RFC 8694

Internet Engineering Task Force (IETF)                           D. King
Request for Comments: 8694                            Old Dog Consulting
Category: Informational                                郑好棉 (H. Zheng)
ISSN: 2070-1721                   华为技术有限公司 (Huawei Technologies)
                                                           December 2019

Applicability of the Path Computation Element to Inter-area and Inter-AS
                   MPLS and GMPLS Traffic Engineering


   The Path Computation Element (PCE) may be used for computing services
   that traverse multi-area and multi-Autonomous System (multi-AS)
   Multiprotocol Label Switching (MPLS) and Generalized MPLS (GMPLS)
   Traffic-Engineered (TE) networks.

   This document examines the applicability of the PCE architecture,
   protocols, and protocol extensions for computing multi-area and
   multi-AS paths in MPLS and GMPLS networks.

Status of This Memo

   This document is not an Internet Standards Track specification; it is
   published for informational purposes.

   This document is a product of the Internet Engineering Task Force
   (IETF).  It represents the consensus of the IETF community.  It has
   received public review and has been approved for publication by the
   Internet Engineering Steering Group (IESG).  Not all documents
   approved by the IESG are candidates for any level of Internet
   Standard; see Section 2 of RFC 7841.

   Information about the current status of this document, any errata,
   and how to provide feedback on it may be obtained at

Copyright Notice

   Copyright (c) 2019 IETF Trust and the persons identified as the
   document authors.  All rights reserved.

   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents
   (https://trustee.ietf.org/license-info) in effect on the date of
   publication of this document.  Please review these documents
   carefully, as they describe your rights and restrictions with respect
   to this document.  Code Components extracted from this document must
   include Simplified BSD License text as described in Section 4.e of
   the Trust Legal Provisions and are provided without warranty as
   described in the Simplified BSD License.

Table of Contents

   1.  Introduction
     1.1.  Domains
     1.2.  Path Computation
       1.2.1.  PCE-Based Path Computation Procedure
     1.3.  Traffic Engineering Aggregation and Abstraction
     1.4.  Traffic-Engineered Label Switched Paths
     1.5.  Inter-area and Inter-AS-capable PCE Discovery
     1.6.  Objective Functions
   2.  Terminology
   3.  Issues and Considerations
     3.1.  Multihoming
     3.2.  Destination Location
     3.3.  Domain Confidentiality
   4.  Domain Topologies
     4.1.  Selecting Domain Paths
     4.2.  Domain Sizes
     4.3.  Domain Diversity
     4.4.  Synchronized Path Computations
     4.5.  Domain Inclusion or Exclusion
   5.  Applicability of the PCE to Inter-area Traffic Engineering
     5.1.  Inter-area Routing
       5.1.1.  Area Inclusion and Exclusion
       5.1.2.  Strict Explicit Path and Loose Path
       5.1.3.  Inter-Area Diverse Path Computation
   6.  Applicability of the PCE to Inter-AS Traffic Engineering
     6.1.  Inter-AS Routing
       6.1.1.  AS Inclusion and Exclusion
     6.2.  Inter-AS Bandwidth Guarantees
     6.3.  Inter-AS Recovery
     6.4.  Inter-AS PCE Peering Policies
   7.  Multi-domain PCE Deployment Options
     7.1.  Traffic Engineering Database and Synchronization
       7.1.1.  Applicability of BGP-LS to PCE
     7.2.  Pre-planning and Management-Based Solutions
   8.  Domain Confidentiality
     8.1.  Loose Hops
     8.2.  Confidential Path Segments and Path-Keys
   9.  Point to Multipoint
   10. Optical Domains
     10.1.  Abstraction and Control of TE Networks (ACTN)
   11. Policy
   12. Manageability Considerations
     12.1.  Control of Function and Policy
     12.2.  Information and Data Models
     12.3.  Liveness Detection and Monitoring
     12.4.  Verifying Correct Operation
     12.5.  Impact on Network Operation
   13. Security Considerations
     13.1.  Multi-domain Security
   14. IANA Considerations
   15. References
     15.1.  Normative References
     15.2.  Informative References

   Authors' Addresses

1.  Introduction

   Computing paths across large multi-domain environments may require
   special computational components and cooperation between entities in
   different domains capable of complex path computation.

   Issues that may exist when routing in multi-domain networks include
   the following:

   *  There is often a lack of full topology and TE information across

   *  No single node has the full visibility to determine an optimal or
      even feasible end-to-end path across domains.

   *  Knowing how to evaluate and select the exit point and next domain
      boundary from a domain.

   *  Understanding how the ingress node determines which domains should
      be used for the end-to-end path.

   An information exchange across multiple domains is often limited due
   to the lack of trust relationship, security issues, or scalability
   issues, even if there is a trust relationship between domains.

   The Path Computation Element (PCE) [RFC4655] provides an architecture
   and a set of functional components to address the problem space and
   the issues highlighted above.

   A PCE may be used to compute end-to-end paths across multi-domain
   environments using a per-domain path computation technique [RFC5152].
   The so-called backward recursive PCE-based computation (BRPC)
   mechanism [RFC5441] defines a path computation procedure to compute
   inter-domain constrained Multiprotocol Label Switching (MPLS) and
   Generalized MPLS (GMPLS) Traffic-Engineered (TE) networks.  However,
   both per-domain and BRPC techniques assume that the sequence of
   domains to be crossed from source to destination is known, either
   fixed by the network operator or obtained by other means.

   In more advanced deployments (including multi-area and multi-
   Autonomous System (multi-AS) environments), the sequence of domains
   may not be known in advance, and the choice of domains in the end-to-
   end domain sequence might be critical to the determination of an
   optimal end-to-end path.  In this case, the use of the hierarchical
   PCE [RFC6805] architecture and mechanisms may be used to discover the
   intra-area path and select the optimal end-to-end domain sequence.

   This document describes the processes and procedures available when
   using the PCE architecture and protocols for computing inter-area and
   inter-AS MPLS and GMPLS Traffic-Engineered paths.

