From: Jongsoo.Kim@Intelsat.com
Date: Fri Feb 18 2005 - 21:40:59 GMT-3
I think this is one of many IPv6 bugs as cisco empire are always using us
for their debugging.
If I see this on the test, I always manually put link-local address because
of your email.
In fact, if possible, I would always manually put link-local because it is
easy to use it.
Thanks
Jongsoo
-----Original Message-----
From: Matt Mullen [mailto:mullenm@gmail.com]
Sent: Friday, 18 February, 2005 4:38 PM
To: ccielab@groupstudy.com
Subject: BGP Peering using IPv6 link-local
Hi guys,
Just wondering if anyone else has run into this. I am trying to peer
R2 to R5 using link-local addresses and the session will not come up.
R2 and R5 can ping each other's link-local address. BGP debug
commands don't yield any useful information, however the debug ipv6
packet detail on R5 shows this:
*Mar 1 02:04:06.827: IPV6: source FE80::260:5CFF:FEF4:15D0 (local)
*Mar 1 02:04:06.831: dest FE80::210:7BFF:FE35:CC72 (Ethernet0)
*Mar 1 02:04:06.835: traffic class 192, flow 0x0, len 64+0,
prot 6, hops 64, originating
*Mar 1 02:04:06.835: IPv6: Encapsulation failed
R5 should be using Serial0 to reach the link-local address of R2, but
it appears to be trying to use Ethernet0. Here's the configuration:
R2#show run | b router bgp
router bgp 246
no synchronization
bgp router-id 2.2.2.2
no bgp default ipv4-unicast
bgp log-neighbor-changes
neighbor FE80::260:5CFF:FEF4:15D0 remote-as 50
neighbor FE80::260:5CFF:FEF4:15D0 ebgp-multihop 255
neighbor FE80::260:5CFF:FEF4:15D0 update-source Serial0
no auto-summary
!
address-family ipv4 multicast
no auto-summary
no synchronization
exit-address-family
!
address-family ipv6
neighbor FE80::260:5CFF:FEF4:15D0 activate
neighbor FE80::260:5CFF:FEF4:15D0 route-map SETNEXTHOP out
exit-address-family
!
R5#show run | b router bgp
router bgp 50
no synchronization
bgp router-id 5.5.5.5
no bgp default ipv4-unicast
bgp log-neighbor-changes
neighbor FE80::210:7BFF:FE35:CC72 remote-as 246
neighbor FE80::210:7BFF:FE35:CC72 ebgp-multihop 255
neighbor FE80::210:7BFF:FE35:CC72 update-source Serial0
no auto-summary
!
address-family ipv4 multicast
no auto-summary
no synchronization
exit-address-family
!
address-family ipv6
neighbor FE80::210:7BFF:FE35:CC72 activate
exit-address-family
What is interesting is that if I hard-code the link-local addresses on
both of the routers and use the hard-coded addresses for the peering,
the session comes up fine. So is it necessary to hard-code the
link-local addresses when using them to set up BGP peering?
Thanks,
Matt
This archive was generated by hypermail 2.1.4 : Thu Mar 03 2005 - 08:51:23 GMT-3