[quagga-users 13231] Re: Ripng propagate global link address

richard xiao xrich1 at yahoo.com
Sat Apr 6 01:53:54 BST 2013

Again, this is what I find:
With quagga 0.99.20, when Zebra redistribute sripng routes to BGP, it does not contain next hop. When iBGP node sends the UPDATE, it inserts the global address of the iBGP node and sends to the peer.
With quagga 0.99.21, when zebra redistribute ripng routes to BGP, it inserts the next hop info. This actually is a fix from 0.99.20. http://nongnu.askapache.com/quagga/quagga-0.99.21.changelog.txt
Author: Stephen Hemminger <shemminger at vyatta.com>
Date:   Tue Dec 6 14:51:10 2011 +0400

    bgpd: store nexthop info for redistributed IPV6 routes
    BGP was ignoring nexthop info for static and other redistributed
    routes for IPv6.  Build extra attribute info to store the nexthop.
    See also:

In IPv6 case, this next hop will be local link address when redistributed from ripngd. However, we do have problem in some cases, if we use iBGP to propagate this route. The other end will not understand this local link address at all. Up to now, I am not able to find any way to set up the next-hop for this ripng route
One may argue eBGP should be used in this case, however, I think "set next-hop" should be functional for iBGP when getting redistributed from Ripngd(or other protocols). 
In my case when eBGP is used, IPV6 recursive routing is not triggered from Zebra, which defeats my routing purpose.
Hopefully someone can give any strong answer here.

 From: Juliusz Chroboczek <jch at pps.jussieu.fr>
To: Nick Hilliard <nick at inex.ie> 
Cc: quagga-users at lists.quagga.net 
Sent: Friday, April 5, 2013 3:37 PM
Subject: [quagga-users 13229] Re: Ripng propagate global link address
> I've never heard of anyone attempting to use RIPng as an IGP for iBGP
> before.  It's a ridiculous protocol and is unsuitable for production use.

Strong words.  Softly spoken?

RIPng is a perfectly good protocol in stable topologies of less than
100 routers or so.  The universities of Paris 6 and 7 have been using
it for years (we've been using OSPF for IPv4 and RIPng for IPv6).

Granted, the implementation of RIPng in Quagga is somewhat deficient
(it doesn't implement triggered updates, and I'm not sure about
explicit retractions), but that has nothing to do with the protocol itself.

> Do yourself a favour and use ospf instead.  Or isis.

With the resulting administrative overhead, and the inability to
perform filtering and aggregation except at area boundaries.

-- Juliusz

Quagga-users mailing list
Quagga-users at lists.quagga.net
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quagga.net/pipermail/quagga-users/attachments/20130405/89ae0674/attachment-0001.html>

More information about the Quagga-users mailing list