RFC 1364
This document is obsolete. Please refer to RFC 1403.






Network Working Group                                        K. Varadhan
Request for Comments: 1364                                        OARnet
                                                          September 1992


                          BGP OSPF Interaction

Status of this Memo



   This RFC specifies an IAB standards track protocol for the Internet
   community, and requests discussion and suggestions for improvements.
   Please refer to the current edition of the "IAB Official Protocol
   Standards" for the standardization state and status of this protocol.
   Distribution of this memo is unlimited.

Abstract



   This memo defines the various criteria to be used when designing
   Autonomous System Border Routers (ASBR) that will run BGP with other
   ASBRs external to the AS and OSPF as its IGP.

Table of Contents



   1.  Introduction .................................................  2
   2.  Route Exchange ...............................................  2
   2.1.  Exporting OSPF routes into BGP .............................  3
   2.2.  Importing BGP routes into OSPF .............................  4
   3.  BGP Identifier and OSPF router ID ............................  5
   4.  Setting OSPF tags, BGP ORIGIN and AS_PATH attributes .........  5
   4.1.  Semantics of the characteristics bits ......................  7
   4.2.  Configuration parameters for setting the OSPF tag ..........  8
   4.3.  Manually configured tags ...................................  9
   4.4.  Automatically generated tags ................................ 9
   4.4.1.  Routes with incomplete path information, pl = 0 ........... 9
   4.4.2.  Routes with incomplete path information, pl = 1 ........... 9
   4.4.3.  Routes with incomplete path information, pl >= 1 ..........10
   4.4.4.  Routes with complete path information, pl = 0 .............10
   4.4.5.  Routes with complete path information, pl = 1 .............11
   4.4.6.  Routes with complete path information, pl >= 1 ............11
   4.5.  Miscellaneous tag settings ..................................12
   4.6.  Summary of the TagType field setting ........................12
   5.  Setting OSPF Forwarding Address and BGP NEXT_HOP attribute ....12
   6.  Security Considerations .......................................13
   7.  Acknowledgements ..............................................13
   8.  Bibliography ..................................................14
   9.  Author's Address ..............................................14





Varadhan                                                        [Page 1]

RFC 1364                  BGP OSPF Interaction            September 1992


1.  Introduction



   This document defines the various criteria to be used when designing
   Autonomous System Border Routers (ASBR) that will run BGP [RFC1267]
   with other ASBRs external to the AS, and OSPF [RFC1247] as its IGP.

   This document defines how the following fields in OSPF and attributes
   in BGP are to be set when interfacing between BGP and OSPF at an
   ASBR:

      OSPF cost and type      vs. BGP INTER-AS METRIC
      OSPF tag                vs. BGP ORIGIN and AS_PATH
      OSPF Forwarding Address vs. BGP NEXT_HOP

   For a more general treatise on routing and route exchange problems,
   please refer to [ROUTE-LEAKING] and [NEXT-HOP] by Philip Almquist.

   This document uses the two terms "Autonomous System" and "Routing
   Domain".  The definitions for the two are below:

   The term Autonomous System is the same as is used in the BGP-3 RFC
   [RFC1267], given below:

      "The use of the term Autonomous System here stresses the fact
      that, even when multiple IGPs and metrics are used, the
      administration of an AS appears to other ASs to have a single
      coherent interior routing plan and presents a consistent picture
      of what networks are reachable through it.  From the standpoint of
      exterior routing, an AS can be viewed as monolithic: reachability
      to networks directly connected to the AS must be equivalent from
      all border gateways of the AS."

   The term Routing Domain was first used in [ROUTE-LEAKING] and is
   given below:

      "A Routing Domain is a collection of routers which coordinate
      their routing knowledge using a single (instance of) a routing
      protocol."

2.  Route Exchange



   This section discusses the constraints that must be met to exchange
   routes between an external BGP session with a peer from another AS
   and internal OSPF routes.

   BGP does not carry subnet information in routing updates.  Therefore,
   when referring to a subnetted network in the OSPF routing domain, we
   consider the equivalent network route in the context of BGP.



Varadhan                                                        [Page 2]

RFC 1364                  BGP OSPF Interaction            September 1992


   Multiple subnet routes for a subnetted network in OSPF are collapsed
   into one network route when exported into BGP.

