Re: [Mip4] Prefix Alloc with DHCPv6-PD? Re: AD review of mip4-dsmipv4

[Basavaraj Patil <basavaraj.patil at nokia.com>]

Hi George,

Comments inline:


On 9/9/08 8:16 AM, "ext George Tsirtsis" <tsirtsis at googlemail.com> wrote:

> Hi Raj,
> 
> I think the distinction you are trying to make no longer applies.
> 
> Based on Jari's feedback we are now only going to support IPv6
> prefixes (no addresses).

Why? Is the link model for IPv6 here being assumed to be a P2P link?
Why not assign an address? (Maybe I missed the discussion on this topic and
if so please point to the appropriate email in the archives).

> In that sense the DSMIPv4 MN can now register
> IPv6 Prefixes of any length it wants to its HAv4, but it can NOT
> register an IPv6 address.

What prevents the MN from registering an IPv6 HoA with the HAv4? I am
missing something here.

> Presumably then the DSMIPv4 "owns" these
> prefixes i.e., it does not share them with anyone else. Also note that
> here is nothing that prevents DSMIPv4 to be used in combination with
> NEMOv4.

Well, if you assign the prefix to the MN, it owns it. But why assign a
prefix and not an address? Agree about the applicabiltty to NEMOv4.

> 
> So, when we are trying to solve the issue of how the MN gets ownership
> of such IPv6 prefixes, it is clear (to me) that we are talking about
> IPv4 delegation, rather than allocation.

I guess you meant to say IPv6 prefix delegation above.
But the fundamental question is : Why is the MN being assigned a prefix
instead of an IPv6 address (starting to be repetetive here). It is okay to
assign a prefix and that's not the problem. But I think what you are saying
here is that with DSMIP4 the MN is assigned a prefix only and not just an
address. Is it the IPv6 link which is created over the (M)IPv4 tunnel that
is the problem? Or are you trying to avoid the issues of neighbor
solicitations etc. that exist if the link is considered as shared?

-Raj

> 
> Am I missing something?
> 
> Regards
> George
> 
> 
> 
> On Mon, Sep 8, 2008 at 5:58 PM, Basavaraj Patil
> <basavaraj.patil at nokia.com> wrote:
>> 
>> Hi George,
>> 
>> The I-D in MEXT proposing PD is applicable to NEMO (mobile routers).
>> I would expect that this mechanism would als o make sense for Proxy MIP6
>> where the MAG is assigning prefixes to an MN. However I don't see how this
>> would be applicable in the case of dsmip4. It would work but the prefix is
>> being assigned to the MN, i.e not being delegated to an MR or LMA type of
>> entity.
>> 
>> -Raj
>> 
>> 
>> On 9/8/08 8:40 AM, "ext George Tsirtsis" <tsirtsis at googlemail.com> wrote:
>> 
>>> Folks,
>>> 
>>> Any response to my question below?
>>> 
>>> Thanks
>>> George
>>> 
>>> On Tue, Sep 2, 2008 at 4:42 AM, George Tsirtsis <tsirtsis at googlemail.com>
>>> wrote:
>>>> I have one more question/suggestion as I am in the process of editing
>>>> the draft to incorporate Jari's earlier comments.
>>>> 
>>>> Section 4.6. "Dynamic IPv6 Prefix allocation", introduces a mechanism
>>>> for IPv6 prefix allocation using MIPv4 extension; similar to IPv4 HoA
>>>> address allocation. In light of the decision made recently in MEXT,
>>>> however, to do IPv6 Prefix allocation in NEMOv6, using DHCPv6-Prefix
>>>> Delegation (draft-ietf-mext-nemo-pd-00), should we change this?
>>>> 
>>>> I think it is the right thing to do. In this case we would only have
>>>> to indicate that DHCPv6 PD runs over the IPv4 HoA to HA tunnel, and
>>>> then when the IPv6 prefix is allocated, and assuming explicit mode, it
>>>> has to be registered via the extensions defined in this spec. I assume
>>>> that we can point directly to the DHCPv6 PD RFC and we do not need to
>>>> depend on draft-ietf-mext-nemo-pd.
>>>> 
>>>> Thoughts?
>>>> 
>>>> George
>>>> 
>>>> On Tue, Sep 2, 2008 at 8:38 AM, George Tsirtsis <tsirtsis at googlemail.com>
>>>> wrote:
>>>>> OK, I received some positive responses and no negative ones. I will
>>>>> remove support for IPv6 addresses from the spec.
>>>>> 
>>>>> Thanks
>>>>> George
>>>>> 
>>>>> On Fri, Aug 29, 2008 at 6:15 PM, Henrik Levkowetz <henrik at levkowetz.com>
>>>>> wrote:
>>>>>> 
>>>>>> 
>>>>>> On 2008-08-29 11:30 George Tsirtsis said the following:
>>>>>>> Hi All,
>>>>>>> 
>>>>>>> I reviewed Jari's comments and I think we need to decide on at least
>>>>>>> one of these as a group before I can create a new version.
>>>>>>> 
>>>>>>> The current version of the draft supports both IPv6 Prefixes and IPv6
>>>>>>> addresses. Jari questions why we need to support addresses and I do
>>>>>>> not have a good answer for that at the moment. Jari, specifically
>>>>>>> points out that a number of issues in the document would be simplified
>>>>>>> if we dropped support for addresses i.e., we would:
>>>>>>>> - get rid of DAD rules
>>>>>>>> - get rid of ND proxying
>>>>>>>> - avoid having to say something about SEND in security considerations
>>>>>>>> section
>>>>>>>> - simplify authorization and overlap rules
>>>>>>>> - not have any multilink subnet-style issues
>>>>>>>> etc
>>>>>>> 
>>>>>>> Is the WG in agreement with Jari's proposal to only support IPv6
>>>>>>> Prefixes in this specification?
>>>>>> 
>>>>>> Works well for me.
>>>>>> 
>>>>>> 
>>>>>>        Henrik (no hat)
>>>>>> 
>>>>>> 
>>>>> 
>>>> --
>>>> Mip4 mailing list: Mip4 at ietf.org
>>>>    Web interface: https://www.ietf.org/mailman/listinfo/mip4
>>>>     Charter page: http://www.ietf.org/html.charters/mip4-charter.html
>>>> Supplemental site: http://www.mip4.org/
>>>> 
>> 
>> 

-- 
Mip4 mailing list: Mip4 at ietf.org
    Web interface: https://www.ietf.org/mailman/listinfo/mip4
     Charter page: http://www.ietf.org/html.charters/mip4-charter.html
Supplemental site: http://www.mip4.org/