[6bone] semi-newbie Q on IPv6 address planning

Joao Luis Silva Damas joao@ripe.net
Fri, 2 Aug 2002 10:39:40 +0200


At 8:26 -0700 1/8/02, Michel Py wrote:
>  > And by the way, a /126 is perfect for Point-to-point
>
>No, it is as illegal with draft-ietf-ipngwg-addr-arch-v3-08.txt than it
>is with RFC2374, read draft-ietf-ipngwg-addr-arch-v3-08.txt again.

I geuss you refer to section 2.5.1 where it says

For all unicast addresses, except those that start with binary value
    000, Interface IDs are required to be 64 bits long and to be
    constructed in Modified EUI-64 format.

which is used together with: (for the first 3 octets)

           0       0 0       1 1       2
          |0       7 8       5 6       3|
          +----+----+----+----+----+----+
          |cccc|ccug|cccc|cccc|cccc|cccc|
          +----+----+----+----+----+----+

and

The motivation for inverting the "u" bit when forming an interface
    identifier is to make it easy for system administrators to hand
    configure non-global identifiers when hardware tokens are not
    available.  This is expected to be case for serial links, tunnel end-
    points, etc.  The alternative would have been for these to be of the
    form 0200:0:0:1, 0200:0:0:2, etc., instead of the much simpler 1, 2,
    etc.


Fine, and I will normally use 1 and 2 as the last 64 bits of the 
address for poi nt-to-point but I will also use a /126 mask for this 
stuff because it reduces the potential for mistakes, because if 
someone makes a mistake and types something inside the same /64 I am 
using for this interface, if it is also inside the /126 it will be 
easy to detect and if it is outside at least it won't screw my 
routing in a way that is hard to debug because some router starts 
thinking it has a direct connection to another one when it doesn't.
I chose a /126 (instead of a /127, even before reading Pekka's draft) 
also because that's what a lot of network engineers are used to in 
IPv4 (a /31) and it minimises human error (which is far more frequent 
than a hardware failure).

And I wish protocol design wasn't a game of designing nice bit 
templates and took operational practice into account as a starting 
point.

Joao