[rrg] BGP scaling limit?

[Brian E Carpenter]

On 2009-01-05 15:02, William Herrin wrote:
...
> When we asked the RIRs obstruct PI prefixes for multihomed IPv4 and
> IPv6 users, we upgraded them from cap guns to bullets.

That was in early 1994, I believe. If you were looking at the DFZ
growth curve back then, the bullets seemed inadequate, but they worked
for a while.

> It's a very
> *harmful* practice which should be discouraged absent testable and
> quantitatively substantiated growth predictions projecting routing
> table growth which outstrips our ability to build hardware. The IPv6
> side is especially harmful, given the looming free pool crisis whose
> projections are very well grounded. 

As I've always understood things, the challenge is to find the best
possible distribution of the harm; if the RIRs hadn't succeeded in
damping prefix growth, who knows when things would have ground to a
painful halt? My conclusion is exactly the opposite of yours: absent
substantiated growth predictions, the priority must continue to be
precautionary damping of the growth, as a matter of public importance.

> As a group, we should withdraw
> that recommendation and discourage obstructing PI space for
> multihomers.

Didn't you forget to add "as soon as there is a proven way to
ensure that the resulting growth in prefixes can be handled by the
routing system"?

(Not that the RRG can withdraw CIDR, anyway.)

    Brian