Skip to main content

The Application of the Path Computation Element Architecture to the Determination of a Sequence of Domains in MPLS and GMPLS

The information below is for an old version of the document that is already published as an RFC.
Document Type
This is an older version of an Internet-Draft that was ultimately published as RFC 6805.
Authors Daniel King , Adrian Farrel
Last updated 2018-12-20 (Latest revision 2012-08-28)
Replaces draft-king-pce-hierarchy-fwk
RFC stream Internet Engineering Task Force (IETF)
Intended RFC status Informational
Additional resources Mailing list discussion
Stream WG state WG Document
Document shepherd (None)
IESG IESG state Became RFC 6805 (Informational)
Action Holders
Consensus boilerplate Unknown
Telechat date (None)
Responsible AD Stewart Bryant
IESG note
Send notices to (None)
Network Working Group                                      D. King (Ed.)
Internet-Draft                                        Old Dog Consulting
Intended Status: Informational                           A. Farrel (Ed.)
Expires: 29 January 2013                              Old Dog Consulting
                                                         29 August 2012   

   The Application of the Path Computation Element Architecture to the
        Determination of a Sequence of Domains in MPLS and GMPLS



   Computing optimum routes for Label Switched Paths (LSPs) across
   multiple domains in MPLS Traffic Engineering (MPLS-TE) and GMPLS
   networks presents a problem because no single point of path
   computation is aware of all of the links and resources in each
   domain. A solution may be achieved using the Path Computation
   Element (PCE) architecture.

   Where the sequence of domains is known a priori, various techniques
   can be employed to derive an optimum path. If the domains are
   simply-connected, or if the preferred points of interconnection are
   also known, the Per-Domain Path Computation technique can be used.
   Where there are multiple connections between domains and there is
   no preference for the choice of points of interconnection, the
   Backward Recursive Path Computation Procedure (BRPC) can be used to
   derive an optimal path.

   This document examines techniques to establish the optimum path when
   the sequence of domains is not known in advance. The document
   shows how the PCE architecture can be extended to allow the optimum
   sequence of domains to be selected, and the optimum end-to-end path
   to be derived through the use of a hierarchical relationship between

Status of this Memo

   This Internet-Draft is submitted to IETF in full conformance with the
   provisions of BCP 78 and BCP 79.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF), its areas, and its working groups.  Note that
   other groups may also distribute working documents as Internet-

   Internet-Drafts are draft documents valid for a maximum of six months
   and may be updated, replaced, or obsoleted by other documents at any
   time.  It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress."

King and Farrel                                                [Page 1]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   The list of current Internet-Drafts can be accessed at

   The list of Internet-Draft Shadow Directories can be accessed at

   This Internet-Draft will expire on 29 August 2012. 

Copyright Notice

   Copyright (c) 2012 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
   ( 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.


   1. Introduction..................................................3
      1.1 Problem Statement.........................................4
      1.2 Definition of a Domain............. ......................5
      1.3 Assumptions and Requirements..............................5
          1.3.1 Metric Objectives...................................6
          1.3.2 Diversity...........................................6
       Physical Diversity..........................6
       Domain Diversity............................7
          1.3.3 Existing Traffic Engineering Constraints............7
          1.3.4 Commercial Constraints..............................7
          1.3.5 Domain Confidentiality..............................7
          1.3.6 Limiting Information Aggregation....................8
          1.3.7 Domain Interconnection Discovery....................8
      1.4 Terminology...............................................8
   2. Examination of Existing PCE Mechanisms........................9
      2.1 Per Domain Path Computation...............................9
      2.2 Backward Recursive Path Computation.......................10
          2.2.1 Applicability of BRPC when the Domain Path is not 
   3. Hierarchical PCE..............................................11
   4. Hierarchical PCE Procedures...................................12
      4.1 Objective Functions and Policy............................12
      4.2 Maintaining Domain Confidentiality........................13
      4.3 PCE Discovery.............................................13

King and Farrel                                                [Page 2]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

      4.4 Parent Domain Traffic Engineering Database................14
      4.5 Determination of Destination Domain ......................15
      4.6 Hierarchical PCE Examples.................................15
          4.6.1 Hierarchical PCE Initial Information Exchange.......17
          4.6.2 Hierarchical PCE End-to-End Path Computation
          Procedure Example.........................................17
      4.7 Hierarchical PCE Error Handling...........................19
      4.8 Hierarchical PCEP Protocol Extensions.....................19
          4.8.1 PCEP Request Qualifiers.............................19
          4.8.2 Indication of H-PCE Capability......................20
          4.8.3 Intention to Utilize Parent PCE Capabilities........20
          4.8.4 Communication of Domain Connectivity Information....20
          4.8.5 Domain Identifiers..................................21
   5. Hierarchical PCE Applicability................................21
      5.1 autonomous Systems and Areas..............................21
      5.2 ASON architecture (G-7715-2)..............................22
          5.2.1 Implicit Consistency Between Hierarchical PCE and
          5.2.2 Benefits of Hierarchical PCEs in ASON...............24
   6. A Note on BGP-TE..............................................24
      6.1 Use of BGP for TED Synchronization........................25
   7. Management Considerations ....................................25
      7.1 Control of Function and Policy............................25
          7.1.1 Child PCE...........................................25
          7.1.2 Parent PCE..........................................26
          7.1.3 Policy Control......................................26
      7.2 Information and Data Models...............................26
      7.3 Liveness Detection and Monitoring.........................26
      7.4 Verifying Correct Operation...............................26
      7.5. Impact on Network Operation..............................27
   8. Security Considerations ......................................27
   9. IANA Considerations ..........................................28
   10. Acknowledgements ............................................28
   11. References ..................................................28
      11.1. Normative References....................................28
      11.2. Informative References .................................29
   12. Authors' Addresses ..........................................12

1. Introduction

   The capability to compute the routes of end-to-end inter-domain MPLS
   Traffic Engineering (TE) and GMPLS Label Switched Paths (LSPs) is
   expressed as requirements in [RFC4105] and [RFC4216]. This capability
   may be realized by a Path Computation Element (PCE). The PCE
   architecture is defined in [RFC4655]. The methods for establishing
   and controlling inter-domain MPLS-TE and GMPLS LSPs are documented in

   In this context, a domain can be defined as a separate

King and Farrel                                                [Page 3]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   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 [RFC4726] and [RFC4655]. Domains are connected
   through ingress and egress boundary nodes (BNs). A more detailed
   definition is given in Section 1.2.

   In a multi-domain environment, the determination of an end-to-end
   traffic engineered path is a problem because no single point of path
   computation is aware of all of the links and resources in each
   domain. PCEs can be used to compute end-to-end paths using a per-
   domain path computation technique [RFC5152]. Alternatively, the
   backward recursive path computation (BRPC) mechanism [RFC5441]
   allows multiple PCEs to collaborate in order to select an optimal
   end-to-end path that crosses multiple domains. Both mechanisms
   assume that the sequence of domains to be crossed between ingress
   and egress is known in advance.

   This document examines techniques to establish the optimum path when
   the sequence of domains is not known in advance. It shows how the PCE
   architecture can be extended to allow the optimum sequence of domains
   to be selected, and the optimum end-to-end path to be derived.

