Tech-invite3GPPspaceIETFspace
96959493929190898887868584838281807978777675747372717069686766656463626160595857565554535251504948474645444342414039383736353433323130292827262524232221201918171615141312111009080706050403020100
in Index   Prev   Next

RFC 3209

RSVP-TE: Extensions to RSVP for LSP Tunnels

Pages: 61
Proposed Standard
Errata
Updated by:  393644204874515154205711678067907274
Part 1 of 3 – Pages 1 to 16
None   None   Next

Top   ToC   RFC3209 - Page 1
Network Working Group                                         D. Awduche
Request for Comments: 3209                          Movaz Networks, Inc.
Category: Standards Track                                      L. Berger
                                                                  D. Gan
                                                  Juniper Networks, Inc.
                                                                   T. Li
                                                  Procket Networks, Inc.
                                                           V. Srinivasan
                                             Cosine Communications, Inc.
                                                              G. Swallow
                                                     Cisco Systems, Inc.
                                                           December 2001


              RSVP-TE: Extensions to RSVP for LSP Tunnels

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 (2001).  All Rights Reserved.

Abstract

This document describes the use of RSVP (Resource Reservation Protocol), including all the necessary extensions, to establish label-switched paths (LSPs) in MPLS (Multi-Protocol Label Switching). Since the flow along an LSP is completely identified by the label applied at the ingress node of the path, these paths may be treated as tunnels. A key application of LSP tunnels is traffic engineering with MPLS as specified in RFC 2702. We propose several additional objects that extend RSVP, allowing the establishment of explicitly routed label switched paths using RSVP as a signaling protocol. The result is the instantiation of label- switched tunnels which can be automatically routed away from network failures, congestion, and bottlenecks.
Top   ToC   RFC3209 - Page 2

Contents

1 Introduction .......................................... 3 1.1 Background ............................................. 4 1.2 Terminology ............................................ 6 2 Overview .............................................. 7 2.1 LSP Tunnels and Traffic Engineered Tunnels ............. 7 2.2 Operation of LSP Tunnels ............................... 8 2.3 Service Classes ........................................ 10 2.4 Reservation Styles ..................................... 10 2.4.1 Fixed Filter (FF) Style ................................ 10 2.4.2 Wildcard Filter (WF) Style ............................. 11 2.4.3 Shared Explicit (SE) Style ............................. 11 2.5 Rerouting Traffic Engineered Tunnels ................... 12 2.6 Path MTU ............................................... 13 3 LSP Tunnel related Message Formats ..................... 15 3.1 Path Message ........................................... 15 3.2 Resv Message ........................................... 16 4 LSP Tunnel related Objects ............................. 17 4.1 Label Object ........................................... 17 4.1.1 Handling Label Objects in Resv messages ................ 17 4.1.2 Non-support of the Label Object ........................ 19 4.2 Label Request Object ................................... 19 4.2.1 Label Request without Label Range ...................... 19 4.2.2 Label Request with ATM Label Range ..................... 20 4.2.3 Label Request with Frame Relay Label Range ............. 21 4.2.4 Handling of LABEL_REQUEST .............................. 22 4.2.5 Non-support of the Label Request Object ................ 23 4.3 Explicit Route Object .................................. 23 4.3.1 Applicability .......................................... 24 4.3.2 Semantics of the Explicit Route Object ................. 24 4.3.3 Subobjects ............................................. 25 4.3.4 Processing of the Explicit Route Object ................ 28 4.3.5 Loops .................................................. 30 4.3.6 Forward Compatibility .................................. 30 4.3.7 Non-support of the Explicit Route Object ............... 31 4.4 Record Route Object .................................... 31 4.4.1 Subobjects ............................................. 31 4.4.2 Applicability .......................................... 34 4.4.3 Processing RRO ......................................... 35 4.4.4 Loop Detection ......................................... 36 4.4.5 Forward Compatibility .................................. 37 4.4.6 Non-support of RRO ..................................... 37 4.5 Error Codes for ERO and RRO ............................ 37 4.6 Session, Sender Template, and Filter Spec Objects ...... 38 4.6.1 Session Object ......................................... 39 4.6.2 Sender Template Object ................................. 40 4.6.3 Filter Specification Object ............................ 42
Top   ToC   RFC3209 - Page 3
   4.6.4  Reroute and Bandwidth Increase Procedure  ...............  42
   4.7    Session Attribute Object  ...............................  43
   4.7.1  Format without resource affinities  .....................  43
   4.7.2  Format with resource affinities  ........................  45
   4.7.3  Procedures applying to both C-Types  ....................  46
   4.7.4  Resource Affinity Procedures   ..........................  48
   5      Hello Extension  ........................................  49
   5.1    Hello Message Format  ...................................  50
   5.2    HELLO Object formats  ...................................  51
   5.2.1  HELLO REQUEST object  ...................................  51
   5.2.2  HELLO ACK object  .......................................  51
   5.3    Hello Message Usage  ....................................  52
   5.4    Multi-Link Considerations  ..............................  53
   5.5    Compatibility  ..........................................  54
   6      Security Considerations  ................................  54
   7      IANA Considerations  ....................................  54
   7.1    Message Types  ..........................................  55
   7.2    Class Numbers and C-Types  ..............................  55
   7.3    Error Codes and Globally-Defined Error Value Sub-Codes  .  57
   7.4    Subobject Definitions  ..................................  57
   8      Intellectual Property Considerations  ...................  58
   9      Acknowledgments  ........................................  58
   10     References  .............................................  58
   11     Authors' Addresses  .....................................  60
   12     Full Copyright Statement  ...............................  61

