RE: [netlmm] MIP6 HoA and PMIPv6 pHoA

Hi Julien, Vidya,

while I can follow the "school of thoughts" taught by Julien and
understand why the separation of MIP6 HoA and PMIP6 pHoA results in
simpler PMIPv6 protocol, we need to keep in mind goal #4 (Limit Overhead
in the Network) from draft-ietf-netlmm-nohost-req. In this I-D it is
written:  "...Therefore, any solutions for localized mobility management
should minimize overhead within the access network."

As it was already mentioned on the mailing list, having a separation of
MIP6 HoA and PMIP6 pHoA leads to IP-in-IP tunnel over the wireless link
for each MIPv6 enabled MN.  Thus, IMHO goal #4 is not fulfilled. My
opinion on this is to keep both options, i.e. separation of MIP6 HoA and
PMIP6 pHoA and both addresses being the same, valid and leave it up to
the network operator to decide which option to deploy. If there is a
"home" PMIP domain for a MN, why not allowing for this MN to feel like
at home and avoid MIPv6 tunneling over the air. Otherwise this MN would
never be at "home". Of course if the MN moves to a "foreign" PMIP
domain, it configures a new PMIP6 pHoA (used only for local MM within
the "foreign" PMIP domain) and registers it as CoA in the MIP6 HA.

Of course there is an impact on the design of the HA if MIP6 HoA and
PMIP6 pHoA are the same in the "home" PMIP domain, but as Vidya has
mentioned ""No changes to MIP6 HA" is NOT a goal of this design".  

regards,
Genadi

> -----Original Message-----
> From: Narayanan, Vidya [mailto:vidyan@qualcomm.com] 
> Sent: Friday, February 02, 2007 8:45 PM
> To: Julien Laganier; netlmm@ietf.org
> Subject: RE: [netlmm] MIP6 HoA and PMIPv6 pHoA
> 
> Hi Julien,
> I completely subscribe to this school of thought :) 
> 
> We have, however, several people on the list who seem to want to allow
> the case of the MIP6 HoA being the same as the PMIP6 pHoA (also agree
> that term is misleading). I would like to understand why they believe
> that scenario is not problematic or why it is okay to impact the MIP6
> operation. 
> 
> Vidya 
> 
> > -----Original Message-----
> > From: julien laganier [mailto:julien.laganier@gmail.com] On 
> > Behalf Of Julien Laganier
> > Sent: Friday, February 02, 2007 6:39 AM
> > To: netlmm@ietf.org
> > Cc: Narayanan, Vidya
> > Subject: Re: [netlmm] MIP6 HoA and PMIPv6 pHoA
> > 
> > Folks,
> > 
> > In other words, what I tried to say below is that IMO:
> > 
> > - It doesn't make sense to have a MIP6 HoA which is
> >   PMIP6 pHoA.
> > 
> > - It makes sense to have a MIP6 CoA which is PMIP6
> >   pHoA.
> > 
> > Hence I think the term pHoA is misleading and should be 
> > change to NMoA (NetLMM Care-of Address)
> > 
> > --julien
> > 
> > On Friday 02 February 2007 15:31, Julien Laganier
> > wrote:
> > > 
> > > [...]
> > >
> > > I don't think it makes sense to have a MIP6 HoA which is 
> PMIP6 pHoA.
> > >
> > > As I see it, the NetLMM fabric should hide L3 movements 
> to *any* MN 
> > > moving within the NetLMM domain, MIPv6 or non-MIPv6 enabled.
> > >
> > > Therefore a MN that is in a NetLMM domain would get an 
> address from 
> > > that domain (I am proposing to call that a NetLMM Mobile 
> Address -- 
> > > NMoA). As long as the MN stays in the domain the NetLMM 
> fabric will 
> > > take care that this NMoA remains constant. When it leaves 
> > the domain 
> > > it will acquire a new address valid at its new point of 
> attachment.
> > >
> > > Now if the MN in question is MIPv6 enabled, it can register 
> > with its 
> > > Home Agent this NMoA as its CoA.
> > > As long as it remains in the NetLMM domain it won't need to 
> > change the 
> > > HoA->CoA binding since NetLMM will take care that this 
> CoA remains 
> > > constant. When it leaves the domain, then it will change the
> > > HoA->CoA binding to point to the address it acquired
> > > at its new point of attachment.
> > >
> > > Thoughts?
> > >
> > > --julien
> > 
> 
> _______________________________________________
> 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