Debian Codegraph @

rfc6805.html

raw
1 <pre>Internet Engineering Task Force (IETF) D. King, Ed.
2 Request for Comments: 6805 A. Farrel, Ed.
3 Category: Informational Old Dog Consulting
4 ISSN: 2070-1721 November 2012
5
6
7 <span class="h1">The Application of the Path Computation Element Architecture to the</span>
8 <span class="h1">Determination of a Sequence of Domains in MPLS and GMPLS</span>
9
10 Abstract
11
12 Computing optimum routes for Label Switched Paths (LSPs) across
13 multiple domains in MPLS Traffic Engineering (MPLS-TE) and GMPLS
14 networks presents a problem because no single point of path
15 computation is aware of all of the links and resources in each
16 domain. A solution may be achieved using the Path Computation
17 Element (PCE) architecture.
18
19 Where the sequence of domains is known a priori, various techniques
20 can be employed to derive an optimum path. If the domains are simply
21 connected, or if the preferred points of interconnection are also
22 known, the Per-Domain Path Computation technique can be used. Where
23 there are multiple connections between domains and there is no
24 preference for the choice of points of interconnection, the Backward-
25 Recursive PCE-based Computation (BRPC) procedure can be used to
26 derive an optimal path.
27
28 This document examines techniques to establish the optimum path when
29 the sequence of domains is not known in advance. The document shows
30 how the PCE architecture can be extended to allow the optimum
31 sequence of domains to be selected, and the optimum end-to-end path
32 to be derived through the use of a hierarchical relationship between
33 domains.
34
35 Status of This Memo
36
37 This document is not an Internet Standards Track specification; it is
38 published for informational purposes.
39
40 This document is a product of the Internet Engineering Task Force
41 (IETF). It represents the consensus of the IETF community. It has
42 received public review and has been approved for publication by the
43 Internet Engineering Steering Group (IESG). Not all documents
44 approved by the IESG are a candidate for any level of Internet
45 Standard; see <a href="./rfc5741#section-2">Section&nbsp;2 of RFC 5741</a>.
46
47
48
49
50
51
52 <span class="grey">King &amp; Farrel Informational [Page 1]</span></pre>
53 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-2" ></span>
54 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
55
56
57 Information about the current status of this document, any errata,
58 and how to provide feedback on it may be obtained at
59 <a href="http://www.rfc-editor.org/info/rfc6805">http://www.rfc-editor.org/info/rfc6805</a>.
60
61 Copyright Notice
62
63 Copyright (c) 2012 IETF Trust and the persons identified as the
64 document authors. All rights reserved.
65
66 This document is subject to <a href="https://www.rfc-editor.org/bcp/bcp78">BCP 78</a> and the IETF Trust's Legal
67 Provisions Relating to IETF Documents
68 (<a href="http://trustee.ietf.org/license-info">http://trustee.ietf.org/license-info</a>) in effect on the date of
69 publication of this document. Please review these documents
70 carefully, as they describe your rights and restrictions with respect
71 to this document. Code Components extracted from this document must
72 include Simplified BSD License text as described in Section 4.e of
73 the Trust Legal Provisions and are provided without warranty as
74 described in the Simplified BSD License.
75
76 Table of Contents
77
78 <a href="#section-1">1</a>. Introduction ....................................................<a href="#page-4">4</a>
79 <a href="#section-1.1">1.1</a>. Problem Statement ..........................................<a href="#page-5">5</a>
80 <a href="#section-1.2">1.2</a>. Definition of a Domain .....................................<a href="#page-5">5</a>
81 <a href="#section-1.3">1.3</a>. Assumptions and Requirements ...............................<a href="#page-6">6</a>
82 <a href="#section-1.3.1">1.3.1</a>. Metric Objectives ...................................<a href="#page-6">6</a>
83 <a href="#section-1.3.2">1.3.2</a>. Diversity ...........................................<a href="#page-7">7</a>
84 <a href="#section-1.3.2.1">1.3.2.1</a>. Physical Diversity .........................<a href="#page-7">7</a>
85 <a href="#section-1.3.2.2">1.3.2.2</a>. Domain Diversity ...........................<a href="#page-7">7</a>
86 <a href="#section-1.3.3">1.3.3</a>. Existing Traffic Engineering Constraints ............<a href="#page-7">7</a>
87 <a href="#section-1.3.4">1.3.4</a>. Commercial Constraints ..............................<a href="#page-8">8</a>
88 <a href="#section-1.3.5">1.3.5</a>. Domain Confidentiality ..............................<a href="#page-8">8</a>
89 <a href="#section-1.3.6">1.3.6</a>. Limiting Information Aggregation ....................<a href="#page-8">8</a>
90 <a href="#section-1.3.7">1.3.7</a>. Domain Interconnection Discovery ....................<a href="#page-8">8</a>
91 <a href="#section-1.4">1.4</a>. Terminology ................................................<a href="#page-8">8</a>
92 <a href="#section-2">2</a>. Examination of Existing PCE Mechanisms ..........................<a href="#page-9">9</a>
93 <a href="#section-2.1">2.1</a>. Per-Domain Path Computation ................................<a href="#page-9">9</a>
94 <a href="#section-2.2">2.2</a>. Backward-Recursive PCE-Based Computation ..................<a href="#page-10">10</a>
95 2.2.1. Applicability of BRPC When the Domain Path
96 is Not Known .......................................<a href="#page-11">11</a>
97 <a href="#section-3">3</a>. Hierarchical PCE ...............................................<a href="#page-12">12</a>
98 <a href="#section-4">4</a>. Hierarchical PCE Procedures ....................................<a href="#page-13">13</a>