1. Introduction

Section 2.9 of the MPLS architecture [2] defines a label distribution protocol as a set of procedures by which one Label Switched Router (LSR) informs another of the meaning of labels used to forward traffic between and through them. The MPLS architecture does not assume a single label distribution protocol. This document is a specification of extensions to RSVP for establishing label switched paths (LSPs) in MPLS networks. Several of the new features described in this document were motivated by the requirements for traffic engineering over MPLS (see [3]). In particular, the extended RSVP protocol supports the instantiation of explicitly routed LSPs, with or without resource reservations. It also supports smooth rerouting of LSPs, preemption, and loop detection. The LSPs created with RSVP can be used to carry the "Traffic Trunks" described in [3]. The LSP which carries a traffic trunk and a traffic trunk are distinct though closely related concepts. For example, two LSPs between the same source and destination could be load shared to carry a single traffic trunk. Conversely several
Top   ToC   RFC3209 - Page 4
   traffic trunks could be carried in the same LSP if, for instance, the
   LSP were capable of carrying several service classes.  The
   applicability of these extensions is discussed further in [10].

   Since the traffic that flows along a label-switched path is defined
   by the label applied at the ingress node of the LSP, these paths can
   be treated as tunnels, tunneling below normal IP routing and
   filtering mechanisms.  When an LSP is used in this way we refer to it
   as an LSP tunnel.

   LSP tunnels allow the implementation of a variety of policies related
   to network performance optimization.  For example, LSP tunnels can be
   automatically or manually routed away from network failures,
   congestion, and bottlenecks.  Furthermore, multiple parallel LSP
   tunnels can be established between two nodes, and traffic between the
   two nodes can be mapped onto the LSP tunnels according to local
   policy.  Although traffic engineering (that is, performance
   optimization of operational networks) is expected to be an important
   application of this specification, the extended RSVP protocol can be
   used in a much wider context.

   The purpose of this document is to describe the use of RSVP to
   establish LSP tunnels.  The intent is to fully describe all the
   objects, packet formats, and procedures required to realize
   interoperable implementations.  A few new objects are also defined
   that enhance management and diagnostics of LSP tunnels.

   The document also describes a means of rapid node failure detection
   via a new HELLO message.