   The model described in this document introduces a hierarchical
   relationship between domains. It is applicable to environments with
   small groups of domains where visibility from the ingress Label
   Switching Router (LSR) is limited. Applying the hierarchical PCE
   model to large groups of domains such as the Internet, is not
   considered feasible or desirable, and is out of scope for this

   This document does not specify any protocol extensions or
   enhancements. That work is left for future protocol specification
   documents. However, some assumptions are made about which protocols
   will be used to provide specific functions, and guidelines to
   future protocol developers are made in the form of requirements

1.1 Problem Statement

   Using a PCE to compute a path between nodes within a single domain is
   relatively straightforward. Computing an end-to-end path when the
   source and destination nodes are located in different domains
   requires co-operation between multiple PCEs, each responsible for
   its own domain.

   Techniques for inter-domain path computation described so far
   ([RFC5152] and [RFC5441]) assume that the sequence of domains to be

King and Farrel                                                [Page 4]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   crossed from source to destination is well known. No explanation is
   given (for example, in [RFC4655]) of how this sequence is generated
   or what criteria may be used for the selection of paths between
   domains. In small clusters of domains, such as simple cooperation
   between adjacent ISPs, this selection process is not complex. In more
   advanced deployments (such as optical networks constructed from
   multiple sub-domains, or in multi-AS environments) the choice of
   domains in the end-to-end domain sequence can be critical to the
   determination of an optimum end-to-end path.

1.2 Definition of a Domain

   A domain is defined in [RFC4726] as any collection of network
   elements within a common sphere of address management or path
   computational responsibility. Examples of such domains include
   IGP areas and Autonomous Systems. 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, a domain might be an ASON Routing Area
   [G-7715]. Furthermore, computation of an end-to-end path requires
   the selection of nodes and links within a routing area where some
   nodes may, in fact, be subnetworks. A PCE may perform the path 
   computation function of an ASON Routing Controller as described in 
   [G-7715-2]. See Section 5.2 for a further discussion of the 
   applicability to the ASON architecture.

   This document assumes that the selection of a sequence of domains for
   an end-to-end path is in some sense a hierarchical path computation
   problem. That is, where one mechanism is used to determine a path
   across a domain, a separate mechanism (or at least a separate set
   of paradigms) is used to determine the sequence of domains. The
   responsibility for the selection of domain interconnection can belong
   to either or both of the mechanisms.

1.3 Assumptions and Requirements

   Networks are often constructed from multiple domains. These
   domains are often interconnected via multiple interconnect points.
   Its assumed that the sequence of domains for an end-to-end path is
   not always well known; that is, an application requesting end-to-end
   connectivity has no preference for, or no ability to specify, the
   sequence of domains to be crossed by the path.

   The traffic engineering properties of a domain cannot be seen from
   outside the domain. Traffic engineering aggregation or abstraction,

King and Farrel                                                [Page 5]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   hides information and can lead to failed path setup 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.
   See Section 6 for a discussion of the concept of inter-domain traffic
   engineering information exchange known as BGP-TE.

   The primary goal of this document is to define how to derive optimal
   end-to-end, multi-domain paths when the sequence of domains is not
   known in advance. The solution needs to be scalable and to maintain
   internal domain topology confidentiality while providing the optimal
   end-to-end path. It cannot rely on the exchange of TE information
   between domains, and for the confidentiality, scaling, and 
   aggregation reasons just cited, it cannot utilize a computation 
   element that has universal knowledge of TE properties and topology
   of all domains.

   The sub-sections that follow set out the primary objectives and
   requirements to be satisfied by a PCE solution to multi-domain path

1.3.1 Metric Objectives

   The definition of optimality is dependent on policy, and is based on
   a single objective or a group objectives. An objective is expressed
   as an objective function [RFC5541] and may be specified on a path
   computation request. The following objective functions are identified
   in this document. They define how the path metrics and TE link
   qualities are manipulated during inter-domain path computation. The
   list is not proscriptive and may be expanded in other documents.

   o Minimize the cost of the path [RFC5541]
   o Select a path using links with the minimal load [RFC5541]
   o Select a path that leaves the maximum residual bandwidth [RFC5541]
   o Minimize aggregate bandwidth consumption [RFC5541]
   o Minimize the Load of the most loaded Link [RFC5541]
   o Minimize the Cumulative Cost of a set of paths [RFC5541]
   o Minimize or cap the number of domains crossed
   o Disallow domain re-entry

   See Section 4.1 for further discussion of objective functions.

1.3.2 Diversity Physical Diversity

   Within a Carrier's carrier environment MPLS services may share common 
   underlying equipment and resources, including optical fiber and 

King and Farrel                                                [Page 6]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   nodes. An MPLS service request may require a path for traffic that is 
   physically disjointed from another service. Thus, if a physical link 
   or node fails on one of the disjoint paths, not all traffic is lost. Domain Diversity

   A pair of paths are domain-diverse if they do not transit any of the
   same domains. A pair of paths that share a common ingress and egress
   are domain-diverse if they only share the same domains at the ingress
   and egress (the ingress and egress domains). Domain diversity may be
   maximized for a pair of paths by selecting paths that have the
   smallest number of shared domains. (Note that this is not the same
   as finding paths with the greatest number of distinct domains!)

   Path computation should facilitate the selection of paths that share
   ingress and egress domains, but do not share any transit domains.
   This provides a way to reduce the risk of shared failure along any
   path, and automatically helps to ensure path diversity for most of
   the route of a pair of LSPs.

   Thus, domain path selection should provide the capability to include
   or exclude specific domains and specific boundary nodes.

1.3.3 Existing Traffic Engineering Constraints

   Any solution should take advantage of typical traffic engineering
   constraints (hop count, bandwidth, lambda continuity, path cost,
   etc.) to meet the service demands expressed in the path computation
   request [RFC4655].

1.3.4 Commercial Constraints

   The solution should provide the capability to include commercially
   relevant constraints such as policy, SLAs, security, peering
   preferences, and monetary costs.

   Additionally it may be necessary for the service provider to
   request that specific domains are included or excluded based on
   commercial relationships, security implications, and reliability.

1.3.5 Domain Confidentiality

   A key requirement is the ability to maintain domain confidentiality
   when computing inter-domain end-to-end paths. It should be possible
   for local policy to require that a PCE not disclose to any other PCE
   the intra-domain paths it computes or the internal topology of the
   domain it serves. This requirement should have no impact in the
   optimality or quality of the end-to-end path that is derived.

King and Farrel                                                [Page 7]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

1.3.6 Limiting Information Aggregation

   In order to reduce processing overhead and to not sacrifice
   computational detail, there should be no requirement to aggregate or
   abstract traffic engineering link information.

1.3.7 Domain Interconnection Discovery

   To support domain mesh topologies, the solution should allow the
   discovery and selection of domain inter-connections. Pre-
   configuration of preferred domain interconnections should also be
   supported for network operators that have bilateral agreement, and
   preference for the choice of points of interconnection.

