automatic ipv4 compatible tunnel supports RIPng?

Unanswered Question
Jun 3rd, 2010

http://www.cisco.com/en/US/tech/tk872/technologies_configuration_example09186a00800b49a5.shtml

with reference to this link.

They have given an example for automatic ipv4 compatible tunnel configuration.


interface Tunnel0
no ip address
no ip redirects
ipv6 rip 6bone enable
tunnel source Serial1/5
tunnel mode ipv6ip auto-tunnel

I wondered if this configuration can support RIPng with an ipv4 address in its link layer.

Can you suggest on this please.

Thanks in advance

I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Average Rating: 0 (0 ratings)
Giuseppe Larosa Fri, 06/04/2010 - 06:58

Hello Ramanan,

you can always use a different tunnel for IPv4 if you need it, these are logical interfaces

Hope to help

Giuseppe

ayyapanrcisco Fri, 06/04/2010 - 21:37

Hello Giuslar,

Thanks for your reply. but my question is how can two routers can establish ripng neighbourship with ipv4 address instead of mac address in its linklayer. The webpage I referred to is of cisco and it suggests RIPng configured on a tunnel interface. In my understanding this is not possible in either automatic6to4 tunnel or ipv4 compatible tunnel.

If this is possible, please suggest me the configuration. I can upload my configuration details.

Thanks,

Ayyappan

Giuseppe Larosa Sat, 06/05/2010 - 08:05

Hello Ayappan,

to build a RIPng adjacency the two devices need to be able to exchange IPv6 packets with link local source and multicast IPv6 link local destination.

If these IPv6 packets are encapsulated into an IPv4 packet (this is the meaning of tunneling) this should be possible.

However, looking at the example you have provided and at the IPv6 configuration guide it is not possible:

a BGP session using  IPv6 automatic addresses is used over the tunnel that is treated as NBMA (non broadcast multiaccess) and so it should not be capable of supporting IPv6 multicast natively.

the command ipv4 rip enable that is present in the tunnel configuration is not there to build a RIP adjacency, but to make reachable and known in RIP domain the IPv6 BGP next-hop it is used to advertise the IPv6 prefix in the autotunnel.

So in the example you have two RIPng domains interconnected with a BGP session over the tunnel.

http://www.cisco.com/en/US/tech/tk872/technologies_configuration_example09186a00800b49a5.shtml#configsipv6

also you see BGP redistributed into RIP

Hope to help

Giuseppe

ayyapanrcisco Sun, 06/06/2010 - 14:07

Hello,

I totally agree with your answer. I tried some of the tunnelling techniques in which gre and manual tunnels gives link local address reachability between the routers and hence form adjacencies without a problem. The Automatic6to4 and ipv4 compatible tunnels dont allow the routers to reach each other's link local address and hence not the adjacencies. There are 2 possibilities.

1) The tunnels which know their destination addresses ie point-to-point tunnels make link local reachability feasbile while others dont

2) The Manual and gre tunnels are treated point-to-point links whereas automatic6to4 and ipv4-compatible tunnels are treated as NBMA links by the routing protocols and hence need special configuration like neighbour command to form adjacencies. ( i tried but routers still cannot reach each other's link local address)

Kindly clarify which perspective is right.

Regards,

Ayyappan

Actions

Login or Register to take actions

This Discussion

Posted June 3, 2010 at 11:54 PM
Stats:
Replies:4 Avg. Rating:
Views:1357 Votes:0
Shares:0

Related Content

Discussions Leaderboard