Internet DRAFT - draft-francois-idr-addpath-limit

draft-francois-idr-addpath-limit






Network Working Group                                    Pierre Francois
Internet-Draft                                            Camilo Cardona
Expires: August 11, 2014                        Institute IMDEA Networks
                                                            Adam Simpson
                                                          Alcatel-Lucent
                                                            Jeffrey Haas
                                                        Juniper Networks
                                                        February 7, 2014


                       ADD-PATH limit capability
                  draft-francois-idr-addpath-limit-00

Abstract

   In this draft, we propose a new capability that allows BGP speakers
   supporting ADD-PATH to announce a limit on the number of paths they
   want to receive from their peers.

Status of this Memo

   This Internet-Draft is submitted in full conformance with the
   provisions of BCP 78 and BCP 79.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF).  Note that other groups may also distribute
   working documents as Internet-Drafts.  The list of current Internet-
   Drafts is at http://datatracker.ietf.org/drafts/current/.

   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."

   This Internet-Draft will expire on August 11, 2014.

Copyright Notice

   Copyright (c) 2014 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



Pierre Francois, et al.  Expires August 11, 2014                [Page 1]

Internet-Draft          ADD-PATH limit capability          February 2014


   the Trust Legal Provisions and are provided without warranty as
   described in the Simplified BSD License.


Table of Contents

   1.  Introduction  . . . . . . . . . . . . . . . . . . . . . . . . . 3
   2.  The ADD-PATH Limit capability . . . . . . . . . . . . . . . . . 3
   3.  Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
   4.  References  . . . . . . . . . . . . . . . . . . . . . . . . . . 5
   Authors' Addresses  . . . . . . . . . . . . . . . . . . . . . . . . 5








































Pierre Francois, et al.  Expires August 11, 2014                [Page 2]

Internet-Draft          ADD-PATH limit capability          February 2014


1.  Introduction

   ADD-PATH is an extension to BGP that allows a BGP speaker to send and
   receive more than one path for the same NLRI [AddPath].  The ADD-PATH
   capability does not include any mechanism for restricting the number
   or type of paths that a peer can receive from others.  Thus, a
   receiving BGP speaker has no control on the number of paths sent by
   its peer, which depends only on the mode the operator desires to
   implement in the transmitting side [AddPathGuidelines].  This
   restriction can make network operations more difficult in some
   situations:
   o  In cases in which all network devices are managed by the same
      group, operators select the ADD-PATH mode that best fits the
      resources of each of their devices.  Operators must configure
      manually each speaker with a mode that does not overload the
      resources of the other devices.  The overhead of this procedure
      can be high in some networks, as this configuration must be
      performed at the session level.  If a ADD-PATH router could signal
      a limit in the number of paths it wants to receive, operators
      could achieve the same resource control by performing a more
      simple configuration.
   o  In cases in which devices are managed by different operators, such
      as in networks spanning large geographical regions or in Internet
      Exchange Points, operators must agree on the ADD-PATH mode to use
      between BGP speakers.  If one of the devices has constraints on
      the number of paths it can receive, the other party must configure
      an ADD-PATH mode that does not overload the memory of other
      device.  Under these circumstances, allowing the receiving side to
      limit the number of paths can ease the management process for all
      administration sides.

   In this document, we propose a new capability that allows an ADD-PATH
   capable BGP speaker to set an explicit upperbound on the number of
   paths it wants to receive from its peer.


2.  The ADD-PATH Limit capability

          +------------------------------------------------+
          | Address Family Identifier (2 octets)           |
          +------------------------------------------------+
          | Subsequent Address Family Identifier (1 octet) |
          +------------------------------------------------+
          | Flags(1 octet)                                 |
          +------------------------------------------------+
          | Receive bound (2 octet)                              |
          +------------------------------------------------+




Pierre Francois, et al.  Expires August 11, 2014                [Page 3]

Internet-Draft          ADD-PATH limit capability          February 2014


                           Figure 1: Capability

   The meaning and use of the fields are as follows:

      Address Family Identifier (AFI):

         This field is the same as the one used in [RFC4760].

