Re: [6lo] Request for Ethertype for 6lowpan-encoded IPv6

["Pascal Thubert (pthubert)" <[e-mail]>]

Hello Don:

Thanks for all this. About your question:

> I just scanned your 6loRH draft.  Do you re-use the RFC 4944/RFC 6282
> Protocol Dispatch?  Is it possible to just use a new Dispatch from
> 4944/6282 along with a 6LoWPAN/6lo EtherType to cover what you want?

It appears that finding encoding space is a lot harder and longer than describing the encoding itself. 6LoRH uses encoding space that is taken from the 6LoWPAN dispatch space and is declared in https://www.ietf.org/id/draft-ietf-6lo-paging-dispatch-01.txt. It took the group more than one year to reach that point and I would not reopen that box.

>From what I'm reading below, I gather that 6lo should be free to extend what it does within its dispatch space, and that should still be valid within the ethertype. I also read that we need to declare to the IEEE what our process is. In this particular case, I think that the IEEE understands very well how the IETF updates RFCs with other RFCs, the IANA registry, etc...

Do I have it right?

Pascal

> -----Original Message-----
> From: Don Sturek [mailto:[e-mail]]
> Sent: mardi 23 février 2016 22:03
> To: Pascal Thubert (pthubert) <[e-mail]>; Paul Duffy (paduffy)
> <[e-mail]>
> Cc: [e-mail]; Ralph Droms (rdroms) <[e-mail]>
> Subject: Re: [6lo] Request for Ethertype for 6lowpan-encoded IPv6
> 
> Hi Pascal,
> 
> For the past several IEEE 802 meetings, the IEEE 802.15 group has held a
> joint meeting with IEEE 802.1 (who seem to be the main point of contact for
> EtherTypes).  This topic of a 6LoWPAN EtherType has been one of the points
> of discussion (along with IEEE 802.15's proposal to finally create a link layer
> interface over IEEE 802.15.4)
> 
> Here is what IEEE 802.1 told us with regard to an EtherType for 6LoWPAN
> (which we asked for via IEEE 802.15 and were turned down, asked for from
> Wi-SUN and were turned down then were told to ask IETF to request it):
> 1.   They won't allocate one unless your protocol provides for sub-typing
> (e.g. The dispatch code in 6LoWPAN/6lo)
> 2.   They have very few so they really no longer allocate EtherTypes easily
> 3.   If they were to allocate an EtherType, they want the receiving
> organization to be able to cover some clearly defined set of uses (e.g.
> Existing 6LoWPAN, 6lo as it is currently defined plus some idea of how to
> extend it assuming you are unlikely to get other EtherTypes).
> 
> I just scanned your 6loRH draft.  Do you re-use the RFC 4944/RFC 6282
> Protocol Dispatch?  Is it possible to just use a new Dispatch from
> 4944/6282 along with a 6LoWPAN/6lo EtherType to cover what you want?
> 
> Feel free to propose 2 EtherTypes but you might want to attend an
> upcoming
> IEEE 802 meeting to defend that decision.   I think with Ralph's draft
> there will be clear reasons to provide a 6LoWPAN/6lo EtherType. If that can
> be made to cover what you need that would the best solution that is the
> ideal situation (and if not, we should discuss since that means the dispatch
> mechanism in 4944/6282 should be re-thought to ensure we have enough
> expansion capability going forward)
> 
> One concern I have on your draft is the intermingling of 6lo and RPL by the
> way.  Not having read the draft in detail it could be that a 6loWPAN/6lo
> EtherType really doesn't well cover what you are trying to do here anyway.
> 
> Don
> 
> 
> 
> On 2/23/16 12:31 PM, "6lo on behalf of Pascal Thubert (pthubert)"
> <[e-mail] on behalf of [e-mail]> wrote:
> 
> >I'm not following you Paul
