Re: [manet] draft-ietf-olsrv2-multipath-09

[Jiazi YI <[e-mail]>]

Hi,

Regarding the use of source routing header in IPv6 and the reference to
RFC6554, we have contacted the 6man chairs.

The answer from Ole (one of the chairs) was that he didn't see why it can't
be used, and suggested the experimental code points of the IPv6 Routing
Types.

So in the next revision of the olsrv2-multipath, we will:

       - Use the experimental code point of IPv6 Routing Types for SRH.

       - Have a well scoped reference to RFC6554: for example, for SRH
format, use section 3, for SRH processing, use section 4, etc. Other than
that, the olsrv2-multipath would be self-contained.

       - Have a well defined scope of the use of the SRH: only the
OLSRv2/MP-OLSRv2 routing domain.

Any other comments on the approach?

best

Jiazi

On Fri, Jun 24, 2016 at 1:40 PM, Alvaro Retana (aretana) <[e-mail]>
wrote:

> On 6/24/16, 5:16 AM, "[e-mail] on behalf of Jiazi YI" <
> [e-mail] on behalf of [e-mail]> wrote:
>
> Hi!
>
> Of course, As a WG, the priority is achieving the milestones, but I don't
> think it would preclude having extensions to OLSRv2, which is part of the
> charter also. In fact, IIRC, during the recharter discussion, several
> extensions for OLSRv2 were proposed. The WG finished with using the term
> "The MANET WG is responsible for the *maintenance* of OLSRv2 [RFC 7181],
> NHDP [RFC 6130] and the Generalized MANET Packet/Message Format [RFC5444],
> and their *extensions*.", instead of listing possible extensions
> explicitly.
>
>
> As I said: "To be clear: I'm not asking you not to do this work, I want
> the WG Chairs to prioritize appropriately…"  That's it!  If the WG wants to
> take this work on, I don't have a problem — it is with in the charter.
> Let's just not forget about the other work we have…
>
> The second part of the point I tried to make was, I think, more important:
>  extensions to IPv6 are not within the manet WG's charter, but they
> belong to 6man.  It would probably be more effective if the source routing
> work was in the Standards Track.  I'm assuming that Multi-path is just one
> use case for source routing, but there are probably more.  I also wrote:
> "[*] I doubt that 6man will want to put in an extension (even if it's
> already defied in RFC6554) for an Experimental effort.  I of course may be
> wrong and the Experimental nature may make it easier in their eyes to
> simply reference RFC6554."
>
> Note that what I'm suggesting is not to put the Multi-path work in the
> Standards Track (that is a separate discussion) — but to justify source
> routing for MANET networks as a larger need (including multi-path, but
> bigger than that).
>
> Thanks!
>
> Alvaro.
>