[arin-ppml] ARIN-prop-127: Shared Transition Space for IPv4 Address Extension
Jimmy Hess
mysidia at gmail.com
Fri Jan 21 00:45:56 EST 2011
On Thu, Jan 20, 2011 at 6:48 PM, Owen DeLong <owen at delong.com> wrote:
> The one amendment I would make to your suggestion, David would be that rather than specify legacy vs. Virgin, I would specify the dirtiest block(s) in the IPv4 free pool that total a /10.
>
> After all, it seems to me that there is no critical need for this space to be a single aggregate.
A single aggregate helps with bogon filtering and human recognition
of "shared IP ranges" for troubleshooting, sane design/addressing of
ISP network, and other purposes. Picking a hundred disjoint /24s -
/16s that add up to a /10 equivalent is not a very palatable idea.
Eliminating human recognition of well-known 'shared private' ranges
also eliminates one way humans can recognize human errors, such as IP
address typos, or obvious bogons such as a 'shared private range'
winding up in EGP announcements.
I would say a single contiguous /10 should be chosen so it overlaps
the portion of unallocated space with the highest density of "dirty"
IP addresses.
--
-JH
More information about the ARIN-PPML
mailing list