   All objects and messages described in this specification are optional
   with respect to RSVP.  This document discusses what happens when an
   object described here is not supported by a node.

   Throughout this document, the discussion will be restricted to
   unicast label switched paths.  Multicast LSPs are left for further
   study.

1.1. Background

Hosts and routers that support both RSVP [1] and Multi-Protocol Label Switching [2] can associate labels with RSVP flows. When MPLS and RSVP are combined, the definition of a flow can be made more flexible. Once a label switched path (LSP) is established, the traffic through the path is defined by the label applied at the ingress node of the LSP. The mapping of label to traffic can be accomplished using a number of different criteria. The set of packets that are assigned the same label value by a specific node are
Top   ToC   RFC3209 - Page 5
   said to belong to the same forwarding equivalence class (FEC) (see
   [2]), and effectively define the "RSVP flow."  When traffic is mapped
   onto a label-switched path in this way, we call the LSP an "LSP
   Tunnel".  When labels are associated with traffic flows, it becomes
   possible for a router to identify the appropriate reservation state
   for a packet based on the packet's label value.

   The signaling protocol model uses downstream-on-demand label
   distribution.  A request to bind labels to a specific LSP tunnel is
   initiated by an ingress node through the RSVP Path message.  For this
   purpose, the RSVP Path message is augmented with a LABEL_REQUEST
   object.  Labels are allocated downstream and distributed (propagated
   upstream) by means of the RSVP Resv message.  For this purpose, the
   RSVP Resv message is extended with a special LABEL object.  The
   procedures for label allocation, distribution, binding, and stacking
   are described in subsequent sections of this document.

   The signaling protocol model also supports explicit routing
   capability.  This is accomplished by incorporating a simple
   EXPLICIT_ROUTE object into RSVP Path messages.  The EXPLICIT_ROUTE
   object encapsulates a concatenation of hops which constitutes the
   explicitly routed path.  Using this object, the paths taken by
   label-switched RSVP-MPLS flows can be pre-determined, independent of
   conventional IP routing.  The explicitly routed path can be
   administratively specified, or automatically computed by a suitable
   entity based on QoS and policy requirements, taking into
   consideration the prevailing network state.  In general, path
   computation can be control-driven or data-driven.  The mechanisms,
   processes, and algorithms used to compute explicitly routed paths are
   beyond the scope of this specification.

   One useful application of explicit routing is traffic engineering.
   Using explicitly routed LSPs, a node at the ingress edge of an MPLS
   domain can control the path through which traffic traverses from
   itself, through the MPLS network, to an egress node.  Explicit
   routing can be used to optimize the utilization of network resources
   and enhance traffic oriented performance characteristics.

   The concept of explicitly routed label switched paths can be
   generalized through the notion of abstract nodes.  An abstract node
   is a group of nodes whose internal topology is opaque to the ingress
   node of the LSP.  An abstract node is said to be simple if it
   contains only one physical node.  Using this concept of abstraction,
   an explicitly routed LSP can be specified as a sequence of IP
   prefixes or a sequence of Autonomous Systems.
Top   ToC   RFC3209 - Page 6
   The signaling protocol model supports the specification of an
   explicit path as a sequence of strict and loose routes.  The
   combination of abstract nodes, and strict and loose routes
   significantly enhances the flexibility of path definitions.

   An advantage of using RSVP to establish LSP tunnels is that it
   enables the allocation of resources along the path.  For example,
   bandwidth can be allocated to an LSP tunnel using standard RSVP
   reservations and Integrated Services service classes [4].

   While resource reservations are useful, they are not mandatory.
   Indeed, an LSP can be instantiated without any resource reservations
   whatsoever.  Such LSPs without resource reservations can be used, for
   example, to carry best effort traffic.  They can also be used in many
   other contexts, including implementation of fall-back and recovery
   policies under fault conditions, and so forth.