99 <a href="#section-4.1">4.1</a>. Objective Functions and Policy ............................<a href="#page-13">13</a>
100 <a href="#section-4.2">4.2</a>. Maintaining Domain Confidentiality ........................<a href="#page-14">14</a>
101 <a href="#section-4.3">4.3</a>. PCE Discovery .............................................<a href="#page-14">14</a>
102 <a href="#section-4.4">4.4</a>. Traffic Engineering Database for the Parent Domain ........<a href="#page-15">15</a>
103 <a href="#section-4.5">4.5</a>. Determination of Destination Domain .......................<a href="#page-16">16</a>
104 <a href="#section-4.6">4.6</a>. Hierarchical PCE Examples .................................<a href="#page-16">16</a>
105
106
107
108 <span class="grey">King &amp; Farrel Informational [Page 2]</span></pre>
109 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-3" ></span>
110 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
111
112
113 <a href="#section-4.6.1">4.6.1</a>. Hierarchical PCE Initial Information Exchange ......<a href="#page-18">18</a>
114 4.6.2. Hierarchical PCE End-to-End Path
115 Computation Procedure ..............................<a href="#page-19">19</a>
116 <a href="#section-4.7">4.7</a>. Hierarchical PCE Error Handling ...........................<a href="#page-20">20</a>
117 <a href="#section-4.8">4.8</a>. Requirements for Hierarchical PCEP Protocol Extensions ....<a href="#page-20">20</a>
118 <a href="#section-4.8.1">4.8.1</a>. PCEP Request Qualifiers ............................<a href="#page-21">21</a>
119 <a href="#section-4.8.2">4.8.2</a>. Indication of Hierarchical PCE Capability ..........<a href="#page-21">21</a>
120 <a href="#section-4.8.3">4.8.3</a>. Intention to Utilize Parent PCE Capabilities .......<a href="#page-21">21</a>
121 <a href="#section-4.8.4">4.8.4</a>. Communication of Domain Connectivity Information ...<a href="#page-22">22</a>
122 <a href="#section-4.8.5">4.8.5</a>. Domain Identifiers .................................<a href="#page-22">22</a>
123 <a href="#section-5">5</a>. Hierarchical PCE Applicability .................................<a href="#page-23">23</a>
124 <a href="#section-5.1">5.1</a>. Autonomous Systems and Areas ..............................<a href="#page-23">23</a>
125 <a href="#section-5.2">5.2</a>. ASON Architecture .........................................<a href="#page-24">24</a>
126 5.2.1. Implicit Consistency between Hierarchical
127 PCE and G.7715.2 ...................................<a href="#page-25">25</a>
128 <a href="#section-5.2.2">5.2.2</a>. Benefits of Hierarchical PCEs in ASON ..............<a href="#page-26">26</a>
129 <a href="#section-6">6</a>. A Note on BGP-TE ...............................................<a href="#page-26">26</a>
130 <a href="#section-6.1">6.1</a>. Use of BGP for TED Synchronization ........................<a href="#page-27">27</a>
131 <a href="#section-7">7</a>. Management Considerations ......................................<a href="#page-27">27</a>
132 <a href="#section-7.1">7.1</a>. Control of Function and Policy ............................<a href="#page-27">27</a>
133 <a href="#section-7.1.1">7.1.1</a>. Child PCE ..........................................<a href="#page-27">27</a>
134 <a href="#section-7.1.2">7.1.2</a>. Parent PCE .........................................<a href="#page-27">27</a>
135 <a href="#section-7.1.3">7.1.3</a>. Policy Control .....................................<a href="#page-28">28</a>
136 <a href="#section-7.2">7.2</a>. Information and Data Models ...............................<a href="#page-28">28</a>
137 <a href="#section-7.3">7.3</a>. Liveness Detection and Monitoring .........................<a href="#page-28">28</a>
138 <a href="#section-7.4">7.4</a>. Verifying Correct Operation ...............................<a href="#page-28">28</a>
139 <a href="#section-7.5">7.5</a>. Impact on Network Operation ...............................<a href="#page-29">29</a>
140 <a href="#section-8">8</a>. Security Considerations ........................................<a href="#page-29">29</a>
141 <a href="#section-9">9</a>. Acknowledgements ...............................................<a href="#page-30">30</a>
142 <a href="#section-10">10</a>. References ....................................................<a href="#page-30">30</a>
143 <a href="#section-10.1">10.1</a>. Normative References .....................................<a href="#page-30">30</a>
144 <a href="#section-10.2">10.2</a>. Informative References ...................................<a href="#page-31">31</a>
145 <a href="#section-11">11</a>. Contributors ..................................................<a href="#page-32">32</a>
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164 <span class="grey">King &amp; Farrel Informational [Page 3]</span></pre>
165 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-4" ></span>
166 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
167
168
169 <span class="h2"><a class="selflink" id="section-1" href="#section-1">1</a>. Introduction</span>
170
171 The capability to compute the routes of end-to-end inter-domain MPLS
172 Traffic Engineering (MPLS-TE) and GMPLS Label Switched Paths (LSPs)
173 is expressed as requirements in [<a href="./rfc4105" title="&quot;Requirements for Inter-Area MPLS Traffic Engineering&quot;">RFC4105</a>] and [<a href="./rfc4216" title="&quot;MPLS Inter- Autonomous System (AS) Traffic Engineering (TE) Requirements&quot;">RFC4216</a>]. This
174 capability may be realized by a Path Computation Element (PCE). The
175 PCE architecture is defined in [<a href="./rfc4655" title="&quot;A Path Computation Element (PCE)-Based Architecture&quot;">RFC4655</a>]. The methods for
176 establishing and controlling inter-domain MPLS-TE and GMPLS LSPs are
177 documented in [<a href="./rfc4726" title="&quot;A Framework for Inter-Domain Multiprotocol Label Switching Traffic Engineering&quot;">RFC4726</a>].
