RE: HBH Obsolete? (was Review of draft-ietf-6man-hbh-header-handling-01)

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

Hello Brian:

> >
> > As we arrange to compress RFC6554/6553, we could change the HbH code
> > point From 63 to 43, which would permit non-RPL nodes (such as 6lowpan
> > only leafs, and cloud data collection devices) to skip the RPI header.
> 
> That makes a lot of sense. It might have made even more sense for RFC 7045
> to have said:
> 
>    The IPv6 Hop-by-Hop Options header MAY be processed by
>    intermediate forwarding nodes as described in [RFC2460].
>    Other forwarding nodes MUST ignore it.
> I-----------------------------

 I read that RFC7045 would override RFC2460 to make the *router* behaviour optional. Our intent was that the *hosts* would not drop a packet with a HbH header with options that they do not understand, which is an additional step. The reasoning being that the HbH is mostly for intermediate nodes and that the role of the header is probably consumed at the time the host gets it. 

For RPL,  we need that hosts that do not participate to RPL to ignore a HbH header with a RPL option in it.
But the RPI in the RPL option may be useful to another type of host that we call leaves; leaves are RPL-aware non-forwarding nodes.
Leaves will set the information in the RPI, and may process incoming packet differently (e.g. priority and or security) depending on the RPI.
This is why the flow label was of such interest for us, only nodes which care would have been impacted by the setting.

Cheers,

Pascal