1.4 Terminology

   This document uses PCE terminology defined in [RFC4655], [RFC4726],
   and [RFC5440]. Additional terms are defined below.

   Domain Path: The sequence of domains for a path.

   Ingress Domain: The domain that includes the ingress LSR of a path.

   Transit Domain: A domain that has an upstream and downstream
   neighbor domain for a specific path.

   Egress Domain: The domain that includes the egress LSR of a path.

   Boundary Nodes: Each Domain has entry LSRs and exit LSRs that could
   be Area Border Routers (ABRs) or Autonomous System Border Routers
   (ASBRs) depending on the type of domain. They are defined here more
   generically as Boundary Nodes (BNs).

   Entry BN of domain(n): a BN connecting domain(n-1) to domain(n)
   on a path.

   Exit BN of domain(n): a BN connecting domain(n) to domain(n+1)
   on a path.

   Parent Domain: A domain higher up in a domain hierarchy such
   that it contains other domains (child domains) and potentially other
   links and nodes.

   Child Domain: A domain lower in a domain hierarchy such that it has
   a parent domain.

   Parent PCE: A PCE responsible for selecting a path across a parent
   domain and any number of child domains by coordinating with child
   PCEs and examining a topology map that shows domain inter-

King and Farrel                                                [Page 8]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   Child PCE: A PCE responsible for computing the path across one or
   more specific (child) domains. A child PCE maintains a relationship
   with at least one parent PCE.

   OF: Objective Function: A set of one or more optimization
   criteria used for the computation of a single path (e.g., path cost
   minimization), or the synchronized computation of a set of paths
   (e.g., aggregate bandwidth consumption minimization). See [RFC4655]
   and [RFC5541].

2. Examination of Existing PCE Mechanisms

   This section provides a brief overview of two existing PCE
   cooperation mechanisms called the per-domain path computation method,
   and the backward recursive path computation method. It describes the
   applicability of these methods to the multi-domain problem.

2.1 Per-Domain Path Computation

   The per-domain path computation method for establishing inter-domain
   TE-LSPs [RFC5152] defines a technique whereby the path is computed
   during the signalling process on a per-domain basis. The entry BN of
   each domain is responsible for performing the path computation for
   the section of the LSP that crosses the domain, or for requesting
   that a PCE for that domain computes that piece of the path.

   During per-domain path computation, each computation results in the
   best path across the domain to provide connectivity to the next
   domain in the domain sequence (usually indicated in signalling by an
   identifier of the next domain or the identity of the next entry BN).

   Per-domain path computation may lead to sub-optimal end-to-end paths
   because the most optimal path in one domain may lead to the choice of
   an entry BN for the next domain that results in a very poor path
   across that next domain.

   In the case that the domain path (in particular, the sequence of
   boundary nodes) is not known, the path computing entity must select
   an exit BN based on some determination of how to reach the 
   destination that is outside the domain for which the path computing 
   entity has computational responsibility. [RFC5152] suggest that this 
   might be achieved using the IP shortest path as advertise by BGP. 
   Note, however, that the existence of an IP forwarding path does not 
   guarantee the presence of sufficient bandwidth, let alone an optimal 
   TE path. Furthermore, in many GMPLS systems inter-domain IP routing 
   will not be present. Thus, per-domain path computation may require a 
   significant number of crankback routing attempts to establish even a 
   sub-optimal path.

 King and Farrel                                                [Page 9]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   Note also that the path computing entities in each domain may have 
   different computation capabilities, may run different path 
   computation algorithms, and may apply different sets of constraints 
   and optimization criteria, etc.
   This can result in the end-to-end path being inconsistent and sub-

   Per-domain path computation can suit simply-connected domains where
   the preferred points of interconnection are known.

2.2 Backward Recursive Path Computation

   The Backward Recursive Path Computation (BRPC) [RFC5441] procedure
   involves cooperation and communication between PCEs in order to
   compute an optimal end-to-end path across multiple domains. The
   sequence of domains to be traversed can either be determined before
   or during the path computation. In the case where the sequence of
   domains is known, the ingress Path Computation Client (PCC) sends a
   path computation request to a PCE responsible for the ingress
   domain. This request is forwarded between PCEs, domain-by-domain, to
   a PCE responsible for the egress domain. The PCE in the egress    
   domain creates a set of optimal paths from all of the domain entry
   BNs to the egress LSR. This set is represented as a tree of potential
   paths called a Virtual Shortest Path Tree (VSPT), and the PCE passes
   it back to the previous PCE on the domain path. As the VSPT is passed
   back toward the ingress domain, each PCE computes the optimal paths
   from its entry BNs to its exit BNs that connect to the rest of the
   tree. It adds these paths to the VSPT and passes the VSPT on until
   the PCE for the ingress domain is reached and computes paths from the
   ingress LSR to connect to the rest of the tree. The ingress PCE then
   selects the optimal end-to-end path from the tree, and returns the
   path to the initiating PCC.

   BRPC may suit environments where multiple connections exist between
   domains and there is no preference for the choice of points of
   interconnection. It is best suited to scenarios where the domain
   path is known in advance, but can also be used when the domain path
   is not known.

2.2.1. Applicability of BRPC when the Domain Path is Not Known

   As described above, BRPC can be used to determine an optimal inter-
   domain path when the domain sequence is known. Even when the sequence
   of domains is not known BRPC could be used as follows.

   o The PCC sends a request to a PCE for the ingress domain (the
     ingress PCE).

King and Farrel                                               [Page 10]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   o The ingress PCE sends the path computation request direct to a
     PCE responsible for the domain containing the destination node (the
     egress PCE).

   o The egress PCE computes an egress VSPT and passes it to a PCE
     responsible for each of the adjacent (potentially upstream)

   o Each PCE in turn constructs a VSPT and passes it on to all of its
     neighboring PCEs.

   o When the ingress PCE has received a VSPT from each of its
     neighboring domains it is able to select the optimum path.

   Clearly this mechanism (which could be called path computation
   flooding) has significant scaling issues. It could be improved by
   the application of policy and filtering, but such mechanisms are not
   simple and would still leave scaling concerns.

3. Hierarchical PCE

   In the hierarchical PCE architecture, a parent PCE maintains a domain
   topology map that contains the child domains (seen as vertices in the
   topology) and their interconnections (links in the topology). The
   parent PCE has no information about the content of the child domains;
   that is, the parent PCE does not know about the resource availability
   within the child domains, nor about the availability of connectivity
   across each domain because such knowledge would violate the
   confidentiality requirement and would either require flooding of full
   information to the parent (scaling issue) or would necessitate some
   form of aggregation. The parent PCE is aware of the TE capabilities
   of the interconnections between child domains as these
   interconnections are links in its own topology map. 

   Note that in the case that the domains are IGP areas, there is no
   link between the domains (the ABRs have a presence in both
   neighboring areas). The parent domain may choose to represent this in
   its TED as a virtual link that is unconstrained and has zero cost,
   but this is entirely an implementation issue.

