[6bone] proposal for transfer of 6bone address management
responsibilities to RIRs
Bob Fink
fink@es.net
Wed, 21 Aug 2002 10:03:23 -0700
Ronald,
At 06:21 PM 8/21/2002 +0200, Ronald van der Pol wrote:
>On Wed, Aug 21, 2002 at 08:23:27 -0700, Bob Fink wrote:
>
> > On keeping the 6bone separate from the production IPv6 Internet, I believe
> > that to be seldom if ever necessary, and that the decision to do so is a
> > local one for any network based on what is being done.
>
>No, it is not a local decision. That is the problem. When I choose to
>serarate them, I can. But my packets travel via several AS-es over
>production links until somewhere, far away, it's routed over an
>experimental service and gets blackholed. I have a problem when
>the policy in that place far away is: well, we are just experimenting
>with IPv6. This can only be solved when we all agree on a minimal
>quality.
I don't disagree that other's casual or irresponsible behavior can hurt
someone else far away, but it can only be their local decision to decide
what their behavior will be. The rest of us then can decide if we like that
behavior and try to filter, block, or whatever to avoid the problems caused
for the rest of us.
The IPv4 Internet has definitely had this problem and the collective
efforts of other operators has dealt with this, as well as the market place
deciding what to support or not. Unfortunately it isn't (and likely will
never be) a perfect process. Many of us have had to live with very
unacceptable behavior from supposedly very professionally operated large
ISPs. However, I stray. Your next comment is the important one here.
>I think we have a very big problem. Some people say: current IPv6
>transport is bad. Other people say: there is no problem.
I agree that we need to do more to police ourselves. Myself, I've come to
be in favor of revoking pTLAs from unresponsive and poorly managed pTLA
holders. However, we need a good process as well as strong 6bone community
support to do this.
I would very much like to hear proposals, as well as newer versions of
RFC2772, that address this. To state the obvious, the RIRs won't do this
for us (nor should they) and if we don't do it the life of the 6bone is
likely to be shortened considerably.
Thanks,
Bob