Table of Contents

Chapter 1 MBGP Multicast Extension Configuration. 1-1

1.1 MBGP Multicast Extension Overview. 1-1

1.1.1 MBGP Extension Attributes for Multicast 1-1

1.1.2 MBGP Operating Mode and Message Type. 1-2

1.2 MBGP Multicast Extension Configuration. 1-3

1.2.1 Enabling MBGP Multicast Extension Protocol 1-3

1.2.2 Configuring Network Routes Notified by MBGP Multicast Extension. 1-4

1.2.3 Configuring the MED Value for an AS. 1-4

1.2.4 Comparing MED Values from Different AS Neighbor Paths. 1-4

1.2.5 Configuring Local Preference. 1-4

1.2.6 Configuring MBGP Timer 1-5

1.2.7 Configuring MBGP Peer (Group) 1-5

1.2.8 Configuring the Maximum Number of Routes a Peer Group can Receive. 1-8

1.2.9 Configuring MBGP Route Aggregation. 1-9

1.2.10 Configuring an MBGP Route Reflector 1-9

1.2.11 Configure MBGP Community Attributes. 1-10

1.2.12 Importing IGP Routing Information into MBGP. 1-10

1.2.13 Defining AS Path List and Routing Policy. 1-10

1.2.14 Configuring MBGP Route Filtering. 1-10

1.2.15 Resetting BGP Connections. 1-11

1.3 Displaying and Maintaining MBGP Configuration. 1-11

1.4 MBGP Multicast Extension Configuration Example. 1-12

 

Chapter 1  MBGP Multicast Extension Configuration

When configuring MBGP, go to these sections for information you are interested in:

l           MBGP Multicast Extension Overview

l           MBGP Multicast Extension Configuration

l           Displaying and Maintaining MBGP Configuration

l           MBGP Multicast Extension Configuration Example

1.1  MBGP Multicast Extension Overview

At present, the most widely used inter-domain unicast routing protocol is BGP-4. Because the multicast topology may be different from the unicast topology, BGP-4 must be modified in order to implement the transmission of inter-domain multicast routing information. Some routers in the network may only support unicast rather than multicast and may not forward multicast packets since the particular policy requires that. To construct inter-domain multicast routing trees, you need to know the unicast routing information as well as the information of multicast-supporting parts of the network, namely, the multicast network topology.

BGP-4 has been proved to be an effective and stable inter-domain unicast routing protocol. Therefore, it is more rational to enhance and extend the BGP-4 protocol than to construct a new protocol. RFC2858 defines an extended BGP-4, Multi-Protocol BGP (MP-BGP), to provide routing information in use for multiple routing applications. MBGP can not only carry IPv4 unicast routing information but also the routing information of other network layer protocols (such as multicast, IPv6). Carrying multicast routing information is only one of the extended functions.

MBGP enables unicast and multicast routing information to be exchanged through the same process but stored in different routing tables. As MBGP is an enhanced version of BGP-4, all the common policies and configuration methods that BGP-4 supports can be applied to multicast.

This chapter describes mainly MBGP extension for multicast.

1.1.1  MBGP Extension Attributes for Multicast

To make MBGP support multicast, RFC2858 defines two new route attributes in the UPDATE message: MP_REACH_NLRI (multiprotocol reachable NLRI) and MP_UNREACH_NLRI (multiprotocol unreachable NLRI). They are all optional non-transitive attributes, that is, routers that do not support MBGP can ignore the attributes and do not forward the attributes.

Among the information carried by MP_REACH_NLRI and MP_UNREACH_NLRI, address family identifier (AFI) and subsequent address family identifier (SAFI) can identify the address family of the information. SAFI is a complement to network layer reachability information (NLRI), with the value 1 for the unicast mode of NLRI, and the value 2 for the multicast mode of NLRI.

I. MP_REACH_NLRI attribute

MP_REACH_NLRI is an optional non-transitive attribute, and can be used to:

l           Send the routing information of a new reachable protocol.

l           Send the next hop information about the new protocol with the same coding mode as that of NLRI.

l           Enable the router to report part or all of the sub-network points of attachment (SNPAs) saved in the local system.