1.2. Terminology

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 RFC2119 [6]. The reader is assumed to be familiar with the terminology in [1], [2] and [3]. Abstract Node A group of nodes whose internal topology is opaque to the ingress node of the LSP. An abstract node is said to be simple if it contains only one physical node. Explicitly Routed LSP An LSP whose path is established by a means other than normal IP routing. Label Switched Path The path created by the concatenation of one or more label switched hops, allowing a packet to be forwarded by swapping labels from an MPLS node to another MPLS node. For a more precise definition see [2]. LSP A Label Switched Path
Top   ToC   RFC3209 - Page 7
   LSP Tunnel

      An LSP which is used to tunnel below normal IP routing and/or
      filtering mechanisms.

   Traffic Engineered Tunnel (TE Tunnel)

      A set of one or more LSP Tunnels which carries a traffic trunk.

   Traffic Trunk

      A set of flows aggregated by their service class and then placed
      on an LSP or set of LSPs called a traffic engineered tunnel.  For
      further discussion see [3].

2. Overview

2.1. LSP Tunnels and Traffic Engineered Tunnels

According to [1], "RSVP defines a 'session' to be a data flow with a particular destination and transport-layer protocol." However, when RSVP and MPLS are combined, a flow or session can be defined with greater flexibility and generality. The ingress node of an LSP can use a variety of means to determine which packets are assigned a particular label. Once a label is assigned to a set of packets, the label effectively defines the "flow" through the LSP. We refer to such an LSP as an "LSP tunnel" because the traffic through it is opaque to intermediate nodes along the label switched path. New RSVP SESSION, SENDER_TEMPLATE, and FILTER_SPEC objects, called LSP_TUNNEL_IPv4 and LSP_TUNNEL_IPv6 have been defined to support the LSP tunnel feature. The semantics of these objects, from the perspective of a node along the label switched path, is that traffic belonging to the LSP tunnel is identified solely on the basis of packets arriving from the PHOP or "previous hop" (see [1]) with the particular label value(s) assigned by this node to upstream senders to the session. In fact, the IPv4(v6) that appears in the object name only denotes that the destination address is an IPv4(v6) address. When we refer to these objects generically, we use the qualifier LSP_TUNNEL. In some applications it is useful to associate sets of LSP tunnels. This can be useful during reroute operations or to spread a traffic trunk over multiple paths. In the traffic engineering application such sets are called traffic engineered tunnels (TE tunnels). To enable the identification and association of such LSP tunnels, two identifiers are carried. A tunnel ID is part of the SESSION object. The SESSION object uniquely defines a traffic engineered tunnel. The
Top   ToC   RFC3209 - Page 8
   SENDER_TEMPLATE and FILTER_SPEC objects carry an LSP ID.  The
   SENDER_TEMPLATE (or FILTER_SPEC) object together with the SESSION
   object uniquely identifies an LSP tunnel

2.2. Operation of LSP Tunnels

This section summarizes some of the features supported by RSVP as extended by this document related to the operation of LSP tunnels. These include: (1) the capability to establish LSP tunnels with or without QoS requirements, (2) the capability to dynamically reroute an established LSP tunnel, (3) the capability to observe the actual route traversed by an established LSP tunnel, (4) the capability to identify and diagnose LSP tunnels, (5) the capability to preempt an established LSP tunnel under administrative policy control, and (6) the capability to perform downstream-on-demand label allocation, distribution, and binding. In the following paragraphs, these features are briefly described. More detailed descriptions can be found in subsequent sections of this document. To create an LSP tunnel, the first MPLS node on the path -- that is, the sender node with respect to the path -- creates an RSVP Path message with a session type of LSP_TUNNEL_IPv4 or LSP_TUNNEL_IPv6 and inserts a LABEL_REQUEST object into the Path message. The LABEL_REQUEST object indicates that a label binding for this path is requested and also provides an indication of the network layer protocol that is to be carried over this path. The reason for this is that the network layer protocol sent down an LSP cannot be assumed to be IP and cannot be deduced from the L2 header, which simply identifies the higher layer protocol as MPLS. If the sender node has knowledge of a route that has high likelihood of meeting the tunnel's QoS requirements, or that makes efficient use of network resources, or that satisfies some policy criteria, the node can decide to use the route for some or all of its sessions. To do this, the sender node adds an EXPLICIT_ROUTE object to the RSVP Path message. The EXPLICIT_ROUTE object specifies the route as a sequence of abstract nodes. If, after a session has been successfully established, the sender node discovers a better route, the sender can dynamically reroute the session by simply changing the EXPLICIT_ROUTE object. If problems are encountered with an EXPLICIT_ROUTE object, either because it causes a routing loop or because some intermediate routers do not support it, the sender node is notified. By adding a RECORD_ROUTE object to the Path message, the sender node can receive information about the actual route that the LSP tunnel traverses. The sender node can also use this object to request
Top   ToC   RFC3209 - Page 9
   notification from the network concerning changes to the routing path.
   The RECORD_ROUTE object is analogous to a path vector, and hence can
   be used for loop detection.

