IPv6 PI allocation

[David Conrad]

On May 17, 2007, at 6:22 AM, Sascha Lenz wrote:
> THERE IS NO ROUTING TABLE PROBLEM, FULL STOP.

True.  There is a _potential_ routing system _information_ problem.

> If you refer to RIPE Presentations, you might want to read
>
> http://www.ripe.net/ripe/meetings/ripe-54/presentations/ 
> Router_Scaling_Trends.pdf
> http://www.ripe.net/ripe/meetings/ripe-54/presentations/ 
> Moderate_BGP.pdf
>
> too, or watch the Presentation Webcasts for the routing-wg when  
> they are available in the RIPE54 Webcast Archive.

Odd how at least some routing folks at large scale ISPs (O(marketsize) 
=100) have been complaining that the router vendors are not able to  
meet their scaling demands (specifically, having hardware that they  
can deploy before having to upgrade again).  Perhaps those folks were  
mistaken and/or they're just not talking to the right router vendor?

> PA-Multihoming IS _NOT_ a solution, it is NOT "provider independant".

True.  Right now, the only answer we have is to flat route everyone  
on the Internet who wants to multi-home.  And who wants to be  
provider independent.  And who wants to traffic engineer longer  
prefixes to external peers.

Oh, and as IPv4 exhaustion approaches, I suspect people are going to  
find bits and pieces of IPv4 prefixes they aren't using and those  
prefixes will start magically appearing (perhaps announced by folks  
not originally associated with the original registrant).

Do we have a ROUTING TABLE PROBLEM now?  Nope.  Don't worry, be happy.

Rgds,
-drc