2.1.  Exporting OSPF routes into BGP



  1.   The administrator must be able to selectively export routes
       into BGP via an appropriate filter mechanism.

       This filter mechanism must support such control with the
       granularity of a single network.

       Additionally, the administrator must be able to filter based
       on the OSPF tag and the various sub-fields of the OSPF tag.
       The settings of the tag and the sub-fields are defined in
       section 4 in more detail.

       o    By default, no routes must be exported from OSPF into
            BGP.  A single mechanism must permit all OSPF inter-area
            and intra-area routes to be exported into BGP.

       OSPF external routes of type 1 and type 2 must never be
       exported into BGP unless they are explicitly configured.

  2.   When configured to export a network, the ASBR must advertise
       a network route for a subnetted network, as long as at least
       one subnet in the subnetted network is reachable via OSPF.

  3.   The network administrator must be able to statically
       configure the BGP attribute INTER-AS METRIC to be used for
       any network route.



       o    By default, the INTER_AS METRIC must default to 1.

       Explanatory text: The OSPF cost and the BGP INTER-AS METRIC
       are of different widths.  The OSPF cost is a two level
       metric.  The BGP INTER-AS METRIC is only an optional non-
       transitive attribute.  Hence, a more complex BGP INTER-AS
       METRIC-OSPF cost mapping scheme is not necessary.

   4.   When an ASBR is advertising an OSPF route to network Y to
        external BGP neighbours and learns that the route has become
        unreachable, the ASBR must immediately propogate this
        information to the external BGP neighbours.

   5.   An implementation of BGP and OSPF on an ASBR must have a
        mechanism to set up a minimum amount of time that must elapse
        between the learning of a new route via OSPF and subsequent
        advertisement of the route via BGP to the external



Varadhan                                                        [Page 3]

RFC 1364                  BGP OSPF Interaction            September 1992


        neighbours.

        o    The default value for this setting must be 0, indicating
             that the route is to be advertised to the neighbour BGP
             peers instantly.

             Note that [RFC1267] mandates a mechanism to dampen the
             inbound advertisements from adjacent neighbours.

2.2.  Importing BGP routes into OSPF



   1.   BGP implementations should allow an AS to control
        announcements of BGP-learned routes into OSPF.
        Implementations should support such control with the
        granularity of a single network.  Implementations should also
        support such control with the granularity of an autonomous
        system, where the autonomous system may be either the
        autonomous system that originated the route or the autonomous
        system that advertised the route to the local system
        (adjacent autonomous system).

         o    By default, no routes must be imported from BGP into
              OSPF.  Administrators must configure every route they
              wish to import.

              A mechanism may allow an administrator to configure an
              ASBR to import all the BGP routes into the OSPF routing
              domain.

   2.   The administrator must be able to configure the OSPF cost and
        the OSPF metric type of every route imported into OSPF.

        o    The OSPF cost must default to 1; the OSPF metric type
             must default to type 2.

   3.   Routes learned via IBGP must not be imported into OSPF.

   4.   The ASBR must never generate a default route into the OSPF
        routing domain unless explicitly configured to do so.



        A possible criterion for generating default into an IGP is to
        allow the administrator to specify a set of (network route,
        AS_PATH, default route cost, default route type) tuples.  If
        the ASBR learns of the network route for an element of the
        set, with the corresponding AS_PATH, then it generates a
        default route into the OSPF routing domain, with cost
        "default route cost" and type, "default route type".  The
        lowest cost default route will then be injected into the OSPF



Varadhan                                                        [Page 4]

RFC 1364                  BGP OSPF Interaction            September 1992


        routing domain.

        This is the recommended method for originating default routes
        in the OSPF routing domain.

3.  BGP Identifier and OSPF router ID

   The BGP identifier must be the same as the OSPF router id at all
   times that the router is up.

   This characteristic is required for two reasons.

      i.   Consider the scenario in which 3 routers, RT1, RT2, and RT3,
           belong to the same autonomous system.

