[arin-ppml] ARIN-prop-127: Shared Transition Space for IPv4 Address Extension
owen at delong.com
Sat Jan 22 15:02:39 EST 2011
On Jan 22, 2011, at 10:29 AM, Jack Bates wrote:
> On 1/22/2011 12:07 PM, Chris Grundemann wrote:
>> Thanks for clarifying Jack, I think I now understand your suggestion.
>> Basically add a sentence or two stating that internal LSN addressing
>> is not a valid justification of need for new IPv4 requests. I believe
>> that it's worth considering including that - if it were included,
>> would you support the proposal?
> I'd disagree with it still (as I'd much prefer to have that /10 open for other uses such as content providers), but I WOULD support it for two reasons:
It's not like content providers would be unable to use this /10, but most content providers don't need that much non-routable space.
> 1) It would at least be better than losing a /10 plus the other space LSN would cost ARIN (not concerned with other regions use of the /10 or policies, only ours); ie, it already has decent support, and I'd rather a better policy go into effect than what this currently is.
> 2) It makes it a much grittier policy. It says, take it as a whole or we don't implement it at all. This is not freebie space to be abused.
I support the idea of amending the proposal to prohibit ARIN from granting requests for other space for this aspect of LSN implementation.
> This policy can't have a form which reduces my concerns of the eyeball networks regaining large chunks of space while the content providers will continue to dwindle. Unfortunately, it's too late in the game to fix policies to favor protecting content networks (ie, eyeballs don't have to utilize NAT444 if they don't want to and can request address space until we are out, at which time they can convert, but the content providers do not have any new tools to deal with migration on their side). Trying to implement policy to deal with this unbalanced set of tools at this point would only cause a fast rush to ARIN by eyeball networks prior to policy ratification and defeat the purpose. As such, there will be a time that content providers cannot offer IPv4, and their competitors (especially eyeball networks who sideline content) will have IPv4. We will have effectively killed the little guy.
I'm not sure that dwindling addresses for CPs is a bad thing. The needs of CPs should dwindle as end users gain more access to IPv6.
CPs, as you have pointed out, could be the hardest group to motivate to IPv6.
More information about the ARIN-PPML