Threads that never end (was: Waste will kill ipv6 too)
owen at delong.com
Tue Jan 2 21:59:46 UTC 2018
I agree we all have a responsibility to hold the line on addresses being network identifiers and to some extent network locators (unfortunately). I agree we have a responsibility to sparsely and liberally allocate within reason (where /8 to ITU isn’t within reason, but a /12 might be, and even if we have every country that wanted one a /16 to play with, that’s not likely to hurt much).
You’re absolutely right that if we get completely stupid beyond the bounds of current address planning we can waste IPv6 into oblivion. However, in a realistic discussion of whether it’s harmful or not to allocate /48s to residences and allocate /64s to point to point links, I don’t think there’s any valid argument that being stingy and putting /127s densely on point to point links with residences getting only a /60 each will somehow miraculously save us from any such actions.
Part of the reason threads like this don’t die is because people get so focused on their (poorly expressed) ideology that they often are agreeing without realizing it.
> On Jan 2, 2018, at 10:37, bzs at theworld.com wrote:
>> On January 1, 2018 at 22:09 trelane at trelane.net (Andrew Kirch) wrote:
>> Lets say the worst case scenario is that we exhaust IPv6 at a rate
>> MASSIVELY higher than planned. Can't we all just do this again in like 80
>> years? I don't get why anyone cares so much that this thread won't die.
>> Speaking of dying, I'll be dead by then anyway.
> One more time, the concern is not running out of ~2^128 addresses per se.
> The concern is running out of 128 bits due to segmentation and sparse
> allocations. A few bits for this (my unfounded example was handing the
> ITU a /8 for re-allocation as they see fit), a few bits for that, etc.
> Who was it who owned 2 x /8s of IPv4 space? AT&T? HP? Someone, I could
> look it up. What was the utilization of those blocks? And multicast,
> and 1914 space, and on and on.
> When one thinks of it like that, as chunks of the 128 bits, it doesn't
> look so vast, and it feels more vulnerable to politics, for example
> some nation demanding they act as their own RIR with a large
> allocation block, or just some clever new use, address blocks as
> cryptocurrency, address blocks with special, magical security
> policies, experimental uses, etc.
> Time, and howlings of pain should it come to that, will tell.
> At this point in time it's just dark speculation.
> -Barry Shein
> Software Tool & Die | bzs at TheWorld.com | http://www.TheWorld.com
> Purveyors to the Trade | Voice: +1 617-STD-WRLD | 800-THE-WRLD
> The World: Since 1989 | A Public Information Utility | *oo*
More information about the NANOG