RFC 3219






Network Working Group                                       J. Rosenberg
Request for Comments: 3219                                   dynamicsoft
Category: Standards Track                                      H. Salama
                                                           Cisco Systems
                                                               M. Squire
                                                       Hatteras Networks
                                                            January 2002


                    Telephony Routing over IP (TRIP)

Status of this Memo



   This document specifies an Internet standards track protocol for the
   Internet community, and requests discussion and suggestions for
   improvements.  Please refer to the current edition of the "Internet
   Official Protocol Standards" (STD 1) for the standardization state
   and status of this protocol.  Distribution of this memo is unlimited.

Copyright Notice



   Copyright (C) The Internet Society (2002).  All Rights Reserved.

Abstract



   This document presents the Telephony Routing over IP (TRIP).  TRIP is
   a policy driven inter-administrative domain protocol for advertising
   the reachability of telephony destinations between location servers,
   and for advertising attributes of the routes to those destinations.
   TRIP's operation is independent of any signaling protocol, hence TRIP
   can serve as the telephony routing protocol for any signaling
   protocol.

   The Border Gateway Protocol (BGP-4) is used to distribute routing
   information between administrative domains.  TRIP is used to
   distribute telephony routing information between telephony
   administrative domains.  The similarity between the two protocols is
   obvious, and hence TRIP is modeled after BGP-4.

Table of Contents



   1    Terminology and Definitions  ..............................   3
   2    Introduction  .............................................   4
   3    Summary of Operation  .....................................   5
   3.1  Peering Session Establishment and Maintenance  ............   5
   3.2  Database Exchanges  .......................................   6
   3.3  Internal Versus External Synchronization  .................   6
   3.4  Advertising TRIP Routes  ..................................   6



Rosenberg, et. al.          Standards Track                     [Page 1]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   3.5  Telephony Routing Information Bases  ......................   7
   3.6  Routes in TRIP  ...........................................   9
   3.7  Aggregation  ..............................................   9
   4    Message Formats  ..........................................  10
   4.1  Message Header Format  ....................................  10
   4.2  OPEN Message Format  ......................................  11
   4.3  UPDATE Message Format  ....................................  15
   4.4  KEEPALIVE Message Format   ................................  22
   4.5  NOTIFICATION Message Format   .............................  23
   5    TRIP Attributes   .........................................  24
   5.1  WithdrawnRoutes  ..........................................  24
   5.2  ReachableRoutes  ..........................................  28
   5.3  NextHopServer   ...........................................  29
   5.4  AdvertisementPath   .......................................  31
   5.5  RoutedPath  ...............................................  35
   5.6  AtomicAggregate   .........................................  36
   5.7  LocalPreference   .........................................  37
   5.8  MultiExitDisc  ............................................  38
   5.9  Communities  ..............................................  39
   5.10 ITAD Topology    ..........................................  41
   5.11 ConvertedRoute  ...........................................  43
   5.12 Considerations for Defining New TRIP Attributes   .........  44
   6    TRIP Error Detection and Handling   .......................  44
   6.1  Message Header Error Detection and Handling   .............  45
   6.2  OPEN Message Error Detection and Handling   ...............  45
   6.3  UPDATE Message Error Detection and Handling   .............  46
   6.4  NOTIFICATION Message Error Detection and Handling   .......  48
   6.5  Hold Timer Expired Error Handling   .......................  48
   6.6  Finite State Machine Error Handling   .....................  48
   6.7  Cease   ...................................................  48
   6.8  Connection Collision Detection   ..........................  48
   7    TRIP Version Negotiation   ................................  49
   8    TRIP Capability Negotiation   .............................  50
   9    TRIP Finite State Machine   ...............................  50
   10   UPDATE Message Handling   .................................  55
   10.1 Flooding Process   ........................................  56
   10.2 Decision Process   ........................................  58
   10.3  Update-Send Process   ..................................... 62
   10.4  Route Selection Criteria   ................................ 67
   10.5  Originating TRIP Routes   ................................. 67
   11    TRIP Transport   .......................................... 68
   12    ITAD Topology   ........................................... 68
   13    IANA Considerations  ...................................... 68
   13.1  TRIP Capabilities   ....................................... 68
   13.2  TRIP Attributes    ........................................ 69
   13.3  Destination Address Families   ............................ 69
   13.4  TRIP Application Protocols   .............................. 69
   13.5  ITAD Numbers   ............................................ 70



Rosenberg, et. al.          Standards Track                     [Page 2]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   14    Security Considerations   ................................. 70
   A1    Appendix 1: TRIP FSM State Transitions and Actions   ...... 71
   A2    Appendix 2: Implementation Recommendations   .............. 73
   Acknowledgments  ................................................ 75
   References  ..................................................... 75
   Intellectual Property Notice  ................................... 77
   Authors' Addresses  ............................................. 78
   Full Copyright Statement  ....................................... 79

1. Terminology and Definitions



   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
   document are to be interpreted as described in RFC 2119 [1].

   A framework for Telephony Routing over IP (TRIP) is described in [2].
   We assume the reader is familiar with the framework and terminology
   of [2].  We define and use the following terms in addition to those
   defined in [2].

   Telephony Routing Information Base (TRIB): The database of reachable
   telephony destinations built and maintained at an LS as a result of
   its participation in TRIP.

   IP Telephony Administrative Domain (ITAD): The set of resources
   (gateways, location servers, etc.) under the control of a single
   administrative authority.  End users are customers of an ITAD.

   Less/More Specific Route: A route X is said to be less specific than
   a route Y if every destination in Y is also a destination in X, and X
   and Y are not equal.  In this case, Y is also said to be more
   specific than X.

   Aggregation: Aggregation is the process by which multiple routes are
   combined into a single less specific route that covers the same set
   of destinations.  Aggregation is used to reduce the size of the TRIB
   being synchronized with peer LSs by reducing the number of exported
   TRIP routes.

   Peers: Two LSs that share a logical association (a transport
   connection).  If the LSs are in the same ITAD, they are internal
   peers.  Otherwise, they are external peers.  The logical association
   between two peer LSs is called a peering session.








Rosenberg, et. al.          Standards Track                     [Page 3]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   Telephony Routing Information Protocol (TRIP): The protocol defined
   in this specification.  The function of TRIP is to advertise the
   reachability of telephony destinations, attributes associated with
   the destinations, as well as the attributes of the path towards those
   destinations.

   TRIP destination: TRIP can be used to manage routing tables for
   multiple protocols (SIP, H323, etc.).  In TRIP, a destination is the
   combination of (a) a set of addresses (given by an address family and
   address prefix), and (b) an application protocol (SIP, H323, etc).

2. Introduction



   The gateway location and routing problem has been introduced in [2].
   It is considered one of the more difficult problems in IP telephony.
   The selection of an egress gateway for a telephony call, traversing
   an IP network towards an ultimate destination in the PSTN, is driven
   in large part by the policies of the various parties along the path,
   and by the relationships established between these parties.  As such,
   a global directory of egress gateways in which users look up
   destination phone numbers is not a feasible solution.  Rather,
   information about the availability of egress gateways is exchanged
   between providers, and subject to policy, made available locally and
   then propagated to other providers in other ITADs, thus creating
   routes towards these egress gateways.  This would allow each provider
   to create its own database of reachable phone numbers and the
   associated routes - such a database could be very different for each
   provider depending on policy.

   TRIP is an inter-domain (i.e., inter-ITAD) gateway location and
   routing protocol.  The primary function of a TRIP speaker, called a
   location server (LS), is to exchange information with other LSs.
   This information includes the reachability of telephony destinations,
   the routes towards these destinations, and information about gateways
   towards those telephony destinations residing in the PSTN.  The TRIP
   requirements are set forth in [2].

   LSs exchange sufficient routing information to construct a graph of
   ITAD connectivity so that routing loops may be prevented.  In
   addition, TRIP can be used to exchange attributes necessary to
   enforce policies and to select routes based on path or gateway
   characteristics.  This specification defines TRIP's transport and
   synchronization mechanisms, its finite state machine, and the TRIP
   data.  This specification defines the basic attributes of TRIP.  The
   TRIP attribute set is extendible, so additional attributes may be
   defined in future documents.





Rosenberg, et. al.          Standards Track                     [Page 4]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   TRIP is modeled after the Border Gateway Protocol 4 (BGP-4) [3] and
   enhanced with some link state features, as in the Open Shortest Path
   First (OSPF) protocol [4], IS-IS [5], and the Server Cache
   Synchronization Protocol (SCSP) [6].  TRIP uses BGP's inter-domain
   transport mechanism, BGP's peer communication, BGP's finite state
   machine, and similar formats and attributes as BGP.  Unlike BGP
   however, TRIP permits generic intra-domain LS topologies, which
   simplifies configuration and increases scalability in contrast to
   BGP's full mesh requirement of internal BGP speakers.  TRIP uses an
   intra-domain flooding mechanism similar to that used in OSPF [4],
   IS-IS [5], and SCSP [6].

   TRIP permits aggregation of routes as they are advertised through the
   network.  TRIP does not define a specific route selection algorithm.

   TRIP runs over a reliable transport protocol.  This eliminates the
   need to implement explicit fragmentation, retransmission,
   acknowledgment, and sequencing.  The error notification mechanism
   used in TRIP assumes that the transport protocol supports a graceful
   close, i.e., that all outstanding data will be delivered before the
   connection is closed.

   TRIP's operation is independent of any particular telephony signaling
   protocol.  Therefore, TRIP can be used as the routing protocol for
   any of these protocols, e.g., H.323 [7] and SIP [8].

   The LS peering topology is independent of the physical topology of
   the network.  In addition, the boundaries of an ITAD are independent
   of the boundaries of the layer 3 routing autonomous systems.  Neither
   internal nor external TRIP peers need to be physically adjacent.

3. Summary of Operation



   This section summarizes the operation of TRIP.  Details are provided
   in later sections.

3.1. Peering Session Establishment and Maintenance



   Two peer LSs form a transport protocol connection between one
   another.  They exchange messages to open and confirm the connection
   parameters, and to negotiate the capabilities of each LS as well as
   the type of information to be advertised over this connection.

   KeepAlive messages are sent periodically to ensure adjacent peers are
   operational.  Notification messages are sent in response to errors or
   special conditions.  If a connection encounters an error condition, a
   Notification message is sent and the connection is closed.




Rosenberg, et. al.          Standards Track                     [Page 5]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


3.2. Database Exchanges



   Once the peer connection has been established, the initial data flow
   is a dump of all routes relevant to the new peer (In the case of an
   external peer, all routes in the LS's Adj-TRIB-Out for that external
   peer.  In the case of an internal peer, all routes in the Ext-TRIB
   and all Adj-TRIBs-In).  Note that the different TRIBs are defined in
   Section 3.5.

   Incremental updates are sent as the TRIP routing tables (TRIBs)
   change.  TRIP does not require periodic refresh of the routes.
   Therefore, an LS must retain the current version of all routing
   entries.

   If a particular ITAD has multiple LSs and is providing transit
   service for other ITADs, then care must be taken to ensure a
   consistent view of routing within the ITAD.  When synchronized the
   TRIP routing tables, i.e., the Loc-TRIBs, of all internal peers are
   identical.

3.3. Internal Versus External Synchronization



   As with BGP, TRIP distinguishes between internal and external peers.
   Within an ITAD, internal TRIP uses link-state mechanisms to flood
   database updates over an arbitrary topology.  Externally, TRIP uses
   point-to-point peering relationships to exchange database
   information.

   To achieve internal synchronization, internal peer connections are
   configured between LSs of the same ITAD such that the resulting
   intra-domain LS topology is connected and sufficiently redundant.
   This is different from BGP's approach that requires all internal
   peers to be connected in a full mesh topology, which may result in
   scaling problems.  When an update is received from an internal peer,
   the routes in the update are checked to determine if they are newer
   than the version already in the database.  Newer routes are then
   flooded to all other peers in the same domain.

3.4. Advertising TRIP Routes



   In TRIP, a route is defined as the combination of (a) a set of
   destination addresses (given by an address family indicator and an
   address prefix), and (b) an application protocol (e.g. SIP, H323,
   etc.).  Generally, there are additional attributes associated with
   each route (for example, the next-hop server).






Rosenberg, et. al.          Standards Track                     [Page 6]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   TRIP routes are advertised between a pair of LSs in UPDATE messages.
   The destination addresses are included in the ReachableRoutes
   attribute of the UPDATE, while other attributes describe things like
   the path or egress gateway.

   If an LS chooses to advertise a TRIP route, it may add to or modify
   the attributes of the route before advertising it to a peer.  TRIP
   provides mechanisms by which an LS can inform its peer that a
   previously advertised route is no longer available for use.  There
   are three methods by which a given LS can indicate that a route has
   been withdrawn from service:

      -  Include the route in the WithdrawnRoutes Attribute in an UPDATE
         message, thus marking the associated destinations as being no
         longer available for use.
      -  Advertise a replacement route with the same set of destinations
         in the ReachableRoutes Attribute.
      -  For external peers where flooding is not in use, the LS-to-LS
         peer connection can be closed, which implicitly removes from
         service all routes which the pair of LSs had advertised to each
         other over that peer session.  Note that terminating an
         internal peering session does not necessarily remove the routes
         advertised by the peer LS as the same routes may have been
         received from multiple internal peers because of flooding.  If
         an LS determines that another internal LS is no longer active
         (from the ITAD Topology attributes of the UPDATE messages from
         other internal peers), then it MUST remove all routes
         originated into the LS by that LS and rerun its decision
         process.

3.5. Telephony Routing Information Bases



   A TRIP LS processes three types of routes:

      -  External routes: An external route is a route received from an
         external peer LS
      -  Internal routes: An internal route is a route received from an
         internal LS in the same ITAD.
      -  Local routes: A local route is a route locally injected into
         TRIP, e.g. by configuration or by route redistribution from
         another routing protocol.

   The Telephony Routing Information Base (TRIB) within an LS consists
   of four distinct parts:







Rosenberg, et. al.          Standards Track                     [Page 7]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


      -  Adj-TRIBs-In: The Adj-TRIBs-In store routing information that
         has been learned from inbound UPDATE messages.  Their contents
         represent TRIP routes that are available as an input to the
         Decision Process.  These are the "unprocessed" routes received.
         The routes from each external peer LS and each internal LS are
         maintained in this database independently, so that updates from
         one peer do not affect the routes received from another LS.
         Note that there is an Adj-TRIB-In for every LS within the
         domain, even those with which the LS is not directly peered.
      -  Ext-TRIB: There is only one Ext-TRIB database per LS.  The LS
         runs the route selection algorithm on all external routes
         (stored in the Adj-TRIBs-In of the external peers) and local
         routes (may be stored in an Adj-TRIB-In representing the local
         LS) and selects the best route for a given destination and
         stores it in the Ext-TRIB.  The use of Ext-TRIB will be
         explained further in Section 10.3.1
      -  Loc-TRIB: The Loc-TRIB contains the local TRIP routing
         information that the LS has selected by applying its local
         policies to the routing information contained in its Adj-
         TRIBs-In of internal LSs and the Ext-TRIB.
      -  Adj-TRIBs-Out:  The Adj-TRIBs-Out store the information that
         the local LS has selected for advertisement to its external
         peers.  The routing information stored in the Adj-TRIBs-Out
         will be carried in the local LS's UPDATE messages and
         advertised to its peers.

   Figure 1 illustrates the relationship between the four parts of the
   routing information base.

                            Loc-TRIB
                                ^
                                |
                        Decision Process
                         ^      ^      |
                         |      |      |
                Adj-TRIBs-In    |      V
               (Internal LSs)   |   Adj-TRIBs-Out
                                |
                                |
                                |
                             Ext-TRIB
                            ^        ^
                            |        |
                   Adj-TRIB-In      Local Routes
               (External Peers)

                     Figure 1: TRIB Relationships




Rosenberg, et. al.          Standards Track                     [Page 8]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   Although the conceptual model distinguishes between Adj-TRIBs-In,
   Ext-TRIB, Loc-TRIB, and Adj-TRIBs-Out, this neither implies nor
   requires that an implementation must maintain four separate copies of
   the routing information.  The choice of implementation (for example,
   4 copies of the information vs. 1 copy with pointers) is not
   constrained by the protocol.

3.6. Routes in TRIP



   A route in TRIP specifies a range of numbers by being a prefix of
   those numbers (the exact definition & syntax of route are in 5.1.1).
   Arbitrary ranges of numbers are not atomically representable by a
   route in TRIP.  A prefix range is the only type of range supported
   atomically.  An arbitrary range can be accomplished by using multiple
   prefixes in a ReachableRoutes attribute (see Section 5.1 & 5.2).  For
   example, 222-xxxx thru 999-xxxx could be represented by including the
   prefixes 222, 223, 224,...,23,24,...,3,4,...,9 in a ReachableRoutes
   attribute.

3.7. Aggregation



   Aggregation is a scaling enhancement used by an LS to reduce the
   number of routing entries that it has to synchronize with its peers.
   Aggregation may be performed by an LS when there is a set of routes
   {R1, R2, ...} in its TRIB such that there exists a less specific
   route R where every valid destination in R is also a valid
   destination in {R1, R2, ...} and vice-versa.  Section 5 includes a
   description of how to combine each attribute (by type) on the {R1,
   R2, ...} routes into an attribute for R.

   Note that there is no mechanism within TRIP to communicate that a
   particular address prefix is not used or valid within a particular
   address family, and thus that these addresses could be skipped during
   aggregation.  LSs may use methods outside of TRIP to learn of invalid
   prefixes that may be ignored during aggregation.

   An LS is not required to perform aggregation, however it is
   recommended whenever maintaining a smaller TRIB is important.  An LS
   decides based on its local policy whether or not to aggregate a set
   of routes into a single aggregate route.

   Whenever an LS aggregates multiple routes where the NextHopServer is
   not identical in all aggregated routes, the NextHopServer attribute
   of the aggregate route must be set to a signalling server in the
   aggregating LS's domain.






Rosenberg, et. al.          Standards Track                     [Page 9]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   When an LS resets the NextHopServer of any route, and this may be
   performed because of aggregation or other reasons, it has the effect
   of adding another signalling server along the signalling path to
   these destinations.  The end result is that the signalling path
   between two destinations may consist of multiple signalling servers
   across multiple domains.