178
179 In this context, a domain can be defined as a separate
180 administrative, geographic, or switching environment within the
181 network. A domain may be further defined as a zone of routing or
182 computational ability. Under these definitions, a domain might be
183 categorized as an Autonomous System (AS) or an Interior Gateway
184 Protocol (IGP) area [<a href="./rfc4726" title="&quot;A Framework for Inter-Domain Multiprotocol Label Switching Traffic Engineering&quot;">RFC4726</a>] [<a href="./rfc4655" title="&quot;A Path Computation Element (PCE)-Based Architecture&quot;">RFC4655</a>]. Domains are connected
185 through ingress and egress boundary nodes (BNs). A more detailed
186 definition is given in <a href="#section-1.2">Section 1.2</a>.
187
188 In a multi-domain environment, the determination of an end-to-end
189 traffic engineered path is a problem because no single point of path
190 computation is aware of all of the links and resources in each
191 domain. PCEs can be used to compute end-to-end paths using a per-
192 domain path computation technique [<a href="./rfc5152" title="&quot;A Per-Domain Path Computation Method for Establishing Inter-Domain Traffic Engineering (TE) Label Switched Paths (LSPs)&quot;">RFC5152</a>]. Alternatively, the
193 Backward-Recursive PCE-based Computation (BRPC) mechanism [<a href="./rfc5441" title="&quot;A Backward-Recursive PCE-Based Computation (BRPC) Procedure to Compute Shortest Constrained Inter- Domain Traffic Engineering Label Switched Paths&quot;">RFC5441</a>]
194 allows multiple PCEs to collaborate in order to select an optimal
195 end-to-end path that crosses multiple domains. Both mechanisms
196 assume that the sequence of domains to be crossed between ingress and
197 egress is known in advance.
198
199 This document examines techniques to establish the optimum path when
200 the sequence of domains is not known in advance. It shows how the
201 PCE architecture can be extended to allow the optimum sequence of
202 domains to be selected, and the optimum end-to-end path to be
203 derived.
204
205 The model described in this document introduces a hierarchical
206 relationship between domains. It is applicable to environments with
207 small groups of domains where visibility from the ingress Label
208 Switching Router (LSR) is limited. Applying the hierarchical PCE
209 model to large groups of domains such as the Internet, is not
210 considered feasible or desirable, and is out of scope for this
211 document.
212
213
214
215
216
217
218
219
220 <span class="grey">King &amp; Farrel Informational [Page 4]</span></pre>
221 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-5" ></span>
222 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
223
224
225 This document does not specify any protocol extensions or
226 enhancements. That work is left for future protocol specification
227 documents. However, some assumptions are made about which protocols
228 will be used to provide specific functions, and guidelines to future
229 protocol developers are made in the form of requirements statements.
230
231 <span class="h3"><a class="selflink" id="section-1.1" href="#section-1.1">1.1</a>. Problem Statement</span>
232
233 Using a PCE to compute a path between nodes within a single domain is
234 relatively straightforward. Computing an end-to-end path when the
235 source and destination nodes are located in different domains
236 requires co-operation between multiple PCEs, each responsible for its
237 own domain.
238
239 Techniques for inter-domain path computation described so far
240 ([<a href="./rfc5152" title="&quot;A Per-Domain Path Computation Method for Establishing Inter-Domain Traffic Engineering (TE) Label Switched Paths (LSPs)&quot;">RFC5152</a>] and [<a href="./rfc5441" title="&quot;A Backward-Recursive PCE-Based Computation (BRPC) Procedure to Compute Shortest Constrained Inter- Domain Traffic Engineering Label Switched Paths&quot;">RFC5441</a>]) assume that the sequence of domains to be
241 crossed from source to destination is well known. No explanation is
242 given (for example, in [<a href="./rfc4655" title="&quot;A Path Computation Element (PCE)-Based Architecture&quot;">RFC4655</a>]) of how this sequence is generated
243 or what criteria may be used for the selection of paths between
244 domains. In small clusters of domains, such as simple cooperation
245 between adjacent ISPs, this selection process is not complex. In
246 more advanced deployments (such as optical networks constructed from
247 multiple sub-domains, or in multi-AS environments), the choice of
248 domains in the end-to-end domain sequence can be critical to the
249 determination of an optimum end-to-end path.
250
251 <span class="h3"><a class="selflink" id="section-1.2" href="#section-1.2">1.2</a>. Definition of a Domain</span>
252
253 A domain is defined in [<a href="./rfc4726" title="&quot;A Framework for Inter-Domain Multiprotocol Label Switching Traffic Engineering&quot;">RFC4726</a>] as any collection of network
254 elements within a common sphere of address management or path
255 computational responsibility. Examples of such domains include IGP
256 areas and Autonomous Systems. Wholly or partially overlapping
257 domains are not within the scope of this document.
258
259 In the context of GMPLS, a particularly important example of a domain
260 is the Automatically Switched Optical Network (ASON) subnetwork
261 [<a href="#ref-G-8080">G-8080</a>]. In this case, a domain might be an ASON Routing Area
262 [<a href="#ref-G-7715">G-7715</a>]. Furthermore, computation of an end-to-end path requires
263 the selection of nodes and links within a routing area where some
264 nodes may, in fact, be subnetworks. A PCE may perform the path
265 computation function of an ASON Routing Controller as described in
266 [<a href="#ref-G-7715-2">G-7715-2</a>]. See <a href="#section-5.2">Section 5.2</a> for a further discussion of the
267 applicability to the ASON architecture.
