Re: [MEXT] Review of draft draft-patil-mext-mip6issueswithipsec-01

[[e-mail] (Arnaud Ebalard)]

Hi,

<[e-mail]> writes:

> I will respond to your other comments as soon as I get some time..

ok.

> Raj> MIP6 by itself is useless today. Maybe 10 years from now when we
> hopefully have some deployment of IPv6 networks and is available
> natively on links, it would be an option. 

Well, you should come to France ;-) At work, I have IPv6 (Nerim as a
provider). At Home, I have IPv6 (Free as a provider, which has more
than 5 millions clients who can activate IPv6 at home via a simple
clic (/64)).

> The only practical solution at this time is DSMIP6 because it deals
> with the reality of the fact that MNs will be mostly attached via IPv4
> and NATed networks.

In order to be more useful than MIPv6 (using Teredo when in IPv4
networks) or a direct use of the available private IPv4 address in a
common network, I think one needs a public IPv4 address as HoA. IMHO,
this is a non starter in term of deployment.

For me, MIPv6 provides address stability and e2e.

>> Thank you. I think you will realize the complexities of the
>> implementation when you go through an exercise of doing so.
>
> I won't. I just don't support DSMIPv6. Reread my email address more
> carefully ;-)
>
> Raj> Right... I noticed that already. I dont understand your reasons
> for not supporting DSMIP6.

 - Complexity
 - No availability of the protocol as an open source implementation
 - No public IPv4 address to use for a deployment (or expensive)
 - MIPv6 + Teredo works fine with me

> Is it only because of the fact that DSMIP6 has to deal with NATs ;) ?
> As I said for all practical purposes, implementing only MIP6 by itself
> is purely an academic exercise.

It may depend where you live. 

a+