Re: [Roll] WG Last Call draft-ietf-roll-minrank-hysteresis-of-01

[JP Vasseur <[e-mail]>]

Sorry for the typo, this is a WG on the -01 version. Thanks !

JP.

On Apr 1, 2011, at 4:47 PM, Phoebus Chen wrote:

> JP,
> 
> 	I suppose draft-ietf-roll-minrank-hysteresis-of-02 will be made available online soon?  Or is this a last call regarding draft-ietf-roll-minrank-hysteresis-of-01?
> 
> 
> 
> 
> Omprakash and Phil,
> 
> 	I made a pass through draft-ietf-roll-minrank-hysteresis-of-01.
> 
> General Comments:
> *****************
> 1) As I understand from Table 1 and the mention of node metrics in the second paragraph of the introduction, MRHOF also supports node metrics.  However, in various places throughout the document, you refer to the "selected metric for the link..." .  I point out where I spot this below.  When using node metrics, I presume the node adds it's own node metric value to the advertised metric from its candidate neighbor to get the path cost?
> 
> 
> 2) In Section 3, you write "MRHOF cannot be used if the routing objective is to maximize the metric."  Those from the optimization community may be a little puzzled, because you can just negate the objective function to convert it from a maximization problem to a minimization problem.
> 
> Let's assume you wish to maximizing the product of link probabilities along a path.  Then, you can use log(p) which results in only negative link metrics (so still monotonic).  And instead of MIN_PATH_COST, you start with a MAX_PATH_REWARD of 0 and add the negative link metrics.  I don't see any of the MRHOF mechanisms breaking here (with some minor tweaks of switching "minimize" to "maximize").  Am I missing something?
> 
> As I understand, limiting MRHOF to minimization makes the explanations easier and the terminology / constant names more consistent.  But are you restricting the scope of the OF too much here?
> 
> 
> 3) Are the parent selection rules in any way affected by minHopRankIncrease, mentioned in (draft-ietf-roll-rpl-19, Section 3.5.2. Rank Relationships)?  Should minHopRankIncrease always be set to 1 for MRHOF?  In Table 1, you discuss how to convert various metrics into rank, but make no mention of minHopRankIncrease. (draft-ietf-roll-rpl-19, Section 3.5.1 Rank Comparison (DAGRank()) ) says:
> 
> "MinHopRankIncrease is the minimum increase in rank between a node and any of its DODAG parents."
> 
> As I understand, PARENT_SWITCH_THRESHOLD plays the role of "coarsening" the metric, which was originally done by the DAGRank() function for rank comparisons.  The way I understand the properties of rank described in (draft-ietf-roll-rpl-19, Section 3.5. Rank Properties) and from earlier discussions on the mailing list, coarsening the rank is for stability. DAGRank() may not do such a good job when two path-cost-to-root's advertised by candidate neighbors are less than PARENT_SWITCH_THRESHOLD apart but they result in different rank (e.g., DAGRank(A) > DAGRank(B)).  This is why we have MRHOF, right?
> ************
> 
> 
> Nits / editorial comments follow below, preceded by "PC>".
> 
> Of course, don't forget to update the version numbers of the informative references for last call (or is this done after last call?  I never remember.).
> 
> 
> -- 
> Phoebus Chen
> Postdoctoral Researcher
> Automatic Control Lab, KTH (Royal Institute of Technology)
> http://www.ee.kth.se/~phoebus
> 
> 
> 
> 
> 
> 
> 
> 
> Networking Working Group                                      O. Gnawali
> Internet-Draft                                                  P. Levis
> Intended status: Standards Track                     Stanford University
> Expires: September 1, 2011                             February 28, 2011
> 
> 
> 
> The Minimum Rank Objective Function with Hysteresis
> 
> 
> draft-ietf-roll-minrank-hysteresis-of-01
> 
> 
> Abstract
> 
>   Hysteresis delays the effect of changes in link metric on parent
>   selection.  Such delay makes the topology stable despite jitters in
>   link metrics.  The Routing Protocol for Low Power and Lossy Networks
>   (RPL) allows the use of objective functions to construct routes that
>   optimize or constrain a routing metric on the paths.  This
>   specification describes the Minimum Rank Objective Function with
>   Hysteresis (MRHOF), an objective function that minimizes the node
>   rank in terms of a given metric, while using hysteresis to prevent
>   excessive rank churn.  The use of MRHOF with RPL results in nodes
>   selecting stable paths that minimize the given routing metric to the
>   roots of a Directed Acyclic Graph (DAG).
