RFC 5243 INFORMATIONAL OSPF · ROUTING · OPTIMIZATION

OSPF Database Exchange Summary List Optimization

A backward-compatible way to send fewer OSPF database summary headers.

On rfc-editor.org ↗
PUBLISHED
May 2008
AUTHORS
R. Ogier
STREAM
IETF
UPDATED BY
RFC 9454 — OSPF terminology update
PROTOCOLS COVERED
OSPFv2 and OSPFv3
OVERHEAD REDUCTION
About 50% in large networks
DOI
10.17487/RFC5243

RFC text

Contents
Original text · RFC Editor View on rfc-editor.org ↗
Network Working Group                                         R. Ogier
Request for Comments: 5243                           SRI International
Category: Informational                                       May 2008


            OSPF Database Exchange Summary List Optimization

Status of This Memo

   This memo provides information for the Internet community.  It does
   not specify an Internet standard of any kind.  Distribution of this
   memo is unlimited.

Abstract

   This document describes a backward-compatible optimization for the
   Database Exchange process in OSPFv2 and OSPFv3.  In this
   optimization, a router does not list a Link State Advertisement (LSA)
   in Database Description packets sent to a neighbor, if the same or a
   more recent instance of the LSA was listed in a Database Description
   packet already received from the neighbor.  This optimization reduces
   Database Description overhead by about 50% in large networks.  This
   optimization does not affect synchronization, since it only omits
   unnecessary information from Database Description packets.

1.  Introduction

   In OSPFv2 [RFC2328] and OSPFv3 [RFC2740], when two neighboring
   routers become adjacent, they synchronize their link-state databases
   via the Database Exchange process.  Each router sends the other
   router a set of Database Description (DD) packets that describes the
   router's link-state database.  This is done by listing each LSA
   (i.e., including the header of each LSA) in one of the sent DD
   packets.  This procedure allows each router to determine whether the
   other router has newer LSA instances that should be requested via
   Link State Request packets.

   The optimization simply observes that it is not necessary for a
   router (master or slave) to list an LSA in a DD packet if it knows
   the neighbor already has an instance of the LSA that is the same or
   more recent (and therefore will not request the LSA).  To avoid
   listing such LSAs in DD packets, when an LSA is listed in a DD packet
   received from the neighbor, and the Database summary list for the
   neighbor has an instance of the LSA that is the same as or less
   recent than the one received, the LSA is removed from the summary
   list.

   The optimization, called the Database Exchange summary list
   optimization, does not affect synchronization, since the LSAs that
   are omitted from DD packets are unnecessary.  The optimization is
   fully backward compatible with OSPF.  The optimization reduces
   Database Description overhead by about 50% in large networks in which
   routers are usually already nearly synchronized when they become
   adjacent, since it reduces the total number of LSA headers exchanged
   by about one-half in such networks.  The optimization is especially
   beneficial in large networks with limited bandwidth, such as large
   mobile ad hoc networks.

2.  Specification of Optimization

   The Database Exchange summary list optimization is defined by
   modifying Section 10.6 "Receiving Database Description Packets" of
   RFC 2328 as follows.  The second-to-last paragraph of Section 10.6 is
   replaced with the following augmented paragraph:

   When the router accepts a received Database Description Packet as the
   next in sequence, the packet contents are processed as follows.  For
   each LSA listed, the LSA's LS type is checked for validity.  If the
   LS type is unknown (e.g., not one of the LS types 1-5 defined by this
   specification), or if this is an AS-external-LSA (LS type = 5) and
   the neighbor is associated with a stub area, generate the neighbor
   event SeqNumberMismatch and stop processing the packet.  Otherwise,
   the router looks up the LSA in its database to see whether it also
   has an instance of the LSA.  If it does not, or if the database copy
   is less recent, the LSA is put on the Link state request list so that
   it can be requested (immediately or at some later time) in Link State
   Request Packets.  In addition, if the Database summary list contains
   an instance of the LSA that is the same as or less recent than the
   listed LSA, the LSA is removed from the Database summary list.

   The above additional step (which updates the Database summary list)
   may be performed either before or after the router looks up the
   listed LSA in its database and possibly adds the LSA to the Link
   state request list.  However, to implement the optimization, the
   additional step must be performed for each LSA listed in the received
   DD packet (to fully update the Database summary list) before the next
   DD packet is sent in response.

   Although the optimization does not require that LSAs be listed in DD
   packets in any particular order, faster lookup of LSAs in the
   Database summary list may be possible if LSAs are listed in the same
   order by all routers.  If such an ordering is used, then to encourage
   different implementations to use the same ordering, this document
   recommends that LSAs be listed in lexicographically increasing order
   of (LS type, Link State ID, Advertising Router) for OSPFv2 and (LS
   type, Advertising Router, Link State ID) for OSPFv3.

3.  Example

   This section describes an example to illustrate the Database Exchange
   summary list optimization.  Assume that routers RT1 and RT2 already
   have identical databases when they start Database Exchange.  Also
   assume that the list of LSA headers for the database fits into two DD
   packets.  Then, the standard Database Exchange is as follows when RT1
   is the first to change the neighbor state to ExStart.  Note that each
   router sends two full DD packets.

