Establishing the MP-EBGP Peer Relationship Between PEs

By importing extended community attributes to the Border Gateway Protocol (BGP), Multiprotocol Extensions for External Border Gateway Protocol (MP-EBGP) can advertise virtual private network version 4 (VPNv4) routes between provider edges (PEs). PEs of different autonomous systems (ASs) are generally not directly connected. Therefore, to set up the External Border Gateway Protocol (EBGP) connection between the PEs of different ASs, configure the permitted maximum hops between the PEs.

Procedure

Configure an autonomous system boundary router (ASBR) to send the loopback interface IP addresses of a PE used for peer relationship establishment to the ASBRs of other ASs and peer PEs. You can also configure a PE to advertise its loopback interface IP addresses used for peer relationship establishment to the peer PEs.

If you want to use inter-AS traffic engineering (TE) tunnels to transmit traffic in inter-AS Option C networking, perform the following steps on PEs so that the loopback interface IP addresses of the PEs used for peer relationship establishment can be advertised to peer PEs.
1. Access the system view.
```
system-view
```
2. Access the BGP view.
```
bgp { as-number-plain | as-number-dot }
```
3. Advertise the loopback address of the PE in the local AS to the remote ASBR.
```
network ip-address [ mask | mask-length ] [ route-policy route-policy-name ]
```
(Optional) Disable an ASBR from advertising BGP supernet labeled routes.
In an inter-AS VPN Option C scenario, a PE uses a routing policy to assign a label to its loopback address route and advertises this route as a BGP labeled route. When an ASBR receives the route, the route is a BGP supernet labeled route in which the destination address and next hop address are the same or the destination address is more detailed than the next hop address.
1. Access the system view.
```
system-view
```
2. Access the BGP view.
```
bgp { as-number-plain | as-number-dot }
```
3. Disable the ASBR from advertising BGP supernet labeled routes.
```
supernet label-route advertise disable
```
  After you disable the ASBR from advertising BGP supernet labeled routes, to advertise the loopback address route of a PE in the local AS to a PE in another AS, run the network command on the ASBR to advertise the BGP route to the loopback address of the PE in the same AS.
Perform the following steps on the PE that is connected to a customer edge (CE).
1. Access the system view.
```
system-view
```
2. Access the BGP view.
```
bgp { as-number-plain | as-number-dot }
```
3. Specify the peer PE as the EBGP peer.
```
peer ipv4-address as-number { as-number-plain | as-number-dot }
```
4. Use an interface to set up a TCP connection with the EBGP peer.
```
peer ipv4-address connect-interface loopback interface-number
```
5. Set the maximum number of hops of the EBGP peer.
```
peer ipv4-address ebgp-max-hop [ hop-count ]
```
  PEs of different ASs are generally not directly connected. To set up the EBGP peer between PEs of different ASs, configure the maximum hop between PEs and ensure that the PEs are reachable.
6. Access the BGP VPNv4 address family view.
```
ipv4-family vpnv4 [ unicast ]
```
7. Enable the exchange of VPN IPv4 routes with the peer PE.
```
peer ipv4-address enable
```
(Optional) Configure a route reflector (RR).
To improve scalability, specify an RR in each AS and establish MP-EBGP peer relationships between the RRs in ASs to save all VPNv4 routes on the RRs. Then configure PEs in each AS as the RR's clients to exchange VPNv4 routing information with the RR. Perform the following steps on a device that needs to be configured as an RR:
1. Access the system view.
```
system-view
```
2. Access the BGP view.
```
bgp { as-number-plain | as-number-dot }
```
3. Specify the RR in another AS as its EBGP peer.
```
peer ipv4-address as-number { as-number-plain | as-number-dot }
```
4. Specify the source interface that sends BGP packets.
```
peer ipv4-address connect-interface loopback interface-number
```
5. Set the maximum number of hops between RRs for which an EBGP peer relationship is to be configured.
```
peer ipv4-address ebgp-max-hop [ hop-count ]
```
6. Access the BGP VPNv4 address family view.
```
ipv4-family vpnv4 [ unicast ]
```
7. Enable the exchange of VPN IPv4 routes with the peer RR.
```
peer ipv4-address enable
```
8. Configure the device as an RR and specify PEs as its clients.
```
peer ipv4-address reflect-client
```
9. Do not change the next hop when the route is advertised to the EBGP peer and Internal Border Gateway Protocol (IBGP) peer.
```
peer ipv4-address next-hop-invariable
```