Re: [GROW] LISP at GROW

[Scott Brim <swb at employees.org>]

On 8/5/08 11:23 PM, Tony Li allegedly wrote:
>  
>
> |> | > If a site wants to change its
> |> |> EIDs, it's forced to renumber.  
> |> |
> |> |That's a tautology.
> |> 
> |> So the implication is that LISP doesn't really protect the 
> |end user site
> |> from having to renumber.
> |
> |If a site wants to change its EIDs it renumbers, by definition.  Under 
> |what conditions do you believe that "wants to" and "has to" are the 
> |same?  I don't see any way you can get from the first statement to the 
> |second.
>
>
> Sorry, I wasn't being that precise.
>
> My understanding is that for a site to have its EIDs aggregate, it must be
> part of the virtual topology associated with a given 'mapping provider'.
> And if they want to change mapping providers, they would need to change EIDs
> and renumber their site.

Aha, the 'mapping provider' shibboleth.  I thought we dealt with that 
quite a while ago.  I wouldn't make any of those assumptions.  There are 
no restrictions on the business model for how ALT will be operated.  I 
cannot predict.  One would assume that the higher levels will be run by 
RIRs or something like them, but the lower levels are wide open and I 
expect some kind of small business competition there.
_______________________________________________
GROW mailing list
GROW at ietf.org
https://www.ietf.org/mailman/listinfo/grow