268
269 This document assumes that the selection of a sequence of domains for
270 an end-to-end path is in some sense a hierarchical path computation
271 problem. That is, where one mechanism is used to determine a path
272 across a domain, a separate mechanism (or at least a separate set of
273
274
275
276 <span class="grey">King &amp; Farrel Informational [Page 5]</span></pre>
277 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-6" ></span>
278 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
279
280
281 paradigms) is used to determine the sequence of domains. The
282 responsibility for the selection of domain interconnection can belong
283 to either or both of the mechanisms.
284
285 <span class="h3"><a class="selflink" id="section-1.3" href="#section-1.3">1.3</a>. Assumptions and Requirements</span>
286
287 Networks are often constructed from multiple domains. These domains
288 are often interconnected via multiple interconnect points. It's
289 assumed that the sequence of domains for an end-to-end path is not
290 always well known; that is, an application requesting end-to-end
291 connectivity has no preference for, or no ability to specify, the
292 sequence of domains to be crossed by the path.
293
294 The traffic engineering properties of a domain cannot be seen from
295 outside the domain. Traffic engineering aggregation or abstraction,
296 hides information and can lead to failed path setup or the selection
297 of suboptimal end-to-end paths [<a href="./rfc4726" title="&quot;A Framework for Inter-Domain Multiprotocol Label Switching Traffic Engineering&quot;">RFC4726</a>]. The aggregation process
298 may also have significant scaling issues for networks with many
299 possible routes and multiple TE metrics. Flooding TE information
300 breaks confidentiality and does not scale in the routing protocol.
301 See <a href="#section-6">Section 6</a> for a discussion of the concept of inter-domain traffic
302 engineering information exchange known as BGP-TE.
303
304 The primary goal of this document is to define how to derive optimal
305 end-to-end, multi-domain paths when the sequence of domains is not
306 known in advance. The solution needs to be scalable and to maintain
307 internal domain topology confidentiality while providing the optimal
308 end-to-end path. It cannot rely on the exchange of TE information
309 between domains, and for the confidentiality, scaling, and
310 aggregation reasons just cited, it cannot utilize a computation
311 element that has universal knowledge of TE properties and topology of
312 all domains.
313
314 The sub-sections that follow set out the primary objectives and
315 requirements to be satisfied by a PCE solution to multi-domain path
316 computation.
317
318 <span class="h4"><a class="selflink" id="section-1.3.1" href="#section-1.3.1">1.3.1</a>. Metric Objectives</span>
319
320 The definition of optimality is dependent on policy and is based on a
321 single objective or a group of objectives. An objective is expressed
322 as an objective function [<a href="./rfc5541" title="&quot;Encoding of Objective Functions in the Path Computation Element Communication Protocol (PCEP)&quot;">RFC5541</a>] and may be specified on a path
323 computation request. The following objective functions are
324 identified in this document. They define how the path metrics and TE
325 link qualities are manipulated during inter-domain path computation.
326 The list is not proscriptive and may be expanded in other documents.
327
328
329
330
331
332 <span class="grey">King &amp; Farrel Informational [Page 6]</span></pre>
333 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-7" ></span>
334 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
335
336
337 o Minimize the cost of the path [<a href="./rfc5541" title="&quot;Encoding of Objective Functions in the Path Computation Element Communication Protocol (PCEP)&quot;">RFC5541</a>].
338 o Select a path using links with the minimal load [<a href="./rfc5541" title="&quot;Encoding of Objective Functions in the Path Computation Element Communication Protocol (PCEP)&quot;">RFC5541</a>].
339 o Select a path that leaves the maximum residual bandwidth
340 [<a href="./rfc5541" title="&quot;Encoding of Objective Functions in the Path Computation Element Communication Protocol (PCEP)&quot;">RFC5541</a>].
341 o Minimize aggregate bandwidth consumption [<a href="./rfc5541" title="&quot;Encoding of Objective Functions in the Path Computation Element Communication Protocol (PCEP)&quot;">RFC5541</a>].
342 o Minimize the load of the most loaded link [<a href="./rfc5541" title="&quot;Encoding of Objective Functions in the Path Computation Element Communication Protocol (PCEP)&quot;">RFC5541</a>].
343 o Minimize the cumulative cost of a set of paths [<a href="./rfc5541" title="&quot;Encoding of Objective Functions in the Path Computation Element Communication Protocol (PCEP)&quot;">RFC5541</a>].
344 o Minimize or cap the number of domains crossed.
345 o Disallow domain re-entry.
346
347 See <a href="#section-4.1">Section 4.1</a> for further discussion of objective functions.
348
349 <span class="h4"><a class="selflink" id="section-1.3.2" href="#section-1.3.2">1.3.2</a>. Diversity</span>
350
351 <span class="h5"><a class="selflink" id="section-1.3.2.1" href="#section-1.3.2.1">1.3.2.1</a>. Physical Diversity</span>
352
353 Within a "Carrier's Carrier" environment, MPLS services may share
354 common underlying equipment and resources, including optical fiber
355 and nodes. An MPLS service request may require a path for traffic
356 that is physically disjointed from another service. Thus, if a
357 physical link or node fails on one of the disjoint paths, not all
358 traffic is lost.
