/127 between routers?

[Sam Wilson]

On 7 Jan 2010, at 13:31, <michael.dillon at bt.com>  
<michael.dillon at bt.com> wrote:

>> What worries me in Mark's proposal (if it's that not inflating the
>> strength of what he's saying) is that we might end up with 8+8 by
>> stealth - CPE which only recognises 64-bit prefixes, big-iron routers
>> that route 64-bit prefixes in the fast path and relegate anything  
>> longer
>> to slower processing.
>
>
> Sounds good to me. If hardware designers can provide a way to speed up
> some traffic, then that is a good thing.

And if some mfr of inexpensive CPE assumes that no prefix can be  
longer than /64 and then an ISP starts handing out longer prefixes  
with [whatever mechanism] PD (I'm not saying that's likely, but it's  
technically possible)?  Remember "DNS packets aren't longer than 512  
bytes"?


>>  It feels as though someone should be deciding either that subnet
>> prefixes should never be longer than 64 bits or championing the right
>> for prefixes to be any old length you want.
>
>
> Championing? This is not politics, this is technology. If the  
> technology
> allows for divide and conquer to be applied to separating traffic  
> into a
> fast path and a slow path, then it will happen.

I think you're missing my point.  At present the (proposed) standard  
says you can have any length prefix you want except in the 1/8 of the  
address space that is currently available, where the max is /64.   
This entire thread is about non-/64 addresses in that address space.   
Doesn't that make you uneasy?

>> The same argument applies to management systems - what happens  
>> when the
>> only systems available assume that the longest subnet prefixes is /64
>> because that's what's currently in use (RFC 4291 2.5 vs 2.5.1)?  What
>> about staff training?
>
>
> Then people will renumber their networks to use /64 prefixes on all
> their p-t-p links. Staff training is easy, or maybe we should say that
> any competent operator already has ongoing staff training in place so
> whatever comes along, it can be handled.

Except that the (proposed) standard says that for other blocks of  
address space there is no presumption of maximum prefix length.   
People will forget that if the culture says /64 everywhere.


Sam Wilson
Network Team, IT Infrastructure
Information Services, The University of Edinburgh
Edinburgh, Scotland, UK



-- 
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.