Re: [Roll] useofrplinfo-07

["Turner, Randy" <[e-mail]>]

Hi Michael,

thanks for the reply  - comments below

R.

> On Aug 2, 2016, at 6:44 PM, Michael Richardson <[e-mail]> wrote:
> 
> 
> Turner, Randy <[e-mail]> wrote:
>> In section 5, “Storing Mode”, first paragraph, there is a sentence that
>> says “the sender cannot determine whether the destination is
>> RPL-capable and thus would need an IP-in-IP header.”  The last sentence
>> in this paragraph says "the sender can determine if the destination is
>> inside the LLN by looking if the destination address is matched by the
>> DIO’s PIO option.”
> 
>> This implies that “RPL-capable” does not mean “inside the LLN”, is this
>> implication correct ?
> 
> I don't think so.
> There may well be some editing that has to happen, since 2460bis assumption.
> Maybe you can contribute some text to fix this so that it makes sense to you?

Yes I understand that, post-plenary, some docs may be in flux due to 2460bis resolution, I
could probably suggest/submit something that works the way I believe is the intent.

> 
> One can be inside the LLN, but be RPL-incapable. (a "leaf" or "IPv6 host”)

Yes, the draft recognizes the RPL-incapable leaf

> 
> One assumes that nodes outside the LLN are RPL-incapable, and one can
> tell this buy looking seeing if the destination address matches the PIO.
> 
> {Clearly, with a 6BBR, one can have RPL nodes outside the local LLN
> which are RPL-capable, but we can't tell that}

Right, anything outside the DODAG/instance would probably have a different PIO prefix so comparing prefixes
is no guarantee the destination is not “RPL-capable”.
So “RPL-capable” might not be the right term to use — “outside the LLN”  might work,  in this case I’m inferring that the LLN is the local DODAG/instance

> 
> (or more to the point, before -07, routers would drop packets with RPL
> artifacts on them. Starting at -07 we assume that RPI is okay, but RH3 is not)

Right, I understand that some 2460bis proposals would make life easier, but the discussion on hop-by-hop headers
and how they are handled is still undecided, or at least that was my takeaway from the plenary….is this right?

> 
> --
> Michael Richardson <[e-mail]>, Sandelman Software Works
> -= IPv6 IoT consulting =-
> 
> 
> 
> _______________________________________________
> Roll mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/roll