On Thu, Feb 23, 2017 at 07:00:13PM +0000, Nick Hilliard wrote:
as it's currently worded, draft-ietf-6man-rfc4291bis seems to prohibit
the implementation of any interface netmask != /64:
However, the Interface ID of
all unicast addresses, except those that start with the binary value
000, is required to be 64 bits long.
This has substantial operational consequences in the ipv6 world because
if it's implemented as stated, it will cause production ipv6 networks to
break.
The ipv6 operational community may have opinions on the wisdom of
mandating new behaviour which would cause their networks to fall over,
so it would probably be a good idea to notify v6ops@ietf about the
existence of this draft so that the folks over there get a look-in
before a consensus call is made. As far as I can tell, this notification
never happened.
This is not 'new', but it is a re-iteration of (in my opinion) a
specification of a non-optimal addressing architecture.
Aside from the newness, I agree with you. I would've hoped that when a
document like this is revised after 10+ years, people would think to ask
and include in dialogue the actual ipv6 operators whether reality is
aligned with the documented architecture.
rfc4291 (2006) "For all unicast addresses, except those that start with
the binary value 000, Interface IDs are required to be
64 bits long and to be constructed in Modified EUI-64
format."
rfc3513 (2003) "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.
rfc2373 (1998) "The format prefixes 001 through 111, except for
Multicast Addresses (1111 1111), are all required to have
to have 64-bit interface identifiers in EUI-64 format.
See section 2.5.1 for definitions."
Kind regards,
Job