   The scope of this document does not include discussions of deployment
   scenarios for stateful PCE, active PCE, remotely initiated PCE, or
   PCE as a central controller (PCECC).

1.1.  Domains

   Generally, a domain can be defined as a separate administrative,
   geographic, or switching environment within the network.  A domain
   may be further defined as a zone of routing or computational ability.
   Under these definitions, a domain might be categorized as an
   Autonomous System (AS) or an Interior Gateway Protocol (IGP) area (as
   per [RFC4726] and [RFC4655]).

   For the purposes of this document, a domain is considered to be a
   collection of network elements within an area or AS that has a common
   sphere of address management or path computational responsibility.
   Wholly or partially overlapping domains are not within the scope of
   this document.

   In the context of GMPLS, a particularly important example of a domain
   is the Automatically Switched Optical Network (ASON) subnetwork
   [G-8080].  In this case, computation of an end-to-end path requires
   the selection of nodes and links within a parent domain where some
   nodes may, in fact, be subnetworks.  Furthermore, a domain might be
   an ASON routing area [G-7715].  A PCE may perform the path
   computation function of an ASON Routing Controller as described in

   It is assumed that the PCE architecture is not applied to a large
   group of domains, such as the Internet.

1.2.  Path Computation

   For the purpose of this document, it is assumed that path computation
   is the sole responsibility of the PCE as per the architecture defined
   in [RFC4655].  When a path is required, the Path Computation Client
   (PCC) will send a request to the PCE.  The PCE will apply the
   required constraints, compute a path, and return a response to the
   PCC.  In the context of this document, it may be necessary for the
   PCE to cooperate with other PCEs in adjacent domains (as per BRPC
   [RFC5441]) or with a parent PCE (as per [RFC6805]).

   It is entirely feasible that an operator could compute a path across
   multiple domains without the use of a PCE if the relevant domain
   information is available to the network planner or network management
   platform.  The definition of what relevant information is required to
   perform this network planning operation and how that information is
   discovered and applied is outside the scope of this document.

1.2.1.  PCE-Based Path Computation Procedure

   As highlighted, the PCE is an entity capable of computing an inter-
   domain TE path upon receiving a request from a PCC.  There could be a
   single PCE per domain or a single PCE responsible for all domains.  A
   PCE may or may not reside on the same node as the requesting PCC.  A
   path may be computed by either a single PCE node or a set of
   distributed PCE nodes that collaborate during path computation.

   According to [RFC4655], a PCC should send a path computation request
   to a particular PCE using [RFC5440] (PCC-to-PCE communication).  This
   negates the need to broadcast a request to all the PCEs.  Each PCC
   can maintain information about the computation capabilities of the
   PCEs it is aware of.  The PCC-PCE capability awareness can be
   configured using static configurations or by automatic and dynamic
   PCE discovery procedures.

   If a network path is required, the PCC will send a path computation
   request to the PCE.  A PCE may then compute the end-to-end path if it
   is aware of the topology and TE information required to compute the
   entire path.  If the PCE is unable to compute the entire path, the
   PCE architecture provides cooperative PCE mechanisms for the
   resolution of path computation requests when an individual PCE does
   not have sufficient TE visibility.

   End-to-end path segments may be kept confidential through the
   application of Path-Keys to protect partial or full path information.
   A Path-Key is a token that replaces a path segment in an explicit
   route.  The Path-Key mechanism is described in [RFC5520].

1.3.  Traffic Engineering Aggregation and Abstraction

   Networks are often constructed from multiple areas or ASes that are
   interconnected via multiple interconnect points.  To maintain network
   confidentiality and scalability, the TE properties of each area and
   AS are not generally advertised outside each specific area or AS.

   TE aggregation or abstraction provide a mechanism to hide information
   but may cause failed path setups or the selection of suboptimal end-
   to-end paths [RFC4726].  The aggregation process may also have
   significant scaling issues for networks with many possible routes and
   multiple TE metrics.  Flooding TE information breaks confidentiality
   and does not scale in the routing protocol.

   The PCE architecture and associated mechanisms provide a solution to
   avoid the use of TE aggregation and abstraction.

1.4.  Traffic-Engineered Label Switched Paths

   This document highlights the PCE techniques and mechanisms that exist
   for establishing TE packet and optical Label Switched Paths (LSPs)
   across multiple areas (inter-area TE LSP) and ASes (inter-AS TE LSP).
   In this context and within the remainder of this document, we
   consider all LSPs to be constraint based and traffic engineered.

   Three signaling options are defined for setting up an inter-area or
   inter-AS LSP [RFC4726]:

   *  Contiguous LSP

   *  Stitched LSP

   *  Nested LSP

   All three signaling methods are applicable to the architectures and
   procedures discussed in this document.

1.5.  Inter-area and Inter-AS-capable PCE Discovery

   When using a PCE-based approach for inter-area and inter-AS path
   computation, a PCE in one area or AS may need to learn information
   related to inter-AS-capable PCEs located in other ASes.  The PCE
   discovery mechanism defined in [RFC5088] and [RFC5089] facilitates
   the discovery of PCEs and disclosure of information related to inter-
   area and inter-AS-capable PCEs.

1.6.  Objective Functions

   An Objective Function (OF) [RFC5541] or a set of OFs specifies the
   intentions of the path computation and so defines the "optimality" in
   the context of the computation request.

   An OF specifies the desired outcome of a computation.  It does not
   describe or specify the algorithm to use.  Also, an implementation
   may apply any algorithm or set of algorithms to achieve the result
   indicated by the OF.  A number of general OFs are specified in

   Various OFs may be included in the PCE computation request to satisfy
   the policies encoded or configured at the PCC, and a PCE may be
   subject to policy in determining whether it meets the OFs included in
   the computation request or whether it applies its own OFs.

   During inter-domain path computation, the selection of a domain
   sequence, the computation of each (per-domain) path fragment, and the
   determination of the end-to-end path may each be subject to different
   OFs and policies.

