[arin-ppml] An article of interest to the community....

Owen DeLong owen at delong.com
Fri Sep 2 18:45:42 EDT 2011


Cathy,

With utmost respect, I agree that IPv6 is finite. However, given any allocation policy presently in existence or even proposed as yet, we will not run out of it before the protocol hits scaling limits unrelated to address size.  Therefore, while we could run out by either getting vastly more creative with allocation policy or going to a new protocol that somehow removes those scaling limits while not changing the addressing scheme, I think that either of those events is so distant in the future as to be meaningfully never.

Owen


Sent from my iPad

On Sep 2, 2011, at 12:06, "cja at daydream.com" <packetgrrl at gmail.com> wrote:

> Owen I get that IPv6 is very large but no matter how big it is it is still a finite resource.  I get that we're not in the scarcity mode that we are with IPv4 but there was a time when folks in this community argued that IPv4 was so large we would never run out.  We did run out and at some point we will run out of IPv6 too.  Just because it's so large that a lot of folks feel they'll be dead when it runs out doesn't mean we still shouldn't consider our actions and realize that IPv6 is indeed a finite resource.
> 
> ----Cathy
> 
> On Fri, Sep 2, 2011 at 2:55 AM, Owen DeLong <owen at delong.com> wrote:
> 
> On Sep 2, 2011, at 12:28 AM, Matthew Kaufman wrote:
> 
>> On 9/2/11 6:31 AM, Michael Wallace wrote:
>>> 
>>> I'm new over here.  Been reading for a couple of weeks.
>> 
>> Welcome.
>> 
>>> 
>>> The only issue I see with IP addresses is people not even making an attempt to use IPv6.  We have so many IP addresses in IPv6 that we should never run out. 
>> 
>> That's a myth.
>> 
> Actually, no, it isn't.
> 
>> The myth is that there's more IPv6 addresses than there are grains of sand on every beach in the world. And while that's true, you'll find that even if you put 100 hosts on every IPv6 subnet (which is entirely unlikely) and use all of the space that isn't otherwise reserved, you only have as many hosts supported as there are grains of sand on a single small beach. Which isn't nearly enough to "never run out".
>> 
> 
> The myth is... and while that's true...
> 
> Pick one... Either it's true, or, it's a myth.
> 
> I find it very interesting that you think there are 1,800 quintillion+  beaches. I do not believe that is possibly true.
> 
> There are 18 quintillion+ subnets available at /64, so, 100 hosts on each would be 1,800 quintillion+, so, your single beach claim is actually a myth.
> 
> From a quick scan of http://en.wikipedia.org/wiki/List_of_beaches it appears that there are less than 1,000 beaches listed. Assuming they have
> missed 9 out of 10 beaches (which I think unlikely), you're still at less than 10,000 beaches.
> 
> If we put 100 /64 subnets on each beach, we'd use 1,000,000 subnets or roughly a /44 of address IPv6 address space.
> 
> All the beaches on earth (and then some) with 100 IP addresses per grain of sand (I'm taking your and other people's
> words for the idea that there are approximately 18,000,000,000,000,000,000 grains of sand per beach) would actually
> only require a /44.
> 
> There are 17,592,000,000,000+  more /44s available in IPv6 even after this silly exercise.
> 
> I doubt we will "never" run out. 'Tis true. However, I do suspect that we will not run out within the expected lifetime of the protocol and that address size will not be the scaling limit that causes the eventual demise of IPv6.
> 
>>> 
>>>   I tried to make a pretty big push to use IPv6 but failed doing so.  Only because I'm getting push back from my upstream providers (Ill leave them unnamed). 
>> 
>> Don't leave them unnamed. The best way to "push to use IPv6" is to name who isn't making it possible.
>> 
> 
> This, I completely agree with.
> 
> Owen
> 
> 
> _______________________________________________
> PPML
> You are receiving this message because you are subscribed to
> the ARIN Public Policy Mailing List (ARIN-PPML at arin.net).
> Unsubscribe or manage your mailing list subscription at:
> http://lists.arin.net/mailman/listinfo/arin-ppml
> Please contact info at arin.net if you experience any issues.
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.arin.net/pipermail/arin-ppml/attachments/20110902/4cb148fe/attachment-0001.html>


More information about the ARIN-PPML mailing list