                            +-----+
                            | RT3 |
                            +-----+
                               |

                Autonomous System running OSPF

                        /             \
                    +-----+          +-----+
                    | RT1 |          | RT2 |
                    +-----+          +-----+

   Both RT1 and RT2 have routes to an external network X and import it
   into the OSPF routing domain.  RT3 is advertising the route to
   network X to other external BGP speakers.  RT3 must use the OSPF
   router ID to determine whether it is using RT1 or RT2 to forward
   packets to network X and hence build the correct AS_PATH to advertise
   to other external speakers.

   More precisely, RT3 must use the AS_PATH of the route announced by
   the ASBR, whose BGP Identifier is the same as the OSPF routerID
   corresponding to its route for network X.

      ii.  It will be convenient for the network administrator looking at
           an ASBR to correlate different BGP and OSPF routes based on
           the identifier.

4.  Setting OSPF tags, BGP ORIGIN and AS_PATH attributes

   The OSPF external route tag is a "32-bit field attached to each
   external route . . . It may be used to communicate information
   between AS boundary routers; the precise nature of such information
   is outside the scope of [the] specification." [RFC1247]



Varadhan                                                        [Page 5]

RFC 1364                  BGP OSPF Interaction            September 1992


   OSPF imports information from various routing protocols at all its
   ASBRs.  In some instances, it is possible to use protocols other than
   EGP or BGP across autonomous systems.  It is important, in BGP, to
   differentiate between routes that are external to the OSPF routing
   domain but must be considered internal to the AS, as opposed to
   routes that are external to the AS.

   Routes that are internal to the AS and that may or may not be
   external to the OSPF routing domain will not come to the various BGP
   speakers via IBGP.  Therefore, ASBRs running BGP must have knowledge
   of this class of routes so that they can advertise these routes to
   the various external AS without waiting for IBGP updates about these
   routes.

   Additionally, in the specific instance of an AS intermixing routers
   running EGP and BGP as external gateway routing protocols, using OSPF
   as an IGP, the network administrator does not have to configure IBGP
   on every ASBR running EGP and not running BGP, if this information
   can be carried in the OSPF tag field.

   We use the external route tag field in OSPF to intelligently set the
   ORIGIN and AS_PATH attributes in BGP.  Both the ORIGIN and AS_PATH
   attributes are well-known, mandatory attributes in BGP.  The exact
   mechanism for setting the tags is defined below.

   The tag is broken up into sub-fields shown below.  The various sub-
   fields specify the characteristics of the route imported into the
   OSPF routing domain.

   The high bit of the OSPF tag is known as the "Automatic" bit.  When
   this bit is set to 1, the following sub-fields apply:

       0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3
       0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
      |a|c|p l|     ArbitraryTag      |       AutonomousSystem        |
      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   a    is 1 bit called the Automatic bit, indicating that the
        Completeness and PathLength bits have been generated
        automatically by a router.  The meaning of this characteristic
        and its setting are defined below.

   c    is 1 bit of Completeness information.  The meaning of this
        characteristic and its settings are defined below.

   pl   are 2 bits of PathLength information.  The meaning of this
        characteristic and its setting are defined below.



Varadhan                                                        [Page 6]

RFC 1364                  BGP OSPF Interaction            September 1992


   ArbitraryTag (or "at")
        is 12 bits of tag information, which defaults to 0 but can be
        configured to anything else.

   AutonomousSystem (or "as")
        is 16 bits, indicating the AS number corresponding to the
        route, 0 if the route is to be considered as part of the local
        AS.

   When the Automatic bit is set to 0, the following sub-fields apply:

       0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3
       0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
      |a|                          LocalInfo                          |
      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   a    is 1 bit called the Automatic bit, set to 0.

   LocalInfo (or "li")
        is 31 bits of an arbitrary value, manually configured by the
        network administrator.

   The format of the tag for various values of the characteristics bits
   is defined below.

4.1.  Semantics of the characteristics bits



   The Completeness and PathLength characteristics bits define the
   characteristic of the route imported into OSPF from other ASBRs in
   the autonomous system.  This setting is then used to set the ORIGIN
   and NEXT_HOP attributes when re-exporting these routes to an external
   BGP speaker.

   o    The "a" bit or the Automatic characteristic bit is set when
        the Completeness and PathLength characteristics bits are
        automatically set by a border router.