2.  Terminology

   This document also uses the terminology defined in [RFC4655] and
   [RFC5440].  Additional terminology is defined below:

   ABR:  IGP Area Border Router -- a router that is attached to more
         than one IGP area.

   ASBR:  Autonomous System Border Router -- a router used to connect
         together ASes of a different or the same Service Provider via
         one or more inter-AS links.

   Inter-area TE LSP:  A TE LSP whose path transits through two or more
         IGP areas.

   Inter-AS MPLS TE LSP:  A TE LSP whose path transits through two or
         more ASes or sub-ASes (BGP confederations)

   SRLG:  Shared Risk Link Group.

   TED:  Traffic Engineering Database, which contains the topology and
         resource information of the domain.  The TED may be fed by
         Interior Gateway Protocol (IGP) extensions or potentially by
         other means.

3.  Issues and Considerations

3.1.  Multihoming

   Networks constructed from multi-areas or multi-AS environments may
   have multiple interconnect points (multihoming).  End-to-end path
   computations may need to use different interconnect points to avoid a
   single-point failure disrupting both the primary and backup services.

3.2.  Destination Location

   A PCC asking for an inter-domain path computation is typically aware
   of the identity of the destination node.  If the PCC is aware of the
   destination domain, it may supply the destination domain information
   as part of the path computation request.  However, if the PCC does
   not know the destination domain, this information must be determined
   by another method.

3.3.  Domain Confidentiality

   When the end-to-end path crosses multiple domains, it may be possible
   that each domain (AS or area) is administered by separate Service
   Providers.  Thus, if a PCE supplies a path segment to a PCE in
   another domain, it may break confidentiality rules and could disclose
   AS-internal topology information.

   If confidentiality is required between domains (ASes and areas)
   belonging to different Service Providers, then cooperating PCEs
   cannot exchange path segments; otherwise, the receiving PCE or PCC
   will be able to see the individual hops through another domain.

   This topic is discussed further in Section 8 of this document.

4.  Domain Topologies

   Constraint-based inter-domain path computation is a fundamental
   requirement for operating traffic-engineered MPLS [RFC3209] and GMPLS
   [RFC3473] networks in inter-area and inter-AS (multi-domain)
   environments.  Path computation across multi-domain networks is
   complex and requires computational cooperational entities like the

4.1.  Selecting Domain Paths

   Where the sequence of domains is known a priori, various techniques
   can be employed to derive an optimal multi-domain path.  If the
   domains are connected to a simple path with no branches and single
   links between all domains or if the preferred points of
   interconnection are also known, the per-domain path computation
   [RFC5152] technique may be used.  Where there are multiple
   connections between domains and there is no preference for the choice
   of points of interconnection, BRPC [RFC5441] can be used to derive an
   optimal path.

   When the sequence of domains is not known in advance or the end-to-
   end path will have to navigate a mesh of small domains (especially
   typical in optical networks), the optimum path may be derived through
   the application of a hierarchical PCE [RFC6805].

4.2.  Domain Sizes

   Very frequently, network domains are composed of dozens or hundreds
   of network elements.  These network elements are usually
   interconnected in a partial-mesh fashion to provide survivability
   against dual failures and to benefit from the traffic-engineering
   capabilities of MPLS and GMPLS protocols.  Network operator feedback
   in the development of the document highlighted that the node degree
   (the number of neighbors per node) typically ranges from 3 to 10 (4-5
   is quite common).

4.3.  Domain Diversity

   Domain and path diversity may also be required when computing end-to-
   end paths.  Domain diversity should facilitate the selection of paths
   that share ingress and egress domains but do not share transit
   domains.  Therefore, there must be a method allowing the inclusion or
   exclusion of specific domains when computing end-to-end paths.

4.4.  Synchronized Path Computations

   In some scenarios, it would be beneficial for the operator to rely on
   the capability of the PCE to perform synchronized path computation.

   Synchronized path computations, known as Synchronization VECtors
   (SVECs), are used for dependent path computations.  SVECs are defined
   in [RFC5440], and [RFC6007] provides an overview of the use of the
   PCE SVEC list for synchronized path computations when computing
   dependent requests.

   In hierarchical PCE (H-PCE) deployments, a child PCE will be able to
   request both dependent and synchronized domain-diverse end-to-end
   paths from its parent PCE.

4.5.  Domain Inclusion or Exclusion

   A domain sequence is an ordered sequence of domains traversed to
   reach the destination domain.  A domain sequence may be supplied
   during path computation to guide the PCEs or are derived via the use
   of hierarchical PCE (H-PCE).

   During multi-domain path computation, a PCC may request specific
   domains to be included or excluded in the domain sequence using the
   Include Route Object (IRO) [RFC5440] and Exclude Route Object (XRO)
   [RFC5521].  The use of Autonomous Number (AS) as an abstract node
   representing a domain is defined in [RFC3209].  [RFC7897] specifies
   new subobjects to include or exclude domains such as an IGP area or a
   4-byte AS number.

   An operator may also need to avoid a path that uses specified nodes
   for administrative reasons.  If a specific connectivity service is
   required to have a 1+1 protection capability, two separate disjoint
   paths must be established.  A mechanism known as Shared Risk Link
   Group (SRLG) information may be used to ensure path diversity.

5.  Applicability of the PCE to Inter-area Traffic Engineering

   As networks increase in size and complexity, it may be required to
   introduce scaling methods to reduce the amount of information flooded
   within the network and make the network more manageable.  An IGP
   hierarchy is designed to improve IGP scalability by dividing the IGP
   domain into areas and limiting the flooding scope of topology
   information to within area boundaries.  This restricts visibility of
   the area to routers in a single area.  If a router needs to compute
   the route to a destination located in another area, a method would be
   required to compute a path across area boundaries.

   In order to support multiple vendors in a network in cases where data
   or control-plane technologies cannot interoperate, it is useful to
   divide the network into vendor domains.  Each vendor domain is an IGP
   area, and the flooding scope of the topology (as well as any other
   relevant information) is limited to the area boundaries.

