[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: WPD-6, WAA-8, and WAA-9 of draft-ietf-v6ops-ipv6-cpe-router-04

Dumb question. In an IPv6 world, we don't actually have unnumbered links; we have link-local-addressed links. The link do in fact have addresses, but they are not externally addressable/ddos-able. If you want to manage a router, you use its loop-back address, which is perhaps a ULA and therefore unroutable from neighboring (customer, upstream, or peer) networks. 

Is there a problem with the link-local-address model if an ISP would rather use it? 

Or did I miss something?

On Apr 5, 2010, at 2:15 PM, Philip Homburg wrote:

> In your letter dated Mon, 5 Apr 2010 16:01:02 -0500 you wrote:
>>> I prefer to avoid unnumbered links.
>> 
>> Unnumbered links are a deployment scenario that is highly desirable from
>> some operators who wish to administratively separate the link between
> the SP and the CPE and the home...  They are also, in some cases, more
>> easily administrated from an addressing prespective than numbered links.
>> 
>> Your proposal would deny the SP the ability to administer their network
>> the way they want to.
> 
> That is a separate issue.
> 
> The first issue is that the DHCP PD RFC doesn't allow what I'm proposing,
> whether the ISP agrees or not.
> 
> The second issue, ISPs who like unnumbered links. I respect that. However,
> as I said in a previous mail. The difference between assigning an address
> to a virtual interface or to the WAN interface itself is almost completely
> invisible to the ISP. 
> 
> So if ISPs essentially cannot tell the difference, how can it matter to them?
> 
> 
> 

http://www.ipinc.net/IPv4.GIF