--On Wednesday, October 15, 2003 12:10:39 +0200 Iljitsch van Beijnum
<[EMAIL PROTECTED]> wrote:

> A very good property of IPv6 is that we get to avoid some of the mistakes
> that were made with IPv4. One of those mistakes was giving out addresses
> in ways that didn't scale. History teaches us that once something like
> this happens the cats don't want to be herded back into the bag, to abuse
> some metaphors here.

Agree. But: The main problem is that people and organisations in the
sunrise phase were given too small allocations, leading to disjunct
allocations and a mess in the routing table. If we give small organisations
a /32 they won't  need another prefix for the forseeable future. This
"wasteful" policy was not possible in v4, but now it is. So, I concur that
there are things to be learnt from v4. 

> There are now less than 35000 free AS numbers. If such a policy would be
> adopted there would be a huge land rush, depleting the AS number supply
> and forever polluting the IPv6 routing table with 64000 or so routes,
> most of which don't need to be there. The fact that this also uses up
> 0.0016 percent of the IPv6 address space isn't a significant issue, of
> course.

The fix for this is 32-bit AS numbers. Those 35000 ASen will suffice while
we look at multihoming problems and routing table growth. Providing we get
people to use v6 instead of talking about it here -- we can't know unless
we test this, and the experience gained is worth some of the risk. 

-- 
M�ns Nilsson            Systems Specialist
+46 70 681 7204         KTHNOC  MN1334-RIPE

We're sysadmins. To us, data is a protocol-overhead.

Attachment: pgp00000.pgp
Description: PGP signature

Reply via email to