   Per-domain path computation [RFC5152] exists to provide a method of
   inter-area path computation.  The per-domain solution is based on
   loose hop routing with an Explicit Route Object (ERO) expansion on
   each Area Border Router (ABR).  This allows an LSP to be established
   using a constrained path.  However, at least two issues exist:

   *  This method does not guarantee an optimal constrained path.

   *  The method may require several crankback signaling messages, as
      per [RFC4920], increasing signaling traffic and delaying the LSP

   PCE-based architecture [RFC4655] is designed to solve inter-area path
   computation problems.  The issue of limited topology visibility is
   resolved by introducing path computation entities that are able to
   cooperate in order to establish LSPs with the source and destinations
   located in different areas.

5.1.  Inter-area Routing

   An inter-area TE-LSP is an LSP that transits through at least two IGP
   areas.  In a multi-area network, topology visibility remains local to
   a given area for scaling and privacy purposes.  A node in one area
   will not be able to compute an end-to-end path across multiple areas
   without the use of a PCE.

5.1.1.  Area Inclusion and Exclusion

   The BRPC method [RFC5441] of path computation provides a more optimal
   method to specify inclusion or exclusion of an ABR.  Using the BRPC
   procedure, an end-to-end path is recursively computed in reverse from
   the destination domain towards the source domain.  Using this method,
   an operator might decide if an area must be included or excluded from
   the inter-area path computation.

5.1.2.  Strict Explicit Path and Loose Path

   A strict explicit path is defined as a set of strict hops, while a
   loose path is defined as a set of at least one loose hop and zero or
   more strict hops.  It may be useful to indicate whether a strict
   explicit path is required during the path computation request.  An
   inter-area path may be strictly explicit or loose (e.g., a list of
   ABRs as loose hops).

   A PCC request to a PCE does allow indication of whether a strict
   explicit path across specific areas ([RFC7897]) is required or
   desired or whether the path request is loose.

5.1.3.  Inter-Area Diverse Path Computation

   It may be necessary to compute a path that is partially or entirely
   diverse from a previously computed path to avoid fate sharing of a
   primary service with a corresponding backup service.  There are
   various levels of diversity in the context of an inter-area network:

   *  Per-area diversity (the intra-area path segments are a link, node,
      or SRLG disjoint).

   *  Inter-area diversity (the end-to-end inter-area paths are a link,
      node, or SRLG disjoint).

   Note that two paths may be disjointed in the backbone area but non-
   disjointed in peripheral areas.  Also, two paths may be node
   disjointed within areas but may share ABRs, in which case path
   segments within an area are node disjointed but end-to-end paths are
   not node disjointed.  Per-domain [RFC5152], BRPC [RFC5441], and H-PCE
   [RFC6805] mechanisms all support the capability to compute diverse
   paths across multi-area topologies.

6.  Applicability of the PCE to Inter-AS Traffic Engineering

   As discussed in Section 5 (Applicability of the PCE to Inter-area
   Traffic Engineering), it is necessary to divide the network into
   smaller administrative domains, or ASes.  If an LSR within an AS
   needs to compute a path across an AS boundary, it must also use an
   inter-AS computation technique.  [RFC5152] defines mechanisms for the
   computation of inter-domain TE LSPs using network elements along the
   signaling paths to compute per-domain constrained path segments.

   The PCE was designed to be capable of computing MPLS and GMPLS paths
   across AS boundaries.  This section outlines the features of a PCE-
   enabled solution for computing inter-AS paths.

6.1.  Inter-AS Routing

6.1.1.  AS Inclusion and Exclusion

   [RFC5441] allows the specification of AS or ASBR inclusion or
   exclusion.  Using this method, an operator might decide whether an AS
   must be included or excluded from the inter-AS path computation.
   Exclusion and/or inclusion could also be specified at any step in the
   LSP path computation process by a PCE (within the BRPC algorithm),
   but the best practice would be to specify them at the edge.  In
   opposition to the strict and loose path, AS inclusion or exclusion
   doesn't impose topology disclosure as ASes and their interconnection
   are public entities.

6.2.  Inter-AS Bandwidth Guarantees

   Many operators with multi-AS domains will have deployed the MPLS-TE
   Diffserv either across their entire network or at the domain edges on
   CE-PE links.  In situations where strict QoS bounds are required,
   admission control inside the network may also be required.

   When the propagation delay can be bounded, the performance targets,
   such as maximum one-way transit delay, may be guaranteed by providing
   bandwidth guarantees along the Diffserv-enabled path.  These
   requirements are described in [RFC4216].

   One typical example of the requirements in [RFC4216] is to provide
   bandwidth guarantees over an end-to-end path for VoIP traffic
   classified as an EF (Expedited Forwarding) class in a Diffserv-
   enabled network.  In cases where the EF path is extended across
   multiple ASes, an inter-AS bandwidth guarantee would be required.

   Another case for an inter-AS bandwidth guarantee is the requirement
   to guarantee a certain amount of transit bandwidth across one or
   multiple ASes.

6.3.  Inter-AS Recovery

   During a path computation process, a PCC request may contain the
   requirement to compute a backup LSP for protecting the primary LSP,
   such as 1+1 protection.  A single LSP or multiple backup LSPs may
   also be used for a group of primary LSPs; this is typically known as
   m:n protection.

   Other inter-AS recovery mechanisms include [RFC4090], which adds Fast
   Reroute (FRR) protection to an LSP.  So, the PCE could be used to
   trigger computation of backup tunnels in order to protect inter-AS

   Inter-AS recovery clearly requires backup LSPs for service
   protection, but it would also be advisable to have multiple PCEs
   deployed for path computation redundancy, especially for service
   restoration in the event of catastrophic network failure.

6.4.  Inter-AS PCE Peering Policies

   Like BGP peering policies, inter-AS PCE peering policies are required
   for an operator.  In an inter-AS BRPC process, the PCE must cooperate
   in order to compute the end-to-end LSP.  Therefore, the AS path must
   not only follow technical constraints, e.g., bandwidth availability,
   but also the policies defined by the operator.