359
360 <span class="h5"><a class="selflink" id="section-1.3.2.2" href="#section-1.3.2.2">1.3.2.2</a>. Domain Diversity</span>
361
362 A pair of paths are domain-diverse if they do not transit any of the
363 same domains. A pair of paths that share a common ingress and egress
364 are domain-diverse if they only share the same domains at the ingress
365 and egress (the ingress and egress domains). Domain diversity may be
366 maximized for a pair of paths by selecting paths that have the
367 smallest number of shared domains. (Note that this is not the same
368 as finding paths with the greatest number of distinct domains!)
369
370 Path computation should facilitate the selection of paths that share
371 ingress and egress domains but do not share any transit domains.
372 This provides a way to reduce the risk of shared failure along any
373 path and automatically helps to ensure path diversity for most of the
374 route of a pair of LSPs.
375
376 Thus, domain path selection should provide the capability to include
377 or exclude specific domains and specific boundary nodes.
378
379 <span class="h4"><a class="selflink" id="section-1.3.3" href="#section-1.3.3">1.3.3</a>. Existing Traffic Engineering Constraints</span>
380
381 Any solution should take advantage of typical traffic engineering
382 constraints (hop count, bandwidth, lambda continuity, path cost,
383 etc.) to meet the service demands expressed in the path computation
384 request [<a href="./rfc4655" title="&quot;A Path Computation Element (PCE)-Based Architecture&quot;">RFC4655</a>].
385
386
387
388 <span class="grey">King &amp; Farrel Informational [Page 7]</span></pre>
389 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-8" ></span>
390 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
391
392
393 <span class="h4"><a class="selflink" id="section-1.3.4" href="#section-1.3.4">1.3.4</a>. Commercial Constraints</span>
394
395 The solution should provide the capability to include commercially
396 relevant constraints such as policy, Service Level Agreements (SLAs),
397 security, peering preferences, and monetary costs.
398
399 Additionally, it may be necessary for the service provider to request
400 that specific domains are included or excluded based on commercial
401 relationships, security implications, and reliability.
402
403 <span class="h4"><a class="selflink" id="section-1.3.5" href="#section-1.3.5">1.3.5</a>. Domain Confidentiality</span>
404
405 A key requirement is the ability to maintain domain confidentiality
406 when computing inter-domain end-to-end paths. It should be possible
407 for local policy to require that a PCE not disclose to any other PCE
408 the intra-domain paths it computes or the internal topology of the
409 domain it serves. This requirement should have no impact in the
410 optimality or quality of the end-to-end path that is derived.
411
412 <span class="h4"><a class="selflink" id="section-1.3.6" href="#section-1.3.6">1.3.6</a>. Limiting Information Aggregation</span>
413
414 In order to reduce processing overhead and to not sacrifice
415 computational detail, there should be no requirement to aggregate or
416 abstract traffic engineering link information.
417
418 <span class="h4"><a class="selflink" id="section-1.3.7" href="#section-1.3.7">1.3.7</a>. Domain Interconnection Discovery</span>
419
420 To support domain mesh topologies, the solution should allow the
421 discovery and selection of domain interconnections. Pre-
422 configuration of preferred domain interconnections should also be
423 supported for network operators that have bilateral agreement and
424 have a preference for the choice of points of interconnection.
425
426 <span class="h3"><a class="selflink" id="section-1.4" href="#section-1.4">1.4</a>. Terminology</span>
427
428 This document uses PCE terminology defined in [<a href="./rfc4655" title="&quot;A Path Computation Element (PCE)-Based Architecture&quot;">RFC4655</a>], [<a href="./rfc4726" title="&quot;A Framework for Inter-Domain Multiprotocol Label Switching Traffic Engineering&quot;">RFC4726</a>],
429 and [<a href="./rfc5440" title="&quot;Path Computation Element (PCE) Communication Protocol (PCEP)&quot;">RFC5440</a>]. Additional terms are defined below.
430
431 Domain Path: The sequence of domains for a path.
432
433 Ingress Domain: The domain that includes the ingress LSR of a path.
434
435 Transit Domain: A domain that has an upstream and downstream neighbor
436 domain for a specific path.
437
438 Egress Domain: The domain that includes the egress LSR of a path.
439
440
441
442
443
444 <span class="grey">King &amp; Farrel Informational [Page 8]</span></pre>
445 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-9" ></span>
446 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
447
448
449 Boundary Nodes: Each Domain has entry LSRs and exit LSRs that could
450 be Area Border Routers (ABRs) or Autonomous System Border Routers
451 (ASBRs) depending on the type of domain. They are defined here more
452 generically as Boundary Nodes (BNs).
453
454 Entry BN of domain(n): a BN connecting domain(n-1) to domain(n) on a
455 path.
456
457 Exit BN of domain(n): a BN connecting domain(n) to domain(n+1) on a
458 path.
459
460 Parent Domain: A domain higher up in a domain hierarchy such that it
461 contains other domains (child domains) and potentially other links
462 and nodes.
463
464 Child Domain: A domain lower in a domain hierarchy such that it has a
465 parent domain.
466
467 Parent PCE: A PCE responsible for selecting a path across a parent
468 domain and any number of child domains by coordinating with child
469 PCEs and examining a topology map that shows domain inter-
470 connectivity.
471
472 Child PCE: A PCE responsible for computing the path across one or
473 more specific (child) domains. A child PCE maintains a relationship
474 with at least one parent PCE.
