[6bone] RFC2772 rewrite -- bigger scope goals
Pekka Savola
pekkas@netcore.fi
Fri, 15 Nov 2002 09:03:42 +0200 (EET)
On 14 Nov 2002, Robert Kiessling wrote:
> Pekka Savola <pekkas@netcore.fi> writes:
>
> > 3) kill 6bone
>
> Surely this will result in much flaming. Maybe it would help to first
> think about the concrete effects this would have.
>
> 1. Which current services are provided in 3FFE space which cannot
> feasibly move to RIR space?
Services, e.g. addressing, tunnels, free transit etc., provided for
non-customers for free.
Some do not want to move these to their RIR space -- I believe that's
okay.
> 2. Apart from services provided to the public, which other effects
> would it have to put a timeout on current allocations?
>
> 3. Which future developments are hindered if 6bone allocations were
> not available?
The RIR policy has gotten much more flexible lately, so 6bone is no longer
needed that much even for experimenting.
Basically, it seems to me that new 6bone allocations seem to be non-LIR's,
ie. not allowed to have RIR space anyway (anyone can think of
exceptions??) -- we don't really want to give these a pTLA.
Perhaps also some smaller ISP's which may be LIR's (but some are not) but
do not fulfill the letter of 200 /48 allocations in the RIR space.
So it seems to me that (all that many) further allocations shouldn't be
even necessary, the question would mostly be on how to manage the current
ones.
> What are 6bones goals today?
>
> Do we still need a "testbed to assist in the evolution and deployment
> of IPv6?"
I'm not sure if this is necessary.
> I would argue that "deployment" in the sense of "provide address space
> to use IPv6" should no longer be considered part of 6bone's mission.
> RIRs now provide this part. Practically 6bone is in direct competition
> with RIRs here, providing a different set of criteria, essentially:
>
> - RIR: LIR fees & 200 potential customers
> - 6bone: wait 6 months
>
> So what is the concrete benefit of having this cost-free alternative
> to RIR allocations?
6bone address pTLA allocations, for new allocs, seem to be useful mostly
to those that should not get it, and those which are small enough not to
be able to get it through RIR's. (A relative fringe case are those who do
have RIR space but want to offer services like tunnels to outsiders also.)
But I believe the 6bone community _could_ _possibly_ be served by these
slightly smaller ISP's getting 6bone space but not RIR space. The
question is how to wordsmith the rules so that too small ones can't get it
(e.g. one requirement could be AS number of your _own_, not loaned).
--
Pekka Savola "Tell me of difficulties surmounted,
Netcore Oy not those you stumble over and fall"
Systems. Networks. Security. -- Robert Jordan: A Crown of Swords