   Typically, PCE interconnections at an AS level must follow the agreed
   contract obligations, also known as peering agreements.  The PCE
   peering policies are the result of the contract negotiation and
   govern the relation between the different PCEs.

7.  Multi-domain PCE Deployment Options

7.1.  Traffic Engineering Database and Synchronization

   An optimal path computation requires knowledge of the available
   network resources, including nodes and links, constraints, link
   connectivity, available bandwidth, and link costs.  The PCE operates
   on a view of the network topology as presented by a TED.  As
   discussed in [RFC4655], the TED used by a PCE may be learned by the
   relevant IGP extensions.

   Thus, the PCE may operate its TED by participating in the IGP running
   in the network.  In an MPLS-TE network, this would require OSPF-TE
   [RFC3630] or ISIS-TE [RFC5305].  In a GMPLS network, it would utilize
   the GMPLS extensions to OSPF and IS-IS defined in [RFC4203] and
   [RFC5307].  Inter-AS connectivity information may be populated via
   [RFC5316] and [RFC5392].

   An alternative method to providing network topology and resource
   information is offered by [RFC7752], which is described in the
   following section.

7.1.1.  Applicability of BGP-LS to PCE

   The concept of the exchange of TE information between Autonomous
   Systems (ASes) is discussed in [RFC7752].  The information exchanged
   in this way could be the full TE information from the AS, an
   aggregation of that information, or a representation of the potential
   connectivity across the AS.  Furthermore, that information could be
   updated frequently (for example, for every new LSP that is set up
   across the AS) or only at threshold-crossing events.

   In an H-PCE deployment, the parent PCE will require the inter-domain
   topology and link status between child domains.  This information may
   be learned by a BGP-LS speaker and provided to the parent PCE.
   Furthermore, link-state performance, including delay, available
   bandwidth, and utilized bandwidth, may also be provided to the parent
   PCE for optimal path link selection.

7.2.  Pre-planning and Management-Based Solutions

   Offline path computation is performed ahead of time before the LSP
   setup is requested.  That means that it is requested by or performed
   as part of an Operation Support System (OSS) management application.
   This model can be seen in Section 5.5 of [RFC4655].

   The offline model is particularly appropriate for long-lived LSPs
   (such as those present in a transport network) or for planned
   responses to network failures.  In these scenarios, more planning is
   normally a feature of LSP provisioning.

   The management system may also use a PCE and BRPC to pre-plan an AS
   sequence, and the source domain PCE and per-domain path computation
   to be used when the actual end-to-end path is required.  This model
   may also be used where the operator wishes to retain full manual
   control of the placement of LSPs, using the PCE only as a computation
   tool to assist the operator and not as part of an automated network.

   In environments where operators peer with each other to provide end-
   to-end paths, the operator responsible for each domain must agree on
   the extent to which paths must be pre-planned or manually controlled.

8.  Domain Confidentiality

   This section discusses the techniques that cooperating PCEs can use
   to compute inter-domain paths without each domain disclosing
   sensitive internal topology information (such as explicit nodes or
   links within the domain) to the other domains.

   Confidentiality typically applies to inter-provider (inter-AS) PCE
   communication.  Where the TE LSP crosses multiple domains (ASes or
   areas), the path may be computed by multiple PCEs that cooperate
   together, with each local PCE responsible for computing a segment of
   the path.  With each local PCE responsible for computing a segment of
   the path.

   In situations where ASes are administered by separate Service
   Providers, it would break confidentiality rules for a PCE to supply
   path segment details to a PCE responsible for another domain, thus
   disclosing AS-internal or area topology information.

8.1.  Loose Hops

   A method for preserving the confidentiality of the path segment is
   for the PCE to return a path containing a loose hop in place of the
   segment that must be kept confidential.  The concept of loose and
   strict hops for the route of a TE LSP is described in [RFC3209].

   [RFC5440] supports the use of paths with loose hops; whether it
   returns a full explicit path with strict hops or uses loose hops is a
   local policy decision at a PCE.  A path computation request may
   require an explicit path with strict hops or may allow loose hops, as
   detailed in [RFC5440].

8.2.  Confidential Path Segments and Path-Keys

   [RFC5520] defines the concept and mechanism of a Path-Key. A Path-Key
   is a token that replaces the path segment information in an explicit
   route.  The Path-Key allows the explicit route information to be
   encoded and is contained in the Path Computation Element
   Communication Protocol (PCEP) ([RFC5440]) messages exchanged between
   the PCE and PCC.

   This Path-Key technique allows explicit route information to be used
   for end-to-end path computation without disclosing internal topology
   information between domains.

9.  Point to Multipoint

   For inter-domain point-to-multipoint application scenarios using
   MPLS-TE LSPs, the complexity of domain sequences, domain policies,
   and the choice and number of domain interconnects is magnified
   compared to point-to-point path computations.  As the size of the
   network grows, the number of leaves and branches increases, further
   increasing the complexity of the overall path computation problem.  A
   solution for managing point-to-multipoint path computations may be
   achieved using the PCE inter-domain point-to-multipoint path
   computation [RFC7334] procedure.

10.  Optical Domains

   The International Telecommunication Union (ITU) defines the ASON
   architecture in [G-8080].  [G-7715] defines the routing architecture
   for ASON and introduces a hierarchical architecture.  In this
   architecture, the Routing Areas (RAs) have a hierarchical
   relationship between different routing levels, which means a parent
   (or higher level) RA can contain multiple child RAs.  The
   interconnectivity of the lower RAs is visible to the higher-level RA.

   In the ASON framework, a path computation request is termed a route
   query.  This query is executed before signaling is used to establish
   an LSP, which is termed a Switched Connection (SC) or a Soft
   Permanent Connection (SPC).  [G-7715-2] defines the requirements and
   architecture for the functions performed by Routing Controllers (RC)
   during the operation of remote route queries.  An RC is synonymous
   with a PCE.

