[ppml] Policy 2002-3,7 and 9 comment

[Ron da Silva]

Jim,

On Mon, Nov 11, 2002 at 06:10:24PM -0500, McBurnett, Jim wrote:
> Okay, 
> Having read through all of the notes from the meeting, and considering my current situation of "leasing" a class C for Multi-homing and being nearly stuck to one of  my ISPs due to the IP range being theirs, I have but a few questions:
> 
> 1. A new  Class C under these policies may not be globally routable if a single provider chooses not to advertise me. Correct?

Correct.  Actually, for that matter any allocation or assignment from ARIN
guarantees no amount of routability.  The routing of address space is negotiated
between two parties either by purchase of services, peering or other arrangement.
ARIN has no ability to ensure routability of address space.

> Should a new routing standard be examined where as a "verification" of source could be attached and authenticated to pass routes to "core" routers so that non-globally routable blocks do not become a rule instead of an exception?  IE. using IPSEC or CA etc. pass routes to a Route processor for a backbone provider to be able to propagate those routes to the net via summerizable routes?  And Yes I know this is not the correct place to mention this, but I remember seeing several concerns about the global routing issues that can arise from the micro-allocations these policies may cause.
> Comments? Where should this be mentioned? IETF? IANA? 

Good comment..and also currently in discussion in a variety of venues.  Your
best place to start this would be on the NANOG mailing list (though you should
search the mailing list archives online for previous discussions first).
Work in securing BGP is also emerging in the IETF.  Check some of the work
being done in the routing area.

-ron