   Each child domain has at least one PCE capable of computing paths
   across the domain. These PCEs are known as child PCEs and have a
   relationship with the parent PCE. Each child PCE also knows the
   identity of the domains that neighbor its own domain. A child PCE
   only knows the topology of the domain that it serves and does not
   know the topology of other child domains. Child PCEs are also not
   aware of the general domain mesh connectivity (i.e., the domain

King and Farrel                                               [Page 11]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   topology map) beyond the connectivity to the immediate neighbor
   domains of the domain it serves.

   The parent PCE builds the domain topology map either from
   configuration or from information received from each child PCE. This
   tells it how the domains are interconnected including the TE
   properties of the domain interconnections. But the parent PCE does
   not know the contents of the child domains. Discovery of the domain
   topology and domain interconnections is discussed further in Section

   When a multi-domain path is needed, the ingress PCE sends a request
   to the parent PCE (using the path computation element protocol, PCEP
   [RFC5440]). The parent PCE selects a set of candidate domain paths
   based on the domain topology and the state of the inter-domain links.
   It then sends computation requests to the child PCEs responsible for
   each of the domains on the candidate domain paths. These requests may
   be sequential or parallel depending on implementation details.

   Each child PCE computes a set of candidate path segments across its
   domain and sends the results to the parent PCE. The parent PCE uses
   this information to select path segments and concatenate them to
   derive the optimal end-to-end inter-domain path. The end-to-end path
   is then sent to the child PCE which received the initial path request
   and this child PCE passes the path on to the PCC that issued the
   original request.
   Specific deployment and implementation scenarios are out of scope of 
   this document. However the hierarchical PCE architecture described 
   does support the function of parent PCE and child PCE being 
   implemented as a common PCE. 

4. Hierarchical PCE Procedures

4.1 Objective Functions and Policy

   Deriving the optimal end-to-end domain path sequence is dependent on
   the policy applied during domain path computation. An Objective
   Function (OF) [RFC5541], or set of OFs, may be applied to define the
   policy being applied to the domain path computation.

   The OF specifies the desired outcome of the computation. It does
   not describe the algorithm to use. When computing end-to-end inter-
   domain paths, required OFs may include (see Section 1.3.1):

   o Minimum cost path
   o Minimum load path
   o Maximum residual bandwidth path
   o Minimize aggregate bandwidth consumption

King and Farrel                                               [Page 12]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   o Minimize or cap the number of transit domains
   o Disallow domain re-entry

   The objective function may be requested by the PCC, the ingress
   domain PCE (according to local policy), or applied by the parent PCE
   according to inter-domain policy.

   More than one OF (or a composite OF) may be chosen to apply to a
   single computation provided they are not contradictory. Composite OFs
   may include weightings and preferences for the fulfilment of pieces
   of the desired outcome.

4.2 Maintaining Domain Confidentiality

   Information about the content of child domains is not shared for
   scaling and confidentiality reasons. This means that a parent PCE is
   aware of the domain topology and the nature of the connections
   between domains, but is not aware of the content of the domains.
   Similarly, a child PCE cannot know the internal topology of another
   child domain. Child PCEs also do not know the general domain mesh
   connectivity, this information is only known by the parent PCE.

   As described in the earlier sections of this document, PCEs can
   exchange path information in order to construct an end-to-end inter-
   domain path. Each per-domain path fragment reveals information about
   the topology and resource availability within a domain. Some
   management domains or ASes will not want to share this information
   outside of the domain (even with a trusted parent PCE). In order to
   conceal the information, a PCE may replace a path segment with a
   path-key [RFC5520]. This mechanism effectively hides the content of a
   segment of a path.

4.3 PCE Discovery

   It is a simple matter for each child PCE to be configured with the
   address of its parent PCE. Typically, there will only be one or two
   parents of any child.

   The parent PCE also needs to be aware of the child PCEs for all child
   domains that it can see. This information is most likely to be
   configured (as part of the administrative definition of each

   Discovery of the relationships between parent PCEs and child PCEs
   does not form part of the hierarchical PCE architecture. Mechanisms
   that rely on advertising or querying PCE locations across domain or
   provider boundaries are undesirable for security, scaling,
   commercial, and confidentiality reasons.

King and Farrel                                               [Page 13]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   The parent PCE also needs to know the inter-domain connectivity.
   This information could be configured with suitable policy and
   commercial rules, or could be learned from the child PCEs as
   described in Section 4.4.

   In order for the parent PCE to learn about domain interconnection
   the child PCE will report the identity of its neighbor domains. The
   IGP in each neighbor domain can advertise its inter-domain TE
   link capabilities [RFC5316], [RFC5392]. This information can be
   collected by the child PCEs and forwarded to the parent PCE, or the
   parent PCE could participate in the IGP in the child domains.

4.4 Parent Domain Traffic Engineering Database

   The parent PCE maintains a domain topology map of the child domains
   and their interconnectivity. Where inter-domain connectivity is
   provided by TE links the capabilities of those links may also be
   known to the parent PCE. The parent PCE maintains a traffic
   engineering database (TED) for the parent domain in the same way that
   any PCE does.

   The parent domain may just be the collection of child domains and
   their interconnectivity, may include details of the inter-domain TE
   links, and may contain nodes and links in its own right.

   The mechanism for building the parent TED is likely to rely heavily
   on administrative configuration and commercial issues because the
   network was probably partitioned into domains specifically to address
   these issues.

   In practice, certain information may be passed from the child domains
   to the parent PCE to help build the parent TED. In theory, the parent
   PCE could listen to the routing protocols in the child domains, but
   this would violate the confidentiality and scaling issues that may be
   responsible for the partition of the network into domains. So it is
   much more likely that a suitable solution will involve specific
   communication from an entity in the child domain (such as the child
   PCE) to convey the necessary information. As already mentioned, the
   "necessary information" relates to how the child domains are inter-
   connected. The topology and available resources within the child
   domain do not need to be communicated to the parent PCE: doing so
   would violate the PCE architecture. Mechanisms for reporting this
   information are described in the examples in Section 4.6 in abstract
   terms as "a child PCE reports its neighbor domain connectivity to its
   parent PCE"; the specifics of a solution are out of scope of this
   document, but the requirements are indicated in Section 4.8. See
   Section 6 for a brief discussion of BGP-TE.

King and Farrel                                               [Page 14]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   In models such as ASON (see Section 5.2), it is possible to consider
   a separate instance of an IGP running within the parent domain where
   the participating protocol speakers are the nodes directly present in
   that domain and the PCEs (Routing Controllers) responsible for each
   of the child domains.

4.5 Determination of Destination Domain

   The PCC asking for an inter-domain path computation is aware of the
   identity of the destination node by definition. If it knows the
   egress domain it can supply this information as part of the path
   computation request. However, if it does not know the egress domain
   this information must be known by the child PCE or known/determined
   by the parent PCE.

   In some specialist topologies the parent PCE could determine the
   destination domain based on the destination address, for example from
   configuration. However, this is not appropriate for many multi-domain
   addressing scenarios. In IP-based multi-domain networks the
   parent PCE may be able to determine the destination domain by
   participating in inter-domain routing. Finally, the parent PCE could
   issue specific requests to the child PCEs to discover if they contain
   the destination node, but this has scaling implications.