   In the ASON routing environment, an RC responsible for an RA may
   communicate with its neighbor RC to request the computation of an
   end-to-end path across several RAs.  The path computation components
   and sequences are defined as follows:

   *  Remote route query.  An operation where a Routing Controller
      communicates with another Routing Controller, which does not have
      the same set of layer resources, in order to compute a routing
      path in a collaborative manner.

   *  Route query requester.  The connection controller or RC that sends
      a route query message to a Routing Controller that requests one or
      more routing paths satisfying a set of routing constraints.

   *  Route query responder.  An RC that performs the path computation
      upon reception of a route query message from a Routing Controller
      or connection controller, and sends a response back at the end of
      the computation.

   When computing an end-to-end connection, the route may be computed by
   a single RC or multiple RCs in a collaborative manner, and the two
   scenarios can be considered a centralized remote route query model
   and a distributed remote route query model.  RCs in an ASON
   environment can also use the hierarchical PCE [RFC6805] model to
   fully match the ASON hierarchical routing model.

10.1.  Abstraction and Control of TE Networks (ACTN)

   Where a single operator operates multiple TE domains (including
   optical environments), an Abstraction and Control of TE Networks
   (ACTN) framework [RFC8453] may be used to create an abstracted
   (virtualized network) view of underlay-interconnected domains.  This
   underlay connectivity is then exposed to higher-layer control
   entities and applications.

   ACTN describes the method and procedure for coordinating the underlay
   per-domain Provisioning Network Controllers (PNCs), which may be
   PCEs, via a hierarchical model to facilitate setup of end-to-end
   connections across interconnected TE domains.

11.  Policy

   Policy is important in the deployment of new services and the
   operation of the network.  [RFC5394] provides a framework for PCE-
   based policy-enabled path computation.  This framework is based on
   the Policy Core Information Model (PCIM) as defined in [RFC3060] and
   further extended by [RFC3460].

   When using a PCE to compute inter-domain paths, policy may be invoked
   by specifying the following:

   *  Each PCC must select which computations it will request from a

   *  Each PCC must select which PCEs it will use.

   *  Each PCE must determine which PCCs are allowed to use its services
      and for what computations.

   *  The PCE must determine how to collect the information in its TED,
      whom to trust for that information, and how to refresh/update the

   *  Each PCE must determine which objective functions and algorithms
      to apply.

12.  Manageability Considerations

   General PCE management considerations are discussed in [RFC4655].  In
   the case of multi-domains within a single service provider network,
   the management responsibility for each PCE would most likely be
   handled by the same service provider.  In the case of multiple ASes
   within different service provider networks, it will likely be
   necessary for each PCE to be configured and managed separately by
   each participating service provider, with policy being implemented
   based on a previously agreed set of principles.

12.1.  Control of Function and Policy

   As per [RFC5440], PCEP implementation allows the user to configure a
   number of PCEP session parameters.  These are detailed in Section 8.1
   of [RFC5440].

   In H-PCE deployments, the administrative entity responsible for the
   management of the parent PCEs for multi-areas would typically be a
   single service provider.  In multiple ASes (managed by different
   service providers), it may be necessary for a third party to manage
   the parent PCE.

12.2.  Information and Data Models

   A PCEP MIB module is defined in [RFC7420], which describes managed
   objects for modeling PCEP communication, including:

   *  PCEP client configuration and status.

   *  PCEP peer configuration and information.

   *  PCEP session configuration and information.

   *  Notifications to indicate PCEP session changes.

   A YANG module for PCEP has also been proposed [PCEP-YANG].

   An H-PCE MIB module or YANG data model will be required to report
   parent PCE and child PCE information, including:

   *  Parent PCE configuration and status.

   *  Child PCE configuration and information.

   *  Notifications to indicate session changes between parent PCEs and
      child PCEs.

   *  Notification of parent PCE TED updates and changes.

12.3.  Liveness Detection and Monitoring

   PCEP includes a keepalive mechanism to check the liveliness of a PCEP
   peer and a notification procedure allowing a PCE to advertise its
   overloaded state to a PCC.  In a multi-domain environment, [RFC5886]
   provides the procedures necessary to monitor the liveliness and
   performance of a given PCE chain.

12.4.  Verifying Correct Operation

   It is important to verify the correct operation of PCEP.  [RFC5440]
   specifies the monitoring of key parameters.  These parameters are
   detailed in [RFC5520].

12.5.  Impact on Network Operation

   [RFC5440] states that in order to avoid any unacceptable impact on
   network operations, a PCEP implementation should allow a limit to be
   placed on the number of sessions that can be set up on a PCEP speaker
   and that it may also be practical to place a limit on the rate of
   messages sent by a PCC and received by the PCE.

13.  Security Considerations

   PCEP security considerations are discussed in [RFC5440] and
   [RFC6952].  Potential vulnerabilities include spoofing, snooping,
   falsification, and using PCEP as a mechanism for denial of service

   As PCEP operates over TCP, it may make use of TCP security encryption
   mechanisms, such as Transport Layer Security (TLS) and TCP
   Authentication Option (TCP-AO).  Usage of these security mechanisms
   for PCEP is described in [RFC8253], and recommendations and best
   current practices are described in [RFC7525].

13.1.  Multi-domain Security

   Any multi-domain operation necessarily involves the exchange of
   information across domain boundaries.  This represents a significant
   security and confidentiality risk.

   It is expected that PCEP is used between PCCs and PCEs that belong to
   the same administrative authority while also using one of the
   aforementioned encryption mechanisms.  Furthermore, PCEP allows
   individual PCEs to maintain the confidentiality of their domain path
   information using path-keys.

14.  IANA Considerations

   This document has no IANA actions.

