Re: [Roll] Unaware-leaves - ND-Status and RPL-Status linkage

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

That's the point in my first sentence Rahul. If all we want to carry is moved then we can use the 6lo status, no need to invent another value.
We need to invent values for RPL specific stuff only. Here it makes the life of the device easier if there's a single code for moved unless we need to know the difference

Pascal

From: Rahul Jadhav <[e-mail]>
Sent: mardi 10 mars 2020 15:58
To: Pascal Thubert (pthubert) <[e-mail]>; Routing Over Low power and Lossy networks <[e-mail]>
Cc: [e-mail]
Subject: Re: [Roll] Unaware-leaves - ND-Status and RPL-Status linkage

Please find my comment [RJ] inline.

________________________________
From: Pascal Thubert (pthubert) <[e-mail]<mailto:[e-mail]>>
Sent: 10 March 2020 08:12 PM
To: Rahul Jadhav <[e-mail]<mailto:[e-mail]>>; Routing Over Low power and Lossy networks <[e-mail]<mailto:[e-mail]>>
Cc: [e-mail]<mailto:[e-mail]> <[e-mail]<mailto:[e-mail]>>
Subject: RE: [Roll] Unaware-leaves - ND-Status and RPL-Status linkage


True, we could.



Note that we might understand the "A" bit differently. For me the flag means that the definition of the value is inherited from a 6lo spec, it does not mean that in runtime the status was imported from a 6lo message.



Now if the node moves to a different root that is attached to a different 6BBR, it will get 1100 0011 which is 195. If the node moves within the same root, the code will be whatever we decide here. Does the node really care which case that is?



Maybe we could use 195 as opposed to 130?



[RJ] Sorry, but I don't get this. Value 195 is a 6lo value i.e., the 'A' bit (2nd MSB) is set. The DCO-draft talks primarily about values when 'A' bit is not set i.e., about RPL related values only. The DCO draft normatively refers to unaware-leaves for values with 'A' bit set and for the definition of 'A' bit itself.

My understanding is, if a common ancestor node within RPL Instance generates DCO then the value is 130 ('A' bit is unset). If the DCO is generated because of 6lo operations (across 6BBR) then the value could be 195 ('A' bit set and lower 6-bits according to 6lo). Thus 130 value is ok in DCO-draft (in ROLL). Unaware-draft will talk about value 195.









2. RPL status code for DCO (currently 130)
RPL Status code of 130 in DCO indicates that the address has moved i.e., somewhere below/downstream in the sub-dodag the target has changed parents.

It was discussed that 130 may not be an appropriate value but when I checked again it seems all right.

 0
 0 1 2 3 4 5 6 7
+-+-+-+-+-+-+-+-+
|E|A|  Value    |
+-+-+-+-+-+-+-+-+

130 == 1000 0010

The E bit is set indicate rejection and A bit is unset to indicate non-ND value.

Thus 130 seems to be the right value.

Action Point? No change if this rationale is correct.



Pascal >  Moved is 3 in 6lo (https://tools.ietf.org/html/rfc8505)



       3   | Moved: The registration failed because it is not the most |

   |       | recent.  This Status indicates that the registration is   |

   |       | rejected because another more recent registration was     |

   |       | done, as indicated by the same ROVR and a more recent     |

   |       | TID.  One possible cause is a stale registration that has |

   |       | progressed slowly in the network and was passed by a more |

   |       | recent one.  It could also indicate a ROVR collision.



Pascal > If we say that Value is the 6Lo status we want 128 + 3, and get 131..  I think that's the point we wanted to make.

Dec  Hex       Oct        Bin

    131 83         203       10000011



Pascal > Same here, am I missing something here? Any comments?



[RJ] But the 6lo value is applicable only when 'A' bit is set. With 'A' bit unset, as is the case above, RPL is free to choose its own values for the remaining 6bits. Isn't it? I was assuming that the 6-bit registry will not be common between 6lo Status Code and RPL Status Code.





Take care,



Pascal



On Sat, 28 Dec 2019 at 12:36, Rahul Jadhav <[e-mail]<mailto:[e-mail]>> wrote:

On Fri, 27 Dec 2019 at 11:58, Michael Richardson <[e-mail]<mailto:[e-mail]>> wrote:
>
>
> Rahul Jadhav <[e-mail]<mailto:[e-mail]>> wrote:
>     >> > > You mean that we could change the ND status to 6-bits in the 6lo
>     >> WG?  > (I would have parsed ND status as being someing 6man takes care
>     >> of)
>     >>
>     >> Yes, or mandate that the 64 first Status values are reserved for stuff
>     >> that could be carried in RPL
>     >>
>
>     RJ> Wondering how such a mandate would work. IF we do it this way,
>     RJ> future extensions to ND ARO status need to consider whether they are
>     RJ> applicable to RPL and define accordingly. This implies familiarity in
>     RJ> 6lo/6MAN with RPL.
>
>     > I prefer the other approach where we restrict the ND ARO status to
>     > 64bits itself. Anyways using 128 bits in the future seems far-fetched
>     > (as indicated by Pascal before). But to limit ND ARO status to 64bits
>     > we need to convince the 6lo/6MAN group that "because of RPL" we need to
>     > reduce the size. I am not sure if this is easy to digest!
>
> I thought that we had 6-bits to store 64-values, while ND has 8-bits to store
> 256 values.  Please correct me if I mis-understood.
>

[RJ] Currently ND has 8bits to store 256 values ... RPL needs to carry
the same 8-bits but currently it has space only for 6-bits in RPL
Status field. But as of now only few values of ND status has been used
and using up 256 values (or for that matter 64 values) seem
far-stretched.
Thus one of the opinions is to limit the ND status to 64-bits and
carry as-is in the RPL status.