4. Message Formats



   This section describes message formats used by TRIP.  Messages are
   sent over a reliable transport protocol connection.  A message MUST
   be processed only after it is entirely received.  The maximum message
   size is 4096 octets.  All implementations MUST support this maximum
   message size.  The smallest message that MAY be sent consists of a
   TRIP header without a data portion, or 3 octets.

4.1. Message Header Format



   Each message has a fixed-size header.  There may or may not be a data
   portion following the header, depending on the message type.  The
   layout of the header fields is shown in Figure 2.

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

                      Figure 2: TRIP Header

   Length:  This 2-octet unsigned integer indicates the total length of
   the message, including the header, in octets.  Thus, it allows one to
   locate, in the transport-level stream, the beginning of the next
   message.  The value of the Length field must always be at least 3 and
   no greater than 4096, and may be further constrained depending on the
   message type.  No padding of extra data after the message is allowed,
   so the Length field must have the smallest value possible given the
   rest of the message.

   Type:  This 1-octet unsigned integer indicates the type code of the
   message.  The following type codes are defined:

      1 - OPEN
      2 - UPDATE
      3 - NOTIFICATION
      4 - KEEPALIVE





Rosenberg, et. al.          Standards Track                    [Page 10]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


4.2. OPEN Message Format



   After a transport protocol connection is established, the first
   message sent by each side is an OPEN message.  If the OPEN message is
   acceptable, a KEEPALIVE message confirming the OPEN is sent back.
   Once the OPEN is confirmed, UPDATE, KEEPALIVE, and NOTIFICATION
   messages may be exchanged.

   The minimum length of the OPEN message is 17 octets (including
   message header).  OPEN messages not meeting this minimum requirement
   are handled as defined in Section 6.2.

   In addition to the fixed-size TRIP header, the OPEN message contains
   the following fields:

       0                   1                   2                   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
      +---------------+---------------+--------------+----------------+
      |    Version    |    Reserved   |          Hold Time            |
      +---------------+---------------+--------------+----------------+
      |                            My ITAD                            |
      +---------------+---------------+--------------+----------------+
      |                        TRIP Identifier                        |
      +---------------+---------------+--------------+----------------+
      |    Optional Parameters Len    |Optional Parameters (variable)...
      +---------------+---------------+--------------+----------------+

                        Figure 3: TRIP OPEN Header

   Version:
   This 1-octet unsigned integer indicates the protocol version of the
   message.  The current TRIP version number is 1.

   Hold Time:
   This 2-octet unsigned integer indicates the number of seconds that
   the sender proposes for the value of the Hold Timer.  Upon receipt of
   an OPEN message, an LS MUST calculate the value of the Hold Timer by
   using the smaller of its configured Hold Time and the Hold Time
   received in the OPEN message.  The Hold Time MUST be either zero or
   at least three seconds.  An implementation MAY reject connections on
   the basis of the Hold Time.  The calculated value indicates the
   maximum number of seconds that may elapse between the receipt of
   successive KEEPALIVE and/or UPDATE messages by the sender.

   This 4-octet unsigned integer indicates the ITAD number of the
   sender.  The ITAD number must be unique for this domain within this
   confederation of cooperating LSs.




Rosenberg, et. al.          Standards Track                    [Page 11]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   ITAD numbers are assigned by IANA as specified in Section 13.  This
   document reserves ITAD number 0.  ITAD numbers from 1 to 255 are
   designated for private use.

   TRIP Identifier:
   This 4-octet unsigned integer indicates the TRIP Identifier of the
   sender.  The TRIP Identifier MUST uniquely identify this LS within
   its ITAD.  A given LS MAY set the value of its TRIP Identifier to an
   IPv4 address assigned to that LS.  The value of the TRIP Identifier
   is determined on startup and MUST be the same for all peer
   connections.  When comparing two TRIP identifiers, the TRIP
   Identifier is interpreted as a numerical 4-octet unsigned integer.

   Optional Parameters Length:
   This 2-octet unsigned integer indicates the total length of the
   Optional Parameters field in octets.  If the value of this field is
   zero, no Optional Parameters are present.

   Optional Parameters:
   This field may contain a list of optional parameters, where each
   parameter is encoded as a <Parameter Type, Parameter Length,
   Parameter Value> triplet.

       0                   1                   2
       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
      +---------------+---------------+--------------+----------------+
      |       Parameter Type          |       Parameter Length        |
      +---------------+---------------+--------------+----------------+
      |                  Parameter Value (variable)...
      +---------------+---------------+--------------+----------------+

                    Figure 4: Optional Parameter Encoding

   Parameter Type:
   This is a 2-octet field that unambiguously identifies individual
   parameters.

   Parameter Length:
   This is a 2-octet field that contains the length of the Parameter
   Value field in octets.

   Parameter Value:
   This is a variable length field that is interpreted according to the
   value of the Parameter Type field.







Rosenberg, et. al.          Standards Track                    [Page 12]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


4.2.1. Open Message Optional Parameters



   This document defines the following Optional Parameters for the OPEN
   message.

4.2.1.1. Capability Information



   Capability Information uses Optional Parameter type 1.  This is an
   optional parameter used by an LS to convey to its peer the list of
   capabilities supported by the LS.  This permits an LS to learn of the
   capabilities of its peer LSs.  Capability negotiation is defined in
   Section 8.

   The parameter contains one or more triples <Capability Code,
   Capability Length, Capability Value>, where each triple is encoded as
   shown below:

    0                   1                   2
    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
   +---------------+---------------+--------------+----------------+
   |       Capability Code         |       Capability Length       |
   +---------------+---------------+--------------+----------------+
   |       Capability Value (variable)...
   +---------------+---------------+--------------+----------------+

           Figure 5:  Capability Optional Parameter

   Capability Code:
   Capability Code is a 2-octet field that unambiguously identifies
   individual capabilities.

   Capability Length:
   Capability Length is a 2-octet field that contains the length of the
   Capability Value field in octets.

   Capability Value:
   Capability Value is a variable length field that is interpreted
   according to the value of the Capability Code field.

   Any particular capability, as identified by its Capability Code, may
   appear more than once within the Optional Parameter.

   This document reserves Capability Codes 32768-65535 for vendor-
   specific applications (these are the codes with the first bit of the
   code value equal to 1).  This document reserves value 0.  Capability
   Codes (other than those reserved for vendor specific use) are
   controlled by IANA.  See Section 13 for IANA considerations.




Rosenberg, et. al.          Standards Track                    [Page 13]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   The following Capability Codes are defined by this specification:

      Code           Capability
      1              Route Types Supported
      2              Send Receive Capability

4.2.1.1.1. Route Types Supported


   The Route Types Supported Capability Code lists the route types
   supported in this peering session by the transmitting LS.  An LS MUST
   NOT
use route types that are not supported by the peer LS in any
   particular peering session.  If the route types supported by a peer
   are not satisfactory, an LS SHOULD terminate the peering session.
   The format for a Route Type is:

    0                   1                   2
    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
   +---------------+---------------+--------------+----------------+
   |        Address Family         |     Application Protocol      |
   +---------------+---------------+--------------+----------------+

            Figure 6: Route Types Supported Capability

   The Address Family and Application Protocol are as defined in Section
   5.1.1.  Address Family gives the address family being routed (within
   the ReachableRoutes attribute).  The application protocol lists the
   application for which the routes apply.  As an example, a route type
   for TRIP could be <E.164, SIP>, indicating a set of E.164
   destinations for the SIP protocol.

   The Route Types Supported Capability MAY contain multiple route types
   in the capability.  The number of route types within the capability
   is the maximum number that can fit given the capability length.  The
   Capability Code is 1 and the length is variable.

4.2.1.1.2. Send Receive Capability


   This capability specifies the mode in which the LS will operate with
   this particular peer.  The possible modes are: Send Only mode,
   Receive Only mode, or Send Receive mode.  The default mode is Send
   Receive mode.

   In Send Only mode, an LS transmits UPDATE messages to its peer, but
   the peer MUST NOT transmit UPDATE messages to that LS.  If an LS in
   Send Only mode receives an UPDATE message from its peer, it MUST
   discard that message, but no further action should be taken.





Rosenberg, et. al.          Standards Track                    [Page 14]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   The UPDATE messages sent by an LS in Send Only mode to its intra-
   domain peer MUST include the ITAD Topology attribute whenever the
   topology changes.  A useful application of an LS in Send Only mode
   with an external peer is to enable gateway registration services.

   If a service provider terminates calls to a set of gateways it owns,
   but never initiates calls, it can set its LSs to operate in Send Only
   mode, since they only ever need to generate UPDATE messages, not
   receive them.  If an LS in Send Receive mode has a peering session
   with a peer in Send Only mode, that LS MUST set its route
   dissemination policy such that it does not send any UPDATE messages
   to its peer.

   In Receive Only mode, the LS acts as a passive TRIP listener.  It
   receives and processes UPDATE messages from its peer, but it MUST NOT
   transmit any UPDATE messages to its peer.  This is useful for
   management stations that wish to collect topology information for
   display purposes.

   The behavior of an LS in Send Receive mode is the default TRIP
   operation specified throughout this document.

   The Send Receive capability is a 4-octet unsigned numeric value.  It
   can only take one of the following three values:

      1 - Send Receive mode
      2 - Send only mode
      3 - Receive Only mode

   A peering session MUST NOT be established between two LSs if both of
   them are  in Send Only mode or if both of them are in Receive Only
   mode.  If a peer LS detects such a capability mismatch when
   processing an OPEN message, it MUST respond with a NOTIFICATION
   message and close the peer session.  The error code in the
   NOTIFICATION message must be set to "Capability Mismatch."

   An LS MUST be configured in the same Send Receive mode for all peers.

4.3. UPDATE Message Format



   UPDATE messages are used to transfer routing information between LSs.
   The information in the UPDATE packet can be used to construct a graph
   describing the relationships between the various ITADs.  By applying
   rules to be discussed, routing information loops and some other
   anomalies can be prevented.






Rosenberg, et. al.          Standards Track                    [Page 15]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   An UPDATE message is used to both advertise and withdraw routes from
   service.  An UPDATE message may simultaneously advertise and withdraw
   TRIP routes.

   In addition to the TRIP header, the TRIP UPDATE contains a list of
   routing attributes as shown in Figure 7.  There is no padding between
   routing attributes.

         +------------------------------------------------+--...
         | First Route Attribute | Second Route Attribute |  ...
         +------------------------------------------------+--...

                    Figure 7: TRIP UPDATE Format

   The minimum length of an UPDATE message is 3 octets (there are no
   mandatory attributes in TRIP).

4.3.1. Routing Attributes



   A variable length sequence of routing attributes is present in every
   UPDATE message.  Each attribute is a triple <attribute type,
   attribute length, attribute value> of variable length.

       0                   1                   2                   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
      +---------------+---------------+--------------+----------------+
      |  Attr. Flags  |Attr. Type Code|         Attr. Length          |
      +---------------+---------------+--------------+----------------+
      |                   Attribute Value (variable)                  |
      +---------------+---------------+--------------+----------------+

                    Figure 8: Routing Attribute Format

   Attribute Type is a two-octet field that consists of the Attribute
   Flags octet followed by the Attribute Type Code octet.

   The Attribute Type Code defines the type of attribute.  The basic
   TRIP-defined Attribute Type Codes are discussed later in this
   section.  Attributes MUST appear in the UPDATE message in numerical
   order of the Attribute Type Code.  An attribute MUST NOT be included
   more than once in the same UPDATE message.  Attribute Flags are used
   to control attribute processing when the attribute type is unknown.
   Attribute Flags are further defined in Section 4.3.2.








Rosenberg, et. al.          Standards Track                    [Page 16]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   This document reserves Attribute Type Codes 224-255 for vendor-
   specific applications (these are the codes with the first three bits
   of the code equal to 1).  This document reserves value 0.  Attribute
   Type Codes (other than those reserved for vendor specific use) are
   controlled by IANA.  See Section 13 for IANA considerations.

   The third and the fourth octets of the route attribute contain the
   length of the attribute value field in octets.

   The remaining octets of the attribute represent the Attribute Value
   and are interpreted according to the Attribute Flags and the
   Attribute Type Code.  The basic supported attribute types, their
   values, and their uses are defined in this specification.  These are
   the attributes necessary for proper loop free operation of TRIP, both
   inter-domain and intra-domain.  Additional attributes may be defined
   in future documents.

4.3.2. Attribute Flags



   It is clear that the set of attributes for TRIP will evolve over
   time.  Hence it is essential that mechanisms be provided to handle
   attributes with unrecognized types.  The handling of unrecognized
   attributes is controlled via the flags field of the attribute.
   Recognized attributes should be processed according to their specific
   definition.

   The following are the attribute flags defined by this specification:
            Bit       Flag
            0         Well-Known Flag
            1         Transitive Flag
            2         Dependent Flag
            3         Partial Flag
            4         Link-state Encapsulated Flag

   The high-order bit (bit 0) of the Attribute Flags octet is the Well-
   Known Bit.  It defines whether the attribute is not well-known (if
   set to 1) or well-known (if set to 0).  Implementations are not
   required to support not well-known attributes, but MUST support
   well-known attributes.

   The second high-order bit (bit 1) of the Attribute Flags octet is the
   Transitive bit.  It defines whether a not well-known attribute is
   transitive (if set to 1) or non-transitive (if set to 0).  For well-
   known attributes, the Transitive bit MUST be zero on transmit and
   MUST be ignored on receipt.

   The third high-order bit (bit 2) of the Attribute Flags octet is the
   Dependent bit.  It defines whether a transitive attribute is



Rosenberg, et. al.          Standards Track                    [Page 17]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   dependent (if set to 1) or independent (if set to 0).  For well-known
   attributes and for non-transitive attributes, the Dependent bit is
   irrelevant, and MUST be set to zero on transmit and MUST be ignored
   on receipt.

   The fourth high-order bit (bit 3) of the Attribute Flags octet is the
   Partial bit.  It defines whether the information contained in the not
   well-known transitive attribute is partial (if set to 1) or complete
   (if set to 0).  For well-known attributes and for non-transitive
   attributes the Partial bit MUST be set to 0 on transmit and MUST be
   ignored on receipt.

   The fifth high-order bit (bit 4) of the Attribute Flags octet is the
   Link-state Encapsulation bit.  This bit is only applicable to certain
   attributes (ReachableRoutes and WithdrawnRoutes) and determines the
   encapsulation of the routes within those attributes.  If this bit is
   set, link-state encapsulation is used within the attribute.
   Otherwise, standard encapsulation is used within the attribute.  The
   Link-state Encapsulation technique is described in Section 4.3.2.4.
   This flag is only valid on the ReachableRoutes and WithdrawnRoutes
   attributes.  It MUST be cleared on transmit and MUST be ignored on
   receipt for all other attributes.

   The other bits of the Attribute Flags octet are unused.  They MUST be
   zeroed on transmit and ignored on receipt.

4.3.2.1. Attribute Flags and Route Selection



   Any recognized attribute can be used as input to the route selection
   process, although the utility of some attributes in route selection
   is minimal.

4.3.2.2. Attribute Flags and Route Dissemination



   TRIP provides for two variations of transitivity due to the fact that
   intermediate LSs need not modify the NextHopServer when propagating
   routes.  Attributes may be non-transitive, dependent transitive, or
   independent transitive.  An attribute cannot be both dependent
   transitive and independent transitive.

   Unrecognized independent transitive attributes may be propagated by
   any intermediate LS.  Unrecognized dependent transitive attributes
   MAY only be propagated if the LS is NOT changing the next-hop server.
   The transitivity variations permit some unrecognized attributes to be
   carried end-to-end (independent transitive), some to be carried
   between adjacent next-hop servers (dependent transitive), and other
   to be restricted to peer LSs (non-transitive).




Rosenberg, et. al.          Standards Track                    [Page 18]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   An LS that passes an unrecognized transitive attribute to a peer MUST
   set the Partial flag on that attribute.  Any LS along a path MAY
   insert a transitive attribute into a route.  If any LS except the
   originating LS inserts a new independent transitive attribute into a
   route, then it MUST set the Partial flag on that attribute.  If any
   LS except an LS that modifies the NextHopServer inserts a new
   dependent transitive attribute into a route, then it MUST set the
   Partial flag on that attribute.  The Partial flag indicates that not
   every LS along the relevant path has processed and understood the
   attribute.  For independent transitive attributes, the "relevant
   path" is the path given in the AdvertisementPath attribute.  For
   dependent transitive attributes, the relevant path consists only of
   those domains thru which this object has passed since the
   NextHopServer was last modified.  The Partial flag in an independent
   transitive attribute MUST NOT be unset by any other LS along the
   path.  The Partial flag in a dependent transitive attribute MUST be
   reset whenever the NextHopServer is changed, but MUST NOT be unset by
   any LS that is not changing the NextHopServer.

   The rules governing the addition of new non-transitive attributes are
   defined independently for each non-transitive attribute.  Any
   attribute MAY be updated by an LS in the path.

4.3.2.3. Attribute Flags and Route Aggregation



   Each attribute defines how it is to be handled during route
   aggregation.

   The rules governing the handling of unknown attributes are guided by
   the Attribute Flags.  Unrecognized transitive attributes are dropped
   during aggregation.  There should be no unrecognized non-transitive
   attributes during aggregation because non-transitive attributes must
   be processed by the local LS in order to be propagated.

4.3.2.4. Attribute Flags and Encapsulation



   Normally attributes have the simple format as described in Section
   4.3.1.  If the Link-state Encapsulation Flag is set, then the two
   additional fields are added to the attribute header as shown in
   Figure 9.