      Subsequent Address Family Identifier (SAFI):

         This field is the same as the one used in [RFC4760].

      Flag Field

         This field contains different bit flags related to the ADD-PATH
         limitation capability.
         The most significant bit of the field (limit-capable bit) is
         used to communicate if the device supports the limitation of
         paths announced to the peer.  If the router supports the ADD-
         PATH Limit capability, but it is not capable of limiting the
         number of announced paths, it should set the limit-capable bit
         to 0.
         The remaining bits MUST be set to 0 and SHOULD be ignored upon
         receipt.
      Receive Bound

         This field indicates the maximum number of paths the device
         wants to receive from a peer for the <AFI, SAFI>.  If the field
         is set to 0, the device has no restriction on the number of
         paths it can receive.



3.  Operation

   The ADD-PATH Limit capability SHOULD be announced by BGP speakers
   that require their neighbors to send a limited number of paths per
   prefix.  A router that is capable of sending a limited number of
   paths to a BGP ADD-PATH neighbor SHOULD also announce the ADD-PATH
   Limit capability.  For cases in which a router is not capable of
   setting a limit in the number of paths it sends to a peer, it should
   set the limit-capable bit in the add-path capability to 0.

   The ADD-PATH capability is a prerequisite of the ADD-PATH Limit.  A
   BGP peer SHOULD ignore the ADD-PATH Limit capability from a peer that
   did not also announce the ADD-PATH capability.

   The ADD-PATH limit capability is used to set an upper bound on the



Pierre Francois, et al.  Expires August 11, 2014                [Page 4]

Internet-Draft          ADD-PATH limit capability          February 2014


   number of paths that the router wants to receives from a peer.  A
   sender capable of limiting the number of paths per peer SHOULD NOT
   send more paths than requested by the receiver.

   It might be impractical for a BGP speaker to strictly stick to each
   of the upper bounds specified by its peers.  Thus, the sender MAY
   send a lower amount of paths than the upper bound indicated by its
   peer.

   A router SHOULD include the best path in the subset of paths to send
   to a peer, except when the best path is received from that peer.  The
   choice of the rest of paths to be sent is left free to the
   implementation.

   A BGP speaker could receive more paths than the number defined in the
   ADD-PATH capability, even when the BGP peer supports the limitation
   of paths.  This event should be logged, but the session with the peer
   should be preserved.  The receiving speaker should implement the
   required mechanism to deal with more paths that it can support.


4.  References

   [AddPath]  D. Walton, E. Chen, A. Retana, and J. Scudder,
              "Advertisement of Multiple Paths in BGP",
              draft-ietf-idr-add-paths-09.txt (work in progress).

   [AddPathGuidelines]
              J. Uttaro, P. Francois, R. Fragassi, A. Simpson, and K.
              Patel, "Best Practices for Advertisement of Multiple Paths
              in IBGP", draft-ietf-idr-add-paths-guidelines-05.txt (work
              in progress).


Authors' Addresses

   Pierre Francois
   Institute IMDEA Networks
   Avda. del Mar Mediterraneo, 22
   Leganes  28918
   ES

   Email: pierre.francois@imdea.org








Pierre Francois, et al.  Expires August 11, 2014                [Page 5]

Internet-Draft          ADD-PATH limit capability          February 2014


   Camilo Cardona
   Institute IMDEA Networks
   Avda. del Mar Mediterraneo, 22
   Leganes  28918
   ES

   Email: juancamilo.cardona@imdea.org


   Adam Simpson
   Alcatel-Lucent
   600 March Road
   Ontario  K2K 2E6
   CA

   Email: adam.simpson@alcatel-lucent.com


   Jeffrey Haas
   Juniper Networks
   1194 N. Mathilda Ave
   Sunnyvale  94089
   USA

   Email: jhaas@juniper.net


























Pierre Francois, et al.  Expires August 11, 2014                [Page 6]