II. MP_UNREACH_NLRI attribute

The MP_UNREACH_NLRI is an optional non-transitive attribute that can be used for the purpose of withdrawing one or multiple unfeasible routes. It includes the following fields:

l           AFI and SAFI.

l           Withdrawn Routes: Contains one or multiple NLRIs, in which are the unreachable destination addresses.

An UPDATE packet that contains the MP_UNREACH_NLRI is not required to carry any other path attributes.

These two attributes enables MBGP to carry multi-protocol information. MBGP therefore supports both unicast and multicast by constructing different topology maps to implement appropriate policies. Besides, MBGP may construct different inter-domain routes for unicast and multicast under a same policy.

1.1.2  MBGP Operating Mode and Message Type

MBGP runs on a router in the following two modes:

l           Internal BGP (IBGP )

l           External BGP (EBGP )

MBGP running in an autonomous system is called IBGP; MBGP running across autonomous systems is called EBGP.

MBGP offers four types of messages:

l           Open Message: First message sent after the TCP connection is established.

l           Notification Message: Error notification message.

l           Keepalive Message: Message used to check the validity of the connection.

l           Update Message: Most important information in the MBGP system, used to exchange routing information among peers. It consists of three parts at the most: MP_UNREACH_NLRI, Path Attributes, and MP_REACH_NLRI.

1.2  MBGP Multicast Extension Configuration

Basic configuration tasks of MBGP multicast extension include

l           Enabling MBGP Multicast Extension Protocol

l           Configuring Network Routes Notified by MBGP Multicast Extension

Advanced configuration tasks of MBGP multicast extension include

l           Configuring the MED Value for an AS

l           Comparing MED Values from Different AS Neighbor Paths

l           Configuring Local Preference

l           Configuring MBGP Timer

l           Configuring MBGP Peer (Group)

l           Configuring the Maximum Number of Routes a Peer Group can Receive

l           Configuring MBGP Route Aggregation

l           Configuring an MBGP Route Reflector

l           Configure MBGP Community Attributes

l           Importing IGP Routing Information into MBGP

l           Defining AS Path List and Routing Policy

l           Configuring MBGP Route Filtering

l           Resetting BGP Connections

 

 

1.2.1  Enabling MBGP Multicast Extension Protocol

To enable the MBGP multicast extension protocol, enter IPv4 multicast sub-address family view.

A router does not start receiving MBGP connection requests immediately the MBGP multicast extension protocol is enabled. To activate a router to originate MBGP connection requests to neighboring routers, refer to the neighbor configuration.

Follow these steps to enable MBGP multicast extension protocol:

To do...	Use the command...
Enter the MBGP multicast address family view	ipv4-family multicast
Remove the MBGP multicast address family view	undo ipv4-family multicast

 

By default, the system does not run the MBGP multicast extension protocol.

1.2.2  Configuring Network Routes Notified by MBGP Multicast Extension

The network command is used to configure the network routes to be advertised to MBGP peers, as well as the mask and route policy of this network route.

Perform the following operation in IPv4 multicast sub-address family view to configure/remove network routes to be advertised by MBGP multicast extension:

To do...	Use the command...
Configure the network routes to be advertised by the local MBGP	network ip-address [ address-mask ] [ route-policy route-policy-name ]
Remove the network routes to be advertised by the local MBGP	undo network ip-address [ address-mask ] [ route-policy route-policy-name ]

 

By default, no route is advertised by the local MBGP.

The network command advertises only the precisely matched route, the one with prefix and mask completely conforming to the configuration. If no mask is specified, match goes by the natural network segment.

1.2.3  Configuring the MED Value for an AS

The MED configured in BGP view is valid for both unicast and multicast.

For details of this configuration, refer to BGP Configuration in the IP Routing Volume.

1.2.4  Comparing MED Values from Different AS Neighbor Paths

Do not use this configuration unless you are sure that different ASs adopt the same IGP and route selection method. The configuration in BGP view works both in unicast and multicast.

For details of this configuration, refer to BGP Configuration in the IP Routing Volume.