Rosenberg, et. al.          Standards Track                    [Page 19]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


    0                   1                   2                   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
   +---------------+---------------+--------------+----------------+
   |  Attr. Flags  |Attr. Type Code|          Attr. Length         |
   +---------------+---------------+--------------+----------------+
   |                  Originator TRIP Identifier                   |
   +---------------+---------------+--------------+----------------+
   |                        Sequence Number                        |
   +---------------+---------------+--------------+----------------+
   |                   Attribute Value (variable)                  |
   +---------------+---------------+--------------+----------------+

                 Figure 9: Link State Encapsulation

   The Originator TRIP ID and Sequence Number are used to control the
   flooding of routing updates within a collection of servers.  These
   fields are used to detect duplicate and old routes so that they are
   not further propagated to other LSs.  The use of these fields is
   defined in Section 10.1.

4.3.3. Mandatory Attributes



   There are no Mandatory attributes in TRIP.  However, there are
   Conditional Mandatory attributes.  A conditional mandatory attribute
   is an attribute, which MUST be included in an UPDATE message if
   another attribute is included in that message.  For example, if an
   UPDATE message includes a ReachableRoutes attribute, it MUST include
   an AdvertisementPath attribute as well.

   The three base attributes in TRIP are WithdrawnRoutes,
   ReachableRoutes, and ITAD Topology.  Their presence in an UPDATE
   message is entirely optional and independent of any other attributes.

4.3.4. TRIP UPDATE Attributes



   This section summarizes the attributes that may be carried in an
   UPDATE message.  Attributes MUST appear in the UPDATE message in
   increasing order of the Attribute Type Code.  Additional details are
   provided in Section 5.

4.3.4.1. WithdrawnRoutes



   This attribute lists a set of routes that are being withdrawn from
   service.  The transmitting LS has determined that these routes should
   no longer be advertised, and is propagating this information to its
   peers.





Rosenberg, et. al.          Standards Track                    [Page 20]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


4.3.4.2. ReachableRoutes



   This attribute lists a set of routes that are being added to service.
   These routes will have the potential to be inserted into the Adj-
   TRIBs-In of the receiving LS and the route selection process will be
   applied to them.

4.3.4.3. NextHopServer



   This attribute gives the identity of the entity to which messages
   should be sent along this routed path.  It specifies the identity of
   the next hop server as either a host domain name or an IP address.
   It MAY optionally specify the UDP/TCP port number for the next hop
   signaling server.  If not specified, then the default port SHOULD be
   used.  The NextHopServer is specific to the set of destinations and
   application protocol defined in the ReachableRoutes attribute.  Note
   that this is NOT necessarily the address to which media (voice,
   video, etc.)  should be transmitted, it is only for the application
   protocol as given in the ReachableRoutes attribute.

4.3.4.4. AdvertisementPath



   The AdvertisementPath is analogous to the AS_PATH in BGP4 [3].  The
   attribute records the sequence of domains through which this
   advertisement has passed.  The attribute is used to detect when the
   routing advertisement is looping.  This attribute does NOT reflect
   the path through which messages following this route would traverse.
   Since the next-hop need not be modified by each LS, the actual path
   to the destination might not have to traverse every domain in the
   AdvertisementPath.

4.3.4.5. RoutedPath



   The RoutedPath attribute is analogous to the AdvertisementPath
   attribute, except that it records the actual path (given by the list
   of domains) *to* the destinations.  Unlike AdvertisementPath, which
   is modified each time the route is propagated, RoutedPath is only
   modified when the NextHopServer attribute changes.  Thus, it records
   the subset of the AdvertisementPath which signaling messages
   following this particular route would traverse.

4.3.4.6. AtomicAggregate



   The AtomicAggregate attribute indicates that a route may actually
   include domains not listed in the RoutedPath.  If an LS, when
   presented with a set of overlapping routes from a peer LS, selects a
   less specific route without selecting the more specific route, then
   the LS MUST include the AtomicAggregate attribute with the route.  An



Rosenberg, et. al.          Standards Track                    [Page 21]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   LS receiving a route with an AtomicAggregate attribute MUST NOT make
   the set of destinations more specific when advertising it to other
   LSs.

4.3.4.7. LocalPreference



   The LocalPreference attribute is an intra-domain attribute used to
   inform other LSs of the local LS's preference for a given route.  The
   preference of a route is calculated at the ingress to a domain and
   passed as an attribute with that route throughout the domain.  Other
   LSs within the same ITAD use this attribute in their route selection
   process.  This attribute has no significance between domains.

4.3.4.8. MultiExitDisc



   There may be more than one LS peering relationship between
   neighboring domains.  The MultiExitDisc attribute is used by an LS to
   express a preference for one link between the domains over another
   link between the domains.  The use of the MultiExitDisc attribute is
   controlled by local policy.

4.3.4.9. Communities



   The Communities attribute is not a well-known attribute.  It is used
   to facilitate and simplify the control of routing information by
   grouping destinations into communities.

4.3.4.10. ITAD Topology



   The ITAD topology attribute is an intra-domain attribute that is used
   by LSs to indicate their intra-domain topology to other LSs in the
   domain.

4.3.4.11. ConvertedRoute



   The ConvertedRoute attribute indicates that an intermediate LS has
   altered the route by changing the route's Application Protocol.

4.4. KEEPALIVE Message Format



   TRIP does not use any transport-based keep-alive mechanism to
   determine if peers are reachable.  Instead, KEEPALIVE messages are
   exchanged between peers often enough as not to cause the Hold Timer
   to expire.  A reasonable maximum time between KEEPALIVE messages
   would be one third of the Hold Time interval.  KEEPALIVE messages
   MUST NOT be sent more than once every 3 seconds.  An implementation
   SHOULD adjust the rate at which it sends KEEPALIVE messages as a
   function of the negotiated Hold Time interval.



Rosenberg, et. al.          Standards Track                    [Page 22]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   If the negotiated Hold Time interval is zero, then periodic KEEPALIVE
   messages MUST NOT be sent.

   The KEEPALIVE message consists of only a message header and has a
   length of 3 octets.

4.5. NOTIFICATION Message Format



   A NOTIFICATION message is sent when an error condition is detected.
   The TRIP transport connection is closed immediately after sending a
   NOTIFICATION message.

   In addition to the fixed-size TRIP header, the NOTIFICATION message
   contains the following fields:

    0                   1                   2                   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
   +---------------+---------------+--------------+----------------+
   |  Error Code   | Error Subcode |       Data... (variable)
   +---------------+---------------+--------------+----------------+

                Figure 10: TRIP NOTIFICATION Format

   Error Code:
   This 1-octet unsigned integer indicates the type of NOTIFICATION.
   The following Error Codes have been defined:

   Error Code       Symbolic Name               Reference
     1         Message Header Error             Section 6.1
     2         OPEN Message Error               Section 6.2
     3         UPDATE Message Error             Section 6.3
     4         Hold Timer Expired               Section 6.5
     5         Finite State Machine Error       Section 6.6
     6         Cease                            Section 6.7

   Error Subcode:
   This 1-octet unsigned integer provides more specific information
   about the nature of the reported error.  Each Error Code may have one
   or more Error Subcodes associated with it.  If no appropriate Error
   Subcode is defined, then a zero (Unspecific) value is used for the
   Error Subcode field.

   Message Header Error Subcodes:
      1  - Bad Message Length.
      2  - Bad Message Type.






Rosenberg, et. al.          Standards Track                    [Page 23]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   OPEN Message Error Subcodes:
      1  - Unsupported Version Number.
      2  - Bad Peer ITAD.
      3  - Bad TRIP Identifier.
      4  - Unsupported Optional Parameter.
      5  - Unacceptable Hold Time.
      6  - Unsupported Capability.
      7  - Capability Mismatch.

   UPDATE Message Error Subcodes:
      1 - Malformed Attribute List.
      2 - Unrecognized Well-known Attribute.
      3 - Missing Well-known Mandatory Attribute.
      4 - Attribute Flags Error.
      5 - Attribute Length Error.
      6 - Invalid Attribute.

   Data:
   This variable-length field is used to diagnose the reason for the
   NOTIFICATION.  The contents of the Data field depend upon the Error
   Code and Error Subcode.

   Note that the length of the data can be determined from the message
   length field by the formula:

            Data Length = Message Length - 5

   The minimum length of the NOTIFICATION message is 5 octets (including
   message header).

5. TRIP Attributes



   This section provides details on the syntax and semantics of each
   TRIP UPDATE attribute.

5.1. WithdrawnRoutes



   Conditional Mandatory: False.
   Required Flags: Well-known.
   Potential Flags: Link-State Encapsulation (when flooding).
   TRIP Type Code: 1

   The WithdrawnRoutes specifies a set of routes that are to be removed
   from service by the receiving LS(s).  The set of routes MAY be empty,
   indicated by a length field of zero.






Rosenberg, et. al.          Standards Track                    [Page 24]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


5.1.1. Syntax of WithdrawnRoutes



   The WithdrawnRoutes Attribute encodes a sequence of routes in its
   value field.  The format for individual routes is given in Section
   5.1.1.1.  The WithdrawnRoutes Attribute lists the individual routes
   sequentially with no padding as shown in Figure 11.  Each route
   includes a length field so that the individual routes within the
   attribute can be delineated.

            +---------------------+---------------------+...
            |  WithdrawnRoute1... |  WithdrawnRoute2... |...
            +---------------------+---------------------+...

                 Figure 11: WithdrawnRoutes Format

5.1.1.1. Generic TRIP Route Format



   The generic format for a TRIP route is given in Figure 12.

    0                   1                   2                   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
   +---------------+---------------+--------------+----------------+
   |       Address Family          |      Application Protocol     |
   +---------------+---------------+--------------+----------------+
   |            Length             |       Address (variable)     ...
   +---------------+---------------+--------------+----------------+

                Figure 12: Generic TRIP Route Format

   Address Family:
   The address family field gives the type of address for the route.
   Three address families are defined in this Section:

            Code              Address Family
            1                 Decimal Routing Numbers
            2                 PentaDecimal Routing Numbers
            3                 E.164 Numbers

   This document reserves address family code 0.  This document reserves
   address family codes 32768-65535 for vendor-specific applications
   (these are the codes with the first bit of the code value equal to
   1).  Additional address families may be defined in the future.
   Assignment of address family codes is controlled by IANA.  See
   Section 13 for IANA considerations.







Rosenberg, et. al.          Standards Track                    [Page 25]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   Application Protocol:
   The application protocol gives the protocol for which this routing
   table is maintained.  The currently defined application protocols
   are:

            Code              Protocol
            1                 SIP
            2                 H.323-H.225.0-Q.931
            3                 H.323-H.225.0-RAS
            4                 H.323-H.225.0-Annex-G

   This document reserves application protocol code 0.  This document
   reserves application protocol codes 32768-65535 for vendor-specific
   applications (these are the codes with the first bit of the code
   value equal to 1).  Additional application protocols may be defined
   in the future.  Assignment of application protocol codes is
   controlled by IANA.  See Section 13 for IANA considerations.

   Length:
   The length of the address field, in bytes.

   Address:
   This is an address (prefix) of the family type given by Address
   Family.  The octet length of the address is variable and is
   determined by the length field of the route.

5.1.1.2. Decimal Routing Numbers



   The Decimal Routing Numbers address family is a super set of all
   E.164 numbers, national numbers, local numbers, and private numbers.
   It can also be used to represent the decimal routing numbers used in
   conjunction with Number Portability in some countries/regions.  A set
   of telephone numbers is specified by a Decimal Routing Number prefix.
   Decimal Routing Number prefixes are represented by a string of
   digits, each digit encoded by its ASCII character representation.
   This routing object covers all phone numbers starting with this
   prefix.  The syntax for the Decimal Routing Number prefix is:

      Decimal-routing-number  = *decimal-digit
      decimal-digit           = DECIMAL-DIGIT
      DECIMAL-DIGIT           = "0"|"1"|"2"|"3"|"4"|"5"|"6"|"7"|"8"|"9"

   This DECIMAL Routing Number prefix is not bound in length.  This
   format is similar to the format for a global telephone number as
   defined in SIP [8] without visual separators and without the "+"
   prefix for international numbers.  This format facilitates efficient
   comparison when using TRIP to route SIP or H323, both of which use
   character based representations of phone numbers.  The prefix length



Rosenberg, et. al.          Standards Track                    [Page 26]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   is determined from the length field of the route.  The type of
   Decimal Routing Number (private, local, national, or international)
   can be deduced from the first few digits of the prefix.

5.1.1.3. PentaDecimal Routing Numbers



   This address family is used to represent PentaDecimal Routing Numbers
   used in conjunction with Number Portability in some
   countries/regions.  PentaDecimal Routing Number prefixes are
   represented by a string of digits, each digit encoded by its ASCII
   character representation.  This routing object covers all routing
   numbers starting with this prefix.  The syntax for the PentaDecimal
   Routing Number prefix is:

      PentaDecimal-routing-number   = *pentadecimal-digit
      pentadecimal-routing-digit    = PENTADECIMAL-DIGIT
      PENTADECIMAL-DIGIT            = "0"|"1"|"2"|"3"|"4"|"5"|"6"|"7"|
                                      "8"|"9"|"A"|"B"|"C"|"D"|"E"

   Note the difference in alphabets between Decimal Routing Numbers and
   PentaDecimal Routing Numbers.  A PentaDecimal Routing Number prefix
   is not bound in length.

   Note that the address family, which suits the routing numbers of a
   specific country/region depends on the alphabets used for routing
   numbers in that country/region.  For example, North American routing
   numbers SHOULD use the Decimal Routing Numbers address family,
   because their alphabet is limited to the digits "0" through "9".
   Another example, in most European countries routing numbers use the
   alphabet "0" through "9" and "A" through "E", and hence these
   countries SHOULD use the PentaDecimal Routing Numbers address family.

5.1.1.4. E.164 Numbers



   The E.164 Numbers address family is dedicated to fully qualified
   E.164 numbers.  A set of telephone numbers is specified by a E.164
   prefix.  E.164 prefixes are represented by a string of digits, each
   digit encoded by its ASCII character representation.  This routing
   object covers all phone numbers starting with this prefix.  The
   syntax for the E.164 prefix is:

      E164-number          = *e164-digit
      E164-digit           = E164-DIGIT
      E164-DIGIT           = "0"|"1"|"2"|"3"|"4"|"5"|"6"|"7"|"8"|"9"







Rosenberg, et. al.          Standards Track                    [Page 27]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   This format facilitates efficient comparison when using TRIP to route
   SIP or H323, both of which use character based representations of
   phone numbers.  The prefix length is determined from the length field
   of the route.

   The E.164 Numbers address family and the Decimal Routing Numbers
   address family have the same alphabet.  The E.164 Numbers address
   family SHOULD be used whenever possible.  The Decimal Routing Numbers
   address family can be used in case of private numbering plans or
   applications that do not desire to advertise fully expanded, fully
   qualified telephone numbers.  If Decimal Routing Numbers are used to
   advertise non-fully qualified prefixes, the prefixes may have to be
   manipulated (e.g. expanded) at the boundary between ITADs.  This adds
   significant complexity to the ITAD-Border LS, because, it has to map
   the prefixes from the format used in its own ITAD to the format used
   in the peer ITAD.

5.2. ReachableRoutes



   Conditional Mandatory: False.
   Required Flags: Well-known.
   Potential Flags: Link-State Encapsulation (when flooding).
   TRIP Type Code: 2

   The ReachableRoutes attribute specifies a set of routes that are to
   be added to service by the receiving LS(s).  The set of routes MAY be
   empty, as indicated by setting the length field to zero.

5.2.1. Syntax of ReachableRoutes



   The ReachableRoutes Attribute has the same syntax as the
   WithdrawnRoutes Attribute.  See Section 5.1.1.

5.2.2. Route Origination and ReachableRoutes



   Routes are injected into TRIP by a method outside the scope of this
   specification.  Possible methods include a front-end protocol, an
   intra-domain routing protocol, or static configuration.

5.2.3. Route Selection and ReachableRoutes



   The routes in ReachableRoutes are necessary for route selection.

5.2.4. Aggregation and ReachableRoutes



   To aggregate multiple routes, the set of ReachableRoutes to be
   aggregated MUST combine to form a less specific set.




Rosenberg, et. al.          Standards Track                    [Page 28]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   There is no mechanism within TRIP to communicate that a particular
   address prefix is not used and thus that these addresses could be
   skipped during aggregation.  LSs MAY use methods outside of TRIP to
   learn of invalid prefixes that may be ignored during aggregation.

   If an LS advertises an aggregated route, it MUST include the
   AtomicAggregate attribute.

5.2.5. Route Dissemination and ReachableRoutes



   The ReachableRoutes attribute is recomputed at each LS except where
   flooding is being used (e.g., within a domain).  It is therefore
   possible for an LS to change the Application Protocol field of a
   route before advertising that route to an external peer.

   If an LS changes the Application Protocol of a route it advertises,
   it MUST include the ConvertedRoute attribute in the UPDATE message.

5.2.6. Aggregation Specifics for Decimal Routing Numbers, E.164 Numbers,
       and PentaDecimal Routing Numbers



   An LS that has routes to all valid numbers in a specific prefix
   SHOULD advertise that prefix as the ReachableRoutes, even if there
   are more specific prefixes that do not actually exist on the PSTN.
   Generally, it takes 10 Decimal Routing/E.164 prefixes, or 15
   PentaDecimal Routing prefixes, of length n to aggregate into a prefix
   of length n-1.  However, if an LS is aware that a prefix is an
   invalid Decimal Routing/E.164 prefix, or PentaDecimal Routing prefix,
   then the LS MAY aggregate by skipping this prefix.  For example, if
   the Decimal Routing prefix 19191 is known not to exist, then an LS
   can aggregate to 1919 without 19191.  A prefix representing an
   invalid set of PSTN destinations is sometimes referred to as a
   "black-hole."  The method by which an LS is aware of black-holes is
   not within the scope of TRIP, but if an LS has such knowledge, it can
   use the knowledge when aggregating.

5.3. NextHopServer



   Conditional Mandatory: True (if ReachableRoutes and/or
   WithdrawnRoutes attribute is present).
   Required Flags: Well-known.
   Potential Flags: None.
   TRIP Type Code: 3.








Rosenberg, et. al.          Standards Track                    [Page 29]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   Given a route with application protocol A and destinations D, the
   NextHopServer indicates to the next-hop that messages of protocol A
   destined for D should be sent to it.  This may or may not represent
   the ultimate destination of those messages.

