[arin-ppml] Draft Policy 2009-7: Open Access To IPv6 - Last Call

[Owen DeLong]

On Oct 29, 2009, at 11:32 AM, George, Wes E [NTK] wrote:

>
> -----Original Message-----
> From: Jim Weyand [mailto:jweyand at computerdata.com]
> Sent: Thursday, October 29, 2009 2:14 PM
> To: Chris Grundemann; George, Wes E [NTK]
> Cc: Owen DeLong; Member Services; arin-ppml at arin.net
> Subject: RE: [arin-ppml] Draft Policy 2009-7: Open Access To IPv6 -  
> Last Call
>
> So then you are proposing that just because an ISP is small today,  
> it should always and forever be stuck with the same upstream  
> provider in an IPv6 world?
>
> [weg] I'm not saying that at all. At its core, we're trying to  
> modify the definition of what an ISP is, at least as it relates to  
> whether they should get PI space. I'm simply saying that a small ISP  
> should either be a) multihomed or b) large enough to have a credible  
> plan to support 200+ customers over 5 years in order to justify PI  
> space. Which provider(s) you choose to use and how often you change  
> them is not really my concern. I'm not saying renumbering is not a  
> barrier, but by definition we're talking about a network with less  
> than 200 (or 100) end sites, meaning that it's less of an impact.
>
> That makes it tough to negotiate better terms with that provider.
> Or, even worse, if that provider pulls a Northpoint and pulls out of  
> region or goes out of business the ISP has an even bigger problem.
>
> [weg] Those both sound like a fantastic justification for not being  
> single-homed, not for getting PI space. You're still in a world of  
> hurt if your *only* ISP goes TU. Having PI space just means that you  
> have an address block to route when you find another provider.
>
That's a pretty major advantage when you have 7 days notice or less  
that you need to migrate your entire AS to some other connectivity.  
(which has historical precedent)

Owen