> >
> >In one hand you claim that an ethertype is hard to get and in the other
> >you are asking to use that bullet with an incomplete scope? What will
> >the next protocol that needs this new work do?
> >
> >Pascal
> >
> >> Le 23 févr. 2016 à 21:27, Paul Duffy (paduffy) <[e-mail]> a
> >>écrit :
> >>
> >>> On 2/23/2016 3:20 PM, Pascal Thubert (pthubert) wrote:
> >>> Hello Ralph
> >>>
> >>> The 6LoRH work is getting ready for last call.
> >>>
> >>> If the demand comes from environments that use RPL, It should most
> >>>certainly be included.
> >>>
> >>> What is the requirement in terms of time frame?
> >>
> >> Immediate if possible.    This request is coming after months of false
> >>starts with IEEE.
> >>
> >> 6LoRH is not in scope of the impacted effort.
> >>
> >>
> >>>
> >>> Pascal
> >>>
> >>>> Le 22 févr. 2016 à 16:15, Ralph Droms (rdroms) <[e-mail]> a
> >>>>écrit :
> >>>>
> >>>>
> >>>>> On Feb 22, 2016, at 7:59 AM 2/22/16, Pascal Thubert (pthubert)
> >>>>><[e-mail]> wrote:
> >>>>>
> >>>>> Hello Ralph and Paul:
> >>>>>
> >>>>> Substantive:
> >>>>> -----------------
> >>>>> The draft is specific that the ethertype means what it names 6loENC.
> >>>>>
> >>>>> 1) What would be the position of new work such as
> >>>>> - https://tools.ietf.org/html/draft-ietf-6lo-routing-dispatch and
> >>>>> - https://tools.ietf.org/html/draft-thubert-6lo-inner-compression ?
> >>>> I would consider those documents to be out of scope of this
> >>>>document as "future work".  One way to handle those extensions would
> >>>>be to have them include explicit text about applying to IPv6 as
> >>>>carried by the Ethertype assigned through draft-droms-6lo-ethertype-
> request.
> >>>>
> >>>>> 2) The group flirted with Jonathan's idea of having more than one
> >>>>>ethertype. For instance, we could have an ethertype that is more
> >>>>>specific than "anything 6loENC can encode with paging" to more
> >>>>>specific stuff that could be even better compressed. An example
> >>>>>that was discussed was a route-over ethertype and a mesh-under
> >>>>>ethertype so for instance page 1 does not have to be signaled in a
> >>>>>route-over ethertype.
> >>>>> Could that be envisioned?
> >>>> I could imagine defining additional Ethertypes but I would consider
> >>>>those extensions to be out of scope of this document, as well.
> >>>>
> >>>> - Ralph
> >>>>
> >>>>> Typos:
> >>>>> ---------
> >>>>> Datatgrams
> >>>>>
> >>>>> Cheers,
> >>>>>
> >>>>> Pascal
> >>>>>
> >>>>>> -----Original Message-----
> >>>>>> From: Ralph Droms (rdroms)
> >>>>>> Sent: vendredi 19 février 2016 20:55
> >>>>>> To: [e-mail] WG <[e-mail]>
> >>>>>> Subject: Request for Ethertype for 6lowpan-encoded IPv6
> >>>>>>
> >>>>>> (BCC: [e-mail])
> >>>>>>
> >>>>>> Paul Duffy and I recently published
> >>>>>>draft-droms-6lo-ethertype-request-00.
> >>>>>> The purpose of this document is to demonstrate IETF consensus for
> >>>>>>requesting an Ethertype to be used with IPv6 datagrams encoded
> >>>>>>according  to RFC 4944/6282, in L2 protocols that have a protocol
> >>>>>>switch.
> >>>>>>
> >>>>>> We'd like to move this document along quickly, because Wi-SUN
> >>>>>>Alliance  would like to specify the Ethertype in the Wi-SUN FAN
> >>>>>>specification.  Please  take a look at the document (it's barely
> >>>>>>over 3 pages, including all the
> >>>>>> boilerplate) and respond to the 6lo WG mailing list with comments.
> >>>>>>
> >>>>>> - Ralph
> >>> .
> >>
> >
> >_______________________________________________
> >6lo mailing list
> >[e-mail]
> >https://www.ietf.org/mailman/listinfo/6lo
>