Laurent,

6to4 was not designed for "sites" containing tens of millions of hosts. With those
numbers you surely need native addressing. I was addressing Juan-Antonio's question
about 6to4.

Regards
   Brian

[EMAIL PROTECTED] wrote:
> 
> Considering that
>      - a mobile operator may have up to 15-20 M subscribers (e.g. in Medium size 
>country such as European countries) ... or.. 35M subscribers or more (considering 
>bigger countries as...China)
>      - each user attached to IMS (3gpp sip services) has got an IPV6 @
>      - although Not 100% of the users have simultaneously switched on their mobile, 
>we have to take some busy hour worst case
> you may end-up with 20-30M or more IPV6 @ per on the APN used by mobiles to reach 
>IMS services of the operator
> 
> As
>      - the number of GGSN has to be limited to something reasonable (that obviously 
>is related to technology capability)
>      - we have to consider an addressing scheme that copes with foreseenable 
>technological improvements on the GGSN capabilities
> a rough estimate of the number of IPV6 @ a GGSN should handle on this APN should be 
>between 500K and 1M.
>      Best regards
>         Laurent T.
> ---------------------------------------------------------------------------
>       V    Laurent Thiebaut      tel: +33 (0)1 3077 0645
> A L C A T E L                    e.mail:[EMAIL PROTECTED]
> ---------------------------------------------------------------------------
> 
> Brian E Carpenter <[EMAIL PROTECTED]> on 10/12/2001 19:47:17
> 
> 
> 
>  To:      Juan-Antonio Ibanez
>           <[EMAIL PROTECTED]>
> 
>  cc:      [EMAIL PROTECTED](bcc: Laurent
>           THIEBAUT/FR/ALCATEL)
> 
> 
> 
>  Subject: Re: Feedback from 3GPP about draft-wasserman
> 
> 
> Juan-Antonio Ibanez wrote:
> ...
> > 2) Would the /64 prefix allocation preclude the use of the 6to4
> > transition mechanism in 3GPP?
> > 6to4 relies on a /48 prefix, meaning that only 64k mobile nodes (or
> > rather "primary" PDP contexts) would be allowed on a given APN that
> > supports 6to4.
> 
> Are you assuming that the APN has exactly one globally routable IPv4
> address? If so, it can only own one 6to4 /48 prefix, i.e. 64k /64s
> as you say. But if the APN has 2 IPv4 addresses, it can have two
> 6to4 prefixes, etc.
> 
> What is the realistic requirement for the number of primary PDP
> contexts per APN?
> 
>    Brian
--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------

Reply via email to