Re: IPv6 Wow

[Mark Andrews <Mark_Andrews () isc org>]

In message <C52676ED.1C53A%alain_durand () cable comcast com>, Alain Durand writes
:
> On 10/23/08 6:39 PM, "Tony Hain" <alh-ietf () tndh net> wrote:
>
> >  A properly
> > implemented client will do the longest prefix match against that set, so a
> > 6to4 client will go directly to the content provider's 6to4 router, while a
> > native client will take the direct path.
>
> Not quite.
> Say the server has native IPv6 address 2001::1 and 6to4 IPv6 2002::X.
> Say the client has native IPv6 address 2003::1 and 6to4 IPv6 2002::Y.
> Longest prefix match will choose 6to4 over native IPv6. Not good.
>
>   - Alain.

        Longest match to select destination address without knowlege of
        the prefix lengths involved is bogus.

        Applying a /32, /48 and /64 prefix break points to addresses
        in 2001::/16 and 2003::/16 and a /16, /48 and /64 to addresses
        in 2002::/16 will produce reasonable but not perfect results.

        That's ISP, SITE and LINK level prefix break points.  6to4
        can be seen as one ISP with a /16.  Note you only need to
        configure the break points for the address space you are
        using.

        We need automate the dissemination of these values within a
        ISP to the customers so they can correctly configure their
        address selection rules.

        Mark
-- 
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: Mark_Andrews () isc org