Re: [netlmm] Is this scenario covered in netlmm-mip-interactions I-D?

[Sri Gundavelli <sgundave at cisco.com>]

Hi Raj,

Collocation of the LMA and HA with physical home link will be
difficult. In PMIPv6 domain, the prefix is a delegated prefix
and is never hosted on any of the physical links of the LMA/HA.
Secondly, there is no HoA that the LMA defends, there is no
Proxy ND defence by the LMA for the MN's HoA and infact the
LMA will not even know the HoA of the MN. Given this, allowing
the HA to host the MN's prefix on a physical link wont quite
fit. So, I'd think in such LMA/HA collocation scenario's, the
HA should not support physical links, it should be the PMIP/MAG
that should be ob physical links and let HA be for global
mobility, with PMIP supporting local mobility.

Sri

On Wed, 21 Oct 2009, Basavaraj.Patil at nokia.com wrote:

> Just a quick question about the following scenario:
>
> Lets say a (DS)MIP6 capable MN is attached via a PMIP6 domain thru a MAG. The MN is assigned a prefix from the LMA/HA.
> To the MN it appears as if it is on its home link.
> The MN moves and attaches to its physical home link.
> The MN does not see any difference in terms of the home link (virtual or physical) and hence does not initiate any deregistration procedures.
> However the MN is unable to use the HoA/HNP until the binding in the LMA is deprecated.
> The MAG may eventually discover that the MN is no longer attached and send a dereg. However that detection may be delayed significantly.
>
> Is the above scenario and issue discussed in the netlmm-mip interactions I-D? At least I could not notice it.
>
> -Raj