Re: [Bgp-autoconf] Move forward with bgp autoconf requirements and design principle

The following analysis on discovery at layer3 surprised me. See my comments inline.

8.  Discovery at Layer Three

   Discovery at Layer-3 can assume IP addressability, though the IP
   addresses of potential peers are not known a priori and need to be
   discovered before further negotiation.

[Xiaohu] It has been proven that multicast is a good choice to address the above concern. For more details, refer to LDP peer discovery mechanism:)

   The principal problem would appear to be discovery at Layer-3,
   because one neither knows whether to use IPv4 or IPv6.  This is
   exacerbated by the possibility of a potential peer not being on the
   local subnet, and hence broadcast and similar techniques may not be
   applicable.

[Xiaohu] Are you intended to perform peer discovery with untrusted devices which are not under your control? If not, why don't you know whether to use IPv4 or IPv6 within your own network (e.g., the datacenter network or metro network which are the major scenarios for BGP peer auto-discovery)?

   If one can assume point-to-point links , then discovery might try
   IPv6 link-local or even IPv4 link-local.  Or maybe a link broadcast
   protocol.

   For switched or bridged multi-point which is at least on the same
   subnet, VLAN, etc., broadcasts might be a viable approach.

[Xiaohu] It has been proven that multicast is applicable to both p2p and broadcast networks. 

   There will be difficulty if one or both peers wish to use a loopback
   for peering.

[Xiaohu] I would like to know what's the exact difficulty in using loopback addresses for BGP session establishment? 

Best regards,
Xiaohu

Bush                    Expires November 26, 2020               [Page 7]

-- 
Bgp-autoconf mailing list
Bgp-autoconf@ietf.org
https://www.ietf.org/mailman/listinfo/bgp-autoconf