1.2.5  Configuring Local Preference

Different local preferences can be configured as a reference of the MBGP route selection. When an MBGP router gets routes with the same destination but different next hops through different neighbors, it will choose the route with the highest local preference.

The configuration works both in unicast and multicast.

For details of this configuration, refer to BGP Configuration in the IP Routing Volume.

1.2.6  Configuring MBGP Timer

After a router establishes an MBGP connection with a peer, it sends Keepalive messages to the peer periodically to check for the smooth connection. If the router does not receive any Keepalive message or any other kind of message from the peer within the defined connection Holdtime, it will think the MBGP connection broken and exit, and will process the routing information received through this connection as appropriate. Therefore, the Keepalive interval and MBGP connection Holdtime are two parameters of great importance in MBGP mechanism.

The configuration works both in unicast and multicast.

For details of this configuration, refer to BGP Configuration in the IP Routing Volume.

1.2.7  Configuring MBGP Peer (Group)

The use of MBGP peer groups is to simplify configuration. When configuring MBGP peers, you can create and configure a peer group in BGP view, and then add the peers into the group, since all peers in a group have the same configuration with the group. Then, enable this peer group in IPv4 multicast sub-address family view and add peers to this peer group to create MBGP peers and an MBGP peer group. In conclusion, to create MBGP peers/peer groups, you must configure them successfully in BGP view first.

 

 

I. Creating a peer group with members

Configure a MBGP peer group in BGP view and add peers to this peer group. For details, refer to BGP Configuration in the IP Routing Volume.

II. Enabling a peer (group)

Perform the following operation in IPv4 multicast sub-address family view to enable/disable a peer (group):

To do...	Use the command...
Enable the specified peer (group)	peer group-name enable
Disable the specified peer (group)	undo peer group-name enable

 

III. Adding an MBGP peer to the group

Perform the following operation in IPv4 multicast sub-address family view to add/delete an MBGP peer to/from the group:

To do...	Use the command...
Add an MBGP peer to the group	peer peer-address group group-name
Delete the MBGP peer	undo peer peer-address

 

IV. Advertising MBGP community attributes to a peer (group)

Perform the following operation in IPv4 multicast sub-address family view to advertise/not to advertise the community attributes to a peer (group):

To do...	Use the command...
Advertise the community attributes to a peer (group)	peer group-name advertise-community
Configure not to advertise the community attributes to a peer (group)	undo peer group-name advertise-community

 

By default, no community attribute is advertised to any peer (group).

V. Configuring a peer (group) as an MBGP route reflector client

Perform the following operation in IPv4 multicast sub-address family view to configure a peer (group) as an MBGP route reflector client or remove the configuration:

To do...	Use the command...
Configure a peer (group) as an MBGP route reflector client	peer group-name reflect-client
Remove the above configuration	undo peer group-name reflect-client

 

By default, there is no route reflector in an AS.

It is generally unnecessary to configure this command for a peer group. This command is reserved for the occasional compatibility with the network equipments of other vendors.

VI. Configuring the local address as the next hop when advertising routes

This involves removing the next hop configuration in the routing information advertised to a peer (group) and configuring the local address as the next hop address. It is valid only for IBGP peers/peer groups.

Perform the following operation in IPv4 multicast sub-address family view to configure the local address as the next hop when advertising routes or remove the configuration:

To do...	Use the command...
Configure the local address as the next hop when advertising routing information	peer group-name next-hop-local
Remove the above configuration	undo peer group-name next-hop-local

 

VII. Specifying the routing policy for a peer (group)

Perform the following operation in IPv4 multicast sub-address family view to specify/remove the routing policy for a peer (group):

To do...	Use the command...
Configure routing policy for incoming packets	peer { group-name | peer-address } route-policy policy-name import
Remove incoming policy configuration	undo peer { group-name | peer-address } route-policy policy-name import
Configure routing policy for outgoing packets	peer group-name route-policy policy-name export
Remove outgoing policy configuration	undo peer group-name route-policy policy-name export

 

By default, no routing policy is specified for any peer (group).

