Cisco Support Community
cancel
Showing results for 
Search instead for 
Did you mean: 

Design Inter-Domain Multicast redundancy using MBGP and MSDP

Overview:

This document will discuss the design concept and configurations of redundant Multicast paths between two multi-homed BGP autonomous systems, by using multicast source discovery protocol MSDP and Multicast BGP MBGP.

Please refer the following document for more details about MSDP configurations between two BGP AS(s):

https://supportforums.cisco.com/docs/DOC-8293

Inter domain Multicast network Design Example:

As illustrated in the bellow diagram, we have two BGP AS(s) , AS 100 has one Edge router that uses two links to connect to AS 200 each link connect to a different router in AS 200, the main goal of this design is to have the IP multicast receivers in AS 100 receive IP Multicast traffic from the Sender in AS 200 using  multicast group 225.1.1.1.

Furthermore, Multicast traffic must flow over the link with subnet 172.16.1.0/24 and the other link will be used as back up or redundant link, while unicast traffic must use the link with subnet 192.168.1.0 as the primary link:

mbgp.jpg

Design Considerations:

To achieve the above requirements we need first to see the IPv4 unicast traffic path in the above network

Lets start with AS 100 :

R1#show ip bgp

BGP table version is 5, local router ID is 1.1.1.1

   Network          Next Hop            Metric LocPrf Weight Path

*> 3.3.3.0/24       192.168.1.3              0             0 200 i

*  4.4.4.0/24       172.16.1.2          156160             0 200 i

*>                  192.168.1.3         156160             0 200 i

*> 10.1.1.0/24      0.0.0.0                  0         32768 i

*  20.1.1.0/24      172.16.1.2               0             0 200 i

*>                  192.168.1.3              0             0 200 i

As we can see R1 ( AS100 ) uses R3 192.168.1.3 as the primary IPv4 unicast next hop

This is important information to be known for multicast RPF check.

MSDP has already been configured between BGP peers

R1#show ip msdp summary

MSDP Peer Status Summary

Peer Address     AS    State    Uptime/  Reset SA    Peer Name

                                Downtime Count Count

172.16.1.2       200   Up       00:23:33 1     0     ?

192.168.1.3      200   Up       00:19:12 1     2     ?

lets check the multicast routing table in AS 100 were the receivers reside:

R1#show ip mroute

(*, 225.1.1.1), 00:00:19/stopped, RP 1.1.1.1, flags: S

  Incoming interface: Null, RPF nbr 0.0.0.0

  Outgoing interface list:

    FastEthernet2/0, Forward/Sparse, 00:00:19/00:03:10

In the mroute table of R1 we have the group 225.1.1.1, which mean that an igmp host is interested to join this multicast group.

Lets generate ICMP traffic from host 4.4.4.4 to group 225.1.1.1 to simulate a multicast traffic, then we will check R1 mroute table again to see the difference:

R1# show ip mroute

IP Multicast Routing Table

(*, 225.1.1.1), 00:04:59/00:03:27, RP 1.1.1.1, flags: S

  Incoming interface: Null, RPF nbr 0.0.0.0

  Outgoing interface list:

    FastEthernet2/0, Forward/Sparse, 00:04:59/00:03:27

(4.4.4.4, 225.1.1.1), 00:00:08/00:02:52, flags: M

  Incoming interface: FastEthernet1/1, RPF nbr 192.168.1.3

  Outgoing interface list:

    FastEthernet2/0, Forward/Sparse, 00:00:08/00:03:27

Now we have the traffic flowing normally and we have the (S,G) in the IP Multicast routing Table.

Issue:

Although the multicast traffic if flowing, we have a problem here, the RPF neighbor shown in the mroute table is 192.168.1.3 which is R3 while in the design requirements we have to have the primary multicast path through R2 !!

Resolution:

This issue can be solved in different ways in our example we will see how we will influence multicast path selection and RPF check by using MBGP

All we need here is to make R1 and R2 as BGP multicast neighbors and advertise AS 200 internal subnet and the IP multicast sender address through BGP multicast address family configuration mode:

R2:

Router bgp 200

address-family ipv4 multicast

neighbor 172.16.1.1 activate

network 4.4.4.0 mask 255.255.255.0

network 20.1.1.0 mask 255.255.255.0

R1:

R1#show ip bgp ipv4 multicast

BGP table version is 11, local router ID is 1.1.1.1

   Network          Next Hop            Metric LocPrf Weight Path

*> 4.4.4.0/24       172.16.1.2          156160             0 200 i

*> 20.1.1.0/24      172.16.1.2               0             0 200 i

lets generate traffic again and see the multicast routing table of R1 after the above changes:

R1#show ip mroute

IP Multicast Routing Table

(*, 225.1.1.1), 00:15:49/00:03:28, RP 1.1.1.1, flags: S

  Incoming interface: Null, RPF nbr 0.0.0.0

  Outgoing interface list:

    FastEthernet2/0, Forward/Sparse, 00:15:49/00:03:28

(4.4.4.4, 225.1.1.1), 00:00:28/00:03:25, flags: MT

  Incoming interface: FastEthernet1/0, RPF nbr 172.16.1.2, Mbgp

  Outgoing interface list:

    FastEthernet2/0, Forward/Sparse, 00:00:28/00:03:28

As it shown above the RPF neighbor now is R2 and also notice the word ( Mbgp), which means that the RPF check has been done through MBGP routing table. Furthermore, if the link connected to R2 fails, the multicast traffic will use the redundant path which is through R3, while the unicast traffic is using R3 path as the primary path.

Thank you

Marwan Alshawi

Version history
Revision #:
1 of 1
Last update:
‎01-02-2010 06:54 PM
Updated by:
 
Labels (1)