   For the purposes of this document, the precise mechanism of the
   discovery of the destination domain is left out of scope. Suffice to
   say that for each multi-domain path computation some mechanism will
   be required to determine the location of the destination.

4.6 Hierarchical PCE Examples

   The following example describes the generic hierarchical domain
   topology. Figure 1 demonstrates four interconnected domains within a
   fifth, parent domain. Each domain contains a single PCE:

   o Domain 1 is the ingress domain and child PCE 1 is able to compute
     paths within the domain. Its neighbors are Domain 2 and Domain 4.
     The domain also contains the source LSR (S) and three egress
     boundary nodes (BN11, BN12, and BN13).

   o Domain 2 is served by child PCE 2. Its neighbors are Domain 1 and
     Domain 3. The domain also contains four boundary nodes (BN21, BN22,
     BN23, and BN24).

   o Domain 3 is the egress domain and is served by child PCE 3. Its
     neighbors are Domain 2 and Domain 4. The domain also contains the
     destination LSR (D) and three ingress boundary nodes (BN31, BN32,
     and BN33).

   o Domain 4 is served by child PCE 4. Its neighbors are Domain 2 and
     Domain 3. The domain also contains two boundary nodes (BN41 and

King and Farrel                                               [Page 15]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   All of these domains are contained within Domain 5 which is served
   by the parent PCE (PCE 5).

   |   Domain 5                                                      |
   |                              -----                              |
   |                             |PCE 5|                             |
   |                              -----                              |
   |                                                                 |
   |    ----------------     ----------------     ----------------   |
   |   | Domain 1       |   | Domain 2       |   | Domain 3       |  |
   |   |                |   |                |   |                |  |
   |   |        -----   |   |        -----   |   |        -----   |  |
   |   |       |PCE 1|  |   |       |PCE 2|  |   |       |PCE 3|  |  |
   |   |        -----   |   |        -----   |   |        -----   |  |
   |   |                |   |                |   |                |  |
   |   |            ----|   |----        ----|   |----            |  |
   |   |           |BN11+---+BN21|      |BN23+---+BN31|           |  |
   |   |   -        ----|   |----        ----|   |----        -   |  |
   |   |  |S|           |   |                |   |           |D|  |  |
   |   |   -        ----|   |----        ----|   |----        -   |  |
   |   |           |BN12+---+BN22|      |BN24+---+BN32|           |  |
   |   |            ----|   |----        ----|   |----            |  |
   |   |                |   |                |   |                |  |
   |   |         ----   |   |                |   |   ----         |  |
   |   |        |BN13|  |   |                |   |  |BN33|        |  |
   |    -----------+----     ----------------     ----+-----------   |
   |                \                                /               |
   |                 \       ----------------       /                |
   |                  \     |                |     /                 |
   |                   \    |----        ----|    /                  |
   |                    ----+BN41|      |BN42+----                   |
   |                        |----        ----|                       |
   |                        |                |                       |
   |                        |        -----   |                       |
   |                        |       |PCE 4|  |                       |
   |                        |        -----   |                       |
   |                        |                |                       |
   |                        | Domain 4       |                       |
   |                         ----------------                        |
   |                                                                 |

                 Figure 1 : Sample Hierarchical Domain Topology

   Figure 2, shows the view of the domain topology as seen by the parent
   PCE (PCE 5). This view is an abstracted topology; PCE 5 is aware of
   domain connectivity, but not of the internal topology within each

King and Farrel                                               [Page 16]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

                      | Domain 5                   |
                      |            ----            |
                      |           |PCE5|           |
                      |            ----            |
                      |                            |
                      |   ----     ----     ----   |
                      |  |    |---|    |---|    |  |
                      |  | D1 |   | D2 |   | D3 |  |
                      |  |    |---|    |---|    |  |
                      |   ----     ----     ----   |
                      |     \      ----      /     |
                      |      \    |    |    /      |
                      |       ----| D4 |----       |
                      |           |    |           |
                      |            ----            |
                      |                            |

      Figure 2 : Abstract Domain Topology as Seen by the Parent PCE

4.6.1 Hierarchical PCE Initial Information Exchange

   Based on the Figure 1 topology, the following is an illustration of
   the initial hierarchical PCE information exchange.

   1. Child PCE 1, the PCE responsible for Domain 1, is configured
      with the location of its parent PCE (PCE5).

   2. Child PCE 1 establishes contact with its parent PCE. The parent
      applies policy to ensure that communication with PCE 1 is allowed.

   3. Child PCE 1 listens to the IGP in its domain and learns its
      inter-domain connectivity. That is, it learns about the links
      BN11-BN21, BN12-BN22, and BN13-BN41.

   4. Child PCE 1 reports its neighbor domain connectivity to its parent

   5. Child PCE 1 reports any change in the resource availability on its
      inter-domain links to its parent PCE.

   Each child PCE performs steps 1 through 5 so that the parent PCE can
   create a domain topology view as shown in Figure 2.

4.6.2 Hierarchical PCE End-to-End Path Computation Procedure

   The procedure below is an example of a source PCC requesting an
   end-to-end path in a multi-domain environment. The topology is

King and Farrel                                               [Page 17]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   represented in Figure 1. It is assumed that the each child PCE has
   connected to its parent PCE and exchanged the initial information
   required for the parent PCE to create its domain topology view as
   described in Section 4.6.1.

   1.  The source PCC (the ingress LSR in our example), sends a request
       to the PCE responsible for its domain (PCE 1) for a path to the
       destination LSR (D).

   2.  PCE 1 determines the destination is not in domain 1.

   3.  PCE 1 sends a computation request to its parent PCE (PCE 5).

   4.  The parent PCE determines that the destination is in Domain 3.
       (See Section 4.5).

   5.  PCE 5 determines the likely domain paths according to the domain
       interconnectivity and TE capabilities between the domains. For
       example, assuming that the link BN12-BN22 is not suitable for the
       requested path, three domain paths are determined:

   6.  PCE 5 sends edge-to-edge path computation requests to PCE 2
       which is responsible for Domain 2 (i.e., BN21-to-BN23 and BN21-
       to-BN24), and to PCE 4 for Domain 4 (i.e., BN41-to-BN42).

   7.  PCE 5 sends source-to-edge path computation requests to PCE 1
       which is responsible for Domain 1 (i.e., S-to-BN11 and S-to-

   8.  PCE 5 sends edge-to-egress path computation requests to PCE3
       which is responsible for Domain 3 (i.e., BN31-to-D, BN32-to-D,
       and BN33-to-D).

   9.  PCE 5 correlates all the computation responses from each child
       PCE, adds in the information about the inter-domain links, and
       applies any requested and locally configured policies.

   10. PCE 5 then selects the optimal end-to-end multi-domain path
       that meets the policies and objective functions, and supplies the
       resulting path to PCE 1.

   11. PCE 1 forwards the path to the PCC (the ingress LSR).

   Note that there is no requirement for steps 6, 7, and 8 to be carried
   out in parallel or in series. Indeed, they could be overlapped with
   step 5. This is an implementation issue.