VIII. Configuring IP-ACL-based route filtering policy for a peer (group)

Perform the following operation in IPv4 multicast sub-address family view to configure/remove IP-ACL-based route filtering policy for a peer (group):

To do...	Use the command...
Configure filtering policy for incoming packets	peer { group-name | peer-address } filter-policy acl-number import
Remove incoming policy configuration	undo peer { group-name | peer-address } filter-policy acl-number import
Configure routing policy for outgoing packets	peer group-name filter-policy acl-number export
Remove outgoing policy configuration	undo peer group-name filter-policy acl-number export

 

By default, a peer (group) does not perform route filtering based on the IP ACL.

IX. Configuring AS-path-list-based route filtering policy for a peer (group)

Perform the following operation in IPv4 multicast sub-address family view to configure/remove the AS-path-list-based route filtering policy for a peer (group):

To do...	Use the command...
Configure filtering policy for incoming packets	peer { group-name | peer-address } as-path-acl acl-number import
Remove incoming policy configuration	undo peer { group-name | peer-address } as-path-acl acl-number import
Configure routing policy for outgoing packets	peer { group-name | peer-address } as-path-acl acl-number export
Remove outgoing policy configuration	undo peer { group-name | peer-address }as-path-acl acl-number export

 

By default, a peer (group) does not perform route filtering based on the AS path list.

X. Configuring prefix-list-based route filtering policy for a peer (group)

Perform the following operation in IPv4 multicast sub-address family view to configure/remove prefix-list-based route filtering policy for a peer (group):

To do...	Use the command...
Configure filtering policy for incoming packets	peer { group-name | peer-address } ip-prefix prefixname import
Remove incoming policy configuration	undo peer { group-name | peer-address } ip-prefix prefixname import
Configure routing policy for outgoing packets	peer {group-name ip-prefix prefixname export
Remove outgoing policy configuration	undo peer group-name ip-prefix prefixname export

 

By default, a peer (group) does not perform route filtering based on the prefix list.

1.2.8  Configuring the Maximum Number of Routes a Peer Group can Receive

Perform the following operation in IPv4 multicast sub-address family view to configure/restore the maximum number of routes a peer group can receive:

To do...	Use the command...
Configure the maximum number of routes that can be received by a peer group	peer { group-name | peer-address } route-limit route-limit-value [ route-limit-percent ] [ alter-only | idle-forever | idle-timeout timeout-value ]
Restore the maximum number of routes that can be received by a peer group	undo peer { group-name | peer-address } route-limit

 

1.2.9  Configuring MBGP Route Aggregation

MBGP supports the manual aggregation of routes. Manual aggregation aggregates the local MBGP routes. A series of parameters can be configured during manual route aggregation.

Perform the following operation in IPv4 multicast sub-address family view to configure/remove MBGP route aggregation:

To do...	Use the command...
Configure the aggregation of local routes	aggregate address mask [ as-set | attribute-policy route-policy-name | detail-suppressed | origin-policy route-policy-name | suppress-policy route-policy-name ] *
Remove the aggregation of local routes	undo aggregate address mask [ as-set | attribute-policy route-policy-name | detail-suppressed | origin-policy route-policy-name | suppress-policy route-policy-name ] *

 

By default, MBGP does not aggregate local routes.

1.2.10  Configuring an MBGP Route Reflector

To ensure the interconnectivity among MBGP peers, it is necessary to establish fully-closed network among IBGP multicast peers. However, some internal MBGP multicast networks are very large, and it costs a good sum to establish a fully-closed network.

Route reflector solves this problem. The core is to specify a router as the focus of the internal sessions. Multiple MBGP multicast routers can be peers of one central point, namely a multiple route reflector, which in turn creates peer relationship with other reflectors. The route reflector is the focus of other routers. The routers other than those reflectors are called clients. The clients are in peer with route reflects and exchange routing information with them. The route reflectors transfer (reflect) information between the clients in turn.

For the details of the principles and configurations, refer to BGP Configuration in the IP Routing Volume.