        For backward compatibility, the Automatic bit must default to
        0 and the network administrator must have a mechanism to
        enable automatic tag generation.  Nothing must be inferred
        about the characteristics of the OSPF route from the tag
        bits, unless the tag has been automatically generated.

   o    The "c" bit of the Completeness characteristic bit is set
        when the source of the incoming route is known precisely, for
        instance, from an IGP within the local autonomous system or
        EGP at one of the autonomous system's boundaries.  It refers



Varadhan                                                        [Page 7]

RFC 1364                  BGP OSPF Interaction            September 1992


        to the status of the path information carried by the routing
        protocol.

   o    The "pl" or the PathLength characteristic sub-field is set
        depending on the length of the AS_PATH that the protocol
        could have carried when importing the route into the OSPF
        routing domain.  The length bits will indicate whether the
        AS_PATH attribute for the length is zero, one, or greater
        than one.

        Routes imported from an IGP will usually have an AS_PATH of
        length of 0, routes imported from an EGP will have an AS_PATH
        of length 1, BGP and routing protocols that support complete
        path information, either as AS_PATHs or routing domain paths,
        will indicate a path greater than 1.

        The OSPF tag is not wide enough to carry path information
        about routes that have an associated PathLength greater than
        one.  Path information about these routes will have to be
        carried via IBGP.  Such routes must not be exported from OSPF
        into BGP.

   For brevity in the following sections, the keywords O and P refer to
   the BGP ORIGIN and AS_PATH attributes respectively.  Likewise, we use
   the abbreviations , "l" and "nh" for the local_AS and next_hop_AS
   respectively in the following sections.

4.2.  Configuration parameters for setting the OSPF tag



   o    There must be a mechanism to enable automatic generation of
        the tag characteristic bits.

   o    Configuration of an ASBR running OSPF must include the
        capability to associate a tag value, for the ArbitraryTag, or
        LocalInfo sub-field of the OSPF tag, with each instance of a
        routing protocol.

   o    Configuration of an ASBR running OSPF must include the
        capability to associate an AS number with each instance of a
        routing protocol.

        Associating an AS number with an instance of an IGP is
        equivalent to flagging those set of routes imported from the
        IGP to be external routes outside the local autonomous
        system.

        Specifically, when the IGP is RIP [RFC1058], it should be
        possible to associate a tag and/or an AS number with every



Varadhan                                                        [Page 8]

RFC 1364                  BGP OSPF Interaction            September 1992


        interface running RIP on the ASBR.

4.3.  Manually configured tags



    0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |0|                          LocalInfo                          |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   This tag setting corresponds to the administrator manually setting
   the tag bits.  Nothing shall be inferred about the characteristics of
   the route corresponding to this tag setting.

   For backward compatibility with existing implementations of OSPF
   currently deployed in the field, this must be the default setting for
   importing routes into the OSPF routing domain.  There must be a
   mechanism to enable automatic tag generation for imported routes.

   The OSPF tag to BGP attribute mappings for these routes must be
   a=0, li=Arbitrary_Value  =>  O=<INCOMPLETE>, P=<l>

4.4.  Automatically generated tags



   4.4.1.  Routes with incomplete path information, pl = 0.



     0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3
     0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    |1|0|0|0|     ArbitraryTag      |       AutonomousSystem        |
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

      These are routes imported from routing protocols with incomplete
      path information and cannot or may not carry the neighbour AS or
      AS path as part of the routing information.

      The OSPF tag to BGP attribute mappings for these routes must be
                   a=1,c=0,pl=00,as=0 => O=<EGP>, P=<l>

   4.4.2  Routes with incomplete path information, pl = 1.



        0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3
        0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
       +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
       |1|0|0|1|     ArbitraryTag      |       AutonomousSystem        |
       +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

      These are routes imported from routing protocols with incomplete



Varadhan                                                        [Page 9]

RFC 1364                  BGP OSPF Interaction            September 1992


      path information and carry the neighbour AS as part of the routing
      information.

