Re: [RRG] Feasibility of transport based approaches?

[Brian E Carpenter <brian.e.carpenter@gmail.com>]

Hannu,

SCTP is probably the model to look at. In shim6, we concluded
that since SCTP is locator-agile for a given connection, it's
basically pointless and probably harmful to run both shim6
and SCTP on the same packet stream - so SCTP needs to be able to
tell the network layer to disable shim6 for an SCTP socket.

I think we'll find the same here - since SCTP is locator-agile,
it would solve the PA-based multihoming problem nicely without any
kind of dynamic EID-RLOC mapping in the routing system.
(Like shim6, it takes control of path selection out of the routing
system.)

Of course, there is the small problem that the network actually
uses TCP and UDP, but that is intrinsic to any transport based
solution.

    Brian

On 2008-05-19 04:07, hannu.flinck@nsn.com wrote:
> Hello (if there is anybody? Has been quite silent recently.)
> 
> Recently we have been discussing (e.g. RRG process clarification
> discussion) about how to classify different approaches.
> 
>> branching structure:
>>
>> 	- Map-n-encap
>> 	- Translation
>> 	- Transport
> 
> Is there a "Transport" based proposal on the table? I know that Mark
> Handly's work has been referred earlier, but is it really meeting the
> design goals? Particularly I am curious to see how a transport (or a
> solution that is in the end points) can possibly solve the routing
> scalability problem that is in the core of the network. How is the
> number of the routing entries going to be decreased with transport level
> solution? I can clearly see that multihoming is be solved but not the
> primary goal of scalability. Any thoughts? 
> 
> If there is not idea how the scalability is improved then it looks to me
> that "transport" solutions are not addressing the problem space and we
> can drop them at least until there is a proposal that also addresses the
> scalability part.
> 
> Regards 
> Hannu
> 

--
to unsubscribe send a message to rrg-request@psg.com with the
word 'unsubscribe' in a single line as the message text body.
archive: <http://psg.com/lists/rrg/> & ftp://psg.com/pub/lists/rrg