1.2.11  Configure MBGP Community Attributes

Within the MBGP, a community is a set of destinations with some characteristics in common. A community is not limited to a network, or an AS has no physical boundary.

For details, refer to BGP Configuration in the IP Routing Volume.

1.2.12  Importing IGP Routing Information into MBGP

MBGP can advertise intra-area network information to other ASs. To this end, you can use MBGP to advertise the intra-area network information that local router gets through IGP routing protocol.

Perform the following operation in IPv4 multicast sub-address family view to import/delete IGP routing information:

To do...	Use the command...
Import IGP Routing Information into MBGP	import-route protocol [ route-policy policy-name ] [ med med-value ] *
Delete the imported IGP routing information	undo import-route protocol

 

By default, MBGP does not import any route of other protocols.

Parameter Protocol specifies the source routing protocols of import, which can be direct, static, rip, isis, ospf, ospf-ase or ospf-nssa at present.

1.2.13  Defining AS Path List and Routing Policy

To configure AS path list and routing polity you need to:

l           Configure the regular expression of autonomous systems (in system view);

The UPDATE information of MBGP contains an AS_PATH domain. The autonomous system paths for MBGP routing information exchange is recorded in this domain.

l           Define the routing policy (in system view);

l           Define matching rules (in routing policy view);

l           Define value assigning rules (in routing policy view)

For the detailed configuration of regular expression of AS, refer to BGP Configuration in the IP Routing Volume. For other configurations, refer to Route Policy Configuration in the IP Routing Volume.

1.2.14  Configuring MBGP Route Filtering

The route filtering configuration of MBGP is the same as that of unicast BGP.

For details, refer to BGP Configuration in the IP Routing Volume.

1.2.15  Resetting BGP Connections

After changing the MBGP policy or protocol configuration, users must disconnect the present BGP connection to make the new configuration effective.

For details, refer to BGP Configuration in the IP Routing Volume.

1.3  Displaying and Maintaining MBGP Configuration

To do...	Use the command...	Remarks
Display an MBGP routing table	display bgp multicast routing-table [ ip-address [ mask ] ]	Available in any view
Display CIDR (classless inter-domain routing)	display bgp multicast routing-table cidr
Display the routing information about the specified MBGP community	display bgp multicast routing-table community [ aa:nn | no-export-subconfed | no-advertise | no-export ] * [ whole-match ]
Display the routes permitted by the specified MBGP community list	display bgp multicast routing-table community-list community-list-number [ whole-match ]
Display the routes with inconsistent source autonomous systems	display bgp multicast routing-table different-origin-as
Display the routing information to or from a specified multicast neighbor	display bgp multicast peer [ peer-address ] [ verbose ]
Display the routing information advertised by MBGP	display bgp multicast network
Display the peer group information	display bgp multicast group [ group-name ]
Display the AS path information matching the AS regular expression	display bgp multicast routing-table regular-expression as-regular-expression
Enable debugging MBGP update packets	debugging bgp mp-update [ receive | send ] [ verbose ]	Available in user view

 

1.4  MBGP Multicast Extension Configuration Example

I. Network requirement

This example describes how the administrator uses the MBGP attributes to manage route selection. All switches are configured with MBGP. The IGP in AS200 uses OSPF. Switch A is in AS100 and serves as the MBGP neighbor of Switch B and Switch C in AS200. Switch B and Switch C run IBGP for Switch D in AS200.

II. Network diagram

Figure 1-1 Network diagram for MBGP path selection configuration

III. Configuration procedure

1)         Configure Switch A

<SwitchA> system-view

System View: return to User View with Ctrl+Z.

