_config_l0# ip address 102.0.0.1 255.255.255.255
S2_config_l0# exit

5. Configuring the physical interface

S2_config# interface GigaEthernet3/5

S2_config_g3/5# switchport pvid 2
S2_config_g3/5# exit
S2_config# interface GigaEthernet3/9
S2_config_g3/9# switchport pvid 3
S2_config_g3/9# exit

- 30 -
 MPLS Configuration

6. Configuring the VLAN interface

S2_config# interface VLAN2

S2_config_v2# ip vrf forwarding vpn1
S2_config_v2# ip address 92.0.0.2 255.0.0.0
S2_config_v2# exit
S2_config# interface VLAN3
S2_config_v3# ip address 93.0.0.1 255.0.0.0
S2_config_v3# mpls ip
S2_config_v3# mpls ldp enable
S2_config_v3# exit

7. Creating VLAN

S2_config# vlan 1-3

8. Configuring an OSPF route between PE and P

S2_config# router ospf 1

S2_config_ospf_1# network 93.0.0.0 255.0.0.0 area 0
S2_config_ospf_1# network 102.0.0.1 255.255.255.255 area 0
S2_config_ospf_1# exit

9. Configuring a VPN route between PE and CE

S2_config# router ospf 2 vrf vpn1

S2_config_ospf_2# network 92.0.0.0 255.0.0.0 area 0
S2_config_ospf_2# redistribute bgp 1
S2_config_ospf_2# exit

10. Configuring MP-BGP

S2_config# router bgp 1

S2_config_bgp# bgp log-neighbor-changes
S2_config_bgp# neighbor 104.0.0.1 remote-as 1
S2_config_bgp# neighbor 104.0.0.1 update-source Loopback0
S2_config_bgp#
S2_config_bgp# address-family vpnv4
S2_config_bgp_vpnv4# neighbor 104.0.0.1 activate
S2_config_bgp_vpnv4# bgp redistribute-internal
S2_config_bgp_vpnv4# exit-address-family
S2_config_bgp#
S2_config_bgp# address-family ipv4 vrf vpn1
S2_config_bgp_vpn1# no synchronization
S2_config_bgp_vpn1# bgp redistribute-internal
S2_config_bgp_vpn1# redistribute ospf 2
S2_config_bgp_vpn1# exit-address-family
S2_config_bgp# exit

- 31 -
 MPLS Configuration

11. Enabling the routing of ip hardware subnet

S2_config# ip exf

3.3.3 Configuring S3 (P)

1. Configuring MPLS and LDP in global mode

S3_config# mpls ip
S3_config# mpls ldp router-id 103.0.0.1

2. Configuring loopback interface

S3_config# interface Loopback0

S3_config_l0# ip address 103.0.0.1 255.255.255.255
S3_config_l0# exit

3. Configuring the physical interface

S3_config# interface FastEthernet0/31

S3_config_f0/31# switchport pvid 3
S3_config_f0/31# exit

4. Configuring the VLAN interface and VLAN

S3_config# interface VLAN3

S3_config_v3# ip address 93.0.0.2 255.0.0.0
S3_config_v3# mpls ip
S3_config_v3# mpls ldp enable
S3_config_v3# exit
S3_config# vlan 1-4

5. Configuring the OSPF route

S3_config# router ospf 1

S3_config_ospf_1# network 93.0.0.0 255.0.0.0 area 0
S3_config_ospf_1# network 103.0.0.1 255.255.255.255 area 0
S3_config_ospf_1# exit

- 32 -