>>>>> On Wed, 10 Mar 2004 16:09:49 +0100,
>>>>> Mattias Pettersson <[EMAIL PROTECTED]> said:
>> I think this is not broken at all. The host should select the correct
>> prefix according to the source address selection rules. I tried this
>> scenario approximately 3 years ago on a KAME stack and it was working
>> correctly.
> But what source address selection rules do you refer to? I saw Alper's
> response and I'm happy he pointed it out because I haven't been aware of
> it, but the wording there is really weak.
> Other than that I don't understand what rules in 3484 help here. I got
> rather familiar with the KAME stack during 1999-2001 and to my knowledge
> the default router selection is made once and the host sticks to that
> router as long as it is considered reachable. What source address being
> used doesn't matter to what next-hop is used.
Just to clarify a few things about a particular implementation,
KAME implemented most part of RFC3484 (in 2001), but does not support
the capability cited in this thread.
As far as I can see, the source address selection algorithm (based on
RFC3484) implemented in KAME can only solve some part of the problem.
For example, if longest matching against the destination address
suggests the correct source address, it will work fine. In general,
however, the implementation can choose an undesirable source address
that can be filtered in an ISP (or can cause other unpleasant
problems).
JINMEI, Tatuya
Communication Platform Lab.
Corporate R&D Center, Toshiba Corp.
[EMAIL PROTECTED]
--------------------------------------------------------------------
IETF IPv6 working group mailing list
[EMAIL PROTECTED]
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