475
476 Objective Function (OF): A set of one or more optimization criteria
477 used for the computation of a single path (e.g., path cost
478 minimization), or the synchronized computation of a set of paths
479 (e.g., aggregate bandwidth consumption minimization). See [<a href="./rfc4655" title="&quot;A Path Computation Element (PCE)-Based Architecture&quot;">RFC4655</a>]
480 and [<a href="./rfc5541" title="&quot;Encoding of Objective Functions in the Path Computation Element Communication Protocol (PCEP)&quot;">RFC5541</a>].
481
482 <span class="h2"><a class="selflink" id="section-2" href="#section-2">2</a>. Examination of Existing PCE Mechanisms</span>
483
484 This section provides a brief overview of two existing PCE
485 cooperation mechanisms called the Per-Domain Path Computation method
486 and the BRPC method. It describes the applicability of these methods
487 to the multi-domain problem.
488
489 <span class="h3"><a class="selflink" id="section-2.1" href="#section-2.1">2.1</a>. Per-Domain Path Computation</span>
490
491 The Per-Domain Path Computation method for establishing inter-domain
492 TE-LSPs [<a href="./rfc5152" title="&quot;A Per-Domain Path Computation Method for Establishing Inter-Domain Traffic Engineering (TE) Label Switched Paths (LSPs)&quot;">RFC5152</a>] defines a technique whereby the path is computed
493 during the signaling process on a per-domain basis. The entry BN of
494 each domain is responsible for performing the path computation for
495 the section of the LSP that crosses the domain, or for requesting
496 that a PCE for that domain computes that piece of the path.
497
498
499
500 <span class="grey">King &amp; Farrel Informational [Page 9]</span></pre>
501 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-10" ></span>
502 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
503
504
505 During per-domain path computation, each computation results in a
506 path that crosses the domain to provide connectivity to the next
507 domain in the sequence. The chosen path across the domain will be
508 selected as best according to the optimization characteristics of the
509 computation. The next domain in the sequence is usually indicated in
510 signaling by an identifier of the next domain or the identity of the
511 next entry BN.
512
513 Per-domain path computation may lead to suboptimal end-to-end paths
514 because the most optimal path in one domain may lead to the choice of
515 an entry BN for the next domain that results in a very poor path
516 across that next domain.
517
518 In the case that the domain path (in particular, the sequence of
519 boundary nodes) is not known, the path computing entity must select
520 an exit BN based on some determination of how to reach the
521 destination that is outside the domain for which the path computing
522 entity has computational responsibility. [<a href="./rfc5152" title="&quot;A Per-Domain Path Computation Method for Establishing Inter-Domain Traffic Engineering (TE) Label Switched Paths (LSPs)&quot;">RFC5152</a>] suggest that this
523 might be achieved using the IP shortest path as advertised by BGP.
524 Note, however, that the existence of an IP forwarding path does not
525 guarantee the presence of sufficient bandwidth, let alone an optimal
526 TE path. Furthermore, in many GMPLS systems, inter-domain IP routing
527 will not be present. Thus, per-domain path computation may require a
528 significant number of crankback routing attempts to establish even a
529 suboptimal path.
530
531 Note also that the path computing entities in each domain may have
532 different computation capabilities, may run different path
533 computation algorithms, and may apply different sets of constraints
534 and optimization criteria, etc.
535
536 This can result in the end-to-end path being inconsistent and
537 suboptimal.
538
539 Per-domain path computation can suit simply connected domains where
540 the preferred points of interconnection are known.
541
542 <span class="h3"><a class="selflink" id="section-2.2" href="#section-2.2">2.2</a>. Backward-Recursive PCE-Based Computation</span>
543
544 The Backward-Recursive PCE-based Computation (BRPC) [<a href="./rfc5441" title="&quot;A Backward-Recursive PCE-Based Computation (BRPC) Procedure to Compute Shortest Constrained Inter- Domain Traffic Engineering Label Switched Paths&quot;">RFC5441</a>]
545 procedure involves cooperation and communication between PCEs in
546 order to compute an optimal end-to-end path across multiple domains.
547 The sequence of domains to be traversed can be determined either
548 before or during the path computation. In the case where the
549 sequence of domains is known, the ingress Path Computation Client
550 (PCC) sends a path computation request to a PCE responsible for the
551 ingress domain. This request is forwarded between PCEs, domain-by-
552 domain, to a PCE responsible for the egress domain. The PCE in the
553
554
555
556 <span class="grey">King &amp; Farrel Informational [Page 10]</span></pre>
557 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-11" ></span>
558 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
559
560
561 egress domain creates a set of optimal paths from all of the domain
562 entry BNs to the egress LSR. This set is represented as a tree of
563 potential paths called a Virtual Shortest Path Tree (VSPT), and the
564 PCE passes it back to the previous PCE on the domain path. As the
565 VSPT is passed back toward the ingress domain, each PCE computes the
566 optimal paths from its entry BNs to its exit BNs that connect to the
567 rest of the tree. It adds these paths to the VSPT and passes the
568 VSPT on until the PCE for the ingress domain is reached and computes
569 paths from the ingress LSR to connect to the rest of the tree. The
570 ingress PCE then selects the optimal end-to-end path from the tree,
571 and returns the path to the initiating PCC.
572
573 BRPC may suit environments where multiple connections exist between
574 domains and there is no preference for the choice of points of
575 interconnection. It is best suited to scenarios where the domain
576 path is known in advance, but it can also be used when the domain
577 path is not known.
578
579 <span class="h4"><a class="selflink" id="section-2.2.1" href="#section-2.2.1">2.2.1</a>. Applicability of BRPC When the Domain Path is Not Known</span>
580
581 As described above, BRPC can be used to determine an optimal inter-
582 domain path when the domain sequence is known. Even when the
583 sequence of domains is not known, BRPC could be used as follows.
