Re: [netext] Additional comments for mMAG presentation

[Alexandru Petrescu <[e-mail]>]

Le 06/11/2012 14:28, Seil Jeon a écrit :
> Hi folks,
>
> On the mMAG presentation in the meeting, I think allocated slot might
> not be enough to have full discussion.
>
> Let me give explanation about the substance of mMAG proposed in our
> slide again.
>
> As I presented, the mMAG is ‘SEEN’ as one of MNs in MAG perspective
> but this is not a ‘_MN_’ like mobile device you simply have. That is
>  a ‘ROUTER’ managed by the operators. So you can imagine same level
> of security association between mMAG and LMA as the MAG and LMA has.

Except that, when LMA sends a HNP (Home Network Prefix) to MAG to mMAG
such as this HNP to be used by a Host in the moving network - there is
no security between this Host and the LMA.

There is a sort of delegation of authority between LMA and MAG and mMAG
to trust about this Host, but obviously this is not 'end-to-end'
security as one considers typically between a fixed Host and
Correspondent in the Internet.

> And this use case and the need are usually reasonable when
> considering real deployment. Of course, we will update this use case
>  of NEMO in next version. But I think the need or the use case has
> implicitly been assumed in prefix delegation WG draft, right?

Well, one reads that as one wishes.

In a ideal world, the problem statement and requirements about network
mobility support with PMIP would be written first, separately, and allow
time for competing solutions to root and grow.  And later make some
selection or join solutions.  In a real world what happens is that some
solutions grow much faster than others and they practically influence a
problem statement, which is somehow more or less assummed to fit what
that solution needs.

Implementation is a key aspect as well.  If one knows that some
particular solution is safely planned to be implemented in a widely
availabble router platform, then only thing one has to do is to agree to
that solution, since deployment is guaranteed.

Alex

>
> Regards,
>
> Seil
>
>
>
> _______________________________________________ netext mailing list
> [e-mail] https://www.ietf.org/mailman/listinfo/netext
>