Re: IID length text

[Suresh Krishnan <[e-mail]>]

Hi Fernando,

On Jan 16, 2017, at 5:09 PM, Fernando Gont <[e-mail]<mailto:[e-mail]>> wrote:

On 01/16/2017 06:22 PM, Behcet Sarikaya wrote:
On Mon, Jan 16, 2017 at 2:46 PM, Fernando Gont <[e-mail]<mailto:[e-mail]>> wrote:
On 01/16/2017 05:42 PM, Behcet Sarikaya wrote:
On Mon, Jan 16, 2017 at 2:27 PM, Alexandre Petrescu
<[e-mail]<mailto:[e-mail]>> wrote:
Le 14/01/2017 à 20:49, Brian E Carpenter a écrit :

A modest suggestion:

OLD
  For all unicast addresses, except those that start with the binary
  value 000, Interface IDs are required to be 64 bits long.  Background
  on the 64 bit boundary in IPv6 addresses can be found in [RFC7421].

NEW
  IPv6 routing is based on prefixes of any valid length up to 128
[BCP198].
  For example, [RFC6164] standardises 127 bit  prefixes on point-to-point
  links. However, consistent use of Stateless Address Autoconfiguration
  (SLAAC)[RFC4862] requires that all interfaces on a link use the same
length
  of Interface ID. In practice, this means that to guarantee
interoperability
  of SLAAC, a fixed length of Interface ID is necessary. For all
currently
  allocated unicast addresses, except those that start with the binary
  value 000, that length is 64 bits. Note that this value is an arbitrary
  choice and might be changed for some future allocation of unicast
address
  space. Background on the 64 bit boundary in IPv6 addresses can be found
  in [RFC7421].


I agree with the change suggestion.  The new text and references are enough
motivation to clarify that that 64bit limit is an arbitrary choice and might
change in the future.


3GPP assigns 64 bit prefixes to each UE.
Extended Unique Identifiers defined are EUI-48 and EUI-64.

What does a layer-2 EUI have to do with a layer-3 address?

I don't know, you tell me :-)

Answer: Nothing. The fact that we've been embedding layer-2 "addresses"
into layer-3 addresses should not be taken as a reason for the two of
them of having to do anything with each other. So, yes, the 64-bit IID
length is, in reality, arbitrary. They could have been e.g. 32-bits or
48-bits (yes, multiples or 64 or 32 tend to be nicer)

Kind of agree but not entirely. There are some protocols in the constrained space that depend on some form of relationship between an L2 address and the L3 IPv6 address to achieve compression.

Thanks
Suresh