   Finally, a SESSION_ATTRIBUTE object can be added to Path messages to
   aid in session identification and diagnostics.  Additional control
   information, such as setup and hold priorities, resource affinities
   (see [3]), and local-protection, are also included in this object.

   Routers along the path may use the setup and hold priorities along
   with SENDER_TSPEC and any POLICY_DATA objects contained in Path
   messages as input to policy control.  For instance, in the traffic
   engineering application, it is very useful to use the Path message as
   a means of verifying that bandwidth exists at a particular priority
   along an entire path before preempting any lower priority
   reservations.  If a Path message is allowed to progress when there
   are insufficient resources, then there is a danger that lower
   priority reservations downstream of this point will unnecessarily be
   preempted in a futile attempt to service this request.

   When the EXPLICIT_ROUTE object (ERO) is present, the Path message is
   forwarded towards its destination along a path specified by the ERO.
   Each node along the path records the ERO in its path state block.
   Nodes may also modify the ERO before forwarding the Path message.  In
   this case the modified ERO SHOULD be stored in the path state block
   in addition to the received ERO.

   The LABEL_REQUEST object requests intermediate routers and receiver
   nodes to provide a label binding for the session.  If a node is
   incapable of providing a label binding, it sends a PathErr message
   with an "unknown object class" error.  If the LABEL_REQUEST object is
   not supported end to end, the sender node will be notified by the
   first node which does not provide this support.

   The destination node of a label-switched path responds to a
   LABEL_REQUEST by including a LABEL object in its response RSVP Resv
   message.  The LABEL object is inserted in the filter spec list
   immediately following the filter spec to which it pertains.

   The Resv message is sent back upstream towards the sender, following
   the path state created by the Path message, in reverse order.  Note
   that if the path state was created by use of an ERO, then the Resv
   message will follow the reverse path of the ERO.

   Each node that receives a Resv message containing a LABEL object uses
   that label for outgoing traffic associated with this LSP tunnel.  If
   the node is not the sender, it allocates a new label and places that
   label in the corresponding LABEL object of the Resv message which it
Top   ToC   RFC3209 - Page 10
   sends upstream to the PHOP.  The label sent upstream in the LABEL
   object is the label which this node will use to identify incoming
   traffic associated with this LSP tunnel.  This label also serves as
   shorthand for the Filter Spec.  The node can now update its "Incoming
   Label Map" (ILM), which is used to map incoming labeled packets to a
   "Next Hop Label Forwarding Entry" (NHLFE), see [2].

   When the Resv message propagates upstream to the sender node, a
   label-switched path is effectively established.

2.3. Service Classes

This document does not restrict the type of Integrated Service requests for reservations. However, an implementation SHOULD support the Controlled-Load service [4] and the Null Service [16].