King and Farrel                                               [Page 18]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

4.7 Hierarchical PCE Error Handling

   In the event that a child PCE in a domain cannot find a suitable
   path to the egress, the child PCE should return the relevant
   error to notify the parent PCE. Depending on the error response the
   parent PCE can elect to:

   o Cancel the request and send the relevant response back to the
     initial child PCE that requested an end-to-end path;
   o Relax some of the constraints associated with the initial path   
   o Select another candidate domain and send the path request to the
     child PCE responsible for the domain.

   If the parent PCE does not receive a response from a child PCE within
   an allotted time period. The parent PCE can elect to:

   o Cancel the request and send the relevant response back to the
     initial child PCE that requested an end-to-end path;  
   o Send the path request to another child PCE in the same domain, if a
     secondary child PCE exists;
   o Select another candidate domain and send the path request to the
     child PCE responsible for that domain. 
   The parent PCE may also want to prune any unresponsive child PCE 
   domain paths from the candidate set.

4.8 Requirements for Hierarchical PCEP Protocol Extensions

   This section lists the high-level requirements for extensions to the
   PCEP to support the hierarchical PCE model. It is provided to offer
   guidance to PCEP protocol developers in designing a solution suitable
   for use in a hierarchical PCE framework.

4.8.1 PCEP Request Qualifiers

   PCEP request (PCReq) messages are used by a PCC or a PCE to make a
   computation request or enquiry to a PCE. The requests are qualified
   so that the PCE knows what type of action is required.

   Support of the hierarchical PCE architecture will introduce two new
   qualifications as follows:

   o It must be possible for a child PCE to indicate that the response
     it receives from the parent PCE should consist of a domain sequence
     only (i.e., not a fully-specified end-to-end path). This allows the
     child PCE to initiate per-domain or backward recursive path

   o A parent PCE may need to be able to ask a child PCE whether a
     particular node address (the destination of an end-to-end path) is
     present in the domain that the child PCE serves.

King and Farrel                                               [Page 19]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   In PCEP, such request qualifications are carried as bit-flags in the
   RP object within the PCReq message.

4.8.2 Indication of Hierarchical PCE Capability

   Although parent/child PCE relationships are likely configured, it
   will assist network operations if the parent PCE is able to indicate
   to the child that it really is capable of acting as a parent PCE.
   This will help to trap misconfigurations.

   In PCEP, such capabilities are carried in the Open Object within the
   Open message.

4.8.3 Intention to Utilize Parent PCE Capabilities

   A PCE that is capable of acting as a parent PCE might not be
   configured or willing to act as the parent for a specific child PCE.
   This fact could be determined when the child sends a PCReq that
   requires parental activity (such as querying other child PCEs), and
   could result in a negative response in a PCEP Error (PCErr) message.

   However, the expense of a poorly targeted PCReq can be avoided if
   the child PCE indicates that it might wish to use the parent-capable
   as a parent (for example, on the Open message), and if the 
   parent-capable determines at that time whether it is willing to act
   as a parent to this child.

4.8.4 Communication of Domain Connectivity Information

   Section 4.4 describes how the parent PCE needs a parent TED and
   indicates that the information might be supplied from the child PCEs
   in each domain. This requires a mechanism whereby information about
   inter-domain links can be supplied by a child PCE to a parent PCE,
   for example on a PCEP Notify (PCNtf) message.

   The information that would be exchanged includes:

   o Identifier of advertising child PCE
   o Identifier of PCE's domain
   o Identifier of the link
   o TE properties of the link (metrics, bandwidth)
   o Other properties of the link (technology-specific)
   o Identifier of link end-points
   o Identifier of adjacent domain

   It may be desirable for this information to be periodically updated,
   for example, when available bandwidth changes. In this case, the
   parent PCE might be given the ability to configure thresholds in the
   child PCE to prevent flapping of information.

King and Farrel                                               [Page 20]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

4.8.5 Domain Identifiers

   Domain identifiers are already present in PCEP to allow a PCE to
   indicate which domains it serves, and to allow the representation of
   domains as abstract nodes in paths. The wider use of domains in the
   context of this work on hierarchical PCE will require that domains
   can be identified in more places within objects in PCEP messages.
   This should pose no problems.

   However, more attention may need to be applied to the precision of
   domain identifier definitions to ensure that it is always possible to
   unambiguously identify a domain from its identifier. This work will
   be necessary in configuration, and also in protocol specifications
   (for example, an OSPF area identifier is sufficient within an
   Autonomous System, but becomes ambiguous in a path that crosses
   multiple Autonomous Systems).

5. Hierarchical PCE Applicability

   As per [RFC4655], PCE can inherently support inter-domain path
   computation for any definition of a domain as set out in Section 1.2
   of this document.

   Hierarchical PCE can be applied to inter-domain environments,
   including autonomous Systems and IGP areas. The hierarchical PCE
   procedures make no distinction between, autonomous Systems and IGP
   area applications, although it should be noted that the TED
   maintained by a parent PCE must be able to support the concept of
   child domains connected by inter-domain links or directly connected
   at boundary nodes (see Section 3).

   This section sets out the applicability of hierarchical PCE to three

   o MPLS traffic engineering across multiple Autonomous Systems
   o MPLS traffic engineering across multiple IGP areas
   o GMPLS traffic engineering in the ASON architecture

5.1 autonomous Systems and Areas

   Networks are comprised of domains. A domain can be considered to be
   a collection of network elements within an AS or area that has a
   common sphere of address management or path computational

   As networks increase in size and complexity it may be required to
   introduce scaling methods to reduce the amount information flooded
   within the network and make the network more manageable. An IGP

King and Farrel                                               [Page 21]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   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 a router's
   visibility to information about links and other routers within the
   single area. If a router needs to compute a route to destination
   located in another area, a method is required to compute a path
   across the area boundary.

   When an LSR within an AS or area needs to compute a path across an
   area or AS boundary it must also use an inter-AS computation
   technique. Hierarchical PCE is equally applicable to computing
   inter-area and inter-AS MPLS and GMPLS paths across domain

5.2 ASON Architecture

   The International Telecommunications 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.
   Note that the RA hierarchy can be recursive.

   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 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. For an end-to-end connection, the route may be computed
   by a single RC or multiple RCs in a collaborative manner (i.e., RC
   federations). In the case of RC federations, [G-7715-2] describes
   three styles during remote route query operation:

   o Step-by-step remote path computation
   o Hierarchical remote path computation
   o A combination of the above.

   In a hierarchical ASON routing environment, a child RC may
   communicate with its parent RC (at the next higher level of the ASON
   routing hierarchy) to request the computation of an end-to-end path
   across several RAs. It does this using a route query message (known
   as the abstract message RI_QUERY). The corresponding parent RC may
   communicate with other child RCs that belong to other child RAs at
   the next lower hierarchical level. Thus, a parent RC can act as
   either a Route Query Requester or Route Query Responder.

King and Farrel                                               [Page 22]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   It can be seen that the hierarchical PCE architecture fits the
   hierarchical ASON routing architecture well. It can be used to
   provide paths across subnetworks, and to determine end-to-end paths
   in networks constructed from multiple subnetworks or RAs.

