OSPFv2, herein abbreviated simply as OSPF, is an IPv4 routing protocol documented in [Ref8]. OSPF is a link-state routing protocol. It is designed to be run internal to a single Autonomous System. Each OSPF router maintains an identical database describing the Autonomous System's topology. From this database, a routing table is calculated by constructing a shortest-path tree. OSPF features include the following:
o OSPF responds quickly to topology changes, expending a minimum of network bandwidth in the process.
o Support for CIDR addressing.
o OSPF routing exchanges can be authenticated, providing routing security.
o Equal-cost multipath.
o An area routing capability is provided, enabling an Autonomous system to be split into a two level hierarchy to further reduce the amount of routing protocol traffic.
o OSPF allows import of external routing information into the Autonomous System, including a tagging feature that can be exploited to exchange extra information at the AS boundary (see [Ref7]).
An analysis of OSPF together with a more detailed description of OSPF features was originally provided in [Ref6], as a part of promoting OSPF to Draft Standard status. The analysis of OSPF remains unchanged. Two additional major features have been developed for OSPF since the protocol achieved Draft Standard status: the Point-to-MultiPoint interface and Cryptographic Authentication. These features are described in Sections 2.1 and 2.2 respectively of this memo.
The OSPF MIB is documented in [Ref4]. It is currently at Draft Standard status.
OSPF became a Draft Standard with the release of RFC 1583 [Ref3]. Implementations of the new specification in [Ref8] are backward- compatible with RFC 1583. The differences between the two documents are described in the Appendix Gs of [Ref1] and [Ref8]. These differences are listed briefly below. Two major features were also added, the Point-to-MultiPoint interface and Cryptographic Authentication, which are described in separate sections.
o Configuration requirements for OSPF area address ranges have been relaxed to allow greater flexibility in area assignment. See Section G.3 of [Ref1] for details.
o The OSPF flooding algorithm was modified to a) improve database convergence in networks with low speed links b) resolve a problem where unnecessary LSA retransmissions could occur as a result of differing clock granularities, c) remove race conditions between the flooding of MaxAge LSAs and the Database Exchange process, d) clarify the use of the MinLSArrival constant, and e) rate-limit the response to less recent LSAs received via flooding. See Sections G.4 and G.5 of [Ref1] and Section G.1 of [Ref8] for details.
o To resolve the long-standing confusion regarding representation of point-to-point links in OSPF, the specification now optionally allows advertisement of a stub link to a point-to- point link's subnet, ala RIP. See Section G.6 of [Ref1].
o Several problems involving advertising the same external route from multiple areas were found and fixed, as described in Section G.7 of [Ref1] and Section G.2 of [Ref8]. Without the fixes, persistent routing loops could form in certain such configurations. Note that one of the fixes was not backward- compatible, in that mixing routers implementing the fixes with those implementing just RFC 1583 could cause loops not present in an RFC 1583-only configuration. This caused an RFC1583Compatibility global configuration parameter to be added, as described in Section C.1 of [Ref1].
Moy Informational [Page 3]
RFC 2329 OSPF Standardization Report April 1998
o In order to deal with high delay links, retransmissions of initial Database Description packets no longer reset an OSPF adjacency.
o In order to detect link MTU mismatches, which can cause problems both in IP forwarding and in the OSPF routing protocol itself, MTU was added to OSPF's Database Description packets. Neighboring routers refuse to bring up an OSPF adjacency unless they agree on their common link's MTU.
o The TOS routing option was deleted from OSPF. However, for backward compatibility the formats of OSPF's various LSAs remain unchanged, maintaining the ability to specify TOS metrics in router-LSAs, summary-LSAs, ASBR-summary-LSAs, and AS-external- LSAs.
o OSPF's routing table lookup algorithm was changed to reflect current practice. The "best match" routing table entry is now always selected to be the one providing the most specific (longest) match. See Section G.4 of [Ref8] for details.
The Point-to-MultiPoint interface was added as an alternative to OSPF's NBMA interface when running OSPF over non-broadcast subnets. Unlike the NBMA interface, Point-to-MultiPoint does not require full mesh connectivity over the non-broadcast subnet. Point-to-MultiPoint is less efficient than NBMA, but is easier to configure (in fact, it can be self-configuring) and is more robust than NBMA, tolerating all failures within the non- broadcast subnet. For more information on the Point-to- MultiPoint interface, see Section G.2 of [Ref1].