2.4. Reservation Styles

The receiver node can select from among a set of possible reservation styles for each session, and each RSVP session must have a particular style. Senders have no influence on the choice of reservation style. The receiver can choose different reservation styles for different LSPs. An RSVP session can result in one or more LSPs, depending on the reservation style chosen. Some reservation styles, such as FF, dedicate a particular reservation to an individual sender node. Other reservation styles, such as WF and SE, can share a reservation among several sender nodes. The following sections discuss the different reservation styles and their advantages and disadvantages. A more detailed discussion of reservation styles can be found in [1].

2.4.1. Fixed Filter (FF) Style

The Fixed Filter (FF) reservation style creates a distinct reservation for traffic from each sender that is not shared by other senders. This style is common for applications in which traffic from each sender is likely to be concurrent and independent. The total amount of reserved bandwidth on a link for sessions using FF is the sum of the reservations for the individual senders. Because each sender has its own reservation, a unique label is assigned to each sender. This can result in a point-to-point LSP between every sender/receiver pair.
Top   ToC   RFC3209 - Page 11

2.4.2. Wildcard Filter (WF) Style

With the Wildcard Filter (WF) reservation style, a single shared reservation is used for all senders to a session. The total reservation on a link remains the same regardless of the number of senders. A single multipoint-to-point label-switched-path is created for all senders to the session. On links that senders to the session share, a single label value is allocated to the session. If there is only one sender, the LSP looks like a normal point-to-point connection. When multiple senders are present, a multipoint-to-point LSP (a reversed tree) is created. This style is useful for applications in which not all senders send traffic at the same time. A phone conference, for example, is an application where not all speakers talk at the same time. If, however, all senders send simultaneously, then there is no means of getting the proper reservations made. Either the reserved bandwidth on links close to the destination will be less than what is required or then the reserved bandwidth on links close to some senders will be greater than what is required. This restricts the applicability of WF for traffic engineering purposes. Furthermore, because of the merging rules of WF, EXPLICIT_ROUTE objects cannot be used with WF reservations. As a result of this issue and the lack of applicability to traffic engineering, use of WF is not considered in this document.

2.4.3. Shared Explicit (SE) Style

The Shared Explicit (SE) style allows a receiver to explicitly specify the senders to be included in a reservation. There is a single reservation on a link for all the senders listed. Because each sender is explicitly listed in the Resv message, different labels may be assigned to different senders, thereby creating separate LSPs. SE style reservations can be provided using multipoint-to-point label-switched-path or LSP per sender. Multipoint-to-point LSPs may be used when path messages do not carry the EXPLICIT_ROUTE object, or when Path messages have identical EXPLICIT_ROUTE objects. In either of these cases a common label may be assigned.
Top   ToC   RFC3209 - Page 12
   Path messages from different senders can each carry their own ERO,
   and the paths taken by the senders can converge and diverge at any
   point in the network topology.  When Path messages have differing
   EXPLICIT_ROUTE objects, separate LSPs for each EXPLICIT_ROUTE object
   must be established.

