IPv6 PI allocation

[Sam Wilson]

On 17 May 2007, at 19:27, Kevin Loch wrote:

> Colm MacCarthaigh wrote:
>
>> Right now, the use of /127's and /126's for link address by some
>> networks are a bigger part of future problems than anything  
>> represented
>> by PI.
>
> Are you suggesting that routers should not route all 128 bits?  If
> we are going to use 128 bit addresses then we need to be able
> to route all 128 bits in silicon.  I'm sure you have a few /32 IPv4
> static routes in your tables from time to time...
>
> I prefer /112 as a default size for links because:
> - It is an ideal boundary for record keeping and visual identification
>   of subnets (unlike /126 or /127)
> - It's large enough to support multiple devices for varying  
> applications
>   and/or reconfiguration/migration on ethernet type links
> - /64 is just an insane waste of addresses (or subnets if you look  
> at it
>   that way).
>
> I do use /64 for subnets that hosts occupy, mainly to make autoconfig
> possible if it is ever desired.

Excuse me while I delurk to make a meta point.  One of the problems  
with IPv6 is that it has too many bits in the address and we don't  
know how to handle that.  Backbone people want to hand out short  
prefixes to keep routing tables short; edge people want to use long  
prefixes so as not to waste bits at the RHS.  It's a fundamental  
problem and it's not going to go away.

-- 
Sam Wilson
Network Team, IT Infrastructure (formerly Computing Services)
Information Services, The University of Edinburgh
Edinburgh, Scotland, UK