584
585 o The PCC sends a request to a PCE for the ingress domain (the
586 ingress PCE).
587
588 o The ingress PCE sends the path computation request direct to a PCE
589 responsible for the domain containing the destination node (the
590 egress PCE).
591
592 o The egress PCE computes an egress VSPT and passes it to a PCE
593 responsible for each of the adjacent (potentially upstream)
594 domains.
595
596 o Each PCE in turn constructs a VSPT and passes it on to all of its
597 neighboring PCEs.
598
599 o When the ingress PCE has received a VSPT from each of its
600 neighboring domains, it is able to select the optimum path.
601
602 Clearly, this mechanism (which could be called path computation
603 flooding) has significant scaling issues. It could be improved by
604 the application of policy and filtering, but such mechanisms are not
605 simple and would still leave scaling concerns.
606
607
608
609
610
611
612 <span class="grey">King &amp; Farrel Informational [Page 11]</span></pre>
613 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-12" ></span>
614 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
615
616
617 <span class="h2"><a class="selflink" id="section-3" href="#section-3">3</a>. Hierarchical PCE</span>
618
619 In the hierarchical PCE architecture, a parent PCE maintains a domain
620 topology map that contains the child domains (seen as vertices in the
621 topology) and their interconnections (links in the topology). The
622 parent PCE has no information about the content of the child domains;
623 that is, the parent PCE does not know about the resource availability
624 within the child domains, nor does it know about the availability of
625 connectivity across each domain because such knowledge would violate
626 the confidentiality requirement and either would require flooding of
627 full information to the parent (scaling issue) or would necessitate
628 some form of aggregation. The parent PCE is aware of the TE
629 capabilities of the interconnections between child domains as these
630 interconnections are links in its own topology map.
631
632 Note that, in the case that the domains are IGP areas, there is no
633 link between the domains (the ABRs have a presence in both
634 neighboring areas). The parent domain may choose to represent this
635 in its Traffic Engineering Database (TED) as a virtual link that is
636 unconstrained and has zero cost, but this is entirely an
637 implementation issue.
638
639 Each child domain has at least one PCE capable of computing paths
640 across the domain. These PCEs are known as child PCEs and have a
641 relationship with the parent PCE. Each child PCE also knows the
642 identity of the domains that neighbor its own domain. A child PCE
643 only knows the topology of the domain that it serves and does not
644 know the topology of other child domains. Child PCEs are also not
645 aware of the general domain mesh connectivity (i.e., the domain
646 topology map) beyond the connectivity to the immediate neighbor
647 domains of the domain it serves.
648
649 The parent PCE builds the domain topology map either from
650 configuration or from information received from each child PCE. This
651 tells it how the domains are interconnected including the TE
652 properties of the domain interconnections. But, the parent PCE does
653 not know the contents of the child domains. Discovery of the domain
654 topology and domain interconnections is discussed further in <a href="#section-4.3">Section</a>
655 <a href="#section-4.3">4.3</a>.
656
657 When a multi-domain path is needed, the ingress PCE sends a request
658 to the parent PCE (using the Path Computation Element Protocol, PCEP
659 [<a href="./rfc5440" title="&quot;Path Computation Element (PCE) Communication Protocol (PCEP)&quot;">RFC5440</a>]). The parent PCE selects a set of candidate domain paths
660 based on the domain topology and the state of the inter-domain links.
661 It then sends computation requests to the child PCEs responsible for
662 each of the domains on the candidate domain paths. These requests
663 may be sequential or parallel depending on implementation details.
664
665
666
667
668 <span class="grey">King &amp; Farrel Informational [Page 12]</span></pre>
669 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-13" ></span>
670 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
671
672
673 Each child PCE computes a set of candidate path segments across its
674 domain and sends the results to the parent PCE. The parent PCE uses
675 this information to select path segments and concatenate them to
676 derive the optimal end-to-end inter-domain path. The end-to-end path
677 is then sent to the child PCE that received the initial path request,
678 and this child PCE passes the path on to the PCC that issued the
679 original request.
680
681 Specific deployment and implementation scenarios are out of scope of
682 this document. However, the hierarchical PCE architecture described
683 does support the function of parent PCE and child PCE being
684 implemented as a common PCE.
685
686 <span class="h2"><a class="selflink" id="section-4" href="#section-4">4</a>. Hierarchical PCE Procedures</span>
687
688 <span class="h3"><a class="selflink" id="section-4.1" href="#section-4.1">4.1</a>. Objective Functions and Policy</span>
689
690 The definition of "optimal" in the context of deriving an optimal
691 end-to-end path is dependent on the choices that are made during the
692 path selection. An Objective Function (OF) [<a href="./rfc5541" title="&quot;Encoding of Objective Functions in the Path Computation Element Communication Protocol (PCEP)&quot;">RFC5541</a>], or set of OFs,
693 specify the intentions of the path computation and so define the
694 "optimality" in the context of that computation.
695
696 An OF specifies the desired outcome of a computation: it does not
697 describe or demand the algorithm to use, and an implementation may
698 apply any algorithm or set of algorithms to achieve the result
699 indicated by the OF. OFs can be included in PCEP computation
700 requests to satisfy the policies encoded or configured at the PCC,
701 and a PCE may be subject to policy in determining whether it meets
702 the OFs included in the computation request, or applies its own OFs.
