[Roll] RPL Design Questions
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Roll] RPL Design Questions
WG,
Please do let us know what you are thinking w/ regard to the design
Questions A, B, & C below. We hope to include your feedback into the next
revision of the document, with FSM descriptions as well.
Thanks,
- RPL Authors
As discussed at the interim meeting, implementer feedback has
indicated that the RPL protocol as proposed appears complex, both in
description and actual FSM.
Our plan of action is to, by end of month, deliver a -04 that will
contain actual (non-editorial) changes. The intent of this mail is
a poll over a period of 2 weeks, to help design the functional
changes.
In this spirit, we would like the WG to work on the following
questions:
Question A: Should RPL make use of the currently proposed local
repair mechanism (DAG splitting and merging)?
[NO implies that DAG repairs shall be coordinated globally with
the use of DAG Sequence Number; the related mechanisms are to
be expanded for -04]
Question B: Should RPL make use of a hold-up timer and related
states/procedures to reduce churn by coordinating the DAG merge?
[NO implies RPL allows nodes to move (jump) between DAGs with
little coordination to reduce complexity of specification/
implementation (perhaps w/ Optional hold-up mechanism)].
Question C: Should RPL make use of a hold-down timer and related
states/procedures to limit flapping when removing DAG Parents /
leaving DAGs
[NO implies RPL allows nodes to freely remove/add DAG Parents
as and when they are able in order to reduce complexity of
specification/implementation (perhaps w/ Optional hold-down
mechanism).
WG, we expect that this thread will contain a lot of interesting
related discussion, but in your comments please do also be sure to
try and address the initial questions A, B, and C so as to help us
better capture the WG position on these issues. Please note that A,
B, and C are to some degree orthogonal, e.g. `No Local Repair' to
Question A does not necessarily imply a particular disposition toward
the Hold-Up mechanism in Question B.
Thanks,
RPL Authors
Some examples, derived from the present draft, are provided below:
------------------------------------------------------------------------
Example: Local Repair with Hold-Up state
: : :
: : :
(A) (A) (A)
|\ | |
| `-----. | |
| \ | |
(B) (C) (B) (C) (B)
| | \
| | `-----.
| | \
(D) (D) (C)
|
|
|
(D)
-1- -2- -3-
Figure 1: DAG Maintenance
Consider the example depicted in Figure 1-1. In this example, Node
(A) is attached to a DAG at some rank d. Node (A) is a DAG parent of
Nodes (B) and (C). Node (C) is a DAG parent of Node (D). There is
also an undirected sibling link between Nodes (B) and (C).
In this example, Node (C) may safely forward to Node (A) without
creating a loop. Node (C) may not safely forward to Node (D),
contained within it's own sub-DAG, without creating a loop. Node (C)
may forward to Node (B) in some cases, e.g. the link (C)->(A) is
temporarily unavailable, but with some chance of creating a loop
(e.g. if multiple nodes in a set of siblings start forwarding
`sideways' in a cycle) and requiring the intervention of additional
mechanisms to detect and break the loop.
Consider the case where Node (C) hears a RA-DIO message from a Node
(Z) at a lesser rank and superior position in the DAG than node (A).
Node (C) may safely undergo the process to evict node (A) from its
DAG parent set and attach directly to Node (Z) without creating a
loop, because its rank will decrease.
Now consider the case where the link (C)->(A) becomes nonviable, and
node (C) must move to a deeper rank within the DAG:
o Node (C) must first detach from the DAG by removing Node (A) from
its DAG parent set, leaving an empty DAG parent set. Node (C)
becomes the root of its own floating, less preferred, DAG.
o Node (D), hearing a modified RA-DIO message from Node (C), follows
Node (C) into the floating DAG. This is depicted in Figure 1-2.
In general, any node with no other options in the sub-DAG of Node
(C) will follow Node (C) into the floating DAG, maintaining the
structure of the sub-DAG.
o Node (C) hears a RA-DIO message from Node (B) and determines it is
able to rejoin the grounded DAG by reattaching at a deeper rank to
Node (B). Node (C) starts a DAG Hop timer (putting Node (B) in
the Hold-Up state) to coordinate this move.
o The timer expires and Node (C) adds Node (B) to its DAG parent
set. Node (C) has now safely moved deeper within the grounded DAG
without creating any loops. Node (D), and any other sub-DAG of
Node (C), will hear the modified RA-DIO message sourced from Node
(C) and follow Node (C) in a coordinated manner to reattach to the
grounded DAG. The final DAG is depicted in Figure 1-3
------------------------------------------------------------------------
Example: Global Repair with Sequence Number / No Held-Up State
: : :
: : :
(A) (A) (A)
|\ | |
| `-----. | |
| \ | |
(B) (C) (B) (C) (B)
| \
| `-----.
| \
(D) (D) (C)
(D)
-1- -2- -3-
Figure 2: DAG Maintenance
Consider the example depicted in Figure 2-1, similar to that depicted
in Figure 1. Let there be a sequence number associated with the DAG,
as originated from the DAG root, with value N. Initially all nodes
have received DAG Sequence Number N. The following example offers one
possible Global Repair scenario to give a high level idea, please
note that there are details that would remain to be worked out if the
WG heads in this direction.
Now consider the case where the link (C)->(A) becomes nonviable, and
node (C) must move to a deeper rank within the DAG:
o Node (C) must first detach from the DAG by removing Node (A) from
its DAG parent set, leaving an empty DAG parent set. At this
point Node (C) is not associated with any DAG [TBD -- Alternately
Node (C) remains associated with the DAG at rank infinity].
o Node (D) may possibly learn that Node (C) is no longer associated
with any DAG and itself becomes unassociated [TBD Node (D) may
also retain a sub-DAG relationship with Node (C)]. Let Node (C)
and (D) both be free from any DAG, but remember the Sequence
Number N associated with their original DAG. This is depicted in
Figure 2-2.
o Node (C) and (D) will now be willing to reattach to the original
DAG at IMMEDIATELY (no hold-up mechanism) at ANY node offering
connectivity to the original DAG and advertising a sequence number
N+1. Since Node (C) and (D) are no longer members of the original
DAG, only a node who is still a member of the original DAG may
possibly present the sequence number N+1. Such a node who
presents N+1 must clearly have another path to the DAG root other
than via (C) and (D) and thus may offer a loop-avoiding attachment
point.
o [TBD] The DAG Root may periodically issue sequence number
increments, causing the issue of N+1
o [TBD] The broken link (C)->(A) may be detected through some other
means, and signalled to or cause a trigger at the DAG Root,
causing the issue of N+1
o Let Node (C) hears a RA-DIO message with N+1 from Node (B) and
determine it is able to rejoin the original DAG by immediately
reattaching j to Node (B) (No hold-up state in this example. The
DAG is now as depicted in Figure 2-3
o Node (C) may then subsequently send RA-DIO with DAG Sequence
number N+1, allowing Node (D) to reattach (not depicted).
------------------------------------------------------------------------
_______________________________________________
Roll mailing list
Roll at ietf.org
https://www.ietf.org/mailman/listinfo/roll
Note: Messages sent to this list are the opinions of the senders and do not imply endorsement by the IETF.
