Re: Review on draft-ietf-v6ops-v6-in-mobile-networks-03

[Rajeev Koodli <[e-mail]>]

Hi Tina,


On 3/3/11 1:25 PM, "Tina TSOU" <[e-mail]> wrote:

>>>> do you mean using some solution like address + port
>> (http://tools.ietf.org/html/draft-ymbk-aplusp-09)?
>>      If yes, where do you think Address + Port solution should be
>> implemented? Maybe one of the node could be GGSN/PDN gateway, how about
> the
>> other node, is it the mobile node or the ANG?
> 
> 
> My personal opinion is that you would have to modify the MN.
> 
> It is not the intent of the ID to suggest location of A+P elements, as you
> probably realize.
> 
> [Tina: 
> Some clarification of A+P used in mobile network is needed. A+P can be used
> on CPE in wireline network. There is no very proper node to support in
> wireless network. Implementing A + P in cell phone may need change of
> Operating System, even though not as much as PNAT. In addition, there maybe
> some ARP issue, for example, cell phones sharing one address can not
> communicate with each other, as the peer address is it self's, unless we are
> going to change ARP.]
> 

I am sorry. This draft is not about describing how A+P protocol or _any_
other protocol is supposed to work in a mobile network. It is about general
considerations that come up when deploying IPv6 in mobile networks.


>>>> now, and in the near future, most of the content would still be
> available
>> in IPv4 network, not sure whether this solution could save private IPv4
>> address.
> 
> Sure. However, by deferring the allocation of an IPv4 address until a PDN
> and/or an application actually needs it, you also conserve the pool.
> [Tina: 
> it could work, which won't save many private IP address. It will also bring
> the complexity of implementation.]
> 

The issues related to host implications are noted in the paragraph.
For instance, see:

" In any case, there need to be appropriate triggers to initiate DHCP
   based on the application and interface usage, as well as DHCP lease
   management based on appropriate address management policies.  These
   considerations may limit the applicability of the address deferring
   option."


Thanks,

-Rajeev
 


>>     And this would also require some modification in the mobile devices.
> 
> Right. The operators who are interested in this already have the necessary
> "bindings" - when a user clicks on an app (that requires IPv4 PDN), it
> invokes the necessary PDN signaling if the PDN is not active already.
> 
> Regards,
> 
> -Rajeev
> 
> 
>> 
>> 
>> We keep our promises with one another - no matter what!
>> 
>> Best Regards,
>> Tina TSOU
>> http://tinatsou.weebly.com/contact.html
>> 
>> 
>> 
> 
>