5.3.1. NextHopServer Syntax



   For generality, the address of the next-hop server may be of various
   types (domain name, IPv4, IPv6, etc).  The NextHopServer attribute
   includes the ITAD number of next-hop server, a length field, and a
   next-hop name or address.

   The syntax for the NextHopServer is given in Figure 13.

    0                   1                   2                   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
   +---------------+---------------+--------------+----------------+
   |                         Next Hop ITAD                         |
   +---------------+---------------+--------------+----------------+
   |             Length            |         Server (variable)    ...
   +---------------+---------------+--------------+----------------+

                  Figure 13: NextHopServer Syntax

   The Next-Hop ITAD indicates the domain of the next-hop.  Length field
   gives the number of octets in the Server field, and the Server field
   contains the name or address of the next-hop server.  The server
   field is represented as a string of ASCII characters.  It is defined
   as follows:

   Server  = host [":" port ]
   host    = <   A legal Internet host domain name
              or an IPv4 address using the textual representation
                 defined in Section 2.1 of RFC 1123 [9]
              or an IPv6 address using the textual representation
                 defined in Section 2.2 of RFC 2373 [10].  The IPv6
                 address MUST be enclosed in "[" and "]"
                 characters.>
   port    = *DIGIT

   If the port is empty or not given, the default port is assumed (e.g.,
   port 5060 if the application protocol is SIP).

5.3.2. Route Origination and NextHopServer



   When an LS originates a routing object into TRIP, it MUST include a
   NextHopServer within its domain.  The NextHopServer could be an
   address of the egress gateway or of a signaling proxy.



Rosenberg, et. al.          Standards Track                    [Page 30]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


5.3.3. Route Selection and NextHopServer



   LS policy may prefer certain next-hops or next-hop domains over
   others.

5.3.4. Aggregation and NextHopServer



   When aggregating multiple routing objects into a single routing
   object, an LS MUST insert a new signaling server from within its
   domain as the new NextHopServer unless all of the routes being
   aggregated have the same next-hop.

5.3.5. Route Dissemination and NextHopServer



   When propagating routing objects to peers, an LS may choose to insert
   a signaling proxy within its domain as the new next-hop, or it may
   leave the next-hop unchanged.  Inserting a new next-hop will cause
   the signaling messages to be sent to that address, and will provide
   finer control over the signaling path.  Leaving the next-hop
   unchanged will yield a more efficient signaling path (fewer hops).
   It is a local policy decision of the LS to decide whether to
   propagate or change the NextHopServer.

5.4. AdvertisementPath



   Conditional Mandatory: True (if ReachableRoutes and/or
   WithdrawnRoutes attribute is present).
   Required Flags: Well-known.
   Potential Flags: None.
   TRIP Type Code: 4.

   This attribute identifies the ITADs through which routing information
   carried in an advertisement has passed.  The AdvertisementPath
   attribute is analogous to the AS_PATH attribute in BGP.  The
   attributes differ in that BGP's AS_PATH also reflects the path to the
   destination.  In TRIP, not every domain need modify the next-hop, so
   the AdvertisementPath may include many more hops than the actual path
   to the destination.  The RoutedPath attribute (Section 5.5) reflects
   the actual signaling path to the destination.

5.4.1. AdvertisementPath Syntax



   AdvertisementPath is a variable length attribute that is composed of
   a sequence of ITAD path segments.  Each ITAD path segment is
   represented by a type-length-value triple.

   The path segment type is a 1-octet long field with the following
   values defined:



Rosenberg, et. al.          Standards Track                    [Page 31]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


      Value      Segment Type
      1          AP_SET: unordered set of ITADs a route in the
                 advertisement message has traversed
      2          AP_SEQUENCE: ordered set of ITADs a route in
                 the advertisement message has traversed

   The path segment length is a 1-octet long field containing the number
   of ITADs in the path segment value field.

   The path segment value field contains one or more ITAD numbers, each
   encoded as a 4-octets long field.  ITAD numbers uniquely identify an
   Internet Telephony Administrative Domain, and must be obtained from
   IANA.  See Section 13 for procedures to obtain an ITAD number from
   IANA.

5.4.2. Route Origination and AdvertisementPath



   When an LS originates a route then:

      -  The originating LS shall include its own ITAD number in the
         AdvertisementPath attribute of all advertisements sent to LSs
         located in neighboring ITADs.  In this case, the ITAD number of
         the originating LS's ITAD will be the only entry in the
         AdvertisementPath attribute.
      -  The originating LS shall include an empty AdvertisementPath
         attribute in all advertisements sent to LSs located in its own
         ITAD.  An empty AdvertisementPath attribute is one whose length
         field contains the value zero.

5.4.3. Route Selection and AdvertisementPath



   The AdvertisementPath may be used for route selection.  Possible
   criteria to be used are the number of hops on the path and the
   presence or absence of particular ITADs on the path.

   As discussed in Section 10, the AdvertisementPath is used to prevent
   routing information from looping.  If an LS receives a route with its
   own ITAD already in the AdvertisementPath, the route MUST be
   discarded.

5.4.4. Aggregation and AdvertisementPath



   The rules for aggregating AdvertisementPath attributes are given in
   the following sections, where the term "path" used in Section 5.4.4.1
   and 5.4.4.2 is understood to mean AdvertisementPath.






Rosenberg, et. al.          Standards Track                    [Page 32]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


5.4.4.1. Aggregating Routes with Identical Paths



   If all routes to be aggregated have identical path attributes, then
   the aggregated route has the same path attribute as the individual
   routes.

5.4.4.2. Aggregating Routes with Different Paths



   For the purpose of aggregating path attributes we model each ITAD
   within the path as a pair <type, value>, where "type" identifies a
   type of the path segment (AP_SEQUENCE or AP_SET), and "value" is the
   ITAD number.  Two ITADs are said to be the same if their
   corresponding <type, value> are the same.

   If the routes to be aggregated have different path attributes, then
   the aggregated path attribute shall satisfy all of the following
   conditions:

      -  All pairs of the type AP_SEQUENCE in the aggregated path MUST
         appear in all of the paths of routes to be aggregated.
      -  All pairs of the type AP_SET in the aggregated path MUST appear
         in at least one of the paths of the initial set (they may
         appear as either AP_SET or AP_SEQUENCE types).
      -  For any pair X of the type AP_SEQUENCE that precedes pair Y in
         the aggregated path, X precedes Y in each path of the initial
         set that contains Y, regardless of the type of Y.
      -  No pair with the same value shall appear more than once in the
         aggregated path, regardless of the pair's type.

   An implementation may choose any algorithm that conforms to these
   rules.  At a minimum, a conformant implementation MUST be able to
   perform the following algorithm that meets all of the above
   conditions:

      -  Determine the longest leading sequence of tuples (as defined
         above) common to all the paths of the routes to be aggregated.
         Make this sequence the leading sequence of the aggregated path.
      -  Set the type of the rest of the tuples from the paths of the
         routes to be aggregated to AP_SET, and append them to the
         aggregated path.
      -  If the aggregated path has more than one tuple with the same
         value (regardless of tuple's type), eliminate all but one such
         tuple by deleting tuples of the type AP_SET from the aggregated
         path.

   An implementation that chooses to provide a path aggregation
   algorithm that retains significant amounts of path information may
   wish to use the procedure of Section 5.4.4.3.



Rosenberg, et. al.          Standards Track                    [Page 33]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


5.4.4.3. Example Path Aggregation Algorithm



   An example algorithm to aggregate two paths works as follows:

      -  Identify the ITADs (as defined in Section 5.4.1) within each
         path attribute that are in the same relative order within both
         path attributes.  Two ITADs, X and Y, are said to be in the
         same order if either X precedes Y in both paths, or if Y
         precedes X in both paths.
      -  The aggregated path consists of ITADs identified in (a) in
         exactly the same order as they appear in the paths to be
         aggregated.  If two consecutive ITADs identified in (a) do not
         immediately follow each other in both of the paths to be
         aggregated, then the intervening ITADs (ITADs that are between
         the two consecutive ITADs that are the same) in both attributes
         are combined into an AP_SET path segment that consists of the
         intervening ITADs from both paths; this segment is then placed
         in between the two consecutive ITADs identified in (a) of the
         aggregated attribute.  If two consecutive ITADs identified in
         (a) immediately follow each other in one attribute, but do not
         follow in another, then the intervening ITADs of the latter are
         combined into an AP_SET path segment; this segment is then
         placed in between the two consecutive ITADs identified in (a)
         of the aggregated path.

   If as a result of the above procedure a given ITAD number appears
   more than once within the aggregated path, all but the last instance
   (rightmost occurrence) of that ITAD number should be removed from the
   aggregated path.

5.4.5. Route Dissemination and AdvertisementPath



   When an LS propagates a route which it has learned from another LS,
   it shall modify the route's AdvertisementPath attribute based on the
   location of the LS to which the route will be sent.

      -  When a LS advertises a route to another LS located in its own
         ITAD, the advertising LS MUST NOT modify the AdvertisementPath
         attribute associated with the route.
      -  When a LS advertises a route to an LS located in a neighboring
         ITAD, then the advertising LS MUST update the AdvertisementPath
         attribute as follows:









Rosenberg, et. al.          Standards Track                    [Page 34]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


         *  If the first path segment of the AdvertisementPath is of
            type AP_SEQUENCE, the local system shall prepend its own
            ITAD number as the last element of the sequence (put it in
            the leftmost position).

         *  If the first path segment of the AdvertisementPath is of
            type AP_SET, the local system shall prepend a new path
            segment of type AP_SEQUENCE to the AdvertisementPath,
            including its own ITAD number in that segment.

5.5. RoutedPath



   Conditional Mandatory: True
   (if ReachableRoutes attribute is present).
   Required Flags: Well-known.
   Potential Flags: None.
   TRIP Type Code: 5.

   This attribute identifies the ITADs through which messages sent using
   this route would pass.  The ITADs in this path are a subset of those
   in the AdvertisementPath.

5.5.1. RoutedPath Syntax



   The syntax of the RoutedPath attribute is the same as that of the
   AdvertisementPath attribute.  See Section 5.4.1.

5.5.2. Route Origination and RoutedPath



   When an LS originates a route it MUST include the RoutedPath
   attribute.

      -  The originating LS shall include its own ITAD number in the
         RoutedPath attribute of all advertisements sent to LSs located
         in neighboring ITADs.  In this case, the ITAD number of the
         originating LS's ITAD will be the only entry in the RoutedPath
         attribute.
      -  The originating LS shall include an empty RoutedPath attribute
         in all advertisements sent to LSs located in its own ITAD.  An
         empty RoutedPath attribute is one whose length field contains
         the value zero.

5.5.3. Route Selection and RoutedPath



   The RoutedPath MAY be used for route selection, and in most cases is
   preferred over the AdvertisementPath for this role.  Some possible
   criteria to be used are the number of hops on the path and the
   presence or absence of particular ITADs on the path.



Rosenberg, et. al.          Standards Track                    [Page 35]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


5.5.4. Aggregation and RoutedPath



   The rules for aggregating RoutedPath attributes are given in Section
   5.4.4.1 and 5.4.4.2, where the term "path" used in Section 5.4.4.1
   and 5.4.4.2 is understood to mean RoutedPath.

5.5.5. Route Dissemination and RoutedPath



   When an LS propagates a route that it learned from another LS, it
   modifies the route's RoutedPath attribute based on the location of
   the LS to which the route is sent.

      -  When an LS advertises a route to another LS located in its own
         ITAD, the advertising LS MUST NOT modify the RoutedPath
         attribute associated with the route.
      -  If the LS has not changed the NextHopServer attribute, then the
         LS MUST NOT change the RoutedPath attribute.
      -  Otherwise, the LS changed the NextHopServer and is advertising
         the route to an LS in another ITAD.  The advertising LS MUST
         update the RoutedPath attribute as follows:

         *  If the first path segment of the RoutedPath is of type
            AP_SEQUENCE, the local system shall prepend its own ITAD
            number as the last element of the sequence (put it in the
            leftmost position).

         *  If the first path segment of the RoutedPath is of type
            AP_SET, the local system shall prepend a new path segment of
            type AP_SEQUENCE to the RoutedPath, including its own ITAD
            number in that segment.

5.6. AtomicAggregate



   Conditional Mandatory: False.
   Required Flags: Well-known.
   Potential Flags: None.
   TRIP Type Code: 6.

   The AtomicAggregate attribute indicates that a route may traverse
   domains not listed in the RoutedPath.  If an LS, when presented with
   a set of overlapping routes from a peer LS, selects the less specific
   route without selecting the more specific route, then the LS includes
   the AtomicAggregate attribute with the routing object.

5.6.1. AtomicAggregate Syntax



   This attribute has length zero (0); the value field is empty.




Rosenberg, et. al.          Standards Track                    [Page 36]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


5.6.2. Route Origination and AtomicAggregate



   Routes are never originated with the AtomicAggregate attribute.

5.6.3. Route Selection and AtomicAggregate



   The AtomicAggregate attribute may be used in route selection - it
   indicates that the RoutedPath may be incomplete.

5.6.4. Aggregation and AtomicAggregate



   If any of the routes to aggregate has the AtomicAggregate attribute,
   then so MUST the resultant aggregate.

5.6.5. Route Dissemination and AtomicAggregate



   If an LS, when presented with a set of overlapping routes from a peer
   LS, selects the less specific route (see Section 0) without selecting
   the more specific route, then the LS MUST include the AtomicAggregate
   attribute with the routing object (if it is not already present).

   An LS receiving a routing object with an AtomicAggregate attribute
   MUST NOT make the set of destinations more specific when advertising
   it to other LSs, and MUST NOT remove the attribute when propagating
   this object to a peer LS.

5.7. LocalPreference



   Conditional Mandatory: False.
   Required Flags: Well-known.
   Potential Flags: None.
   TRIP Type Code: 7.

   The LocalPreference attribute is only used intra-domain, it indicates
   the local LS's preference for the routing object to other LSs within
   the same domain.  This attribute MUST NOT be included when
   communicating to an LS in another domain, and MUST be included over
   intra-domain links.

5.7.1. LocalPreference Syntax



   The LocalPreference attribute is a 4-octet unsigned numeric value.  A
   higher value indicates a higher preference.








Rosenberg, et. al.          Standards Track                    [Page 37]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


5.7.2. Route Origination and LocalPreference



   Routes MUST NOT be originated with the LocalPreference attribute to
   inter-domain peers.  Routes to intra-domain peers MUST be originated
   with the LocalPreference attribute.

5.7.3. Route Selection and LocalPreference



   The LocalPreference attribute allows one LS in a domain to calculate
   a preference for a route, and to communicate this preference to other
   LSs within the domain.

5.7.4. Aggregation and LocalPreference



   The LocalPreference attribute is not affected by aggregation.

5.7.5. Route Dissemination and LocalPreference



   An LS MUST include the LocalPreference attribute when communicating
   with peer LSs within its own domain.  An LS MUST NOT include the
   LocalPreference attribute when communicating with LSs in other
   domains.  LocalPreference attributes received from inter-domain peers
   MUST be ignored.

5.8. MultiExitDisc



   Conditional Mandatory: False.
   Required Flags: Well-known.
   Potential Flags: None.
   TRIP Type Code: 8.

   When two ITADs are connected by more than one set of peers, the
   MultiExitDisc attribute may be used to specify preferences for routes
   received over one of those links versus routes received over other
   links.  The MultiExitDisc parameter is used only for route selection.

5.8.1. MultiExitDisc Syntax



   The MultiExitDisc attribute carries a 4-octet unsigned numeric value.
   A higher value represents a more preferred routing object.

5.8.2. Route Origination and MultiExitDisc



   Routes originated to intra-domain peers MUST NOT be originated with
   the MultiExitDisc attribute.  When originating a route to an inter-
   domain peer, the MultiExitDisc attribute may be included.





Rosenberg, et. al.          Standards Track                    [Page 38]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


5.8.3. Route Selection and MultiExitDisc



   The MultiExitDisc attribute is used to express a preference when
   there are multiple links between two domains.  If all other factors
   are equal, then a route with a higher MultiExitDisc attribute is
   preferred over a route with a lower MultiExitDisc attribute.

5.8.4. Aggregation and MultiExitDisc



   Routes with differing MultiExitDisc parameters MUST NOT be
   aggregated.  Routes with the same value in the MultiExitDisc
   attribute MAY be aggregated and the same MultiExitDisc attribute
   attached to the aggregated object.

5.8.5. Route Dissemination and MultiExitDisc



   If received from a peer LS in another domain, an LS MAY propagate the
   MultiExitDisc to other LSs within its domain.  The MultiExitDisc
   attribute MUST NOT be propagated to LSs in other domains.

   An LS may add the MultiExitDisc attribute when propagating routing
   objects to an LS in another domain.  The inclusion of the
   MultiExitDisc attribute is a matter of policy, as is the value of the
   attribute.

5.9. Communities



   Conditional Mandatory: False.
   Required Flags: Not Well-Known, Independent Transitive.
   Potential Flags: None.
   TRIP Type Code: 9.

   A community is a group of destinations that share some common
   property.

   The Communities attribute is used to group destinations so that the
   routing decision can be based on the identity of the group.  Using
   the Communities attribute should significantly simplify the
   distribution of routing information by providing an administratively
   defined aggregation unit.

   Each ITAD administrator may define the communities to which a
   particular route belongs.  By default, all routes belong to the
   general Internet Telephony community.

   As an example, the Communities attribute could be used to define an
   alliance between a group of Internet Telephony service providers for
   a specific subset of routing information.  In this case, members of



Rosenberg, et. al.          Standards Track                    [Page 39]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   that alliance would accept only routes for destinations in this group
   that are advertised by other members of the alliance.  Other
   destinations would be more freely accepted.  To achieve this, a
   member would tag each route with a designated Community attribute
   value before disseminating it.  This relieves the members of such an
   alliance, from the responsibility of keeping track of the identities
   of all other members of that alliance.

