Re: (resend) [rfc2462bis] prefix length check for existing addresses

[Erik Nordmark <Erik.Nordmark@xxxxxxx>]

> I have one last question.  RFC2462 says just "an address already in
> the list" in the case of no identical (or matched) prefix while it
> says "an *autoconfigured* address in the list" in the case where there
> is an identical prefix.  Does the former mean all addresses including
> stateless-, stateful-, or manually- configured ones?  If so, is this
> intentional?  And if so, we'll still see the problem that DHCPv6 does
> not carry the information of prefix for assigned addresses...

I don't think there was an intent when 2462 was written that
it affect the lifetime for DHCP assigned addresses.
I don't think anybody knew at the time how DHCPv6 would manage lifetimes.
And I'm sure the intent was not to affect some lifetime on manually configured
addresses.

Given this I think it makes sense to explicitly state in RFC 2462 that "the
list" is the (per interface) list of stateless prefixes learned from
router advertisements (i.e. the prefixes with the A-bit set).

  Erik


--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------