Dear Alain, all,

Please see inline.

Cheers,
Med
 

-----Message d'origine-----
De : Alain Durand [mailto:adur...@juniper.net] 
Envoyé : vendredi 7 mai 2010 18:19
À : Behcet Sarikaya
Cc : BOUCADAIR Mohamed NCPI/NAD/TIP; softwires@ietf.org
Objet : Re: [Softwires] GI-DS-lite as working group item?


On May 7, 2010, at 10:55 AM, Behcet Sarikaya wrote:

> 
>> (c) N:1 
>> model: a single CGN serve a group of PGW/GGSN. Indeed, having +16M of 
>> customers 
>> is a valid case. **BUT** which Service Provider will accept to service this 
>> huge 
>> amount of UEs with the same node (if we suppose that a mega centralised CGN 
>> implementation is found in the market)? This is single point of failure 
>> design 
>> which SHOULD NOT BE RECOMMENDED.
> 
> IMHO it is N:1 and that's why GI-DS lite is interesting to the operators. I 
> agree with you that having 16M+ MNs NAT'ed by a single CGN is something I 
> have not heard of, yes there there is a problem.


co-chair hat off:

This is the N:P aspect that I find interesting in GI-DS-lite. And it does not 
have to be limited to 16M+ address being concentrated... nor be limited to 
wireless.

Med: I have several questions here:

1- What do you mean by N:P? (In my first e-mail N refers to the number of gws 
and 1 is number of CGNs). 

2- If applied to fixed networks,  

(*) this would lead to NAT444 design model since nothing prevent to embed a NAT 
function in the AD.
(*) IPv6 is not a MANDATORY building block of the solution.   
(*) What is then the recommendation of the WG: use DS-Lite? Or GI-DS-Lite? Etc.
(*) A last comment, since GI-DS-Lite is a request from the 3GPP which 
recommends DS and IPv6-only models, this solution SHOULD be part of that 
package.

Think about an ISP with a number of access router and a number of centralized 
NATs. Each customer connected to an access router is 

Med: Even if centralised, having several millions of customers behind the same 
"stateful" device is not a viable approach: this node needs to support logging, 
LI, NAT, ALG, etc.    

using his own version of RFC1918.

Med: Yes, I agree with this point.

What GI-DS-lite enable is to implement the NAT function in a different box that 
potentially has different scaling properties than the access router.

Med: Yes but GI-DS-Lite is not the ONLY solution which allows this.  You may 
rely on simple routing configuration from the gateway to its associated NAT(s), 
use DS-Lite ;-), or use A+P to avoid having any stateful device in the core 
network, etc. 


*********************************
This message and any attachments (the "message") are confidential and intended 
solely for the addressees. 
Any unauthorised use or dissemination is prohibited.
Messages are susceptible to alteration. 
France Telecom Group shall not be liable for the message if altered, changed or 
falsified.
If you are not the intended addressee of this message, please cancel it 
immediately and inform the sender.
********************************

_______________________________________________
Softwires mailing list
Softwires@ietf.org
https://www.ietf.org/mailman/listinfo/softwires

Reply via email to