   Another example use of the Communities attribute is with aggregation.
   It is often useful to advertise both the aggregate route and the
   component more-specific routes that were used to form the aggregate.
   These information components are only useful to the neighboring TRIP
   peer, and perhaps the ITAD of the neighboring TRIP peer, so it is
   desirable to filter out the component routes.  This can be achieved
   by specifying a Community attribute value that the neighboring peers
   will match and filter on.  That way it can be assured that the more
   specific routes will not propagate beyond their desired scope.

5.9.1. Syntax of Communities



   The Communities attribute is of variable length.  It consists of a
   set of 8-octet values, each of which specifies a community.  The
   first 4 octets of the Community value are the Community ITAD Number
   and the next 4 octets are the Community ID.

   0                   1                   2                   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
   +---------------+---------------+--------------+----------------+
   |                       Community ITAD Number 1                 |
   +---------------+---------------+--------------+----------------+
   |                         Community ID 1                        |
   +---------------+---------------+--------------+----------------+
   |                       . . . . . . . . .
   +---------------+---------------+--------------+----------------+

                    Figure 14: Communities Syntax

   For administrative assignment, the following assumptions may be made:

      The Community attribute values starting with a Community ITAD
      Number of 0x00000000 are hereby reserved.

   The following communities have global significance and their
   operation MUST be implemented in any Community attribute-aware TRIP
   LS.






Rosenberg, et. al.          Standards Track                    [Page 40]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


      -  NO_EXPORT (Community ITAD Number = 0x00000000 and Community ID
         = 0xFFFFFF01).  Any received route with a community attribute
         containing this value MUST NOT be advertised outside of the
         receiving TRIP ITAD.

   Other community values MUST be encoded using an ITAD number in the
   four most significant octets.  The semantics of the final four octets
   (the Community ID octets) may be defined by the ITAD (e.g., ITAD 690
   may define research, educational, and commercial community IDs that
   may be used for policy routing as defined by the operators of that
   ITAD).

5.9.2. Route Origination and Communities



   The Communities attribute is not well-known.  If a route has a
   Communities attribute associated with it, the LS MUST include that
   attribute in the advertisement it originates.

5.9.3. Route Selection and Communities



   The Communities attribute may be used for route selection.  A route
   that is a member of a certain community may be preferred over another
   route that is not a member of that community.  Likewise, routes
   without a certain community value may be excluded from consideration.

5.9.4. Aggregation and Communities



   If a set of routes is to be aggregated and the resultant aggregate
   does not carry an Atomic_Aggregate attribute, then the resulting
   aggregate should have a Communities attribute that contains the union
   of the Community attributes of the aggregated routes.

5.9.5. Route Dissemination and Communities



   An LS may manipulate the Communities attribute before disseminating a
   route to a peer.  Community attribute manipulation may include adding
   communities, removing communities, adding a Communities attribute (if
   none exists), deleting the Communities attribute, etc.

5.10. ITAD Topology



   Conditional Mandatory: False.
   Required Flags: Well-known, Link-State encapsulated.
   Potential Flags: None.
   TRIP Type Code: 10.






Rosenberg, et. al.          Standards Track                    [Page 41]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   Within an ITAD, each LS must know the status of other LSs so that LS
   failure can be detected.  To do this, each LS advertises its internal
   topology to other LSs within the domain.  When an LS detects that
   another LS is no longer active, the information sourced by that LS
   can be deleted (the Adj-TRIB-In for that peer may be cleared).  The
   ITAD Topology attribute is used to communicate this information to
   other LSs within the domain.

   An LS MUST send a topology update each time it detects a change in
   its internal peer set.  The topology update may be sent in an UPDATE
   message by itself or it may be piggybacked on an UPDATE message which
   includes ReachableRoutes and/or WithdrawnRoutes information.

   When an LS receives a topology update from an internal LS, it MUST
   recalculate which LSs are active within the ITAD via a connectivity
   algorithm on the topology.

5.10.1. ITAD Topology Syntax



   The ITAD Topology attribute indicates the LSs with which the LS is
   currently peering.  The attribute consists of a list of the TRIP
   Identifiers with which the LS is currently peering, the format is
   given in  Figure 15.  This attribute MUST use the link-state
   encapsulation as defined in Section 4.3.2.4.

    0                   1                   2                   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
   +---------------+---------------+--------------+----------------+
   |                        TRIP Identifier 1                      |
   +---------------+---------------+--------------+----------------+
   |                        TRIP Identifier 2 ...                  |
   +---------------+---------------+--------------+----------------+

                   Figure 15: ITAD Topology Syntax

5.10.2. Route Origination and ITAD Topology



   The ITAD Topology attribute is independent of any routes in the
   UPDATE.  Whenever the set of internal peers of an LS changes, it MUST
   create an UPDATE with the ITAD Topology Attribute included listing
   the current set of internal peers.  The LS MUST include this
   attribute in the first UPDATE it sends to a peer after the peering
   session is established.








Rosenberg, et. al.          Standards Track                    [Page 42]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


5.10.3. Route Selection and ITAD Topology



   This attribute is independent of any routing information in the
   UPDATE.  When an LS receives an UPDATE with an ITAD Topology
   attribute, it MUST compute the set of LSs currently active in the
   domain by performing a connectivity test on the ITAD topology as
   given by the set of originated ITAD Topology attributes.  The LS MUST
   locally purge the Adj-TRIB-In for any LS that is no longer active in
   the domain.  The LS MUST NOT propagate this purging information to
   other LSs as they will make a similar decision.

5.10.4. Aggregation and ITAD Topology



   This information is not aggregated.

5.10.5. Route Dissemination and ITAD Topology



   An LS MUST ignore the attribute if received from a peer in another
   domain.  An LS MUST NOT send this attribute to an inter-domain peer.

5.11. ConvertedRoute



   Conditional Mandatory: False.
   Required Flags: Well-known.
   Potential Flags: None.
   TRIP Type Code: 12.

   The ConvertedRoute attribute indicates that an intermediate LS has
   altered the route by changing the route's Application Protocol.  For
   example, if an LS receives a route with Application Protocol X and
   changes the Application Protocol to Y before advertising the route to
   an external peer, the LS MUST include the ConvertedRoute attribute.
   The attribute is an indication that the advertised application
   protocol will not be used end-to-end, i.e., the information
   advertised about this route is not complete.

5.11.1. ConvertedRoute Syntax



   This attribute has length zero (0); the value field is empty.

5.11.2. Route Origination and ConvertedRoute



   Routes are never originated with the ConvertedRoute attribute.








Rosenberg, et. al.          Standards Track                    [Page 43]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


5.11.3. Route Selection and ConvertedRoute



   The ConvertedRoute attribute may be used in route selection - it
   indicates that advertised routing information is not complete.

5.11.4. Aggregation and ConvertedRoute



   If any of the routes to aggregate has the ConvertedRoute attribute,
   then so MUST the resultant aggregate.

5.11.5. Route Dissemination and ConvertedRoute



   If an LS changes the Application Protocol of a route before
   advertising the route to an external peer, the LS MUST include the
   ConvertedRoute attribute.

5.12. Considerations for Defining New TRIP Attributes



   Any proposal for defining new TRIP attributes should specify the
   following:

      -  the use of this attribute,
      -  the attribute's flags,
      -  the attribute's syntax,
      -  how the attribute works with route origination,
      -  how the attribute works with route aggregation, and
      -  how the attribute works with route dissemination and the
         attribute's scope (e.g., intra-domain only like
         LocalPreference)

   IANA will manage the assignment of TRIP attribute type codes to new
   attributes.

6. TRIP Error Detection and Handling



   This section describes errors to be detected and the actions to be
   taken while processing TRIP messages.

   When any of the conditions described here are detected, a
   NOTIFICATION message with the indicated Error Code, Error Subcode,
   and Data fields MUST be sent, and the TRIP connection MUST be closed.
   If no Error Subcode is specified, then a zero Subcode MUST be used.

   The phrase "the TRIP connection is closed" means that the transport
   protocol connection has been closed and that all resources for that
   TRIP connection have been de-allocated.  If the connection was
   inter-domain, then routing table entries associated with the remote
   peer MUST be marked as invalid.  Routing table entries MUST NOT be



Rosenberg, et. al.          Standards Track                    [Page 44]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   marked as invalid if an internal peering session is terminated.  The
   fact that the routes have been marked as invalid is passed to other
   TRIP peers before the routes are deleted from the system.

   Unless specified explicitly, the Data field of the NOTIFICATION
   message that is sent to indicate an error MUST be empty.

6.1. Message Header Error Detection and Handling



   All errors detected while processing the Message Header are indicated
   by sending the NOTIFICATION message with the Error Code Message
   Header Error.  The Error Subcode elaborates on the specific nature of
   the error.  The error checks in this section MUST be performed by
   each LS upon receipt of every message.

   If the Length field of the message header is less than 3 or greater
   than 4096, or if the Length field of an OPEN message is less than the
   minimum length of the OPEN message, or if the Length field of an
   UPDATE message is less than the minimum length of the UPDATE message,
   or if the Length field of a KEEPALIVE message is not equal to 3, or
   if the Length field of a NOTIFICATION message is less than the
   minimum length of the NOTIFICATION message, then the Error Subcode
   MUST be set to Bad Message Length.  The Data field contains the
   erroneous Length field.

   If the Type field of the message header is not recognized, then the
   Error Subcode MUST be set to "Bad Message Type."  The Data field
   contains the erroneous Type field.

6.2. OPEN Message Error Detection and Handling



   All errors detected while processing the OPEN message are indicated
   by sending the NOTIFICATION message with the Error Code "OPEN Message
   Error."  The Error Subcode elaborates on the specific nature of the
   error.  The error checks in this section MUST be performed by each LS
   upon receipt of every OPEN message.

   If the version number contained in the Version field of the received
   OPEN message is not supported, then the Error Subcode MUST be set to
   "Unsupported Version Number."  The Data field is a 1-octet unsigned
   integer, which indicates the largest locally supported version
   number, which is less than the version of the remote TRIP peer bid
   (as indicated in the received OPEN message).

   If the ITAD field of the OPEN message is unacceptable, then the Error
   Subcode MUST be set to "Bad Peer ITAD."  The determination of
   acceptable ITAD numbers is outside the scope of this protocol.




Rosenberg, et. al.          Standards Track                    [Page 45]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   If the Hold Time field of the OPEN message is unacceptable, then the
   Error Subcode MUST be set to "Unacceptable Hold Time."  An
   implementation MUST reject Hold Time values of one or two seconds.
   An implementation MAY reject any proposed Hold Time.  An
   implementation that accepts a Hold Time MUST use the negotiated value
   for the Hold Time.

   If the TRIP Identifier field of the OPEN message is not valid, then
   the Error Subcode MUST be set to "Bad TRIP Identifier."  A TRIP
   identifier is 4-octets in length and can take any value.  An LS
   considers the TRIP Identifier invalid if it already has an open
   connection with another peer LS that has the same ITAD and TRIP
   Identifier.

   Any two LSs within the same ITAD MUST NOT have equal TRIP Identifier
   values.  This restriction does not apply to LSs in different ITADs
   since the purpose is to uniquely identify an LS using its TRIP
   Identifier and its ITAD number.

   If one of the Optional Parameters in the OPEN message is not
   recognized, then the Error Subcode MUST be set to "Unsupported
   Optional Parameters."

   If the Optional Parameters of the OPEN message include Capability
   Information with an unsupported capability (unsupported in either
   capability type or value), then the Error Subcode MUST be set to
   "Unsupported Capability," and the entirety of the unsupported
   capabilities MUST be listed in the Data field of the NOTIFICATION
   message.

   If the Optional Parameters of the OPEN message include Capability
   Information which does not match the receiving LS's capabilities,
   then the Error Subcode MUST be set to "Capability Mismatch," and the
   entirety of the mismatched capabilities MUST be listed in the Data
   field of the NOTIFICATION message.

6.3. UPDATE Message Error Detection and Handling



   All errors detected while processing the UPDATE message are indicated
   by sending the NOTIFICATION message with the Error Code "UPDATE
   Message Error."  The Error Subcode elaborates on the specific nature
   of the error.  The error checks in this section MUST be performed by
   each LS upon receipt of every UPDATE message.  These error checks
   MUST occur before flooding procedures are invoked with internal
   peers.






Rosenberg, et. al.          Standards Track                    [Page 46]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   If any recognized attribute has Attribute Flags that conflict with
   the Attribute Type Code, then the Error Subcode MUST be set to
   "Attribute Flags Error."  The Data field contains the erroneous
   attribute (type, length and value).

   If any recognized attribute has an Attribute Length that conflicts
   with the expected length (based on the attribute type code), then the
   Error Subcode MUST be set to "Attribute Length Error."  The Data
   field contains the erroneous attribute (type, length and value).

   If any of the mandatory (i.e., conditional mandatory attribute and
   the conditions for including it in the UPDATE message are fulfilled)
   well-known attributes are not present, then the Error Subcode MUST be
   set to "Missing Well-known Mandatory Attribute."  The Data field
   contains the Attribute Type Code of the missing well-known
   conditional mandatory attributes.

   If any of the well-known attributes are not recognized, then the
   Error Subcode MUST be set to "Unrecognized Well-known Attribute."
   The Data field contains the unrecognized attribute (type, length and
   value).

   If any attribute has a syntactically incorrect value, or an undefined
   value, then the Error Subcode is set to "Invalid Attribute."  The
   Data field contains the incorrect attribute (type, length and value).
   Such a NOTIFICATION message is sent, for example, when a
   NextHopServer attribute is received with an invalid address.

   The information carried by the AdvertisementPath attribute is checked
   for ITAD loops.  ITAD loop detection is done by scanning the full
   AdvertisementPath, and checking that the ITAD number of the local
   ITAD does not appear in the AdvertisementPath.  If the local ITAD
   number appears in the AdvertisementPath, then the route MAY be stored
   in the Adj-TRIB-In.  However unless the LS is configured to accept
   routes with its own ITAD in the advertisement path, the route MUST
   not be passed to the TRIP Decision Process.  The operation of an LS
   that is configured to accept routes with its own ITAD number in the
   advertisement path are outside the scope of this document.

   If the UPDATE message was received from an internal peer and either
   the WithdrawnRoutes, ReachableRoutes, or ITAD Topology attribute does
   not have the Link-State Encapsulation flag set, then the Error
   Subcode is set to "Invalid Attribute" and the data field contains the
   attribute.  Likewise, the attribute is invalid if received from an
   external peer and the Link-State Flag is set.

   If any attribute appears more than once in the UPDATE message, then
   the Error Subcode is set to "Malformed Attribute List."



Rosenberg, et. al.          Standards Track                    [Page 47]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


6.4. NOTIFICATION Message Error Detection and Handling



   If a peer sends a NOTIFICATION message, and there is an error in that
   message, there is unfortunately no means of reporting this error via
   a subsequent NOTIFICATION message.  Any such error, such as an
   unrecognized Error Code or Error Subcode, should be noticed, logged
   locally, and brought to the attention of the administration of the
   peer.  The means to do this, however, are outside the scope of this
   document.

6.5. Hold Timer Expired Error Handling



   If a system does not receive successive messages within the period
   specified by the negotiated Hold Time, then a NOTIFICATION message
   with a "Hold Timer Expired" Error Code MUST be sent and the TRIP
   connection MUST be closed.

6.6. Finite State Machine Error Handling



   An error detected by the TRIP Finite State Machine (e.g., receipt of
   an unexpected event) MUST result in sending a NOTIFICATION message
   with the Error Code "Finite State Machine Error" and the TRIP
   connection MUST be closed.

6.7. Cease



   In the absence of any fatal errors (that are indicated in this
   section), a TRIP peer MAY choose at any given time to close its TRIP
   connection by sending the NOTIFICATION message with the Error Code
   "Cease."  However, the Cease NOTIFICATION message MUST NOT be used
   when a fatal error indicated by this section exists.

6.8. Connection Collision Detection



   If a pair of LSs try simultaneously to establish a transport
   connection to each other, then two parallel connections between this
   pair of speakers might well be formed.  We refer to this situation as
   connection collision.  Clearly, one of these connections must be
   closed.

   Based on the value of the TRIP Identifier, a convention is
   established for detecting which TRIP connection is to be preserved
   when a collision occurs.  The convention is to compare the TRIP
   Identifiers of the peers involved in the collision and to retain only
   the connection initiated by the LS with the higher-valued TRIP
   Identifier.





Rosenberg, et. al.          Standards Track                    [Page 48]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   Upon receipt of an OPEN message, the local LS MUST examine all of its
   connections that are in the OpenConfirm state.  An LS MAY also
   examine connections in an OpenSent state if it knows the TRIP
   Identifier of the peer by means outside of the protocol.  If among
   these connections there is a connection to a remote LS, whose TRIP
   Identifier equals the one in the OPEN message, then the local LS MUST
   perform the following collision resolution procedure:

   The TRIP Identifier and ITAD of the local LS is compared to the TRIP
   Identifier and ITAD of the remote LS (as specified in the OPEN
   message).  TRIP Identifiers are treated as 4-octet unsigned integers
   for comparison.

   If the value of the local TRIP Identifier is less than the remote
   one, or if the two TRIP Identifiers are equal and the value of the
   ITAD of the local LS is less than value of the ITAD of the remote LS,
   then the local LS MUST close the TRIP connection that already exists
   (the one that is already in the OpenConfirm state), and accept the
   TRIP connection initiated by the remote LS:

      1. Otherwise, the local LS closes the newly created TRIP
         connection and continues to use the existing one (the one that
         is already in the OpenConfirm state).
      2. If a connection collision occurs with an existing TRIP
         connection that is in the Established state, then the LS MUST
         unconditionally close off the newly created connection.  Note
         that a connection collision cannot be detected with connections
         in Idle, Connect, or Active states.
      3. To close the TRIP connection (that results from the collision
         resolution procedure), an LS MUST send a NOTIFICATION message
         with the Error Code "Cease" and the TRIP connection MUST be
         closed.

7. TRIP Version Negotiation



   Peer LSs may negotiate the version of the protocol by making multiple
   attempts to open a TRIP connection, starting with the highest version
   number each supports.  If an open attempt fails with an Error Code
   "OPEN Message Error" and an Error Subcode "Unsupported Version
   Number," then the LS has available the version number it tried, the
   version number its peer tried, the version number passed by its peer
   in the NOTIFICATION message, and the version numbers that it
   supports.  If the two peers support one or more common versions, then
   this will allow them to rapidly determine the highest common version.
   In order to support TRIP version negotiation, future versions of TRIP
   must retain the format of the OPEN and NOTIFICATION messages.