15.  References

15.1.  Normative References

   [RFC3209]  Awduche, D., Berger, L., Gan, D., Li, T., Srinivasan, V.,
              and G. Swallow, "RSVP-TE: Extensions to RSVP for LSP
              Tunnels", RFC 3209, DOI 10.17487/RFC3209, December 2001,

   [RFC3473]  Berger, L., Ed., "Generalized Multi-Protocol Label
              Switching (GMPLS) Signaling Resource ReserVation Protocol-
              Traffic Engineering (RSVP-TE) Extensions", RFC 3473,
              DOI 10.17487/RFC3473, January 2003,

   [RFC4216]  Zhang, R., Ed. and J.-P. Vasseur, Ed., "MPLS Inter-
              Autonomous System (AS) Traffic Engineering (TE)
              Requirements", RFC 4216, DOI 10.17487/RFC4216, November
              2005, <https://www.rfc-editor.org/info/rfc4216>.

   [RFC4655]  Farrel, A., Vasseur, J.-P., and J. Ash, "A Path
              Computation Element (PCE)-Based Architecture", RFC 4655,
              DOI 10.17487/RFC4655, August 2006,

   [RFC4726]  Farrel, A., Vasseur, J.-P., and A. Ayyangar, "A Framework
              for Inter-Domain Multiprotocol Label Switching Traffic
              Engineering", RFC 4726, DOI 10.17487/RFC4726, November
              2006, <https://www.rfc-editor.org/info/rfc4726>.

   [RFC5152]  Vasseur, JP., Ed., Ayyangar, A., Ed., and R. Zhang, "A
              Per-Domain Path Computation Method for Establishing Inter-
              Domain Traffic Engineering (TE) Label Switched Paths
              (LSPs)", RFC 5152, DOI 10.17487/RFC5152, February 2008,

   [RFC5440]  Vasseur, JP., Ed. and JL. Le Roux, Ed., "Path Computation
              Element (PCE) Communication Protocol (PCEP)", RFC 5440,
              DOI 10.17487/RFC5440, March 2009,

   [RFC5441]  Vasseur, JP., Ed., Zhang, R., Bitar, N., and JL. Le Roux,
              "A Backward-Recursive PCE-Based Computation (BRPC)
              Procedure to Compute Shortest Constrained Inter-Domain
              Traffic Engineering Label Switched Paths", RFC 5441,
              DOI 10.17487/RFC5441, April 2009,

   [RFC5520]  Bradford, R., Ed., Vasseur, JP., and A. Farrel,
              "Preserving Topology Confidentiality in Inter-Domain Path
              Computation Using a Path-Key-Based Mechanism", RFC 5520,
              DOI 10.17487/RFC5520, April 2009,

   [RFC5541]  Le Roux, JL., Vasseur, JP., and Y. Lee, "Encoding of
              Objective Functions in the Path Computation Element
              Communication Protocol (PCEP)", RFC 5541,
              DOI 10.17487/RFC5541, June 2009,

   [RFC6805]  King, D., Ed. and A. Farrel, Ed., "The Application of the
              Path Computation Element Architecture to the Determination
              of a Sequence of Domains in MPLS and GMPLS", RFC 6805,
              DOI 10.17487/RFC6805, November 2012,

15.2.  Informative References

   [RFC3060]  Moore, B., Ellesson, E., Strassner, J., and A. Westerinen,
              "Policy Core Information Model -- Version 1
              Specification", RFC 3060, DOI 10.17487/RFC3060, February
              2001, <https://www.rfc-editor.org/info/rfc3060>.

   [RFC3460]  Moore, B., Ed., "Policy Core Information Model (PCIM)
              Extensions", RFC 3460, DOI 10.17487/RFC3460, January 2003,

   [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Engineering
              (TE) Extensions to OSPF Version 2", RFC 3630,
              DOI 10.17487/RFC3630, September 2003,

   [RFC4090]  Pan, P., Ed., Swallow, G., Ed., and A. Atlas, Ed., "Fast
              Reroute Extensions to RSVP-TE for LSP Tunnels", RFC 4090,
              DOI 10.17487/RFC4090, May 2005,

   [RFC4203]  Kompella, K., Ed. and Y. Rekhter, Ed., "OSPF Extensions in
              Support of Generalized Multi-Protocol Label Switching
              (GMPLS)", RFC 4203, DOI 10.17487/RFC4203, October 2005,

   [RFC4920]  Farrel, A., Ed., Satyanarayana, A., Iwata, A., Fujita, N.,
              and G. Ash, "Crankback Signaling Extensions for MPLS and
              GMPLS RSVP-TE", RFC 4920, DOI 10.17487/RFC4920, July 2007,

   [RFC5088]  Le Roux, JL., Ed., Vasseur, JP., Ed., Ikejiri, Y., and R.
              Zhang, "OSPF Protocol Extensions for Path Computation
              Element (PCE) Discovery", RFC 5088, DOI 10.17487/RFC5088,
              January 2008, <https://www.rfc-editor.org/info/rfc5088>.

   [RFC5089]  Le Roux, JL., Ed., Vasseur, JP., Ed., Ikejiri, Y., and R.
              Zhang, "IS-IS Protocol Extensions for Path Computation
              Element (PCE) Discovery", RFC 5089, DOI 10.17487/RFC5089,
              January 2008, <https://www.rfc-editor.org/info/rfc5089>.

   [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
              Engineering", RFC 5305, DOI 10.17487/RFC5305, October
              2008, <https://www.rfc-editor.org/info/rfc5305>.

   [RFC5307]  Kompella, K., Ed. and Y. Rekhter, Ed., "IS-IS Extensions
              in Support of Generalized Multi-Protocol Label Switching
              (GMPLS)", RFC 5307, DOI 10.17487/RFC5307, October 2008,

   [RFC5316]  Chen, M., Zhang, R., and X. Duan, "ISIS Extensions in
              Support of Inter-Autonomous System (AS) MPLS and GMPLS
              Traffic Engineering", RFC 5316, DOI 10.17487/RFC5316,
              December 2008, <https://www.rfc-editor.org/info/rfc5316>.