      The OSPF tag to BGP attribute mappings for these routes must be
                a=1,c=0,pl=01,as=nh => O=<EGP>, P=<l, nh>

      This setting should be used for importing EGP routes into the OSPF
      routing domain.  This setting can also be used when importing BGP
      routes whose origin=<EGP> and AS_PATH=<nh>; if the BGP learned
      route has no other transitive attributes, then its propogation via
      IBGP can be suppressed.

   4.4.3.  Routes with incomplete path information, pl >= 1.



        0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3
        0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
       +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
       |1|0|1|0|     ArbitraryTag      |       AutonomousSystem        |
       +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

      These are routes imported from routing protocols with truncated
      path information.

      The OSPF tag to BGP attribute mappings for these routes must be
                        a=1,c=0,pl=10,as=don't care

      These are imported by a border router, which is running BGP to a
      stub domain, and not running IBGP to other ASBRs.  This causes a
      truncation of the AS_PATH.  These routes must not be re-exported
      into BGP at another ASBR.

   4.4.4.  Routes with complete path information, pl = 0.



        0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3
        0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
       +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
       |1|1|0|0|     ArbitraryTag      |       AutonomousSystem        |
       +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

      These are routes imported from routing protocols with either
      complete path information or are known to be complete through
      means other than that carried by the routing protocol.

      The OSPF tag to BGP attribute mappings for these routes must be
                   a=1,c=1,pl=00,as=0 => O=<IGP>, P=<l>

      This should be used for importing routes into OSPF from an IGP.




Varadhan                                                       [Page 10]

RFC 1364                  BGP OSPF Interaction            September 1992


   4.4.5.  Routes with complete path information, pl = 1.



        0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3
        0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
       +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
       |1|1|0|1|     ArbitraryTag      |       AutonomousSystem        |
       +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

      These are routes imported from routing protocols with either
      complete path information, or are known to be complete through
      means other than that carried by the routing protocol.  The
      routing protocol also has additional information about the
      neighbour AS of the route.

      The OSPF tag to BGP attribute mappings for these routes must be
                 a=1,c=1,pl=01,as=nh => O=<IGP>, P=<l, nh>

      This setting should be used when the administrator explicitly
      associates an AS number with an instance of an IGP.  This setting
      can also be used when importing BGP routes whose origin=<IGP> and
      AS_PATH=<nh>; if the BGP learned route has no other transitive
      attributes, then its propogation via IBGP can be suppressed.

   4.4.6.  Routes with complete path information, pl >= 1.



        0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3
        0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
       +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
       |1|1|1|0|     ArbitraryTag      |       AutonomousSystem        |
       +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

      These are routes imported from routing protocols with complete
      path information and carry the AS path information as part of the
      routing information.

      The OSPF tag must be set to
                        a=1,c=1,pl=10,as=don't care

      These routes must not be exported into BGP because these routes
      are already imported from BGP into the OSPF RD.  Hence, it is
      assumed that the BGP speaker will convey this information to other
      BGP speakers via IBGP.

      Note that an implementation may import BGP routes with a path
      length of 1 and no other transitive attributes directly into OSPF
      and not send these routes via IBGP.  In this situation, it must
      use tag settings corresponding to 4.1.2.2, or 4.1.2.5.




Varadhan                                                       [Page 11]

RFC 1364                  BGP OSPF Interaction            September 1992


4.5.  Miscellaneous tag settings



     0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3
     0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    |1|x|1|1|              Reserved  for  future  use               |
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   The value of pl = 3 is reserved during automatic tag generation.
   Routers must not generate such a tag when importing routes into the
   OSPF routing domain.  ASBRs must ignore tags which indicate a pl = 3.

4.6.  Summary of the tag sub-field setting



   The following table summarises the various combinations of automatic
   tag settings for the Completeness and PathLength sub-field of the
   OSPF tag and the default behaviour permitted for each setting.

      Completeness := 0 | 1 ;
      PathLength := 00 | 01 | 10 | 11;
      ORIGIN := <INCOMPLETE> | <IGP> | <EGP>;

               AS_PATH := valid AS path settings as defined in BGP.

           pl = 00       pl = 01            pl = 10        pl = 11
       +----------------------------------------------------------------
       |
c = 0  |    <EGP><l>    <EGP><l,nh>       never export       reserved
c = 1  |    <IGP><l>    <IGP><l,nh>       out of band        reserved
       |

      The "out of band" in the table above implies that OSPF will not be
      able to carry everything that BGP needs in its routing
      information.  Therefore, some other means must be found to carry
      this information.  In BGP, this is done via IBGP.