[SwitchA] vlan 20

[SwitchA-vlan20] port ethernet1/1/2

[SwitchA-vlan20] quit

[SwitchA] interface vlan-interface 20

[SwitchA-Vlan-interface20] ip address 192.1.1.1 255.255.255.0

[SwitchA-Vlan-interface20] quit

[SwitchA] vlan 30

[SwitchA-vlan30] port ethernet1/1/3

[SwitchA-vlan30] quit

[SwitchA] interface vlan-interface 30

[SwitchA-Vlan-interface30] ip address 193.1.1.1 255.255.255.0

[SwitchA-Vlan-interface30] quit

# Enable MBGP.

[SwitchA] bgp 100

[SwitchA-bgp] ipv4-family multicast

# Specify the target network for MBGP.

[SwitchA-bgp-af-mul] network 1.0.0.0

[SwitchA-bgp-af-mul] network 2.0.0.0

[SwitchA-bgp-af-mul] quit

# Configure peers relationship.

[SwitchA-bgp] bgp 100

[SwitchA-bgp] group a1 external

[SwitchA-bgp] peer 192.1.1.2 group a1 as-number 200

[SwitchA-bgp] group a2 external

[SwitchA-bgp] peer 193.1.1.2 group a2 as-number 200

[SwitchA-bgp] ipv4-family multicast

[SwitchA-bgp-af-mul] peer a1 enable

[SwitchA-bgp-af-mul] peer a2 enable

# Configure the MED attribute of Switch A.

l           Add an ACL on Switch A to permit network 1.0.0.0/8.

[SwitchA] acl number 2000

[SwitchA-acl-basic-2000] rule permit source 1.0.0.0 0.255.255.255

[SwitchA-acl-basic-2000] rule deny source any

l           Define two routing policies: set_med_50 and set_med_100, providing two MED values for network 1.0.0.0 (50 and 100 respectively).

[SwitchA] route-policy set_med_50 permit node 10

[SwitchA-route-policy] if-match acl 2000

[SwitchA-route-policy] apply cost 50

[SwitchA-route-policy] quit

[SwitchA] route-policy set_med_100 permit node 10

[SwitchA-route-policy] if-match acl 2000

[SwitchA-route-policy] apply cost 100

l           Apply the routing policy set_med_50 to the exported route updates of Switch C (193.1.1.2). Apply the routing policy set_med_100 to the exported route updates of Switch B (192.1.1.2).

[SwitchA] bgp 100

[SwitchA-bgp] ipv4-family multicast

[SwitchA-bgp-af-mul] peer a2 route-policy set_med_50 export

[SwitchA-bgp-af-mul] peer a1 route-policy set_med_100 export

2)         Configure Switch B

<SwitchB> system-view

System View: return to User View with Ctrl+Z.

[SwitchB] vlan 20

[SwitchB-vlan20] port ethernet2/1/2

[SwitchB-vlan20] quit

[SwitchB] interface vlan-interface 20

[SwitchB-Vlan-interface20] ip address 192.1.1.2 255.255.255.0

[SwitchB-Vlan-interface20] quit

[SwitchB] vlan 40

[SwitchB-vlan40] port ethernet2/1/4

[SwitchB-vlan40] quit

[SwitchB] interface vlan-interface 40

[SwitchB-Vlan-interface40] ip address 194.1.1.2 255.255.255.0

[SwitchB-Vlan-interface40] quit

[SwitchB] ospf

[SwitchB-ospf-1] area 0

[SwitchB-ospf-1-area-0.0.0.0] network 194.1.1.0 0.0.0.255

[SwitchB-ospf-1-area-0.0.0.0] network 192.1.1.0 0.0.0.255

[SwitchB-ospf-1-area-0.0.0.0] quit

[SwitchB-ospf-1] quit

[SwitchB] bgp 200

[SwitchB-bgp] undo synchronization

[SwitchB-bgp] group b1 external

[SwitchB-bgp] peer 192.1.1.1.1 group b1 as-number 100

[SwitchB-bgp] group b2 internal

[SwitchB-bgp] peer 194.1.1.1.1 group b2

[SwitchB-bgp] peer 195.1.1.1.2 group b2

[SwitchB-bgp] ipv4-family multicast

[SwitchB-bgp-af-mul] peer b1 enable

[SwitchB-bgp-af-mul] peer b2 enable

3)         Configure Switch C

<SwitchC> system-view

System View: return to User View with Ctrl+Z.