   When hierarchical PCE is applied to implement hierarchical remote
   path computation in [G-7715-2], it is very important for operators to
   understand the different terminology and implicit consistency
   between hierarchical PCE and [G-7715-2].

5.2.1 Implicit Consistency Between Hierarchical PCE and G.7715.2

   This section highlights the correspondence between features of the
   hierarchical PCE architecture and the ASON routing architecture.

   (1) RC (Routing Controller) and PCE (Path Computation Element)

   [G-8080] describes the Routing Controller component as an
   abstract entity, which is responsible for responding to requests
   for path (route) information and topology information. It can be
   implemented as a single entity, or as a distributed set of
   entities that make up a cooperative federation.

   [RFC4655] describes PCE (Path Computation Element) is an entity
   (component, application, or network node) that is capable of
   computing a network path or route based on a network graph and
   applying computational constraints.

   Therefore, in the ASON architecture, a PCE can be regarded as a
   realizations of the RC.

   (2) Route Query Requester/Route Query Responder and PCC/PCE

   [G-7715-2] describes the Route Query Requester as a Connection
   Controller or Routing Controller that sends a route query message
   to a Routing Controller requesting one or more paths that
   satisfy a set of routing constraints. The Route Query Responder
   is a Routing Controller that performs path computation upon
   receipt of a route query message from a Route Query Requester,
   sending a response back at the end of the path computation.

   In the context of ASON, a Signaling Controller initiates and
   processes signaling messages and is closely coupled to a
   Signaling Protocol Speaker. A Routing Controller makes routing
   decisions and is usually coupled to configuration entities
   and/or a Routing Protocol Speaker.

   It can be seen that a PCC corresponds to a Route Query Requester,
   and a PCE corresponds to a Route Query Responder. A PCE/RC can

King and Farrel                                               [Page 23]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   also act as a Route Query Requester sending requests to another
   Route Query Responder.

   The PCEP path computation request (PCReq) and path computation
   reply (PCRep) messages between PCC and PCE correspond to the
   RI_QUERY and RI_UPDATE messages in [G-7715-2].

   (3) Routing Area Hierarchy and Hierarchical Domain

   The ASON routing hierarchy model is shown in Figure 6 of
   [G-7715] through an example that illustrates routing area levels.
   If the hierarchical remote path computation mechanism of
   [G-7715-2] is applied in this scenario, each routing area should
   have at least one RC for route query function and there is a
   parent RC for the child RCs in each routing area.

   According to [G-8080], the parent RC has visibility of the
   structure of the lower level, so it knows the interconnectivity
   of the RAs in the lower level. Each child RC can compute edge-to-
   edge paths across its own child RA.

   Thus, an RA corresponds to a domain in the PCE architecture, and
   the hierarchical relationship between RAs corresponds to the
   hierarchical relationship between domains in the hierarchical PCE
   architecture. Furthermore, a parent PCE in a parent domain can be
   regarded as parent RC in a higher routing level, and a child PCE
   in a child domain can be regarded as child RC in a lower routing

5.2.2 Benefits of Hierarchical PCEs in ASON

   RCs in an ASON environment can use the hierarchical PCE model to
   fully match the ASON hierarchical routing model, so the hierarchical
   PCE mechanisms can be applied to fully satisfy the architecture and
   requirements of [G-7715-2] without any changes. If the hierarchical
   PCE mechanism is applied in ASON, it can be used to determine end-to-
   end optimized paths across sub-networks and RAs before initiating
   signaling to create the connection. It can also improve the
   efficiency of connection setup to avoid crankback.

6. A Note on BGP-TE

   The concept of exchange of TE information between Autonomous Systems
   (ASes) is discussed in [BGP-TE]. 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.

King and Farrel                                               [Page 24]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   There are a number of discussion points associated with the use of
   [BGP-TE] concerning the volume of information, the rate of churn of
   information, the confidentiality of information, the accuracy of
   aggregated or potential-connectivity information, and the processing
   required to generate aggregated information. The PCE architecture and
   the architecture enabled by [BGP-TE] make different assumptions about
   the operational objectives of the networks, and this document does
   not attempt to make one of the approaches "right" and the other
   "wrong". Instead, this work assumes that a decision has been made to
   utilize the PCE architecture.

6.1 Use of BGP for TED Synchronization
   Indeed, [BGP-TE] may have some uses within the PCE model. For
   example, [BGP-TE] could be used as a "northbound" TE advertisement
   such that a PCE does not need to listen to an IGP in its domain, but
   has its TED populated by messages received (for example) from a
   Route Reflector. Furthermore, the inter-domain connectivity and
   connectivity capabilities that is required information for a parent
   PCE could be obtained as a filtered subset of the information
   available in [BGP-TE]. This scenario is discussed further in 

7. Management Considerations

   General PCE management considerations are discussed in [RFC4655]. In
   the case of the hierarchical PCE architecture, there are additional
   management considerations.

   The administrative entity responsible for the management of the
   parent PCEs must be determined. In the case of multi-domains (e.g.,
   IGP areas or multiple ASes) within a single service provider
   network, the management responsibility for the parent PCE would most
   likely be handled by the service provider. In the case of multiple
   ASes within different service provider networks, it may be necessary
   for a third-party to manage the parent PCEs according to commercial
   and policy agreements from each of the participating service

7.1 Control of Function and Policy

7.1.1 Child PCE

   Support of the hierarchical procedure will be controlled by the
   management organization responsible for each child PCE. A child PCE
   must be configured with the address of its parent PCE in order for
   it to interact with its parent PCE. The child PCE must also be
   authorized to peer with the parent PCE.

King and Farrel                                               [Page 25]

ddraft-ietf-pce-hierarchy-fwk-05.txt                        August 2012

7.1.2 Parent PCE

   The parent PCE must only accept path computation requests from
   authorized child PCEs. If a parent PCE receives requests from an
   unauthorized child PCE, the request should be dropped.

   This means that a parent PCE must be configured with the identities
   and security credentials of all of its child PCEs, or there must be
   some form of shared secret that allows an unknown child PCE to be
   authorized by the parent PCE.

7.1.3 Policy Control

   It may be necessary to maintain a policy module on the parent PCE
   [RFC5394]. This would allow the parent PCE to apply commercially
   relevant constraints such as SLAs, security, peering preferences, and
   monetary costs.

   It may also be necessary for the parent PCE to limit end-to-end path
   selection by including or excluding specific domains based on
   commercial relationships, security implications, and reliability.

7.2 Information and Data Models

   A PCEP MIB module is defined in [PCEP-MIB] that describes managed
   objects for modeling of PCEP communication. An additional PCEP MIB
   will be required to report parent PCE and child PCE information,

   o Parent PCE configuration and status,

   o Child PCE configuration and information,

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

   o Notification of parent PCE TED updates and changes.

7.3 Liveness Detection and Monitoring

   The hierarchical procedure requires interaction with multiple PCEs.
   Once a child PCE requests an end-to-end path, a sequence of events
   occurs that requires interaction between the parent PCE and each
   child  PCE. If a child PCE is not operational, and an alternate
   transit domain is not available, then a failure must be reported.