Rosenberg, et. al.          Standards Track                    [Page 49]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


8. TRIP Capability Negotiation



   An LS MAY include the Capabilities Option in its OPEN message to a
   peer to indicate the capabilities supported by the LS.  An LS
   receiving an OPEN message MUST NOT use any capabilities that were not
   included in the OPEN message of the peer when communicating with that
   peer.

9. TRIP Finite State Machine



   This section specifies TRIP operation in terms of a Finite State
   Machine (FSM).  Following is a brief summary and overview of TRIP
   operations by state as determined by this FSM.  A condensed version
   of the TRIP FSM is found in Appendix 1.  There is one TRIP FSM per
   peer and these FSMs operate independently.

   Idle state:
   Initially TRIP is in the Idle state for each peer.  In this state,
   TRIP refuses all incoming connections.  No resources are allocated to
   the peer.  In response to the Start event (initiated by either the
   system or the operator), the local system initializes all TRIP
   resources, starts the ConnectRetry timer, initiates a transport
   connection to the peer, starts listening for a connection that may be
   initiated by the remote TRIP peer, and changes its state to Connect.
   The exact value of the ConnectRetry timer is a local matter, but
   should be sufficiently large to allow TCP initialization.

   If an LS detects an error, it closes the transport connection and
   changes its state to Idle.  Transitioning from the Idle state
   requires generation of the Start event.  If such an event is
   generated automatically, then persistent TRIP errors may result in
   persistent flapping of the LS.  To avoid such a condition, Start
   events MUST NOT be generated immediately for a peer that was
   previously transitioned to Idle due to an error.  For a peer that was
   previously transitioned to Idle due to an error, the time between
   consecutive Start events, if such events are generated automatically,
   MUST exponentially increase.  The value of the initial timer SHOULD
   be 60 seconds, and the time SHOULD be at least doubled for each
   consecutive retry up to some maximum value.

   Any other event received in the Idle state is ignored.

   Connect State:
   In this state, an LS is waiting for a transport protocol connection
   to be completed to the peer, and is listening for inbound transport
   connections from the peer.





Rosenberg, et. al.          Standards Track                    [Page 50]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   If the transport protocol connection succeeds, the local LS clears
   the ConnectRetry timer, completes initialization, sends an OPEN
   message to its peer, sets its Hold Timer to a large value, and
   changes its state to OpenSent.  A Hold Timer value of 4 minutes is
   suggested.

   If the transport protocol connect fails (e.g., retransmission
   timeout), the local system restarts the ConnectRetry timer, continues
   to listen for a connection that may be initiated by the remote LS,
   and changes its state to Active state.

   In response to the ConnectRetry timer expired event, the local LS
   cancels any outstanding transport connection to the peer, restarts
   the ConnectRetry timer, initiates a transport connection to the
   remote LS, continues to listen for a connection that may be initiated
   by the remote LS, and stays in the Connect state.

   If the local LS detects that a remote peer is trying to establish a
   connection to it and the IP address of the peer is not an expected
   one, then the local LS rejects the attempted connection and continues
   to listen for a connection from its expected peers without changing
   state.

   If an inbound transport protocol connection succeeds, the local LS
   clears the ConnectRetry timer, completes initialization, sends an
   OPEN message to its peer, sets its Hold Timer to a large value, and
   changes its state to OpenSent.  A Hold Timer value of 4 minutes is
   suggested.

   The Start event is ignored in the Connect state.

   In response to any other event (initiated by either the system or the
   operator), the local system releases all TRIP resources associated
   with this connection and changes its state to Idle.

   Active state:
   In this state, an LS is listening for an inbound connection from the
   peer, but is not in the process of initiating a connection to the
   peer.

   If an inbound transport protocol connection succeeds, the local LS
   clears the ConnectRetry timer, completes initialization, sends an
   OPEN message to its peer, sets its Hold Timer to a large value, and
   changes its state to OpenSent.  A Hold Timer value of 4 minutes is
   suggested.






Rosenberg, et. al.          Standards Track                    [Page 51]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   In response to the ConnectRetry timer expired event, the local system
   restarts the ConnectRetry timer, initiates a transport connection to
   the TRIP peer, continues to listen for a connection that may be
   initiated by the remote TRIP peer, and changes its state to Connect.

   If the local LS detects that a remote peer is trying to establish a
   connection to it and the IP address of the peer is not an expected
   one, then the local LS rejects the attempted connection and continues
   to listen for a connection from its expected peers without changing
   state.

   Start event is ignored in the Active state.

   In response to any other event (initiated by either the system or the
   operator), the local system releases all TRIP resources associated
   with this connection and changes its state to Idle.

   OpenSent state:
   In this state, an LS has sent an OPEN message to its peer and is
   waiting for an OPEN message from its peer.  When an OPEN message is
   received, all fields are checked for correctness.  If the TRIP
   message header checking or OPEN message checking detects an error
   (see Section 6.2) or a connection collision (see Section 6.8), the
   local system sends a NOTIFICATION message and changes its state to
   Idle.

   If there are no errors in the OPEN message, TRIP sends a KEEPALIVE
   message and sets a KeepAlive timer.  The Hold Timer, which was
   originally set to a large value (see above), is replaced with the
   negotiated Hold Time value (see Section 4.2).  If the negotiated Hold
   Time value is zero, then the Hold Time timer and KeepAlive timers are
   not started.  If the value of the ITAD field is the same as the local
   ITAD number, then the connection is an "internal" connection;
   otherwise, it is "external" (this will affect UPDATE processing).
   Finally, the state is changed to OpenConfirm.

   If the local LS detects that a remote peer is trying to establish a
   connection to it and the IP address of the peer is not an expected
   one, then the local LS rejects the attempted connection and continues
   to listen for a connection from its expected peers without changing
   state.

   If a disconnect notification is received from the underlying
   transport protocol, the local LS closes the transport connection,
   restarts the ConnectRetry timer, continues to listen for a connection
   that may be initiated by the remote TRIP peer, and goes into the
   Active state.




Rosenberg, et. al.          Standards Track                    [Page 52]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   If the Hold Timer expires, the local LS sends a NOTIFICATION message
   with the Error Code "Hold Timer Expired" and changes its state to
   Idle.

   In response to the Stop event (initiated by either system or
   operator) the local LS sends a NOTIFICATION message with the Error
   Code "Cease" and changes its state to Idle.

   The Start event is ignored in the OpenSent state.

   In response to any other event the local LS sends a NOTIFICATION
   message with the Error Code "Finite State Machine Error" and changes
   its state to Idle.

   Whenever TRIP changes its state from OpenSent to Idle, it closes the
   transport connection and releases all resources associated with that
   connection.

   OpenConfirm state:
   In this state, an LS has sent an OPEN to its peer, received an OPEN
   from its peer, and sent a KEEPALIVE in response to the OPEN.  The LS
   is now waiting for a KEEPALIVE or NOTIFICATION message in response to
   its OPEN.

   If the local LS receives a KEEPALIVE message, it changes its state to
   Established.

   If the Hold Timer expires before a KEEPALIVE message is received, the
   local LS sends NOTIFICATION message with the Error Code "Hold Timer
   Expired" and changes its state to Idle.

   If the local LS receives a NOTIFICATION message, it changes its state
   to Idle.

   If the KeepAlive timer expires, the local LS sends a KEEPALIVE
   message and restarts its KeepAlive timer.

   If a disconnect notification is received from the underlying
   transport protocol, the local LS closes the transport connection,
   restarts the ConnectRetry timer, continues to listen for a connection
   that may be initiated by the remote TRIP peer, and goes into the
   Active state.

   In response to the Stop event (initiated by either the system or the
   operator) the local LS sends NOTIFICATION message with the Error Code
   "Cease" and changes its state to Idle.

   The Start event is ignored in the OpenConfirm state.



Rosenberg, et. al.          Standards Track                    [Page 53]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   In response to any other event the local LS sends a NOTIFICATION
   message with the Error Code "Finite State Machine Error" and changes
   its state to Idle.

   Whenever TRIP changes its state from OpenConfirm to Idle, it closes
   the transport connection and releases all resources associated with
   that connection.

   Established state:
   In the Established state, an LS can exchange UPDATE, NOTIFICATION,
   and KEEPALIVE messages with its peer.

   If the negotiated Hold Timer is zero, then no procedures are
   necessary for keeping a peering session alive.  If the negotiated
   Hold Time value is non-zero, the procedures of this paragraph apply.
   If the Hold Timer expires, the local LS sends a NOTIFICATION message
   with the Error Code "Hold Timer Expired" and changes its state to
   Idle.  If the KeepAlive Timer expires, then the local LS sends a
   KeepAlive message and restarts the KeepAlive Timer.  If the local LS
   receives an UPDATE or KEEPALIVE message, then it restarts its Hold
   Timer.  Each time the LS sends an UPDATE or KEEPALIVE message, it
   restarts its KeepAlive Timer.

   If the local LS receives a NOTIFICATION message, it changes its state
   to Idle.

   If the local LS receives an UPDATE message and the UPDATE message
   error handling procedure (see Section6.3) detects an error, the local
   LS sends a NOTIFICATION message and changes its state to Idle.

   If a disconnect notification is received from the underlying
   transport protocol, the local LS changes its state to Idle.

   In response to the Stop event (initiated by either the system or the
   operator), the local LS sends a NOTIFICATION message with the Error
   Code "Cease" and changes its state to Idle.

   The Start event is ignored in the Established state.

   In response to any other event, the local LS sends a NOTIFICATION
   message with Error Code "Finite State Machine Error" and changes its
   state to Idle.

   Whenever TRIP changes its state from Established to Idle, it closes
   the transport connection and releases all resources associated with
   that connection.  Additionally, if the peer is an external peer, the
   LS deletes all routes derived from that connection.




Rosenberg, et. al.          Standards Track                    [Page 54]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


10. UPDATE Message Handling



   An UPDATE message may be received only in the Established state.
   When an UPDATE message is received, each field is checked for
   validity as specified in Section 6.3.  The rest of this section
   presumes that the UPDATE message has passed the error-checking
   procedures of Section 6.3.

   If the UPDATE message was received from an internal peer, the
   flooding procedures of Section 10.1 MUST be applied.  The flooding
   process synchronizes the Loc-TRIBs of all LSs within the domain.
   Certain routes within the UPDATE may be marked as old or duplicates
   by the flooding process and are ignored during the rest of the UPDATE
   processing.

   If the UPDATE message contains withdrawn routes, then the
   corresponding previously advertised routes shall be removed from the
   Adj-TRIB-In.  This LS MUST rerun its Decision Process since the
   previously advertised route is no longer available for use.

   If the UPDATE message contains a route, then the route MUST be placed
   in the appropriate Adj-TRIB-In, and the following additional actions
   MUST be taken:

      1. If its destinations are identical to those of a route currently
         stored in the Adj-TRIB-In, then the new route MUST replace the
         older route in the Adj-TRIB-In, thus implicitly withdrawing the
         older route from service.  The LS MUST rerun its Decision
         Process since the older route is no longer available for use.
      2. If the new route is more specific than an earlier route
         contained in the Adj-TRIB-In and has identical attributes, then
         no further actions are necessary.
      3. If the new route is more specific than an earlier route
         contained in the Adj-TRIB-In but does not have identical
         attributes, then the LS MUST run its Decision Process since the
         more specific route has implicitly made a portion of the less
         specific route unavailable for use.
      4. If the new route has destinations that are not present in any
         of the routes currently stored in the Adj-TRIB-In, then the LS
         MUST run its Decision Process.
      5. If the new route is less specific than an earlier route
         contained in the Adj-TRIB-In, the LS MUST run its Decision
         Process on the set of destinations that are described only by
         the less specific route.







Rosenberg, et. al.          Standards Track                    [Page 55]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


10.1. Flooding Process



   When an LS receives an UPDATE message from an internal peer, the LS
   floods the new information from that message to all of its other
   internal peers.  Flooding is used to efficiently synchronize all of
   the LSs within a domain without putting any constraints on the
   domain's internal topology.  The flooding mechanism is based on the
   techniques used in OSPF [4] and SCSP [6].  One may argue that TRIP's
   flooding process is in reality a controlled broadcast mechanism.

10.1.1. Database Information



   The LS MUST maintain the sequence number and originating TRIP
   identifier for each link-state encapsulated attribute in an internal
   Adj-TRIB-In.  These values are included with the route in the
   ReachableRoutes, WithdrawnRoutes, and ITAD Topology attributes.  The
   originating TRIP identifier gives the internal LS that originated
   this route into the ITAD, the sequence number gives the version of
   this route at the originating LS.

10.1.2. Determining Newness



   For each route in the ReachableRoutes or WithdrawnRoutes field, the
   LS decides if the route is new or old.  This is determined by
   comparing the Sequence Number of the route in the UPDATE with the
   Sequence Number of the route saved in the Adj-TRIB-In.  The route is
   new if either the route does not exist in the Adj-TRIB-In for the
   originating LS, or if the route does exist in the Adj-TRIB-In but the
   Sequence Number in the UPDATE is greater than the Sequence Number
   saved in the Adj-TRIBs-In.  Note that the newness test is
   independently applied to each link-state encapsulated attribute in
   the UPDATE (WithdrawnRoutes or ReachableRoutes or ITAD Topology).

10.1.3. Flooding



   Each route in the ReachableRoutes or WithdrawnRoutes field that is
   determined to be old is ignored in further processing.  If the route
   is determined to be new then the following actions occur.

   If the route is being withdrawn, then the LS MUST flood the withdrawn
   route to all other internal peers, and MUST mark the route as
   withdrawn.  An LS MUST maintain routes marked as withdrawn in its
   databases for MaxPurgeTime seconds.

   If the route is being updated, then the LS MUST update the route in
   the Adj-TRIB-In and MUST flood it to all other internal peers.





Rosenberg, et. al.          Standards Track                    [Page 56]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   If these procedures result in changes to the Adj-TRIB-In, then the
   route is also made available for local route processing as described
   early in Section 10.

   To implement flooding, the following is recommended.  All routes
   received in a single UPDATE message that are determined to be new
   should be forwarded to all other internal peers in a single UPDATE
   message.  Other variations of flooding are possible, but the local LS
   MUST ensure that each new route (and any associated attributes)
   received from an internal peer get forwarded to every other internal
   peer.

10.1.4. Sequence Number Considerations



   The Sequence Number is used to determine when one version of a Route
   is newer than another version of a route.  A larger Sequence Number
   indicates a newer version.  The Sequence Number is assigned by the LS
   originating the route into the local ITAD.  The Sequence Number is an
   unsigned 4-octet integer in the range of 1 thru 2^31-1 MinSequenceNum
   thru MaxSequenceNum).  The value 0 is reserved.  When an LS first
   originates a route (including when the LS restarts/reboots) into its
   ITAD, it MUST originate it with a Sequence Number of MinSequenceNum.
   Each time the route is updated within the ITAD by the originator, the
   Sequence Number MUST be increased.

   If it is ever the case that the sequence number is MaxSequenceNum-1
   and it needs to be increased, then the TRIP module of the LS MUST be
   disabled for a period of TripDisableTime so that all routes
   originated by this LS with high sequence numbers can be removed.

10.1.5. Purging a Route Within the ITAD



   To withdraw a route that it originated within the ITAD, an LS
   includes the route in the WithdrawnRoutes field of an UPDATE message.
   The Sequence Number MUST be greater than the last valid version of
   the route.  The LS MAY choose to use a sequence number of
   MaxSequenceNum when withdrawing routes within its ITAD, but this is
   not required.

   After withdrawing a route, an LS MUST mark the route as "withdrawn"
   in its database, and maintain the withdrawn route in its database for
   MaxPurgeTime seconds.  If the LS needs to re-originate a route that
   had been purged but is still in its database, it can either re-
   originate the route immediately using a Sequence Number that is
   greater than that used in the withdraw, or the LS may wait until
   MaxPurgeTime seconds have expired since the route was withdrawn.





Rosenberg, et. al.          Standards Track                    [Page 57]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


10.1.6. Receiving Self-Originated Routes



   It is common for an LS to receive UPDATES for routes that it
   originated within the ITAD via the flooding procedure.  If the LS
   receives an UPDATE for a route that it originated that is newer (has
   a higher sequence number) than the LSs current version, then special
   actions must be taken.  This should be a relatively rare occurrence
   and indicates that a route still exists within the ITAD since the LSs
   last restart/reboot.

   If an LS receives a self-originated route update that is newer than
   the current version of the route at the LS, then the following
   actions MUST be taken.  If the LS still wishes to advertise the
   information in the route, then the LS MUST increase the Sequence
   Number of the route to a value greater than that received in the
   UPDATE and re-originate the route.  If the LS does not wish to
   continue to advertise the route, then it MUST purge the route as
   described in Section 10.1.5.

10.1.7. Removing Withdrawn Routes



   An LS SHOULD ensure that routes marked as withdrawn are removed from
   the database in a timely fashion after the MaxPurgeTime has expired.
   This could be done, for example, by periodically sweeping the
   database, and deleting those entries that were withdrawn more than
   MaxPurgeTime seconds ago.

10.2. Decision Process



   The Decision Process selects routes for subsequent advertisement by
   applying the policies in the local Policy Information Base (PIB) to
   the routes stored in its Adj-TRIBs-In.  The output of the Decision
   process is the set of routes that will be advertised to all peers;
   the selected routes will be stored in the local LS's Adj-TRIBs-Out.

   The selection process is formalized by defining a function that takes
   the attributes of a given route as an argument and returns a non-
   negative integer denoting the degree of preference for the route.
   The function that calculates the degree of preference for a given
   route shall not use as its inputs any of the following:  the
   existence of other routes, the non-existence of other routes, or the
   attributes of other routes.  Route selection then consists of an
   individual application of the degree of preference function to each
   feasible route, followed by the choice of the one with the highest
   degree of preference.