[SwitchC] vlan 30

[SwitchC-vlan30] port ethernet3/1/3

[SwitchC-vlan30] quit

[SwitchC] interface vlan-interface 30

[SwitchC-Vlan-interface30] ip address 193.1.1.2 255.255.255.0

[SwitchC-Vlan-interface30] quit

[SwitchC] vlan 50

[SwitchC-vlan50] port ethernet3/1/5

[SwitchC-vlan50] quit

[SwitchC] interface vlan-interface 50

[SwitchC-Vlan-interface50] ip address 195.1.1.2 255.255.255.0

[SwitchC-Vlan-interface50] quit

[SwitchC] ospf

[SwitchC-ospf-1] area 0

[SwitchC-ospf-1-area-0.0.0.0] network 193.1.1.0 0.0.0.255

[SwitchC-ospf-1-area-0.0.0.0] network 195.1.1.0 0.0.0.255

[SwitchC-ospf-1-area-0.0.0.0] quit

[SwitchC-ospf-1] quit

[SwitchC] bgp 200

[SwitchC-bgp] undo synchronization

[SwitchC-bgp] group c1 external

[SwitchC-bgp] peer 193.1.1.1 group c1 as-number 100

[SwitchC-bgp] group c2 internal

[SwitchC-bgp] peer 194.1.1.2 group c2

[SwitchC-bgp] peer 195.1.1.1 group c2

[SwitchC-bgp] ipv4-family multicast

[SwitchC-bgp-af-mul] peer c1 enable

[SwitchC-bgp-af-mul] peer c2 enable

# Configure the local preference attribute of Switch C.

l           Add ACL 2000 on Switch C to permit network 1.0.0.0.

[SwitchC] acl number 2000

[SwitchC-acl-basic-2000] rule permit source 1.0.0.0 0.255.255.255

[SwitchC-acl-basic-2000] quit

l           Define the routing policy named "localpref". Set the local preference for the routes matching ACL 2000 to 200, and otherwise, to 100.

[SwitchC] route-policy localpref permit node 10

[SwitchC-route-policy] if-match acl 2000

[SwitchC-route-policy] apply local-preference 200

[SwitchC-route-policy] quit

[SwitchC] route-policy localpref permit node 20

[SwitchC-route-policy] apply local-preference 100

l           Apply this routing policy to the inbound traffic from BGP neighbor 193.1.1.1 (Switch A).

[SwitchC] bgp 200

[SwitchC-bgp] ipv4-family multicast

[SwitchC-bgp-af-mul] peer 193.1.1.1 route-policy localpref import

4)         Configure Switch D

<SwitchD> system-view

System View: return to User View with Ctrl+Z.

[SwitchD] vlan 40

[SwitchD-vlan40] port ethernet4/1/4

[SwitchD-vlan40] quit

[SwitchD] interface vlan-interface 40

[SwitchD-Vlan-interface40] ip address 194.1.1.1 255.255.255.0

[SwitchD-Vlan-interface40] quit

[SwitchD] vlan 50

[SwitchD-vlan50] port ethernet4/1/5

[SwitchD-vlan50] quit

[SwitchD] interface vlan-interface 50

[SwitchD-Vlan-interface50] ip address 195.1.1.1 255.255.255.0

[SwitchD-Vlan-interface50] quit

[SwitchD] ospf

[SwitchD-ospf-1] area 0

[SwitchD-ospf-1-area-0.0.0.0] network 194.1.1.0 0.0.0.255

[SwitchD-ospf-1-area-0.0.0.0] network 195.1.1.0 0.0.0.255

[SwitchD-ospf-1-area-0.0.0.0] network 4.0.0.0 0.0.0.255

[SwitchD-ospf-1-area-0.0.0.0] quit

[SwitchD-ospf-1] quit

[SwitchD] bgp 200

[SwitchD-bgp] undo synchronization

[SwitchD-bgp] group d1 internal

[SwitchD-bgp] peer 194.1.1.2 group d1

[SwitchD-bgp] peer 195.1.1.2 group d1

[SwitchD-bgp] ipv4-family multicast

[SwitchD-bgp-af-mul] peer d1 enable

To make the configuration effective, you need to use the reset bgp all command on all MBGP neighbors.