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

RE: draft-azinger-cidrv6-00



Hi Tony,

> -----Original Message-----
> From: owner-v6ops@ops.ietf.org [mailto:owner-v6ops@ops.ietf.org] On Behalf Of Tony Li
> Sent: Wednesday, July 21, 2010 2:12 PM
> To: Leo Vegoda
> Cc: IPv6 Operations
> Subject: Re: draft-azinger-cidrv6-00
> 
> 
> Hi Leo,
> 
> > Firstly, in section 4, I think it would be useful to note that RIRs should aim to aggregate an ISPs
> second and later allocation with earlier allocations. I believe this is already done but it would be
> good to recommend it. I'd like to suggest the second sentence below be added.
> >
> > "RIRs initial and subsequent allocation policy to service providers should allow for a minimum of 2
> years worth of usage based on historical or business plan projections. Further, RIRs should implement
> practices that maximise the potential for subsequent allocations to be made contiguous with previous
> allocations."
> >
> > Also, in section 10 it is recommended that "Internet Registries should severely limit or eliminate
> [...] PI assignments". There are obviously reasons for the current policies allowing IPv6 PI
> assignments but the potential for significant routing table growth has not been removed, so I wonder
> whether it would make sense to request the RIR communities to review the impact of PI assignment
> policies on the rate of routing table growth on a regular basis.
> 
> 
> The authors concur and unless there is some dissent from the list, we'll amend accordingly.

I think there may be some new developments you may want to
consider before casting allocation policies in stone. Fred
mentioned ILNP, but there is also the Internet Routing
Overlay Network (IRON - 'draft-templin-iron') which we
are currently progressing through the publication process.

IRON expects that there will be (at least) two distinct
ranges of IPv6 prefixes: 1) PA prefixes which are delegated
to ISPs, and 2) PI prefixes which are delegated to small
end user sites.

IRON expects that the (highly-aggregated) PA prefixes will
be represented in the global IPv6 routing system while the
(fine-grained) PI prefixes will be kept out of the routing
system. So, the routing system scaling can be kept to easily
manageable levels while billions of end user networks can
multi-home, traffic engineer, etc. using PI.

How is this done you may ask? Please check the IRON proposal
and let me know when and where you'd like to discuss it.

Thanks - Fred
fred.l.templin@boeing.com  

> Tony
>