All things related to multicast

[Caren Litvanyi <>]

Hi folks,
Just thought I'd pass on a little info in case someone else is running
into this.
We found that a Cisco 6509 running E-train IOS code prior to 12.1(26)E
speaking MBGP with a Juniper running JUNOS 6.4 interprets all the
SAFI=2 nexthops as 0.0.0.0.  This was fixed by moving to 12.1(26)E.
The Abilene backbone is running JUNOS 6.4, and I am not sure if any
other connectors are running IOS E-train code.  Not many?
More generally, when it comes to troubleshooting, I see an awfully
lot of MBGP MP_REACH_NLRI nexthop interpretation bugs.  I have had
cases related to it open with every pairwise set of vendors I've
worked with in the past few years.  Usually I have to tcpdump the
bgp updates and look very closely for standards compliance and
work with the vendors on their interpretation.  So, just a place to
look if you are having RPF issues.
Thanks,
Caren
------------
Caren Litvanyi, network engineer
Global Research NOC at Indiana University

o:812-856-4576  c:720-839-7739