7.4 Verifying Correct Operation

   Verifying the correct operation of a parent PCE can be performed by

King and Farrel                                               [Page 26]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   monitoring a set of parameters. The parent PCE implementation should
   provide the following parameters monitored by the parent PCE:

   o  Number of child PCE requests.

   o  Number of successful hierarchical PCE procedures completions on a
      per-PCE-peer basis.

   o  Number of hierarchical PCE procedure completion failures on a per-
      PCE-peer basis.

   o  Number of hierarchical PCE procedure requests from unauthorized
      child PCEs.

7.5. Impact on Network Operation

   The hierarchical PCE procedure is a multiple-PCE path computation
   scheme. Subsequent requests to and from the child and parent PCEs do
   not differ from other path computation requests and should not have
   any significant impact on network operations.

8. Security Considerations

   The hierarchical PCE procedure relies on PCEP and inherits the
   security requirements defined [RFC5440]. As noted in Section 7,
   there is a security relationship between child and parent PCEs.
   This relationship, like any PCEP relationship assumes
   pre-configuration of identities, authority, and keys, or can
   operate through any key distribution mechanism outside the scope of
   PCEP. As PCEP operates over TCP, it may make use of any TCP security

   The hierarchical PCE architecture makes use of PCE policy
   [RFC5394] and the security aspects of the PCE communication protocol
   documented in [RFC5440]. It is expected that the parent PCE will
   require all child PCEs to use full security when communicating with
   the parent and that security will be maintained by not supporting the
   discovery by a parent of child PCEs.

   PCE operation also relies on information used to build the TED.
   Attacks on a PCE system may be achieved by falsifying or impeding
   this flow of information. The child PCE TEDs are constructed as
   described in [RFC4655] and are unchanged in this document: if the PCE
   listens to the IGP for this information, then normal IGP security
   measures may be applied, and it should be noted that an IGP routing
   system is generally assumed to be a trusted domain such that router
   subversion is not a risk. The parent PCE TED is constructed as
   described in this document and may involve:

King and Farrel                                               [Page 27]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   - multiple parent-child relationships using PCEP (as already

   - the parent PCE listening to child domain IGPs (with the same
     security features as a child PCE listening to its IGP)

   - an external mechanism (such as [BGP-TE]) which will need to be
     authorized and secured.

   Any multi-domain operation necessarily involves the exchange of
   information across domain boundaries. This is bound to represent a
   significant security and confidentiality risk especially when the
   child domains are controlled by different commercial concerns. PCEP
   allows individual PCEs to maintain confidentiality of their domain
   path information using Path Keys [RFC5520], and the hierarchical
   PCE architecture is specifically designed to enable as much isolation
   of domain topology and capabilities information as is possible.

   Further considerations of the security issues related to inter-AS
   path computation see [RFC5376].

9. IANA Considerations

   This document makes no requests for IANA action.

10. Acknowledgements

   The authors would like to thank David Amzallag, Oscar Gonzalez de
   Dios, Franz Rambach, Ramon Casellas, Olivier Dugeon, Filippo Cugini, 
   Dhruv Dhody and Julien Meuric for their comments and suggestions.

11. References

11.1 Normative References

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

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

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

King and Farrel                                               [Page 28]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   [RFC5440]  Ayyangar, A., Farrel, A., Oki, E., Atlas, A., Dolganow,
              A., Ikejiri, Y., Kumaki, K., Vasseur, J., and J. Roux,
              "Path Computation Element (PCE) Communication Protocol
              (PCEP)", RFC 5440, March 2009.

   [RFC5441]  Vasseur, J.P., Ed., "A Backward Recursive PCE-based
              Computation (BRPC) procedure to compute shortest inter-
              domain Traffic Engineering Label Switched Paths", RFC
              5441, April 2009.

   [RFC5520]  Brandford, R., Vasseur J.P., and Farrel A., "Preserving
              Topology Confidentiality in Inter-Domain Path
              Computation Using a Key-Based Mechanism
              RFC5520, April 2009.

11.2. Informative References

   [RFC4105]  Le Roux, JL., Vasseur, J., Boyle, J.,
              "Requirements for Inter-Area MPLS Traffic Engineering",
              RFC 4105, June 2005.

   [RFC4216]  Zhang, R., and Vasseur, J., "MPLS Inter-Autonomous
              System (AS) Traffic Engineering (TE) Requirements", RFC
              4216, November 2005.

   [RFC4726]  Farrel, A., Vasseur, J., Ayyangar, A., "A Framework
              for Inter-Domain Multiprotocol Label Switching Traffic
              Engineering", RFC 4726, November 2006.

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

   [RFC5316]  Chen, M., Zhang, R., Duan, X., "ISIS Extensions in
              Support of Inter-Autonomous System (AS) MPLS and GMPLS
              Traffic Engineering", RFC 5316, December 2008.

   [RFC5376]  Bitar, N., et al., "Inter-AS Requirements for the
              Path Computation Element Communication Protocol
              (PCECP)", RFC 5376, November 2008.

   [RFC5392]  Chen, M., Zhang, R., Duan, X., "OSPF Extensions in
              Support of Inter-Autonomous System (AS) MPLS and GMPLS
              Traffic Engineering", RFC 5392, January 2009.

   [RFC5541]  Le Roux, J., Vasseur, J., Lee, Y., "Encoding
              of Objective Functions in the Path Computation Element
              Communication Protocol (PCEP)", RFC5541, December 2008.

King and Farrel                                               [Page 29]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   [G-8080]   ITU-T Recommendation G.8080/Y.1304, Architecture for
              the automatically switched optical network (ASON).

   [G-7715]   ITU-T Recommendation G.7715 (2002), Architecture
              and Requirements for the Automatically
              Switched Optical Network (ASON).

   [G-7715-2] ITU-T Recommendation G.7715.2 (2007), ASON
              routing architecture and requirements for remote route

   [BGP-TE]   Gredler, H., Medved, J, Farrel, A. Previdi, S., 
              "North-Bound Distribution of Link-State and TE 
              Information using BGP", draft-gredler-idr-ls-distribution,
              work in progress.
   [PCE-AREA-AS] King, D., Meuric, J., Dugeon, O., Zhao, Q., Gonzalez de       
              Dios, O., "Applicability of the Path Computation Element
              to Inter-Area and Inter-AS MPLS and GMPLS Traffic 
              Engineering", draft-ietf-pce-inter-area-as-applicability,
              work in progress.   

   [PCEP-MIB] Stephan, E., Koushik, K., Zhao, Q., King, D., "PCE
              Communication Protocol (PCEP) Management Information
              Base", work in progress.

12. Authors' Addresses

   Daniel King
   Old Dog Consulting

   Adrian Farrel
   Old Dog Consulting

   Quintin Zhao
   Huawei Technology
   125 Nagog Technology Park
   Acton, MA  01719

King and Farrel                                               [Page 30]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012

   Fatai Zhang
   Huawei Technologies
   F3-5-B R&D Center, Huawei Base
   Bantian, Longgang District
   Shenzhen 518129 P.R.China

King and Farrel                                               [Page 31]

draft-ietf-pce-hierarchy-fwk-05.txt                         August 2012