2.5. Rerouting Traffic Engineered Tunnels

One of the requirements for Traffic Engineering is the capability to reroute an established TE tunnel under a number of conditions, based on administrative policy. For example, in some contexts, an administrative policy may dictate that a given TE tunnel is to be rerouted when a more "optimal" route becomes available. Another important context when TE tunnel reroute is usually required is upon failure of a resource along the TE tunnel's established path. Under some policies, it may also be necessary to return the TE tunnel to its original path when the failed resource becomes re-activated. In general, it is highly desirable not to disrupt traffic, or adversely impact network operations while TE tunnel rerouting is in progress. This adaptive and smooth rerouting requirement necessitates establishing a new LSP tunnel and transferring traffic from the old LSP tunnel onto it before tearing down the old LSP tunnel. This concept is called "make-before-break." A problem can arise because the old and new LSP tunnels might compete with each other for resources on network segments which they have in common. Depending on availability of resources, this competition can cause Admission Control to prevent the new LSP tunnel from being established. An advantage of using RSVP to establish LSP tunnels is that it solves this problem very elegantly. To support make-before-break in a smooth fashion, it is necessary that on links that are common to the old and new LSPs, resources used by the old LSP tunnel should not be released before traffic is transitioned to the new LSP tunnel, and reservations should not be counted twice because this might cause Admission Control to reject the new LSP tunnel. A similar situation can arise when one wants to increase the bandwidth of a TE tunnel. The new reservation will be for the full amount needed, but the actual allocation needed is only the delta between the new and old bandwidth. If policy is being applied to PATH messages by intermediate nodes, then a PATH message requesting too much bandwidth will be rejected. In this situation simply increasing the bandwidth request without changing the SENDER_TEMPLATE, could result in a tunnel being torn down, depending upon local policy.
Top   ToC   RFC3209 - Page 13
   The combination of the LSP_TUNNEL SESSION object and the SE
   reservation style naturally accommodates smooth transitions in
   bandwidth and routing.  The idea is that the old and new LSP tunnels
   share resources along links which they have in common.  The
   LSP_TUNNEL SESSION object is used to narrow the scope of the RSVP
   session to the particular TE tunnel in question.  To uniquely
   identify a TE tunnel, we use the combination of the destination IP
   address (an address of the node which is the egress of the tunnel), a
   Tunnel ID, and the tunnel ingress node's IP address, which is placed
   in the Extended Tunnel ID field.

   During the reroute or bandwidth-increase operation, the tunnel
   ingress needs to appear as two different senders to the RSVP session.
   This is achieved by the inclusion of the "LSP ID", which is carried
   in the SENDER_TEMPLATE and FILTER_SPEC objects.  Since the semantics
   of these objects are changed, a new C-Types are assigned.

   To effect a reroute, the ingress node picks a new LSP ID and forms a
   new SENDER_TEMPLATE.  The ingress node then creates a new ERO to
   define the new path.  Thereafter the node sends a new Path Message
   using the original SESSION object and the new SENDER_TEMPLATE and
   ERO.  It continues to use the old LSP and refresh the old Path
   message.  On links that are not held in common, the new Path message
   is treated as a conventional new LSP tunnel setup.  On links held in
   common, the shared SESSION object and SE style allow the LSP to be
   established sharing resources with the old LSP.  Once the ingress
   node receives a Resv message for the new LSP, it can transition
   traffic to it and tear down the old LSP.

   To effect a bandwidth-increase, a new Path Message with a new LSP_ID
   can be used to attempt a larger bandwidth reservation while the
   current LSP_ID continues to be refreshed to ensure that the
   reservation is not lost if the larger reservation fails.

2.6. Path MTU

Standard RSVP [1] and Int-Serv [11] provide the RSVP sender with the minimum MTU available between the sender and the receiver. This path MTU identification capability is also provided for LSPs established via RSVP. Path MTU information is carried, depending on which is present, in the Integrated Services or Null Service objects. When using Integrated Services objects, path MTU is provided based on the procedures defined in [11]. Path MTU identification when using Null Service objects is defined in [16].
Top   ToC   RFC3209 - Page 14
   With standard RSVP, the path MTU information is used by the sender to
   check which IP packets exceed the path MTU.  For packets that exceed
   the path MTU, the sender either fragments the packets or, when the IP
   datagram has the "Don't Fragment" bit set, issues an ICMP destination
   unreachable message.  This path MTU related handling is also required
   for LSPs established via RSVP.

   The following algorithm applies to all unlabeled IP datagrams and to
   any labeled packets which the node knows to be IP datagrams, to which
   labels need to be added before forwarding.  For labeled packets the
   bottom of stack is found, the IP header examined.

   Using the terminology defined in [5], an LSR MUST execute the
   following algorithm:

   1. Let N be the number of bytes in the label stack (i.e, 4 times the
      number of label stack entries) including labels to be added by
      this node.