There are at least six independent implementations of the Point-to-MultiPoint interface. Interoperability has been demonstrated between at least two pairs of implementations: between 3com and Bay Networks, and between cisco and Cascade.
Non-trivial authentication was added to OSPF with the development of the Cryptographic Authentication type. This authentication type uses any keyed message digest algorithm, with explicit instructions included for the use of MD5. For more information on OSPF authentication, see Section 4.
There are at least three independent implementations of the OSPF Cryptographic authentication type. Interoperability has been demonstrated between the implementations from cisco and Cascade.
3. Updated implementation and deployment experience
When OSPF was promoted to Draft Standard Status, a report was issued documenting current implementation and deployment experience (see [Ref6]). That report is now quite dated. In an attempt to get more current data, a questionnaire was sent to OSPF mailing list in January 1996. Twelve responses were received, from 11 router vendors and 1 manufacturer of test equipment. These responses represented 6 independent implementations. A tabulation of the results are presented below.
Table 1 indicates the implementation, interoperability and deployment of the major OSPF functions. The number in each column represents the number of responses in the affirmative.
Table 2 indicates the size of the OSPF routing domains that vendors have tested. For each size parameter, the number of responders and the range of responses (minimum, mode, mean and maximum) are listed.
Parameter Responses Min Mode Mean Max _________________________________________________________________ Max routers in domain 7 30 240 460 1600 Max routers in single area 7 20 240 380 1600 Max areas in domain 7 1 10 16 60 Max AS-external-LSAs 9 50 10K 10K 30K
Table 2: OSPF domain sizes tested
Table 3 indicates the size of the OSPF routing domains that vendors have deployed in real networks. For each size parameter, the number of responders and the range of responses (minimum, mode, mean and maximum) are listed.
Moy Informational [Page 6]
RFC 2329 OSPF Standardization Report April 1998
Parameter Responses Min Mode Mean Max _________________________________________________________________ Max routers in domain 8 20 350 510 1000 Max routers in single area 8 20 100 160 350 Max areas in domain 7 1 15 23 60 Max AS-external-LSAs 6 50 1K 2K 5K
Table 3: OSPF domain sizes deployed
In an attempt to ascertain the extent to which OSPF is currently deployed, vendors were also asked in January 1998 to provide deployment estimates. Four vendors of OSPF routers responded, with a total estimate of 182,000 OSPF routers in service, organized into 4300 separate OSPF routing domains.
All OSPF protocol exchanges are authenticated. OSPF supports multiple types of authentication; the type of authentication in use can be configured on a per network segment basis. One of OSPF's authentication types, namely the Cryptographic authentication option, is believed to be secure against passive attacks and provide significant protection against active attacks. When using the Cryptographic authentication option, each router appends a "message digest" to its transmitted OSPF packets. Receivers then use the shared secret key and received digest to verify that each received OSPF packet is authentic.
The quality of the security provided by the Cryptographic authentication option depends completely on the strength of the message digest algorithm (MD5 is currently the only message digest algorithm specified), the strength of the key being used, and the correct implementation of the security mechanism in all communicating OSPF implementations. It also requires that all parties maintain the secrecy of the shared secret key.
None of the OSPF authentication types provide confidentiality. Nor do they protect against traffic analysis. Key management is also not addressed by the OSPF specification.
Moy Informational [Page 7]
RFC 2329 OSPF Standardization Report April 1998
For more information, see Sections 8.1, 8.2, and Appendix D of [Ref1].
[Ref1] Moy, J., "OSPF Version 2", RFC 2178, July 1997.
Copyright (C) The Internet Society (1998). All Rights Reserved.
This document and translations of it may be copied and furnished to others, and derivative works that comment on or otherwise explain it or assist in its implementation may be prepared, copied, published and distributed, in whole or in part, without restriction of any kind, provided that the above copyright notice and this paragraph are included on all such copies and derivative works. However, this document itself may not be modified in any way, such as by removing the copyright notice or references to the Internet Society or other Internet organizations, except as needed for the purpose of developing Internet standards in which case the procedures for copyrights defined in the Internet Standards process must be followed, or as required to translate it into languages other than English.
The limited permissions granted above are perpetual and will not be revoked by the Internet Society or its successors or assigns.
This document and the information contained herein is provided on an "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING TASK FORCE DISCLAIMS 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.