   [RFC5392]  Chen, M., Zhang, R., and X. Duan, "OSPF Extensions in
              Support of Inter-Autonomous System (AS) MPLS and GMPLS
              Traffic Engineering", RFC 5392, DOI 10.17487/RFC5392,
              January 2009, <https://www.rfc-editor.org/info/rfc5392>.

   [RFC5394]  Bryskin, I., Papadimitriou, D., Berger, L., and J. Ash,
              "Policy-Enabled Path Computation Framework", RFC 5394,
              DOI 10.17487/RFC5394, December 2008,

   [RFC5521]  Oki, E., Takeda, T., and A. Farrel, "Extensions to the
              Path Computation Element Communication Protocol (PCEP) for
              Route Exclusions", RFC 5521, DOI 10.17487/RFC5521, April
              2009, <https://www.rfc-editor.org/info/rfc5521>.

   [RFC5886]  Vasseur, JP., Ed., Le Roux, JL., and Y. Ikejiri, "A Set of
              Monitoring Tools for Path Computation Element (PCE)-Based
              Architecture", RFC 5886, DOI 10.17487/RFC5886, June 2010,

   [RFC6007]  Nishioka, I. and D. King, "Use of the Synchronization
              VECtor (SVEC) List for Synchronized Dependent Path
              Computations", RFC 6007, DOI 10.17487/RFC6007, September
              2010, <https://www.rfc-editor.org/info/rfc6007>.

   [G-8080]   ITU-T, "Architecture for the automatically switched
              optical network", ITU-T Recommendation G.8080/Y.1304,
              February 2012.

   [G-7715]   ITU-T, "Architecture and requirements for routing in the
              automatically switched optical networks", ITU-T
              Recommendation G.7715/Y.1706, June 2002.

   [G-7715-2] ITU-T, "ASON routing architecture and requirements for
              remote route query", ITU-T
              Recommendation G.7715.2/Y.1706.2, February 2007.

   [RFC6952]  Jethanandani, M., Patel, K., and L. Zheng, "Analysis of
              BGP, LDP, PCEP, and MSDP Issues According to the Keying
              and Authentication for Routing Protocols (KARP) Design
              Guide", RFC 6952, DOI 10.17487/RFC6952, May 2013,

   [RFC7334]  Zhao, Q., Dhody, D., King, D., Ali, Z., and R. Casellas,
              "PCE-Based Computation Procedure to Compute Shortest
              Constrained Point-to-Multipoint (P2MP) Inter-Domain
              Traffic Engineering Label Switched Paths", RFC 7334,
              DOI 10.17487/RFC7334, August 2014,

   [RFC7420]  Koushik, A., Stephan, E., Zhao, Q., King, D., and J.
              Hardwick, "Path Computation Element Communication Protocol
              (PCEP) Management Information Base (MIB) Module",
              RFC 7420, DOI 10.17487/RFC7420, December 2014,

   [RFC7525]  Sheffer, Y., Holz, R., and P. Saint-Andre,
              "Recommendations for Secure Use of Transport Layer
              Security (TLS) and Datagram Transport Layer Security
              (DTLS)", BCP 195, RFC 7525, DOI 10.17487/RFC7525, May
              2015, <https://www.rfc-editor.org/info/rfc7525>.

   [RFC7752]  Gredler, H., Ed., Medved, J., Previdi, S., Farrel, A., and
              S. Ray, "North-Bound Distribution of Link-State and
              Traffic Engineering (TE) Information Using BGP", RFC 7752,
              DOI 10.17487/RFC7752, March 2016,

   [RFC7897]  Dhody, D., Palle, U., and R. Casellas, "Domain Subobjects
              for the Path Computation Element Communication Protocol
              (PCEP)", RFC 7897, DOI 10.17487/RFC7897, June 2016,

   [RFC8253]  Lopez, D., Gonzalez de Dios, O., Wu, Q., and D. Dhody,
              "PCEPS: Usage of TLS to Provide a Secure Transport for the
              Path Computation Element Communication Protocol (PCEP)",
              RFC 8253, DOI 10.17487/RFC8253, October 2017,

   [RFC8453]  Ceccarelli, D., Ed. and Y. Lee, Ed., "Framework for
              Abstraction and Control of TE Networks (ACTN)", RFC 8453,
              DOI 10.17487/RFC8453, August 2018,

              Dhody, D., Hardwick, J., Beeram, V., and J. Tantsura, "A
              YANG Data Model for Path Computation Element
              Communications Protocol (PCEP)", Work in Progress,
              Internet-Draft, draft-ietf-pce-pcep-yang-13, 31 October


   The author would like to thank Adrian Farrel for his review and Meral
   Shirazipour and Francisco Javier Jiménez Chico for their comments.


   Dhruv Dhody
   Huawei Technologies
   Divyashree Techno Park, Whitefield
   Bangalore 560066

   Email: dhruv.ietf@gmail.com

   Quintin Zhao
   Huawei Technologies
   125 Nagog Technology Park
   Acton, MA 01719
   United States of America

   Email: qzhao@huawei.com

   Julien Meuric
   France Telecom
   2, avenue Pierre-Marzin
   22307 Lannion Cedex

   Email: julien.meuric@orange.com

   Olivier Dugeon
   France Telecom
   2, avenue Pierre-Marzin
   22307 Lannion Cedex

   Email: olivier.dugeon@orange.com

   Jon Hardwick
   Metaswitch Networks
   100 Church Street
   EN2 6BQ
   United Kingdom

   Email: jonathan.hardwick@metaswitch.com

   Óscar González de Dios
   Telefonica I+D
   Emilio Vargas 6

   Email: oscar.gonzalezdedios@telefonica.com

Authors' Addresses

   Daniel King
   Old Dog Consulting

   Email: daniel@olddog.co.uk

   Haomian Zheng
   Huawei Technologies
   H1, Huawei Xiliu Beipo Village, Songshan Lake
   Guangdong, 523808

   Email: zhenghaomian@huawei.com

   Additional contact information:

      广东 东莞