   2. Let M be the smaller of the "Maximum Initially Labeled IP Datagram
      Size" or of (Path MTU - N).

   When the size of an IPv4 datagram (without labels) exceeds the value
      of M,

      If the DF bit is not set in the IPv4 header, then

         (a) the datagram MUST be broken into fragments, each of whose
             size is no greater than M, and

         (b) each fragment MUST be labeled and then forwarded.

      If the DF bit is set in the IPv4 header, then

         (a) the datagram MUST NOT be forwarded

         (b) Create an ICMP Destination Unreachable Message:

              i. set its Code field [12] to "Fragmentation Required and
                 DF Set",
             ii. set its Next-Hop MTU field [13] to M

         (c) If possible, transmit the ICMP Destination Unreachable
             Message to the source of the of the discarded datagram.

      When the size of an IPv6 datagram (without labels) exceeds the
             value of M,
Top   ToC   RFC3209 - Page 15
         (a) the datagram MUST NOT be forwarded

         (b) Create an ICMP Packet too Big Message with the Next-Hop
             link MTU field [14] set to M

         (c) If possible, transmit the ICMP Packet too Big Message to
             the source of the of the discarded datagram.

3. LSP Tunnel related Message Formats

Five new objects are defined in this section: Object name Applicable RSVP messages --------------- ------------------------ LABEL_REQUEST Path LABEL Resv EXPLICIT_ROUTE Path RECORD_ROUTE Path, Resv SESSION_ATTRIBUTE Path New C-Types are also assigned for the SESSION, SENDER_TEMPLATE, and FILTER_SPEC, objects. Detailed descriptions of the new objects are given in later sections. All new objects are OPTIONAL with respect to RSVP. An implementation can choose to support a subset of objects. However, the LABEL_REQUEST and LABEL objects are mandatory with respect to this specification. The LABEL and RECORD_ROUTE objects, are sender specific. In Resv messages they MUST appear after the associated FILTER_SPEC and prior to any subsequent FILTER_SPEC. The relative placement of EXPLICIT_ROUTE, LABEL_REQUEST, and SESSION_ATTRIBUTE objects is simply a recommendation. The ordering of these objects is not important, so an implementation MUST be prepared to accept objects in any order.

3.1. Path Message

The format of the Path message is as follows: <Path Message> ::= <Common Header> [ <INTEGRITY> ] <SESSION> <RSVP_HOP> <TIME_VALUES> [ <EXPLICIT_ROUTE> ] <LABEL_REQUEST> [ <SESSION_ATTRIBUTE> ]
Top   ToC   RFC3209 - Page 16
                               [ <POLICY_DATA> ... ]
                               <sender descriptor>

      <sender descriptor> ::=  <SENDER_TEMPLATE> <SENDER_TSPEC>
                               [ <ADSPEC> ]
                               [ <RECORD_ROUTE> ]

3.2. Resv Message

The format of the Resv message is as follows: <Resv Message> ::= <Common Header> [ <INTEGRITY> ] <SESSION> <RSVP_HOP> <TIME_VALUES> [ <RESV_CONFIRM> ] [ <SCOPE> ] [ <POLICY_DATA> ... ] <STYLE> <flow descriptor list> <flow descriptor list> ::= <FF flow descriptor list> | <SE flow descriptor> <FF flow descriptor list> ::= <FLOWSPEC> <FILTER_SPEC> <LABEL> [ <RECORD_ROUTE> ] | <FF flow descriptor list> <FF flow descriptor> <FF flow descriptor> ::= [ <FLOWSPEC> ] <FILTER_SPEC> <LABEL> [ <RECORD_ROUTE> ] <SE flow descriptor> ::= <FLOWSPEC> <SE filter spec list> <SE filter spec list> ::= <SE filter spec> | <SE filter spec list> <SE filter spec> <SE filter spec> ::= <FILTER_SPEC> <LABEL> [ <RECORD_ROUTE> ] Note: LABEL and RECORD_ROUTE (if present), are bound to the preceding FILTER_SPEC. No more than one LABEL and/or RECORD_ROUTE may follow each FILTER_SPEC.


(next page on part 2)

Next Section