MP-BGP with ipv4 and ipv6 address family

Hello Rene,

I am kind of confused with the topic, in the previous example that you provided regarding MP-BGP with ipv4 ipv6, why we have to use route-map for that? also, the next hop address is an ipv4 address, why we have to change it to an ipv6 address? kind got confused with that, could you please explain it again. Thanks :slight_smile:

Are you referring to this example?

When your routers have an IPv4 adjacency and are exchanging IPv4 prefixes, there are no issues. The same thing applies when you have an IPv6 adjacency and you are exchanging IPv6 prefixes.

You will run into issues when you have an IPv4 adjacency and are trying to exchange IPv6 prefixes or the other way around…IPv6 adjacency with IPv4 prefixes.

In the second example (MP-BGP with IPv4 adjacency & IPv6 prefixes) we have two routers with an IPv4 adjacency but they are exchanging IPv6 prefixes. Our routers don’t have any valid next hop addresses that they can use (::FFFF: and ::FFFF: are no valid addresses). To fix this, I used a route-map to change the next hop address.


Hello Rene,

Yes i am referring to this example:

the output here got me confused : my confusion is with this :

R1# sh ip bgp ipv6 unicast

    Network          Next Hop            Metric LocPrf Weight Path
*> 2001:DB8::1/128  ::                       0         32768 i
*  <strong>2001:DB8::2/128  ::FFFF:</strong>
                                             0             0 2 i

the next hop for 2001:DB8::2/128 is, why we have to change the next hop to an ipv6 address??? can we still
use the ipv4 as the next hop to reach that network?? please clarify for me. Thanks a lot Rene :slight_smile:

continuation for previous question:

in the config below:

R2(config)#<strong>router bgp 2</strong>
R2(config-router)#<strong>address-family ipv6</strong>
R2(config-router-af)#<strong>neighbor route-map IPV6_NEXT_HOP in

why the route-map have to be (in) not (out)
neighbor route-map IPV6_NEXT_HOP in

since we are talking to R1 in this case ??? thanks Rene

I understand your confusion…

When a router receives an IPv6 packet with a destination, it will check its IPv6 routing table. You can’t have an IPv4 address as the next hop, it has to be an IPv6 address.

About the route-map, I changed the next hop inbound when the router receives the prefixes. You can also apply it outbound so that you change the next hop when you advertise the prefixes.