> 
> Status of this Memo
> 
>   This Internet-Draft is submitted to IETF in full conformance with the
>   provisions of BCP 78 and BCP 79.
> 
>   Internet-Drafts are working documents of the Internet Engineering
>   Task Force (IETF), its areas, and its working groups.  Note that
>   other groups may also distribute working documents as Internet-
>   Drafts.
> 
>   Internet-Drafts are draft documents valid for a maximum of six months
>   and may be updated, replaced, or obsoleted by other documents at any
>   time.  It is inappropriate to use Internet-Drafts as reference
>   material or to cite them other than as "work in progress."
> 
>   The list of current Internet-Drafts can be accessed at
>   http://www.ietf.org/ietf/1id-abstracts.txt.
> 
>   The list of Internet-Draft Shadow Directories can be accessed at
>   http://www.ietf.org/shadow.html.
> 
>   This Internet-Draft will expire on September 1, 2011.
> 
> Copyright Notice
> 
> 
> 
> 
> Gnawali & Levis         Expires September 1, 2011               [Page 1]
> 
> 
> Internet-Draft    draft-ietf-roll-minrank-hysteresis-of    February 2011
> 
> 
>   Copyright (c) 2011 IETF Trust and the persons identified as the
>   document authors.  All rights reserved.
> 
>   This document is subject to BCP 78 and the IETF Trust's Legal
>   Provisions Relating to IETF Documents
>   (http://trustee.ietf.org/license-info) in effect on the date of
>   publication of this document.  Please review these documents
>   carefully, as they describe your rights and restrictions with respect
>   to this document.  Code Components extracted from this document must
>   include Simplified BSD License text as described in Section 4.e of
>   the Trust Legal Provisions and are provided without warranty as
>   described in the BSD License.
> 
> 
> Table of Contents
> 
>   1.  Introduction  . . . . . . . . . . . . . . . . . . . . . . . . . 3
>   2.  Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . 3
>   3.  The Minimum Rank Objective Function with Hysteresis . . . . . . 4
>     3.1.  Computing the Path cost . . . . . . . . . . . . . . . . . . 4
>     3.2.  Parent Selection  . . . . . . . . . . . . . . . . . . . . . 5
>     3.3.  Computing Rank  . . . . . . . . . . . . . . . . . . . . . . 6
>     3.4.  Advertising the path cost . . . . . . . . . . . . . . . . . 6
>   4.  MRHOF Variables and Parameters  . . . . . . . . . . . . . . . . 7
>   5.  Acknowledgements  . . . . . . . . . . . . . . . . . . . . . . . 8
>   6.  IANA Considerations . . . . . . . . . . . . . . . . . . . . . . 8
>   7.  Security Considerations . . . . . . . . . . . . . . . . . . . . 8
>   8.  References  . . . . . . . . . . . . . . . . . . . . . . . . . . 8
>     8.1.  Normative References  . . . . . . . . . . . . . . . . . . . 8
>     8.2.  Informative References  . . . . . . . . . . . . . . . . . . 8
>   Authors' Addresses  . . . . . . . . . . . . . . . . . . . . . . . . 8
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Gnawali & Levis         Expires September 1, 2011               [Page 2]
> 
> 
> Internet-Draft    draft-ietf-roll-minrank-hysteresis-of    February 2011
> 
> 
> 1. Introduction
> 
> 
>   An objective function allows RPL [I-D.ietf-roll-rpl] to select paths
>   that are best in terms of a given routing metric or select paths that
>   meet certain constraints in terms of the routing metric.  RPL
>   achieves this goal by selecting the parent among the alternate
>   parents as dictated by that objective function.  For example, if an
>   RPL instance uses an objective function that minimizes hop-count, RPL
>   will select paths with minimum hop count.
> 
>   The nodes running RPL might use a number of metrics to describe a
>   link or a node [I-D.ietf-roll-routing-metrics] and make it available
>   for route selection.  A metric can be used by different objective
>   functions to optimize or constrain the metric in different ways.
> 
>   This specification describes MRHOF, an objective function for RPL.
>   MRHOF uses hysteresis while selecting the path with the smallest
>   metric value.  The path with the minimum cost has different property
> 
> PC> s/has different/has a different/
> 
>   depending on the metric used for path selection.  For example, the
>   use of MRHOF with the latency metric allows RPL to find stable
>   minimum-latency paths from the nodes to a root in the DAG instance.
>   The use of MRHOF with the ETX metric allows RPL to find the stable
>   minimum-ETX paths from the nodes to a root in the DAG instance.