      RT1 (slave)                                      RT2 (master)

      ExStart      Empty DD (Seq=x,I,M,Master)
                 ------------------------------>
                   Empty DD (Seq=y,I,M,Master)         ExStart
                 <------------------------------
      Exchange     Full  DD (Seq=y,M,Slave)
                 ------------------------------>
                   Full  DD (Seq=y+1,M,Master)         Exchange
                 <------------------------------
                   Full  DD (Seq=y+1,Slave)
                 ------------------------------>
                   Full  DD (Seq=y+2, Master)
                 <------------------------------
       Full        Empty DD (Seq=y+2, Slave)
                 ------------------------------>
                                                       Full

   If the optimization is used, when RT2 receives the first full DD
   packet from RT1, it removes from its summary list all LSAs that are
   listed in the DD packet.  Then RT2 sends a DD packet that lists the
   remaining LSAs (since all of the LSA headers fit into two DD
   packets).  When RT1 receives this DD packet, it removes these
   remaining LSAs from its summary list (causing it to be empty) and
   sends an empty DD packet to RT2.

   With the optimization, each router sends only one full DD packet
   instead of two, as shown below.

      RT1 (slave)                                      RT2 (master)

      ExStart      Empty DD (Seq=x,I,M,Master)
                 ------------------------------>
                   Empty DD (Seq=y,I,M,Master)         ExStart
                 <------------------------------
      Exchange     Full  DD (Seq=y,M,Slave)
                 ------------------------------>
                   Full  DD (Seq=y+1,Master)           Exchange
                 <------------------------------
       Full        Empty DD (Seq=y+1, Slave)
                 ------------------------------>
                                                       Full

4.  Security Considerations

   This document does not raise any new security concerns.

5.  IANA Considerations

   This document specifies a simple backward-compatible optimization for
   OSPFv2 and OSPFv3 that does not require any new number assignment.

6.  Normative References

   [RFC2328] Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.


   [RFC2740] Coltun, R., Ferguson, D., and J. Moy, "OSPF for IPv6", RFC
             2740, December 1999.

7.  Acknowledgments

   The recommendation to list LSAs in lexicographical order was
   suggested by Acee Lindem.

Author's Address

   Richard G. Ogier
   EMail: rich.ogier@earthlink.net

Full Copyright Statement

   Copyright (C) The IETF Trust (2008).

   This document is subject to the rights, licenses and restrictions
   contained in BCP 78, and except as set forth therein, the authors
   retain all their rights.

   This document and the information contained herein are provided on an
   "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS
   OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND
   THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS
   OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF
   THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED
   WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.

Intellectual Property

   The IETF takes no position regarding the validity or scope of any
   Intellectual Property Rights or other rights that might be claimed to
   pertain to the implementation or use of the technology described in
   this document or the extent to which any license under such rights
   might or might not be available; nor does it represent that it has
   made any independent effort to identify any such rights.  Information
   on the procedures with respect to rights in RFC documents can be
   found in BCP 78 and BCP 79.

   Copies of IPR disclosures made to the IETF Secretariat and any
   assurances of licenses to be made available, or the result of an
   attempt made to obtain a general license or permission for the use of
   such proprietary rights by implementers or users of this
   specification can be obtained from the IETF on-line IPR repository at
   http://www.ietf.org/ipr.

   The IETF invites any interested party to bring to its attention any
   copyrights, patents or patent applications, or other proprietary
   rights that may cover technology that may be required to implement
   this standard.  Please address the information to the IETF at
   ietf-ipr@ietf.org.

ISO Plain Language

On this page
PROBLEMDuring database exchange, neighboring OSPF routers can list LSA headers that the other router already has.
MECHANISMA router removes a summary entry when the neighbor's Database Description packet already listed the same or a more recent LSA.
WHY IT MATTERSThe optimization reduces Database Description overhead by about 50% in large, nearly synchronized networks. RFC 9454 later updated this memo.

§1The Database Exchange Optimization

When neighboring OSPF routers become adjacent, they synchronize their link-state databases through Database Description (DD) packets. Each packet lists link-state advertisement (LSA) headers so a router can tell whether its neighbor has newer information to request.

A router need not list an LSA if the neighbor already has the same or a more recent instance. The neighbor's DD packet shows this. The router then removes that LSA from the neighbor's Database summary list.

A router can skip describing an LSA when its neighbor has already listed the same or a newer copy. The neighbor's own packet tells it which entries to remove from its summary.

§2Updating the Summary List

When a router accepts the next Database Description packet, it checks each listed LSA. If its own database has no copy or has an older copy, the LSA is added to the Link State Request list. If the Database summary list contains an instance that is the same as or less recent than the one listed by the neighbor, that summary entry is removed.

The removal may happen before or after the database lookup, but it must be done for every LSA in the received packet before the router sends its next DD packet in response. The RFC recommends a common increasing order for LSAs to make lookups faster, with a different field order for OSPFv2 and OSPFv3.

For each LSA in the neighbor's packet, the router requests it only if its own copy is missing or older. It crosses off entries the neighbor already has before sending its next packet.

§3Example and Effects

The example starts with RT1 and RT2 holding identical databases whose LSA headers fit in two DD packets. Without the optimization, each router sends two full DD packets. With it, each sends one full DD packet because it removes headers already listed by the neighbor.

The RFC says the change is backward compatible and does not affect synchronization, since it omits only unnecessary information. It can reduce the number of LSA headers exchanged by about half in large networks where routers are usually nearly synchronized when they become adjacent, and is especially useful where bandwidth is limited.

In the example, each router sends one full packet instead of two. The RFC says this saves repeated headers without changing how the databases synchronize.

FULL DD PACKETS PER ROUTER IN THE RFC EXAMPLE
ExchangeFull DD packets per router
Standard2
With optimization THIS RFC1

§4References