Addressing plan exercise for our IPv6 course

JC Dill jcdill.lists at gmail.com
Fri Jul 30 15:24:07 UTC 2010


Matthew Walster wrote:
> On 29 July 2010 18:08, Leo Vegoda <leo.vegoda at icann.org> wrote:
>   
>> There's a good chance that in the long run multi-subnet home networks will become the norm.
>>     
>
> With all due respect, I can't see it. Why would a home user need
> multiple subnets? Are they really likely to have CPE capable of
> routing between subnets at 21st Century LAN speeds? Isn't that
> needlessly complicating the home environment?

I strongly urge all my competitors to approach IPv6 with this philosophy.

In other words, in the long run it will push up your labor costs (for 
admins, for customer support), and push down your customer satisfaction 
because you were needlessly worried about the "scarcity" of a plentiful 
resource and didn't think ahead to new technologies, new ideas, and 
hampered your network with an allocation scheme that didn't expand 
gracefully to acomodate new uses.

Look at it this way - most (ISPs, businesses, consumers, appliance 
vendors) are going to allocate according to the recommendations or be 
using an allocation according to the recommendations.  Why are you even 
*considering* using a different allocation scheme?  What do *you* gain?  
All I see are headaches from doing it differently.  When you hire you 
will need to retrain admins who are accustomed to the recommended 
system.  When you get new customers, you will have to retrain them to 
use your non-standard system.  When they try to use appliances that are 
pre-configured to use the recommended system, their appliances won't 
work right or will need special configuration.  Etc. 

If - IF the recommendations are not conservative enough (which is 
considered to be a very remote possibility), then we can change the 
recommendations when we put the next 1/8 of the IPv6 IPs into service.  
But consider the possible use case where we actually start to run out of 
IPs in this first 1/8 segment of the IPv6 space.  It's not going to 
happen by IP usage from current services.  It's going to happen by IP 
consumption from new, as yet unimagined, services.  And if we have all 
these new services (devices, appliances) that require IP addresses then 
it means we WILL need to do subnetting at end user premises.

jc





More information about the NANOG mailing list