Re: [Monami6] Fwd: I-D ACTION:draft-bagnulo-shim6-mip-00.txt

[Thierry Ernst <[e-mail]>]

Hi Marcelo,

I did miss that mail you sent on 21, as you pointed to me on 25 - sorry
about this. Please read my answer below:

On Thu, 21 Jul 2005 11:22:52 +0200
marcelo bagnulo braun <[e-mail]> wrote:

> > If I'm wrong in this, or in any case where you think it should be
> > discussed at Paris meeting, please explain how you think this would
> > fit with the objectives of the BOF, and particularly, which of the
> > issues raised in draft-montavont-mobileip-multihoming-pb-statement
> > this is related to. Otherwise, I cannot support the request.
> 
> The point is that the SHIM work is a potential solution to some of the
> problems defined in draft-montavont-mobileip-multihoming-pb-statement 
> (and to some other problems that are not in the draft but need to be 
> included in order to really solve the problem, as i detailed in the 
> mail i sent commenting this draft)
> 
> In particular, the running the shim and MIP is a mechanisms that can
> be used to preserve the established communications through failures in
> the HoA used in the communication (this means that it can be used to 
> provide the fault tolerance capabilities required described in
> sections 5.2, 5.4, 5.7, 5.8 and 5.9 of 
> draft-montavont-mobileip-multihoming-pb-statement)

First, I would like to clarify one thing: the draft is trying to be
exhaustive about the issues; it doesn't mean we intend to solve all the
issues.

> In addition, the mechanisms defined in the SHIM wg to deal with
> ingress filtering could be applied to deal with the ingress filtering
> issue described in section 6.1.2 of 
> draft-montavont-mobileip-multihoming-pb-statement

OK, this is one of the generic issues, on we don't intend to solve this
has claimed by the proposed charter:

 - Seek for help or push the other WGs to standardize solutions not
 specific to mobility,

unless no WG is taking it.

> In addition, you will be needing mechanisms for path failure detection
> (not included in draft-montavont-mobileip-multihoming-pb-statement but
> that are needed to provide proper multihoming support) that the shim 
> can provide, as some mechanisms for exploring alternative available 
> paths in order to rehome to alternative paths after an outage.

Agree with you. The text should express this issue explictly, and I
guess the issue would belong to the "generic" category, so not one of
the issue Monami6 would have to solve directly.

> So, there are clearly multiple issues of the ones describe in 
> draft-montavont-mobileip-multihoming-pb-statement for which the shim 
> seems to be a reasonable candidate for providing a solution

Agree that Shim6 is a valid solution to these issues, but again, it's
not the purpose of Monami6 to solve them; however our mission would be
to examine the interactions with other WGs, i.e. with Shim6 in the
present case.

> The proposal then is to present the shim as a possible solution for 
> some of the problems described in 
> draft-montavont-mobileip-multihoming-pb-statement, just as the
> multiple CoAs drafts or the mmi and so on.

The difference is that MCoA is an issue MIP-explicit issue, so we want
to solve it whereas I guess the other your mention would be solved in
the Shim6 WG, right ? So, Monami6, as a WG, would have to interact with
Shim6 to ensure compatibility with MIP and NEMO.

> Is this a good enough justification for you?

It's a good justification for us to keep a close eye on Shim6 work, but
not appropriate, I think, to discuss the Shim6 approach at the next BOF.

Can we please speak about the Monami6 charter instead ?

Thierry

_______________________________________________
Monami6 mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/monami6