Re: 9/9/2004 IP6.INT Removal (Was: 9/9/2006 : ip6.int shutdown?)

[bmanning@vacation.karoshi.com]

On Wed, Jul 28, 2004 at 05:53:52PM +0000, Paul Vixie wrote:
> > > Considering the DNAME-baed renaming mechanism documented at
> > > <http://www.isc.org/pubs/tn/isc-tn-2002-1.txt>, what I wonder
> > > is, why doesn't Bill Manning just put in a DNAME at IP6.INT
> > > renaming the whole tree to point at IP6.ARPA?
> > 
> > 	in part, because there are entries in ip6.int that have no
> > 	counterpart in ip6.arpa.... otherwise, a really good idea.
> 
> i think that you should spend your efforts trying to get the .INT
> content mirrored in .ARPA somehow.  such as contacting the folks
> who own the delegations you have in .INT and teaching them how to
> re-register their content in .ARPA, such as by going through their
> RIR's.

	not all entries in the .int tree are RIR related. the most
	obvious is 3ffe::/16. It does not help that the IETF has a
	BCP/PS RFC that prohibits any entry in ip6.arpa that is not
	an RIR managed block.

> >       i have been trying to get the good folks at ICANN to effect
> >	such changes as can be done, done.
> 
> icann's not involved at all.  once you get the content re-registered
> in .ARPA, you can put a DNAME at the top of your .INT zone and the
> rest is history.

	er, yes ICANN is involved. And I understand the process here.

> > 	and i do not belive that older implementations will be
> > 	out there forever.  but as long as the number of queries
> > 	for ip6.int is noticable, i plan on serving the data.
> 
> what's the standard for "noticable"?  1qps?  10qps?  100qps?  what
> are you seeing, in qps, today?

	the standard is likely different for different people.
	we did have this same discussion - about 10 years ago
	when there was a proposal to shut down in-addr.arpa and
	move the v4 reverse anchor to ip4.int... :)

	as to qps, not so many, but that may be due to the fact that
	once sites get the RIR, 6bone, and other delegations cached,
	they don't come back too often.

--bill