> 
>   MRHOF can be used only with additive metric that must be minimized on
> 
> PC> s/can be used only with additive/can only be used with an additive/
> 
>   the paths selected for routing.  Although MRHOF can be used with a
>   number of metrics, this draft is based on experiences with the ETX
> 
> PC> s/a number of metrics/different metrics/
> 
>   metric.
> 
> 
> 2. Terminology
> 
> 
>   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
>   "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
>   "OPTIONAL" in this document are to be interpreted as described in RFC
>   2119 [RFC2119].
> 
>   This terminology used in this document is consistent with the
>   terminologies described in [I-D.ietf-roll-terminology],
>   [I-D.ietf-roll-rpl], and [I-D.ietf-roll-routing-metrics].
> 
>   This document introduces two terms:
> 
>   Selected metric:  The metric chosen by the network operator to use
>         for path selection.  This metric can be any additive metric
>         listed in [I-D.ietf-roll-routing-metrics]
> 
> PC> Insert period at end of sentence above.
> 
> 
> 
> 
> 
> Gnawali & Levis         Expires September 1, 2011               [Page 3]
> 
> 
> Internet-Draft    draft-ietf-roll-minrank-hysteresis-of    February 2011
> 
> 
>   Path cost:  Path cost quantifies a property of an end-to-end path.
>         Path cost is composed using the selected metric of the links
>         along the path.  Path cost can be used by RPL to compare
>         different paths.
> 
> PC> s/composed using the/obtained by summing up the/
> 
> 
> 3. The Minimum Rank Objective Function with Hysteresis
> 
> 
>   The Minimum Rank Objective Function with Hysteresis, MRHOF, is
>   designed to find the paths with the smallest path cost while
>   preventing excessive churn in the network.  It does so by switching
>   to the minimum cost path only if the path cost of the current path is
>   larger than the path cost of the minimum cost path by a given
>   threshold.  MRHOF may be used with any additive metric listed in
>   [I-D.ietf-roll-routing-metrics] as long the routing objective is to
>   minimize the given routing metric.  MRHOF cannot be used if the
>   routing objective is to maximize the metric.
> 
> 3.1. Computing the Path cost
> 
> 
>   Nodes compute the path cost for each candidate neighbor reachable on
>   all the interfaces.  The Path cost represents the cost of the path,
> 
> PC> When I read this sentence literally, it means that a candidate
> PC> neighbor must be reachable on ALL the interfaces.  I don't think
> PC> that's what you mean.  This is done at various places throughout
> PC> the document.  May I suggest:
> PC> s/reachable on all the interfaces/that can be reached through an interface/
> 
> 
>   in terms of the selected metric, from a node to the root of the DODAG
>   through the neighbor.
> 
>   Root nodes (Grounded or Floating) set the variable cur_min_path_cost
>   to MIN_PATH_COST.
> 
>   A non-root node computes the path cost for a path to the root through
>   each candidate neighbor by adding these two components:
> 
>   1.  The selected metric for the link to a candidate neighbor.
> 
>   2.  The value of the selected metric in the metric container in the
>       DIO sent by that neighbor.
> 
>   A node SHOULD compute the path cost for the path through each
>   candidate neighbor reachable through all the interfaces.  If a node
> 
> PC> Again, "reachable through all the interfaces".  See comment above.
> 
>   cannot compute the path cost for the path through a candidate
>   neighbor, the node MUST NOT select the candidate neighbor as its
>   preferred parent with one exception.  If the node does not have
> 
> PC> s/parent with one/parent, with one/
> 
>   metrics to compute the path cost through any of the candidate
>   neighbors, it SHOULD join one of the candidate neighbors as a leaf
>   node.
> 
>   If the selected metric of the link to a neighbor is not available,
>   the path cost for the path through that neighbor SHOULD be set to
>   MAX_PATH_COST.  This cost value will prevent this path from being
> 
> 
> 
> Gnawali & Levis         Expires September 1, 2011               [Page 4]
> 
> 
> Internet-Draft    draft-ietf-roll-minrank-hysteresis-of    February 2011
> 
> 
>   considered for path selection.
> 
>   The path cost corresponding to a neighbor SHOULD be re-computed each
>   time:
> 
>   1.  The selected metric of the link to the candidate neighbor is
>       updated.
> 
>   2.  A node receives a new metric advertisement from the candidate
>       neighbor.
> 
>   This computation MAY also be performed periodically.  Deferring the
>   path cost computation for too long after new metric advertisements or
>   updates to the selected link metric results in nodes making parent
>   selection decision based on stale link and path information.