5.  Setting OSPF Forwarding Address and BGP NEXT_HOP attribute



   Forwarding addresses are used to avoid extra hops between multiple
   routers that share a common network and that speak different routing
   protocols with each other.

   Both BGP and OSPF have equivalents of forwarding addresses.  In BGP,
   the NEXT_HOP attribute is a well-known, mandatory attribute.  OSPF
   has a Forwarding address field.  We will discuss how these are to be
   filled in various situations.

   Consider the 4 router situation below:



Varadhan                                                       [Page 12]

RFC 1364                  BGP OSPF Interaction            September 1992


   RT1 and RT2 are in one autonomous system, RT3 and RT4 are in another.
   RT1 and RT3 are talking BGP with each other.
   RT3 and RT4 are talking OSPF with each other.

           +-----+                 +-----+
           | RT1 |                 | RT2 |
           +-----+                 +-----+
              |                       |            common network
           ---+-----------------------+--------------------------
               <BGP> |                       |
                  +-----+     <OSPF>      +-----+
                  | RT3 |                 | RT4 |
                  +-----+                 +-----+


   - Importing network X to OSPF:

        Consider an external network X, learnt via BGP from RT1.

        RT3 must always fill the OSPF Forwarding Address with the BGP

        NEXT_HOP attribute for the route to network X.

   - Exporting network Y to BGP:

        Consider a network Y, internal to the OSPF routing domain,
        RT3's route to network Y is via RT4, and network Y is to be
        exported via BGP to RT1.

        If network Y is not a subnetted network, RT3 must fill the
        NEXT_HOP attribute for network Y with the address of RT4.
        This is to avoid requiring packets to take an extra hop
        through RT3 when traversing the AS boundary.  This is similar
        to the concept of indirect neighbour support in EGP [RFC888,
        RFC827].

6.  Security Considerations



   Security considerations are not discussed in this memo.

7.  Acknowledgements



   I would like to thank Yakov Rekhter, Jeff Honig, John Moy, Tony Li,
   and Dennis Ferguson for their help and suggestions in writing this
   document, without which I could not have written this document.  I
   would also like to thank them for giving me the opportunity to write
   this document, and putting up with my muddlements through various
   phases of this document.



Varadhan                                                       [Page 13]

RFC 1364                  BGP OSPF Interaction            September 1992


   I would also like to thank the countless number of people from the
   OSPF and BGP working groups who have offered numerous suggestions and
   comments on the different stages of this document.

8.  Bibliography



   [RFC827]  Rosen, E., "Exterior Gateway Protocol (EGP)", RFC 827,
             BBN, October 1982.

   [RFC888]  Seamonson, L., and E. Rosen, "STUB Exterior Gateway
             Protocol", RFC 888, BBN, January 1984.

   [RFC1058]  Hedrick, C., "Routing Information Protocol", RFC 1058,
              Rutgers University, June 1988.

   [RFC1267]  Lougheed, K., and Y. Rekhter, "A Border Gateway
              Protocol 3 (BGP-3)", RFC 1267, cisco Systems,
              T.J. Watson Research Center, IBM Corp., October 1991.

   [RFC1268]  Rekhter, Y., and P. Gross, Editors, "Application of the
              Border Gateway Protocol in the Internet", RFC 1268,
              T.J. Watson Research Center, IBM Corp., ANS, October 1991.

   [RFC1247]  Moy, J., "The OSPF Specification - Version 2:", RFC 1247,
              Proteon, January 1991.

   [ROUTE-LEAKING]  Almquist, P., "Ruminations on Route Leaking",
                    Work in Progress.

   [NEXT-HOP]  Almquist, P., "Ruminations on the Next Hop",
               Work in Progress.

9.  Author's  Address



   Kannan Varadhan
   Internet Engineer, OARnet
   1224 Kinnear Road
   Columbus, OH 43212-1136

   EMail: kannan@oar.net











Varadhan                                                       [Page 14]