[arin-ppml] Global Uniqueness vs Global Reachability

David Farmer farmer at umn.edu
Thu Jun 4 10:21:46 EDT 2009


On 4 Jun 2009 Paul Vixie wrote:

> scottleibrand at gmail.com (Scott Leibrand) writes:
> 
> > There was a more recent (and IMO better) version of this called ULA
> > Global. IIRC, it was a lot closer to what you are describing than
> > ULA Cental was.
> >
> > -Scott
> 
> Noting that the URL changed to
> <http://nsa.vix.com/~vixie/ula-global.txt> and that my purported
> co-authors do not actually support this proposal.
> 
> In case it seems odd that IPv6 would need anything scoped in this way,
> the purpose is to make it easy to keep this crud out of the DFZ while
> still permitting metro and neighborhood level experiments in "DFZ
> bypass." -- Paul Vixie KI6YSY

Well I have to say I wasn't realy aware of this draft.  While I 
think this is much better than ULA-Central, it still claims to be 
something other than regular Unicast address space with a 
easily defined routing policy scoping it out of the mythical 
global route table

One example from the draft;

"5.5 IANA and all RIR/LIRs are encouraged to avoid allocating 
aligned blocks of address space under this specification, in 
order to specifically discourage aggregation and wide area 
routing of such address space."

I think I have to go with Randy on this one, if someone 
wants/needs unicast IPv6 address space they should be able 
to get it unless we have a really good reason not to give it to 
them.  

All I'm suggesting is that we can provide it in a way that easily 
allows a coordinated routing policy regarding assignments 
outside an aggregated routing hierarchy.

If people are wary about trying something like this, since we 
are working within ARIN's policy process and within ARIN's 
registry services contracts, we could back stop this whole thing 
by allowing ARIN to withdraw all assignments of this type with 
a 5 year notice or something like that.  If people are worried 
about it turning into a mess, we can try to provide a way 
mitigate that risk.  This is something well with in ARIN's and the 
other RIRs perveiw to manage IP address assignments and 
allocations, and would be much harder to for the IETF to make 
happen.

Today aggregated hierarchical routing is the best way to make 
the Internet work.  However, if that hammer remains the only 
tool in the Internet tool box, then everything will look like nails.


===============================================
David Farmer                                      Email:farmer at umn.edu
Office of Information Technology
Networking & Telecomunication Services
University of Minnesota		       Phone: 612-626-0815
2218 University Ave SE		       Cell: 612-812-9952
Minneapolis, MN 55414-3029	       FAX: 612-626-1818
===============================================




More information about the ARIN-PPML mailing list