RE: [netlmm] The term LMA

Hi,

I think that there is no dependence between "global mobility" and "host" or between "local mobility" and "network".

Kassi 

-----Message d'origine-----
De : Damic, Damjan [mailto:damjan.damic@siemens.com] 
Envoyé : vendredi 11 mai 2007 11:17
À : conny.larsson@ericsson.com; vidyan@qualcomm.com; basavaraj.patil@nsn.com
Cc : netlmm@ietf.org
Objet : Re: [netlmm] The term LMA

Hello Conny et al.,

I believe the notion for using PMIP6 for network-based global-scoped mobility is not essentially in conflict with the WG charter.

For example,
Section 3.10. of RFC4831 specifies: Goal 10: Localized Mobility Management Independent of Global Mobility Management To my understanding this section implies localized mobility management is provided by the network, where global mobility protocol is run directly from the mobile nodes.
This requirement will not be broken if the network is able provide a more advanced service for the nodes which are completely lacking the mobility support. Those MNs comprising the MIP6 mobile client probably could/should override this, and manage global mobility on their own...

Also, placement of HA/LMA within a certain part of the network doesn't categorize protocol local or global by default.
If trying to map the "PMIP6 domain" within an individual SDO architecture it's not as clear what kind of topology this should be.
Relative MAG-LMA distance might be a matter of perception; it surely affects the performance (optimizations could solve this), and does make a constraint to use the same security model, but doesn't seem much more than that. Am I maybe missing something here?

Regards,
	Damjan

>Date: Fri, 11 May 2007 10:05:55 +0200
>From: Conny Larsson <conny.larsson@ericsson.com>
>Subject: Re: [netlmm] The term LMA
>To: "Narayanan, Vidya" <vidyan@qualcomm.com>, 	Basavaraj Patil
>	<basavaraj.patil@nsn.com>
>Cc: netlmm@ietf.org
>Message-ID: <464423E3.1040603@ericsson.com>
>Content-Type: text/plain; charset="iso-8859-1"
>
>Hi Vidya and Raj,
>
>Although I agree with Raj that there is a confusion among the potential 
>customers of PMIP6 as they intend to use PMIP in a global scope whereas 
>the intention of this working group is to use it for local mobility.
>
>However, if the PMIP6 protocol is going to be used for global mobility 
>I believe there are issues that needs to be solved.
>One that comes to my mid is how to solve route optimization in PMIP6. 
>There are others, Vidya mentioned one, and I'm sure other people will 
>find additional issues that needs to be looked into.
>
>For the SDO's I assume that they have to decide the applicability of 
>the
>PMIP6 protocol based on the current scope of PMIP6. If the SDO's would 
>like to use it for global mobility (which I agree is possible) they 
>have to consider potential trade offs when making the decision.
>
>/Conny
>
>
>Narayanan, Vidya wrote:
>>
>> Hi Raj,
>> The IETF did look at this work space and make a conscious decision 
>> that the applicability of a network-based mobility
>management protocol
>> is only appropriate for local mobility.  So, this charter has been 
>> scoped to what was believed to be appropriate.  The intent was never 
>> to have this protocol be applicable for all kinds of global
>mobility -
>> the fact that some systems may be considering this protocol
>for global
>> mobility is something they need to make a conscious decision on, 
>> keeping in mind that the IETF is only developing this for
>local mobility.
>>
>> I also don't believe it is a simple charter revision
>involving removal
>> of the term "local".  If we had to recommend this protocol
>for use in
>> the same scope as MIP6, as seems to be the discussion, we
>will need to
>> tackle how cross-domain security associations can be practical and 
>> other features that are feasible with MIP6 as the MN moves across 
>> various domains and heterogeneous technologies.  Extending the scope 
>> is going to require a number of other aspects to be flushed
>out and I
>> don't believe is appropriate at this time.  We need to
>deliver on the
>> chartered work before we can ask for more things and scope expansion.
>>
>> Regards,
>> Vidya

_______________________________________________
netlmm mailing list
netlmm@ietf.org
https://www1.ietf.org/mailman/listinfo/netlmm

_______________________________________________
netlmm mailing list
netlmm@ietf.org
https://www1.ietf.org/mailman/listinfo/netlmm