> 
> 3.2. Parent Selection
> 
> 
>   After computing the path cost for all the candidate neighbors
>   reachable through all the interfaces for the current DODAG iteration,
>   a node selects the preferred parent.  This process is called parent
>   selection.  Parent Selection SHOULD be performed each time:
> 
>   1.  The path cost for an existing candidate neighbor, including the
>       preferred parent, changes.  This condition can be checked
>       immediately after the path cost is computed.
> 
>   2.  A new candidate neighbor is inserted into the neighbor table.
> 
>   The parent selection MAY be deferred until a later time.  Deferring
>   the parent selection can delay the use of better paths or stopping
> 
> PC> s/stopping/stop/
> 
>   the use of worse paths than what is available in the network.
> 
>   A node MUST select a candidate neighbor as its preferred parent if
>   the path cost corresponding to that neighbor is smaller than the path
>   cost corresponding to the rest of the neighbors, except as indicated
>   below:
> 
>   1.  If the smallest path cost for paths through the candidate
>       neighbors is smaller than cur_min_path_cost by less than
>       PARENT_SWITCH_THRESHOLD, the node MAY continue to use the current
>       preferred parent.
> 
>   2.  If there are multiple paths with the smallest path cost and that
> 
> PC> remove "that"
> 
>       the smallest path cost is smaller than cur_min_path_cost by at
>       least PARENT_SWITCH_THRESHOLD, a node MAY use a different
>       objective function to select the preferred parent among the
>       candidates which are first hop on the path with the minimum cost.
> 
> PC> The last phrase is hard to parse, although it is correct.  I'm not
> PC> sure how to reword it.
> 
> 
> Gnawali & Levis         Expires September 1, 2011               [Page 5]
> 
> 
> Internet-Draft    draft-ietf-roll-minrank-hysteresis-of    February 2011
> 
> 
>   3.  A node MAY declare itself as a Floating root, and hence no
>       preferred parent, depending on the configuration.
> 
>   4.  If the selected metric for a link is greater than
>       MAX_LINK_METRIC, the node SHOULD exclude that link from
>       consideration for parent selection.
> 
>   5.  If cur_min_path_cost is greater than MAX_PATH_COST, the node MAY
>       declare itself as a Floating root.
> 
>   6.  If the configuration disallows a node to be a Floating root and
>       no neighbors are discovered, the node does not have a preferred
>       parent, and MUST set cur_min_path_cost to MAX_PATH_COST.
> 
>   The preferred parent is the only node in the parent set at a given
>   time.  Any candidate neighbor may become the preferred parent as
>   indicated above.
> 
> 3.3. Computing Rank
> 
> 
>   The DAG roots set their rank to MIN_PATH_COST for the selected
>   metric.
> 
>   Once a non-root node selects its preferred parent, it can use the
>   following table to covert its path cost to the DAG root through its
>   preferred parent (written as Cost in the table) to its rank:
> 
>                    +--------------------+------------+
>                    |  Node/link Metric  |    Rank    |
>                    +--------------------+------------+
>                    |     Node Energy    | 255 - Cost |
>                    |      Hop-Count     |    Cost    |
>                    |       Latency      | Cost/65536 |
>                    | Link Quality Level |    Cost    |
>                    |         ETX        |    Cost    |
>                    +--------------------+------------+
> 
>                  Table 1: Conversion of metric to rank.
> 
>   Node rank is undefined for these node/link metrics: Node state and
>   attributes, throughput, and link color.
> 
> 3.4. Advertising the path cost
> 
> 
>   Once the preferred parent is selected, the node sets its
>   cur_min_path_cost variable to the path cost corresponding to the
>   preferred parent.  Thus, cur_min_path_cost is the cost of the minimum
>   cost path from the node to the root.  The value of the
> 
> 
> 
> Gnawali & Levis         Expires September 1, 2011               [Page 6]
> 
> 
> Internet-Draft    draft-ietf-roll-minrank-hysteresis-of    February 2011
> 
> 
>   cur_min_path_cost is carried in the metric container whenever DIO
>   messages are sent.
> 
> 
> 4. MRHOF Variables and Parameters
> 
> 
>   MRHOF uses the following variable:
> 
>      cur_min_path_cost: The cost of the path from a node through its
>      preferred parent to the root computed at the last parent
>      selection.
> 
>   MRHOF uses the following parameters:
> 
>      MAX_LINK_METRIC: Maximum allowed value for the selected link
>      metric for each link on the path.
