Hello Michael: The BBR draft will be repositioned as a number of functions are migrated to other drafts, including efficient ND. What will be left in the BBr draft is probably the function by a RPL root (LBR) to proxy ND over the backbone so as to attract traffic for IPv6 addresses that are located on the LLN. So it is not just a root. The key piece is the mapping (redistribution?) of RPL routes learnt from DAO into a Binding Table state in order to enable ND proxy operation.
> Are the wired part of the 6LBRs always located on the same link? Yes. This is why we call this link the backbone. It is expected to be a transit link where traditional NDP operates. Backbone Router is the term used by ISA100.11a to refer to the routing entity that federates multiple LLNs as a single subnet over a high speed backbone. I'm using that same term to ease communication between ISA and IETF. > It seems to me that you are inventing a protocol to solve the MESH-under > problem, but that we already have such a protocol. The backbone router does not operate mesh routing at all. It is really focused on ND proxy operation over a transit link we use as federating backbone for a multilink subnet. Cheers, Pascal -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Michael Richardson Sent: lundi 26 août 2013 02:37 To: [email protected]; [email protected] Subject: [6lo] 6lowpan-backbone-router-03 Pascal Thubert (pthubert) <[email protected]> wrote: > There is a need for a protocol between the 6LBRs/NEAR routers to > exchange their state. This could be achieved by an extension on ND for > proxy operations as described in dratf-thubert-6lowpan-backbone-router, > a similar extension to a routing protocol running on the backbone or a > richer extension to IPv6 NDP allowing some degree of pub/sub as well. Are the wired part of the 6LBRs always located on the same link? I'm asking because it seems like this is an important assumption. I read 6lowpan-backbone-router-03. Please number the figure in section 3. I'll call it "Figure Plan Network" I don't like the term backbone, and backbone router. What I see are three areas of a DAG connected by a heterogeneous layer-2. The "BBR" are just RPL routers. I don't know why a new protocol is necessary for the mesh-over case. Perhaps the MESH-under case has some requirements, but I'd rather just pretend that it's four links (an ethernet link, and three mesh-under links) connected together to form a single mesh-over network. It seems to me that you are inventing a protocol to solve the MESH-under problem, but that we already have such a protocol. -- Michael Richardson -at the cottage- -------------------------------------------------------------------- IETF IPv6 working group mailing list [email protected] Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
