renumbering & IPv6

Wed Sep 13 21:08:38 UTC 2006

On Sep 13, 2006, at 1:27 PM, David Barak wrote:

> Perhaps a customer who wanted to make IP addresses
> "portable" would pay a fee to the ISP whose addresses
> they are, and maintain redirection equipment to the
> "real" IPs...  And perhaps the price of doing so would
> actually be higher than just keeping a T1 to that
> first provider...

from http://www.ietf.org/rfc/rfc4192.txt

-----

    Some took it on themselves to convince the authors that the concept
    of network renumbering as a normal or frequent procedure is daft.
    Their comments, if they result in improved address management
    practices in networks, may be the best contribution this note has to
    offer.

-----

Without PI space for customers, both renumbering and traffic  
engineering/redundancy for the enterprise customer become a) horribly  
complex and b) subject to the whims of business relationships.   
Neither of these conditions is tolerable for those customers; turning  
every host on the network into a router via a Shim-6-like mechanism  
isn't, either (can you imagine help-desks who can barely cope with  
basic Windows issues trying to support Shim-6, heh?).

Until these issues are resolved, widespread adoption of IPv6 by large  
enterprise customers for general-purpose networking will be  
problematic (note that these aren't the only issues, but they are  
gating issues which render the others moot) at best.  Vendors,  
network operators and those participating in standards bodies must  
understand the seriousness of these issues for customers and work to  
address them (pardon the pun, heh).

------------------------------------------------------------------------
Roland Dobbins <rdobbins at cisco.com> // 408.527.6376 voice

One of the main causes of the fall of the Roman Empire was that, lacking
zero, they had no way to indicate successful termination of their C
programs.

            	         -- Robert Firth