> 
>      MAX_PATH_COST: Maximum allowed value for the path metric of a
>      selected path.
> 
>      MIN_PATH_COST: The minimum allowed value for the path metric of
>      the selected path.
> 
>      PARENT_SWITCH_THRESHOLD: The difference between metric of the path
>      through the preferred parent and the minimum-metric path to
>      trigger new preferred parent selection.
> 
> PC> s/to trigger new preferred parent selection/in order to trigger the selection of a new preferred parent./
> 
>   The parameter values are assigned depending on the selected metric.
>   The best values for these parameters should be experimentally
>   determined.  The working group has long experience routing with the
>   ETX metric.  Based on those experiences, these ETX parameters are
>   known to work in many settings:
> 
>      MAX_LINK_METRIC: 10.  Disallow links with greater than 10 expected
>      transmission count on the selected path.
> 
>      MAX_PATH_COST: 100.  Disallow paths with greater than 100 expected
>      transmission count.
> 
>      MIN_PATH_COST: 0.  At root, the expected transmission count is 0.
> 
>      PARENT_SWITCH_THRESHOLD: 1.5.  Switch to a new path only if it is
>      expected to require at least 1.5 fewer transmission than the
>      current path.
> 
> 
> 
> 
> 
> 
> 
> Gnawali & Levis         Expires September 1, 2011               [Page 7]
> 
> 
> Internet-Draft    draft-ietf-roll-minrank-hysteresis-of    February 2011
> 
> 
> 5. Acknowledgements
> 
> 
>   Thanks to Antonio Grilo, Nicolas Tsiftes, Matteo Paris, JP Vasseur
>   for their comments.
> 
> 
> 6. IANA Considerations
> 
> 
>   This specification requires an allocated OCP.  A value of 1 is
>   requested.
> 
> 
> 7. Security Considerations
> 
> 
>   Security considerations to be developed in accordance to the output
>   of the WG.
> 
> 
> 8. References
> 
> 
> 8.1. Normative References
> 
> 
>   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
>              Requirement Levels", BCP 14, RFC 2119, March 1997.
> 
> 8.2. Informative References
> 
> 
>   [I-D.ietf-roll-routing-metrics]
>              Vasseur, J. and D. Networks, "Routing Metrics used for
>              Path Calculation in Low Power and Lossy Networks",
>              draft-ietf-roll-routing-metrics-01 (work in progress),
>              October 2009.
> 
>   [I-D.ietf-roll-rpl]
>              Winter, T., Thubert, P., and R. Team, "RPL: IPv6 Routing
>              Protocol for Low power and Lossy Networks",
>              draft-ietf-roll-rpl-05 (work in progress), December 2009.
> 
>   [I-D.ietf-roll-terminology]
>              Vasseur, J., "Terminology in Low power And Lossy
>              Networks", draft-ietf-roll-terminology-01 (work in
>              progress), May 2009.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Gnawali & Levis         Expires September 1, 2011               [Page 8]
> 
> 
> Internet-Draft    draft-ietf-roll-minrank-hysteresis-of    February 2011
> 
> 
> Authors' Addresses
> 
>   Omprakash Gnawali
>   Stanford University
>   S255 Clark Center, 318 Campus Drive
>   Stanford, CA  94305
>   USA
> 
>   Phone: [phone]
>   Email: [e-mail]
> 
> 
>   Philip Levis
>   Stanford University
>   358 Gates Hall, Stanford University
>   Stanford, CA  94305
>   USA
> 
>   Email: [e-mail]
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Gnawali & Levis         Expires September 1, 2011               [Page 9]
> 
> 
> Html markup produced by rfcmarkup 1.92, available from http://tools.ietf.org/tools/rfcmarkup/
> 
> 
> 
> 
> On 4/1/11 6:13 AM, JP Vasseur wrote:
>> Dear all,
>> 
>> draft-ietf-roll-minrank-hysteresis-of has been discussed for quite some time,
>> the document is stable and is now ready for Working Group Last Call. As discussed
>> during the WG meeting yesterday, there a very questions to address.
>> 
>> This starts a 2-week WG Last call on draft-ietf-roll-minrank-hysteresis-of-02
>> that will end on April 11 at noon ET (this is a bit more than 2 weeks, considering
>> most of us flying back end of this week).
>> 
>> Please send your comments to the authors and copy the mailing list.
>> 
>> Thanks.
>> 
>> JP.
>> _______________________________________________
>> Roll mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/roll
> _______________________________________________
> Roll mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/roll