Rosenberg, et. al.          Standards Track                    [Page 58]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   All internal LSs in an ITAD MUST run the Decision Process and apply
   the same decision criteria, otherwise it will not be possible to
   synchronize their Loc-TRIBs.

   The Decision Process operates on routes contained in each Adj-TRIBs-
   In, and is responsible for:

      -  selection of routes to be advertised to internal peers
      -  selection of routes to be advertised to external peers
      -  route aggregation and route information reduction

   The Decision Process takes place in three distinct phases, each
   triggered by a different event:

      -  Phase 1 is responsible for calculating the degree of preference
         for each route received from an external peer.
      -  Phase 2 is invoked on completion of phase 1.  It is responsible
         for choosing the best route out of all those available for each
         distinct destination, and for installing each chosen route into
         the Loc-TRIB.
      -  Phase 3 is invoked after the Loc-TRIB has been modified.  It is
         responsible for disseminating routes in the Loc-TRIB to each
         external peer, according to the policies contained in the PIB.
         Route aggregation and information reduction can optionally be
         performed within this phase.

10.2.1. Phase 1: Calculation of Degree of Preference



   The Phase 1 decision function shall be invoked whenever the local LS
   receives from a peer an UPDATE message that advertises a new route, a
   replacement route, or a withdrawn route.

   The Phase 1 decision function is a separate process that is completed
   when it has no further work to do.

   The Phase 1 decision function shall lock an Adj-TRIB-In prior to
   operating on any route contained within it, and shall unlock it after
   operating on all new or replacement routes contained within it.

   The local LS MUST determine a degree of preference for each newly
   received or replacement route.  If the route is learned from an
   internal peer, the value of the LocalPreference attribute MUST be
   taken as the degree of preference.  If the route is learned from an
   external peer, then the degree of preference MUST be computed based
   on pre-configured policy information and used as the LocalPreference
   value in any intra-domain TRIP advertisement.  The exact nature of
   this policy information and the computation involved is a local
   matter.



Rosenberg, et. al.          Standards Track                    [Page 59]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   The output of the degree of preference determination process is the
   local preference of a route.  The local LS computes the local
   preference of routes learned from external peers or originated
   internally at that LS.  The local preference of a route learned from
   an internal peer is included in the LocalPreference attribute
   associated with that route.

10.2.2. Phase 2: Route Selection



   The Phase 2 decision function shall be invoked on completion of Phase
   1.  The Phase 2 function is a separate process that completes when it
   has no further work to do.  Phase 2 consists of two sub-phases: 2a
   and 2b.  The same route selection function is applied in both sub-
   phases, but the inputs to each phase are different.  The Phase 2a
   process MUST consider as inputs all external routes, that are present
   in the Adj-TRIBs-In of external peers, and all local routes.  The
   output of Phase 2a is inserted into the Ext-TRIB.  The Phase 2b
   process shall be invoked upon completion of Phase 2a and it MUST
   consider as inputs all routes in the Ext-TRIB and all routes that are
   present in the Adj-TRIBs-In of internal LSs.  The output of Phase 2b
   is stored in the Loc-TRIB.

   The Phase 2 decision function MUST be blocked from running while the
   Phase 3 decision function is in process.  The Phase 2 function MUST
   lock all Adj-TRIBs-In and the Ext-TRIB prior to commencing its
   function, and MUST unlock them on completion.

   If the LS determines that the NextHopServer listed in a route is
   unreachable, then the route MAY be excluded from the Phase 2 decision
   function.  The means by which such a determination is made is not
   mandated here.

   For each set of destinations for which one or more routes exist, the
   local LS's route selection function MUST identify the route that has:

      -  the highest degree of preference, or
      -  is selected as a result of the tie breaking rules specified in
         10.2.2.1.

   Withdrawn routes MUST be removed from the Loc-TRIB, Ext-TRIB, and the
   Adj-TRIBs-In.

10.2.2.1. Breaking Ties (Phase 2)



   Several routes to the same destination that have the same degree of
   preference may be input to the Phase 2 route selection function.  The
   local LS can select only one of these routes for inclusion in the




Rosenberg, et. al.          Standards Track                    [Page 60]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   associated Ext-TRIB (Phase 2a) or Loc-TRIB (Phase 2b).  The local LS
   considers all routes with the same degrees of preference.  The
   following algorithm shall be used to break ties.

      -  If the local LS is configured to use the MultiExitDisc
         attribute to break ties, and candidate routes received from the
         same neighboring ITAD differ in the value of the MultiExitDisc
         attribute, then select the route that has the larger value of
         MultiExitDisc.
      -  If at least one of the routes was originated by an internal LS,
         select the route route that was advertised by the internal LS
         that has the lowest TRIP ID.
      -  Otherwise, select the route that was advertised by the neighbor
         domain that has the lowest ITAD number.

10.2.3. Phase 3: Route Dissemination



   The Phase 3 decision function MUST be invoked upon completion of
   Phase 2 if Phase 2 results in changes to the Loc-TRIB or when a new
   LS-to-LS peer session is established.

   The Phase 3 function is a separate process that is completed when it
   has no further work to do.  The Phase 3 routing decision function
   MUST be blocked from running while the Phase 2 decision function is
   in process.

   All routes in the Loc-TRIB shall be processed into a corresponding
   entry in the associated Adj-TRIBs-Out.  Route aggregation and
   information reduction techniques (see 10.3.4) MAY optionally be
   applied.

   When the updating of the Adj-TRIBs-Out is complete, the local LS MUST
   run the external update process of 10.3.2.

10.2.4. Overlapping Routes



   When overlapping routes are present in the same Adj-TRIB-In, the more
   specific route shall take precedence, in order, from most specific to
   least specific.

   The set of destinations described by the overlap represents a portion
   of the less specific route that is feasible, but is not currently in
   use.  If a more specific route is later withdrawn, the set of
   destinations described by the more specific route will still be
   reachable using the less specific route.






Rosenberg, et. al.          Standards Track                    [Page 61]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   If an LS receives overlapping routes, the Decision Process MUST take
   into account the semantics of the overlapping routes.  In particular,
   if an LS accepts the less specific route while rejecting the more
   specific route from the same peer, then the destinations represented
   by the overlap may not forward along the domains listed in the
   AdvertisementPath attribute of that route.  Therefore, an LS has the
   following choices:

      1. Install both the less and the more specific routes
      2. Install the more specific route only
      3. Install the non-overlapping part of the less specific route
         only (that implies disaggregation of the less-specific route)
      4. Aggregate the two routes and install the aggregated route
      5. Install the less specific route only
      6. Install neither route

   If an LS chooses 5), then it SHOULD add AtomicAggregate attribute to
   the route.  A route that carries AtomicAggregate attribute MUST NOT
   be de-aggregated.  That is, the route cannot be made more specific.
   Forwarding along such a route does not guarantee that route traverses
   only domains listed in the RoutedPath of the route.  If an LS chooses
   1), then it MUST NOT advertise the less specific route without the
   more specific route.

10.3. Update-Send Process



   The Update-Send process is responsible for advertising UPDATE
   messages to all peers.  For example, it distributes the routes chosen
   by the Decision Process to other LSs that may be located in either
   the same ITAD or a neighboring ITAD.  Rules for information exchange
   between peer LSs located in different ITADs are given in 10.3.2;
   rules for information exchange between peer LSs located in the same
   ITAD are given in 10.3.1.

   Before forwarding routes to peers, an LS MUST determine which
   attributes should be forwarded along with that route.  If a not
   well-known non-transitive attribute is unrecognized, it is quietly
   ignored.  If a not well-known dependent-transitive attribute is
   unrecognized, and the NextHopServer attribute has been changed by the
   LS, the unrecognized attribute is quietly ignored.  If a not well-
   known dependent-transitive attribute is unrecognized, and the
   NextHopServer attribute has not been modified by the LS, the Partial
   bit in the attribute flags octet is set to 1, and the attribute is
   retained for propagation to other TRIP speakers.  Similarly, if an
   not well-known independent-transitive attribute is unrecognized, the
   Partial bit in the attribute flags octet is set to 1, and the
   attribute is retained for propagation to other TRIP speakers.




Rosenberg, et. al.          Standards Track                    [Page 62]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   If a not well-known attribute is recognized, and has a valid value,
   then, depending on the type of the not well-known attribute, it is
   updated, if necessary, for possible propagation to other TRIP
   speakers.

10.3.1. Internal Updates



   The Internal update process is concerned with the distribution of
   routing information to internal peers.

   When an LS receives an UPDATE message from another TRIP LS located in
   its own ITAD, it is flooded as described in Section 10.1.

   When an LS receives a new route from an LS in a neighboring ITAD, or
   if a local route is injected into TRIP, the LS determines the
   preference of that route.  If the new route has the highest degree of
   preference for all external routes and local routes to a given
   destination (or if the route was selected via a tie-breaking
   procedure as specified in 10.3.1.1), the LS MUST insert that new
   route into the Ext-TRIB database and the LS MUST advertise that route
   to all other LSs in its ITAD by means of an UPDATE message.  The LS
   MUST advertise itself as the Originator of that route within the
   ITAD.

   When an LS receives an UPDATE message with a non-empty
   WithdrawnRoutes attribute from an external peer, or if a local route
   is withdrawn from TRIP, the LS MUST remove from its Adj-TRIB-In all
   routes whose destinations were carried in this field.  If the
   withdrawn route was previously selected into the Ext-TRIB, the LS
   MUST take the following additional steps:

      -  If a new route is selected for advertisement for those
         destinations, then the LS MUST insert the replacement route
         into Ext-TRIB to replace the withdrawn route and advertise it
         to all internal LSs.
      -  If a replacement route is not available for advertisement, then
         the LS MUST include the destinations of the route in the
         WithdrawnRoutes attribute of an UPDATE message, and MUST send
         this message to each internal peer.  The LS MUST also remove
         the withdrawn route from the Ext-TRIB.

10.3.1.1. Breaking Ties (Routes Received from External Peers)



   If an LS has connections to several external peers, there will be
   multiple Adj-TRIBs-In associated with these peers.  These databases
   might contain several equally preferable routes to the same





Rosenberg, et. al.          Standards Track                    [Page 63]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   destination, all of which were advertised by external peers.  The
   local LS shall select one of these routes according to the following
   rules:

      -  If the LS is configured to use the MultiExitDisc attribute to
         break ties, and the candidate routes differ in the value of the
         MultiExitDisc attribute, then select the route that has the
         lowest value of MultiExitDisc, else
      -  Select the route that was advertised by the external LS that
         has the lowest TRIP Identifier.

10.3.2. External Updates



   The external update process is concerned with the distribution of
   routing information to external peers.  As part of the Phase 3 route
   selection process, the LS has updated its Adj-TRIBs-Out.  All newly
   installed routes and all newly unfeasible routes for which there is
   no replacement route MUST be advertised to external peers by means of
   UPDATE messages.

   Any routes in the Loc-TRIB marked as withdrawn MUST be removed.
   Changes to the reachable destinations within its own ITAD SHALL also
   be advertised in an UPDATE message.

10.3.3. Controlling Routing Traffic Overhead



   The TRIP protocol constrains the amount of routing traffic (that is,
   UPDATE messages) in order to limit both the link bandwidth needed to
   advertise UPDATE messages and the processing power needed by the
   Decision Process to digest the information contained in the UPDATE
   messages.

10.3.3.1. Frequency of Route Advertisement



   The parameter MinRouteAdvertisementInterval determines the minimum
   amount of time that must elapse between advertisements of routes to a
   particular destination from a single LS.  This rate limiting
   procedure applies on a per-destination basis, although the value of
   MinRouteAdvertisementInterval is set on a per LS peer basis.

   Two UPDATE messages sent from a single LS that advertise feasible
   routes to some common set of destinations received from external
   peers MUST be separated by at least MinRouteAdvertisementInterval.
   Clearly, this can only be achieved precisely by keeping a separate
   timer for each common set of destinations.  This would be unwarranted
   overhead.  Any technique which ensures that the interval between two
   UPDATE messages sent from a single LS that advertise feasible routes




Rosenberg, et. al.          Standards Track                    [Page 64]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   to some common set of destinations received from external peers will
   be at least MinRouteAdvertisementInterval, and will also ensure that
   a constant upper bound on the interval is acceptable.

   Two UPDATE messages, sent from a single LS to an external peer, that
   advertise feasible routes to some common set of destinations received
   from internal peers MUST be separated by at least
   MinRouteAdvertisementInterval.

   Since fast convergence is needed within an ITAD, this rate limiting
   procedure does not apply to routes received from internal peers and
   being broadcast to other internal peers.  To avoid long-lived black
   holes, the procedure does not apply to the explicit withdrawal of
   routes (that is, routes whose destinations explicitly withdrawn by
   UPDATE messages).

   This procedure does not limit the rate of route selection, but only
   the rate of route advertisement.  If new routes are selected multiple
   times while awaiting the expiration of MinRouteAdvertisementInterval,
   the last route selected shall be advertised at the end of
   MinRouteAdvertisementInterval.

10.3.3.2. Frequency of Route Origination



   The parameter MinITADOriginationInterval determines the minimum
   amount of time that must elapse between successive advertisements of
   UPDATE messages that report changes within the advertising LS's own
   ITAD.

10.3.3.3. Jitter



   To minimize the likelihood that the distribution of TRIP messages by
   a given LS will contain peaks, jitter should be applied to the timers
   associated with MinITADOriginationInterval, KeepAlive, and
   MinRouteAdvertisementInterval.  A given LS shall apply the same
   jitter to each of these quantities regardless of the destinations to
   which the updates are being sent; that is, jitter will not be applied
   on a "per peer" basis.

   The amount of jitter to be introduced shall be determined by
   multiplying the base value of the appropriate timer by a random
   factor that is uniformly distributed in the range from 0.75 to 1.0.









Rosenberg, et. al.          Standards Track                    [Page 65]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


10.3.4. Efficient Organization of Routing Information



   Having selected the routing information that it will advertise, a
   TRIP speaker may use methods to organize this information in an
   efficient manner.  These methods are discussed in the following
   sections.

10.3.4.1. Information Reduction



   Information reduction may imply a reduction in granularity of policy
   control - after information has collapsed, the same policies will
   apply to all destinations and paths in the equivalence class.

   The Decision Process may optionally reduce the amount of information
   that it will place in the Adj-TRIBs-Out by any of the following
   methods:

      -  ReachableRoutes: A set of destinations can be usually
         represented in compact form.  For example, a set of E.164 phone
         numbers can be represented in more compact form using E.164
         prefixes.
      -  AdvertisementPath: AdvertisementPath information can be
         represented as ordered AP_SEQUENCEs or unordered AP_SETs.
         AP_SETs are used in the route aggregation algorithm described
         in Section 5.4.4.  They reduce the size of the AP_PATH
         information by listing each ITAD number only once, regardless
         of how many times it may have appeared in multiple
         advertisement paths that were aggregated.

   An AP_SET implies that the destinations advertised in the UPDATE
   message can be reached through paths that traverse at least some of
   the constituent ITADs.  AP_SETs provide sufficient information to
   avoid route looping; however their use may prune potentially feasible
   paths, since such paths are no longer listed individually as in the
   form of AP_SEQUENCEs.  In practice this is not likely to be a
   problem, since once a call arrives at the edge of a group of ITADs,
   the LS at that point is likely to have more detailed path information
   and can distinguish individual paths to destinations.

10.3.4.2. Aggregating Routing Information



   Aggregation is the process of combining the characteristics of
   several different routes in such a way that a single route can be
   advertised.  Aggregation can occur as part of the decision process to
   reduce the amount of routing information that is placed in the Adj-
   TRIBs-Out.





Rosenberg, et. al.          Standards Track                    [Page 66]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   Aggregation reduces the amount of information an LS must store and
   exchange with other LSs.  Routes can be aggregated by applying the
   following procedure separately to attributes of like type.

   Routes that have the following attributes shall not be aggregated
   unless the corresponding attributes of each route are identical:
   MultiExitDisc, NextHopServer.

   Attributes that have different type codes cannot be aggregated.
   Attributes of the same type code may be aggregated.  The rules for
   aggregating each attribute MUST be provided together with attribute
   definition.  For example, aggregation rules for TRIP's basic
   attributes, e.g., ReachableRoutes and AdvertisementPath, are given in
   Section 5.

10.4. Route Selection Criteria



   Generally speaking, additional rules for comparing routes among
   several alternatives are outside the scope of this document.  There
   are two exceptions:

      -  If the local ITAD appears in the AdvertisementPath of the new
         route being considered, then that new route cannot be viewed as
         better than any other route.  If such a route were ever used, a
         routing loop could result (see Section 6.3).
      -  In order to achieve successful distributed operation, only
         routes with a likelihood of stability can be chosen.  Thus, an
         ITAD must avoid using unstable routes, and it must not make
         rapid spontaneous changes to its choice of route.  Quantifying
         the terms "unstable" and "rapid" in the previous sentence will
         require experience, but the principle is clear.

10.5. Originating TRIP Routes



   An LS may originate local routes by injecting routing information
   acquired by some other means (e.g. via an intra-domain routing
   protocol or through manual configuration or some dynamic registration
   mechanism/protocol) into TRIP.  An LS that originates TRIP routes
   shall assign the degree of preference to these routes by passing them
   through the Decision Process (see Section 10.2).  To TRIP, local
   routes are identical to external routes and are subjected to the same
   two phase route selection mechanism.  A local route which is selected
   into the Ext-TRIB MUST be advertised to all internal LSs.  The
   decision whether to distribute non-TRIP acquired routes within an
   ITAD via TRIP or not depends on the environment within the ITAD (e.g.
   type of intra-domain routing protocol) and should be controlled via
   configuration.




Rosenberg, et. al.          Standards Track                    [Page 67]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


11. TRIP Transport



   This specification defines the use of TCP as the transport layer for
   TRIP.  TRIP uses TCP port 6069.  Running TRIP over other transport
   protocols is for further study.

12. ITAD Topology



   There are no restrictions on the intra-domain topology of TRIP LSs.
   For example, LSs in an ITAD can be configured in a full mesh, star,
   or any other connected topology.  Similarly, there are no
   restrictions on the topology of TRIP ITADs.  For example, the ITADs
   can be organized in a flat topology (mesh or ring) or in multi-level
   hierarchy or any other topology.

