Re: [6lo] Generation of IPv6 IIDs

[Fernando Gont <[e-mail]>]

Hi, Esko,

Thanks so much for your feedback! Please find my responses in-line....

On 05/01/2014 03:41 AM, Dijk, Esko wrote:
> 
> I would expect that at least RFC 6282 (and perhaps the RFC 4944 which
> it updates?) should be mentioned, since the 6LoWPAN header
> compression is based on the IID containing the hardware address. This
> RFC is then a notable exception to the 'SHOULD NOT'.

Since this one does not really follow the "SHOULD NOT", I guess we might
opt to:

a) List the RFC as "There are issues with this specification" (as
suggested by Erik), or,

b) Do not mention this RFC at all.


Thoughts?



> Furthermore for the 6lo active WG documents: - draft-ietf-6lo-btle-00
> has the use of hardware address for IID as a 'MAY' - that would need
> to become a 'SHOULD NOT' (unless there is some negative impact on
> performance)

I'd suggest that draft-ietf-6man-default-iids is referenced in that I-D.
Then you either follow the advice in draft-ietf-6man-default-iids, or
explain your reasoning to go against it. (this was suggested by Dave
Thaler earlier on).


> ? - draft-ietf-6lo-lowpanz-04 has a MUST on the use of
> NodeID in the IID, however the NodeID is only 8-bit locally assigned
> and is not an identifiable hardware address.

I'll check the aforementioned I-D when I land (I'm currently on a
plane). I'd note that the issue is not only about IIDs being derived
from hardware addresses, but also about they being constant across
different network attachment points.

Thanks so much!

Best regards,
-- 
Fernando Gont
SI6 Networks
e-mail: [e-mail]
PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492