Re: (ngtrans) Joint DNSEXT & NGTRANS summary

[Jun-ichiro itojun Hagino <itojun@iijlab.net>]

>  | 	you should care about this.
>Why?
>  |	there's no guarantee that AAAA synthesis
>  | 	happen between end client and master/slave nameservers.  AAAA will
>  | 	leak from leaf to the core.
>Yes.   That doesn't tell me why I need to care.   People send queries
>for all kinds of RR types that I don't have in my zone files, my
>servers return NO DATA responses (no error, no answer).  Perfectly
>normal.   That's what is likely to happen if they request an AAAA and
>all I have is A6.   (Just maybe I'll do AAAA synthesis for random people
>in the very early days).
>
>Exactly why is this a problem to care about?

	we (pick any community i'm involved with) are already operating AAAA
	in a lot of places.  we need to keep those configuration to work fine
	while we migrate to A6 + AAAA synthesis.  we have serious users of
	IPv6 using AAAA, and we cannot break their configuration, nor
	transition them into some other record type in one night.  you have
	been proposing a migration path with a hard "flag day", with:
	- all existing zone files must be rewritten to from AAAA to A6 in one
	  night (of which timezone?)
	- all AAAA queriers can stay as is, but there has to be AAAA synthesis
	  server at every leaf
	- no AAAA traffic in core DNS cloud, only A6 traffic.
	this is totally unacceptable to us.

	the story applies only if we are going to move to A6 + AAAA synthesis.
	if we don't migrate, we have no worry and everyone is happy.

itojun@tired.