   The border between two TRIP ITADs may be located either on the link
   between two TRIP LSs or it may coincide on a TRIP LS.  In the latter
   case, the same TRIP LS will be member in more than one ITAD, and it
   appears to be an internal peer to LSs in each ITAD it is member of.

13. IANA Considerations



   This document creates a new IANA registry for TRIP parameters.  The
   following TRIP parameters are included in the registry:

      - TRIP Capabilities
      - TRIP Attributes
      - TRIP Address Families
      - TRIP Application Protocols
      - TRIP ITAD Numbers

   Protocol parameters are frequently initialized/reset to 0.  This
   document reserves the value 0 of each of the above TRIP parameters in
   order to clearly distinguish between an unset parameter and any other
   registered values for that parameter.

   The sub-registries for each of the above parameters are discussed in
   the sections below.

13.1. TRIP Capabilities



   Requests to add TRIP capabilities other than those defined in Section
   4.2.1.1 must be submitted to iana@iana.org.  Following the assigned
   number policies outlined in [11], Capability Codes in the range
   32768-65535 are reserved for Private Use (these are the codes with
   the first bit of the code value equal to 1).  This document reserves
   value 0.  Capability Codes 1 and 2 have been assigned in Section
   4.2.1.1.  Capability Codes in the range 2-32767 are controlled by



Rosenberg, et. al.          Standards Track                    [Page 68]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   IANA, and are allocated subject to the Specification Required (IETF
   RFC or equivalent) condition.  The specification MUST include a
   description of the capability, the possible values it may take, and
   what constitutes a capability mismatch.

13.2. TRIP Attributes



   This document reserves Attribute Type Codes 224-255 for Private Use
   (these are the codes with the first three bits of the code equal to
   1).  This document reserves the value 0.  Attribute Type Codes 1
   through 11 have already been allocated by this document.  Attribute
   Type Codes 1 through 11 are defined in Sections 5.1 through 5.11.

   Attribute Type Codes in the range 12-223 are controlled by IANA, and
   require a Specification document (RFC or equivalent).  The
   specification MUST provide all information required in Section 5.12
   of this document.

   Attribute Type Code registration requests must be sent to
   iana@iana.org.  In addition to the specification requirement, the
   request MUST include an indication of who has change control over the
   attribute and contact information (postal and email address).

13.3. Destination Address Families



   This document reserves address family 0. Requests to add TRIP address
   families other than those defined in Section 5.1.1.1 ( address
   families 1, 2, and 3), i.e., in the range 4-32767, must be submitted
   to iana@iana.org.  The request MUST include a brief description of
   the address family, its alphabet, and special processing rules and
   guidelines, such as guidelines for aggregation, if any.  The requests
   are subject to Expert Review.  This document reserves the address
   family codes 32768-65535 for vendor-specific applications.

13.4. TRIP Application Protocols



   This document creates a new IANA registry for TRIP application
   protocols.  This document reserves the application protocol code 0.
   Requests to add TRIP application protocols other than those defined
   in Section 5.1.1.1 (application protocols 1 through 4), i.e., in the
   range 5-32767, must be submitted to iana@iana.org.  The request MUST
   include a brief background on the application protocol, and a
   description of how TRIP can be used to advertise routes for that
   protocol.  The requests are subject to Expert Review.  This document
   reserves the application protocol codes 32768-65535 for vendor-
   specific applications.





Rosenberg, et. al.          Standards Track                    [Page 69]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


13.5. ITAD Numbers



   This document reserves the ITAD number 0.  ITAD numbers in the range
   1-255 are designated for Private Use.  ITAD numbers in the range from
   256 to (2**32)-1 are allocated by IANA on a First-Come-First-Serve
   basis.  Requests for ITAD numbers must be submitted to iana@iana.org.
   The requests MUST include the following:

      -  Information about the organization that will administer the
         ITAD.
      -  Contact information (postal and email address).

14. Security Considerations



   This section covers security between peer TRIP LSs when TRIP runs
   over TCP in an IP environment.

   A security mechanism is clearly needed to prevent unauthorized
   entities from using the protocol defined in this document for setting
   up unauthorized peer sessions with other TRIP LSs or interfering with
   authorized peer sessions.  The security mechanism for the protocol,
   when transported over TCP in an IP network, is IPsec [12].  IPsec
   uses two protocols to provide traffic security: Authentication Header
   (AH) [13] and Encapsulating Security Payload (ESP) [14].

   The AH header affords data origin authentication, connectionless
   integrity and optional anti-replay protection of messages passed
   between the peer LSs.  The ESP header provides origin authentication,
   connectionless integrity, anti-replay protection, and confidentiality
   of messages.

   Implementations of the protocol defined in this document employing
   the ESP header SHALL comply with section 5 of [14], which defines a
   minimum set of algorithms for integrity checking and encryption.
   Similarly, implementations employing the AH header SHALL comply with
   section 5 of [13], which defines a minimum set of algorithms for
   integrity checking using manual keys.

   Implementations SHOULD use IKE [15] to permit more robust keying
   options.  Implementations employing IKE SHOULD support authentication
   with RSA signatures and RSA public key encryption.

   A Security Association (SA) [12] is a simplex "connection" that
   affords security services to the traffic carried by it.  Security
   services are afforded to a SA by the use of AH, or ESP, but not both.
   Two types of SAs are defined: transport mode and tunnel mode [12].  A
   transport mode SA is a security association between two hosts, and is
   appropriate for protecting the TRIP session between two peer LSs.



Rosenberg, et. al.          Standards Track                    [Page 70]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


A1. Appendix 1: TRIP FSM State Transitions and Actions

   This Appendix discusses the transitions between states in the TRIP
   FSM in response to TRIP events.  The following is the list of these
   states and events when the negotiated Hold Time value is non-zero.

   TRIP States:
      1 - Idle
      2 - Connect
      3 - Active
      4 - OpenSent
      5 - OpenConfirm
      6 - Established

   TRIP Events:
      1 - TRIP Start
      2 - TRIP Stop
      3 - TRIP Transport connection open
      4 - TRIP Transport connection closed
      5 - TRIP Transport connection open failed
      6 - TRIP Transport fatal error
      7 - ConnectRetry timer expired
      8 - Hold Timer expired
      9 - KeepAlive timer expired
      10 - Receive OPEN message
      11 - Receive KEEPALIVE message
      12 - Receive UPDATE messages
      13 - Receive NOTIFICATION message

   The following table describes the state transitions of the TRIP FSM
   and the actions triggered by these transitions.

   Event                Actions              Message Sent    Next State
   --------------------------------------------------------------------
   Idle (1)
    1            Initialize resources            none             2
                 Start ConnectRetry timer
                 Initiate a transport connection
    others               none                    none             1

   Connect(2)
    1                    none                    none             2
    3            Complete initialization         OPEN             4
                 Clear ConnectRetry timer
    5            Restart ConnectRetry timer      none             3
    7            Restart ConnectRetry timer      none             2
                 Initiate a transport connection
    others       Release resources               none             1



Rosenberg, et. al.          Standards Track                    [Page 71]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   Active (3)
    1                    none                    none             3
    3            Complete initialization         OPEN             4
                 Clear ConnectRetry timer
    5            Close connection                                 3
                 Restart ConnectRetry timer
    7            Restart ConnectRetry timer      none             2
                 Initiate a transport connection
    others       Release resources               none             1

   OpenSent(4)
    1                    none                    none             4
    4            Close transport connection      none             3
                 Restart ConnectRetry timer
    6            Release resources               none             1
   10            Process OPEN is OK            KEEPALIVE          5
                 Process OPEN failed           NOTIFICATION       1
   others        Close transport connection    NOTIFICATION       1
                 Release resources

   OpenConfirm (5)
    1                   none                     none             5
    4            Release resources               none             1
    6            Release resources               none             1
    9            Restart KeepAlive timer       KEEPALIVE          5
   11            Complete initialization         none             6
                 Restart Hold Timer
   13            Close transport connection                       1
                 Release resources
   others        Close transport connection    NOTIFICATION       1
                 Release resources

   Established (6)
    1                   none                     none             6
    4            Release resources               none             1
    6            Release resources               none             1
    9            Restart KeepAlive timer       KEEPALIVE          6
   11            Restart Hold Timer              none             6
   12            Process UPDATE is OK          UPDATE             6
                 Process UPDATE failed         NOTIFICATION       1
   13            Close transport connection                       1
                 Release resources
   others        Close transport connection    NOTIFICATION       1
                 Release resources
   -----------------------------------------------------------------






Rosenberg, et. al.          Standards Track                    [Page 72]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   The following is a condensed version of the above state transition
   table.

   Events| Idle | Connect | Active | OpenSent | OpenConfirm | Estab
         | (1)  |   (2)   |  (3)   |    (4)   |     (5)     |   (6)
         |----------------------------------------------------------
    1    |  2   |    2    |   3    |     4    |      5      |    6
         |      |         |        |          |             |
    2    |  1   |    1    |   1    |     1    |      1      |    1
         |      |         |        |          |             |
    3    |  1   |    4    |   4    |     1    |      1      |    1
         |      |         |        |          |             |
    4    |  1   |    1    |   1    |     3    |      1      |    1
         |      |         |        |          |             |
    5    |  1   |    3    |   3    |     1    |      1      |    1
         |      |         |        |          |             |
    6    |  1   |    1    |   1    |     1    |      1      |    1
         |      |         |        |          |             |
    7    |  1   |    2    |   2    |     1    |      1      |    1
         |      |         |        |          |             |
    8    |  1   |    1    |   1    |     1    |      1      |    1
         |      |         |        |          |             |
    9    |  1   |    1    |   1    |     1    |      5      |    6
         |      |         |        |          |             |
   10    |  1   |    1    |   1    |  1 or 5  |      1      |    1
         |      |         |        |          |             |
   11    |  1   |    1    |   1    |     1    |      6      |    6
         |      |         |        |          |             |
   12    |  1   |    1    |   1    |     1    |      1      | 1 or 6
         |      |         |        |          |             |
   13    |  1   |    1    |   1    |     1    |      1      |    1
         |      |         |        |          |             |
         --------------------------------------------------------------

A2. Appendix 2: Implementation Recommendations

   This section presents some implementation recommendations.

A.2.1: Multiple Networks Per Message



   The TRIP protocol allows for multiple address prefixes with the same
   advertisement path and next-hop server to be specified in one
   message.  Making use of this capability is highly recommended.  With
   one address prefix per message there is a substantial increase in
   overhead in the receiver.  Not only does the system overhead increase
   due to the reception of multiple messages, but the overhead of
   scanning the routing table for updates to TRIP peers is incurred
   multiple times as well.  One method of building messages containing



Rosenberg, et. al.          Standards Track                    [Page 73]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   many address prefixes per advertisement path and next hop from a
   routing table that is not organized per advertisement path is to
   build many messages as the routing table is scanned.  As each address
   prefix is processed, a message for the associated advertisement path
   and next hop is allocated, if it does not exist, and the new address
   prefix is added to it.  If such a message exists, the new address
   prefix is just appended to it.  If the message lacks the space to
   hold the new address prefix, it is transmitted, a new message is
   allocated, and the new address prefix is inserted into the new
   message.  When the entire routing table has been scanned, all
   allocated messages are sent and their resources released.  Maximum
   compression is achieved when all the destinations covered by the
   address prefixes share the same next hop server and common
   attributes, making it possible to send many address prefixes in one
   4096-byte message.

   When peering with a TRIP implementation that does not compress
   multiple address prefixes into one message, it may be necessary to
   take steps to reduce the overhead from the flood of data received
   when a peer is acquired or a significant network topology change
   occurs.  One method of doing this is to limit the rate of updates.
   This will eliminate the redundant scanning of the routing table to
   provide flash updates for TRIP peers.  A disadvantage of this
   approach is that it increases the propagation latency of routing
   information.  By choosing a minimum flash update interval that is not
   much greater than the time it takes to process the multiple messages,
   this latency should be minimized.  A better method would be to read
   all received messages before sending updates.

A.2.2: Processing Messages on a Stream Protocol



   TRIP uses TCP as a transport mechanism.  Due to the stream nature of
   TCP, all the data of a received message does not necessarily arrive
   at the same time.  This can make it difficult to process the data as
   messages, especially on systems where it is not possible to determine
   how much data has been received but not yet processed.

   One method that can be used in this situation is to first try to read
   just the message header.  For the KEEPALIVE message type, this is a
   complete message; for other message types, the header should first be
   verified, in particular the total length.  If all checks are
   successful, the specified length, minus the size of the message
   header is the amount of data left to read.  An implementation that
   would "hang" the routing information process while trying to read
   from a peer could set up a message buffer (4096 bytes) per peer and
   fill it with data as available until a complete message has been
   received.




Rosenberg, et. al.          Standards Track                    [Page 74]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


A.2.3: Reducing Route Flapping



   To avoid excessive route flapping an LS which needs to withdraw a
   destination and send an update about a more specific or less specific
   route SHOULD combine them into the same UPDATE message.

A.2.4: TRIP Timers



   TRIP employs seven timers: ConnectRetry, Hold Time, KeepAlive,
   MaxPurgeTime, TripDisableTime, MinITADOriginationInterval, and
   MinRouteAdvertisementInterval.  The suggested value for the
   ConnectRetry timer is 120 seconds.  The suggested value for the Hold
   Time is 90 seconds.  The suggested value for the KeepAlive timer is
   30 seconds.  The suggested value for the MaxPurgeTime timer is 10
   seconds.  The suggested value for the TripDisableTime timer is 180
   seconds.  The suggested value for the MinITADOriginationInterval is
   30 seconds.  The suggested value for the
   MinRouteAdvertisementInterval is 30 seconds.

   An implementation of TRIP MUST allow these timers to be configurable.

A.2.5: AP_SET Sorting



   Another useful optimization that can be done to simplify this
   situation is to sort the ITAD numbers found in an AP_SET.  This
   optimization is entirely optional.

Acknowledgments

   We wish to thank Dave Oran for his insightful comments and
   suggestions.

References

   [1]   Bradner, S., "Keywords for use in RFCs to Indicate Requirement
         Levels", BCP 14, RFC 2119, March 1997.

   [2]   Rosenberg, J. and H. Schulzrinne, "A Framework for a Gateway
         Location Protocol", RFC 2871, June 2000.

   [3]   Rekhter, Y. and T. Li, "Border Gateway Protocol 4 (BGP-4)," RFC
         1771, March 1995.

   [4]   Moy, J., "Open Shortest Path First Version 2", STD 54, RFC
         2328, April 1998.






Rosenberg, et. al.          Standards Track                    [Page 75]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


   [5]   "Intermediate System to Intermediate System Intra-Domain
         Routing Exchange Protocol for use in Conjunction with the
         Protocol for Providing the Connectionless-mode Network Service
         (ISO 8473)," ISO DP 10589, February 1990.

   [6]   Luciani, J., Armitage, G., Halpern, J. and N. Doraswamy,
         "Server Cache Synchronization Protocol (SCSP)", RFC 2334, April
         1998.

   [7]   International Telecommunication Union, "Packet-Based Multimedia
         Communication Systems," Recommendation H.323, Version 3
         Telecommunication Standardization Sector of ITU, Geneva,
         Switzerland, November 2000.

   [8]   Handley, H., Schulzrinne, H., Schooler, E. and J. Rosenberg,
         "SIP:  Session Initiation Protocol", RFC 2543, March 1999.

   [9]   Braden, R., "Requirements for Internet Hosts -- Application and
         Support", STD 3, RFC 1123, October 1989.

   [10]  Hinden, R. and S. Deering, "IP Version 6 Addressing
         Architecture", RFC 2373, July 1998.

   [11]  Narten, T. and H. Alvestrand, "Guidelines for Writing an IANA
         Considerations Section in RFCs", BCP 26, RFC 2434, October
         1998.

   [12]  Kent, S. and R. Atkinson, "Security Architecture for the
         Internet Protocol", RFC 2401, November 1998.

   [13]  Kent, S. and R. Atkinson, "IP Authentication Header", RFC 2402,
         November 1998.

   [14]  Kent, S. and R. Atkinson, "IP Encapsulating Security Payload
         (ESP)", RFC 2406, November 1998.

   [15]  Harkins, D. and D. Carrel, "The Internet Key Exchange (IKE)",
         RFC 2409, November 1998.













Rosenberg, et. al.          Standards Track                    [Page 76]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


Intellectual Property Notice

   The IETF takes no position regarding the validity or scope of any
   intellectual property 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; neither does it represent that it
   has made any effort to identify any such rights.  Information on the
   IETF's procedures with respect to rights in standards-track and
   standards-related documentation can be found in BCP 11.  Copies of
   claims of rights made available for publication 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 Secretariat.

   The IETF invites any interested party to bring to its attention any
   copyrights, patents or patent applications, or other proprietary
   rights which may cover technology that may be required to practice
   this standard.  Please address the information to the IETF Executive
   Director.

   The IETF has been notified of intellectual property rights claimed in
   regard to some or all of the specification contained in this
   document.  For more information consult the online list of claimed
   rights.

























Rosenberg, et. al.          Standards Track                    [Page 77]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


Authors' Addresses



   Jonathan Rosenberg
   dynamicsoft
   72 Eagle Rock Avenue
   First Floor
   East Hanover, NJ 07936

   Phone: 973-952-5000
   EMail: jdrosen@dynamicsoft.com


   Hussein F. Salama
   Cisco Systems
   170 W. Tasman Drive
   San Jose, CA 95134

   Phone: 408-527-7147
   EMail: hsalama@cisco.com


   Matt Squire
   Hatteras Networks
   639 Davis Drive
   Suite 200
   Durham, NC 27713

   EMail: mattsquire@acm.org























Rosenberg, et. al.          Standards Track                    [Page 78]

RFC 3219            Telephony Routing over IP (TRIP)        January 2002


Full Copyright Statement



   Copyright (C) The Internet Society (2002).  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.

Acknowledgement



   Funding for the RFC Editor function is currently provided by the
   Internet Society.



















Rosenberg, et. al.          Standards Track                    [Page 79]