Re: [narten@us.ibm.com: PI addressing in IPv6 advances in ARIN]

["Per Heldal" <heldal@eml.cc>]

On Fri, 14 Apr 2006 11:22:58 +0200, "Iljitsch van Beijnum"
<iljitsch@muada.com> said:
> On 14-apr-2006, at 10:48, Per Heldal wrote:
> 
> >> So I suggest that the IESG either:
> 
> >> 1. Tells ARIN that this policy is incompatible with work inside the
> >> IETF so it shouldn't be adopted, or
> 
> >> 2. Conclude shim6.
> 
> > That is awfully narrow-minded. The new policy indicate that V6- 
> > adoption
> > is growing,
> 
> Well, if you start at zero growing isn't that hard.
> 
> The truth of the matter is that IPv6 use is still 0% if you round to  
> a whole percentage point, and I've seen no indication that whatever  
> is needed, is needed TODAY. Shim6 is moving along well, there was no  
> reason to take action at this point in time.

Neither does anything indicate that usage (in terms of available space)
will grow to 100% anytime soon. Maybe shim6 is a viable solution at 5%
usage. Maybe it or something else yet to be invented takes over at 10%,
to the extent that the PI blocks are abandoned.

> 
> > and that it is time for operational policies to reflect
> > real-life.
> 
> If you mean that in real life, people don't care about the long term  
> they just want to be able to switch ISPs without having to renumber,  
> well, yes, this new policy reflects this. Congratulations on that.

In real life there are requirements and possible solutions. Vapor-ware
and other hallusinations don't qualify as solutions.

> 
> > Multihoming with BGP works fine today
> 
> If it works so fine, how come that only a few thousand people world  
> wide are doing it?

Because it's complex, has administrative overhead, policy restrictions
... pick whatever reason.

> 
> > How can you expect operational policies to be based on technology that
> > doesn't yet exist (feed the starving from next years crops)? From a
> > non-technical perspective, can you somehow
> > eliminate redundant connectivity and provider independence from  
> > people's
> > list of business-requirements?
> 
> You are arguing the question whether we "need" something. I'm trying  
> to argue whether we can afford to give it out. The latter trumps the  
> former.
> 
> > OTOH, this is in no way the end of shim6. The new policy is,
> > as you say, restricted to a limited number of sites.
> 
> It explicitly talks about removing some of these limitations in the  
> future. The point is that the game is now trying to get PI by lying  
> to the RIR or pushing for more relaxed rules rather than deploy shim6.

That's already an old game. A large number (maybe the majority) of
current allocations are made to entities which don't (and never will)
meet the requirements of being a transit provider.

> 
> > An alternative policy would be to make these PI
> > assignments temporary.
> 
> An alternative would be to give the prefixes out based on geography,  
> so there is at least a chance that we get to aggregate them  
> geographically in the future. But despite the fact that doing this is  
> entirely risk-free (geo PI without geo aggregation is no worse than  
> regular PI) people want proof that it works first, while at the same  
> for deploying regular PI the burden of proof is reversed so it's  
> allowed now despite very good indications (but not hard proof) that  
> it will be problematic in the future.

Geo-aggregation doesn't work in a competitive market. It would require
new routing-technology to do so, and thus fall in the same category as
shim6. 


> 
> > When alternative technology is available,
> > PI-sites could be asked to move to alternative solutions within X- 
> > years
> > or so.
> 
> Oh, and those same people that can't even be bothered to aggregate  
> two /21s into a /20 will now give up their portable address block?  
> Yeah right.

The challenge for the technical community is to create better
technology. Turn shim6 into something people want/need. Then you won't
need to force them to use it. Such a decision, made within the RIR
democracy, is easy to enforce should it be necessary. Our job as
researchers/developers/administratiors is to make sure that doesn't
become necessary.

> 
> > I do however not see how PI-assignments can be avoided
> > altogether as there is no viable technical alternative.
> 
> IPv6 is really just repeating the same mistake with longer addresses.

I didn't say it's a perfect solution, but we're not living in a perfect
world. If we're unable to come up with a product the market wants we
have to accept that the market makes the most of what it's got
regardless of what we may think about it.


//per

PS! I belive the probability of something entirely new replacing
IPv<anything> is much greater than the probability of the internet
falling apart due to its growth.
-- 
  Per Heldal
  http://heldal.eml.cc/