703
704 In inter-domain path computation, the selection of a domain sequence,
705 the computation of each (per-domain) path fragment, and the
706 determination of the end-to-end path may each be subject to different
707 OFs and different policy.
708
709 When computing end-to-end paths, OFs may include (see <a href="#section-1.3.1">Section 1.3.1</a>):
710
711 o Minimum cost path
712 o Minimum load path
713 o Maximum residual bandwidth path
714 o Minimize aggregate bandwidth consumption
715 o Minimize or cap the number of transit domains
716 o Disallow domain re-entry
717
718 The objective function may be requested by the PCC, the ingress
719 domain PCE (according to local policy), or applied by the parent PCE
720 according to inter-domain policy.
721
722
723
724 <span class="grey">King &amp; Farrel Informational [Page 13]</span></pre>
725 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-14" ></span>
726 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
727
728
729 More than one OF (or a composite OF) may be chosen to apply to a
730 single computation provided they are not contradictory. Composite
731 OFs may include weightings and preferences for the fulfillment of
732 pieces of the desired outcome.
733
734 <span class="h3"><a class="selflink" id="section-4.2" href="#section-4.2">4.2</a>. Maintaining Domain Confidentiality</span>
735
736 Information about the content of child domains is not shared for
737 scaling and confidentiality reasons. This means that a parent PCE is
738 aware of the domain topology and the nature of the connections
739 between domains but is not aware of the content of the domains.
740 Similarly, a child PCE cannot know the internal topology of another
741 child domain. Child PCEs also do not know the general domain mesh
742 connectivity; this information is only known by the parent PCE.
743
744 As described in the earlier sections of this document, PCEs can
745 exchange path information in order to construct an end-to-end inter-
746 domain path. Each per-domain path fragment reveals information about
747 the topology and resource availability within a domain. Some
748 management domains or ASes will not want to share this information
749 outside of the domain (even with a trusted parent PCE). In order to
750 conceal the information, a PCE may replace a path segment with a
751 path-key [<a href="./rfc5520" title="&quot;Preserving Topology Confidentiality in Inter-Domain Path Computation Using a Path-Key-Based Mechanism&quot;">RFC5520</a>]. This mechanism effectively hides the content of
752 a segment of a path.
753
754 <span class="h3"><a class="selflink" id="section-4.3" href="#section-4.3">4.3</a>. PCE Discovery</span>
755
756 It is a simple matter for each child PCE to be configured with the
757 address of its parent PCE. Typically, there will only be one or two
758 parents of any child.
759
760 The parent PCE also needs to be aware of the child PCEs for all child
761 domains that it can see. This information is most likely to be
762 configured (as part of the administrative definition of each domain).
763
764 Discovery of the relationships between parent PCEs and child PCEs
765 does not form part of the hierarchical PCE architecture. Mechanisms
766 that rely on advertising or querying PCE locations across domain or
767 provider boundaries are undesirable for security, scaling,
768 commercial, and confidentiality reasons.
769
770 The parent PCE also needs to know the inter-domain connectivity.
771 This information could be configured with suitable policy and
772 commercial rules, or could be learned from the child PCEs as
773 described in <a href="#section-4.4">Section 4.4</a>.
774
775
776
777
778
779
780 <span class="grey">King &amp; Farrel Informational [Page 14]</span></pre>
781 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-15" ></span>
782 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
783
784
785 In order for the parent PCE to learn about domain interconnection,
786 the child PCE will report the identity of its neighbor domains. The
787 IGP in each neighbor domain can advertise its inter-domain TE link
788 capabilities [<a href="./rfc5316" title="&quot;ISIS Extensions in Support of Inter-Autonomous System (AS) MPLS and GMPLS Traffic Engineering&quot;">RFC5316</a>] [<a href="./rfc5392" title="&quot;OSPF Extensions in Support of Inter-Autonomous System (AS) MPLS and GMPLS Traffic Engineering&quot;">RFC5392</a>]. This information can be collected
789 by the child PCEs and forwarded to the parent PCE, or the parent PCE
790 could participate in the IGP in the child domains.
791
792 <span class="h3"><a class="selflink" id="section-4.4" href="#section-4.4">4.4</a>. Traffic Engineering Database for the Parent Domain</span>
793
794 The parent PCE maintains a domain topology map of the child domains
795 and their interconnectivity. Where inter-domain connectivity is
796 provided by TE links, the capabilities of those links may also be
797 known to the parent PCE. The parent PCE maintains a TED for the
798 parent domain in the same way that any PCE does.
799
800 The parent domain may just be the collection of child domains and
801 their interconnectivity, may include details of the inter-domain TE
802 links, and may contain nodes and links in its own right.
803
804 The mechanism for building the parent TED is likely to rely heavily
805 on administrative configuration and commercial issues because the
806 network was probably partitioned into domains specifically to address
807 these issues.
808
809 In practice, certain information may be passed from the child domains
810 to the parent PCE to help build the parent TED. In theory, the
811 parent PCE could listen to the routing protocols in the child
812 domains, but this would violate the confidentiality and scaling
813 principles that may be responsible for the partition of the network
814 into domains. So, it is much more likely that a suitable solution
815 will involve specific communication from an entity in the child
816 domain (such as the child PCE) to convey the necessary information.
817 As already mentioned, the "necessary information" relates to how the
818 child domains are inter-connected. The topology and available
819 resources within the child domain do not need to be communicated to
820 the parent PCE: doing so would violate the PCE architecture.
821 Mechanisms for reporting this information are described in the
822 examples in <a href="#section-4.6">Section 4.6</a> in abstract terms as a child PCE "reports its
823 neighbor domain connectivity to its parent PCE"; the specifics of a
824 solution are out of scope of this document, but the requirements are
825 indicated in <a href="#section-4.8">Section 4.8</a>. See <a href="#section-6">Section 6</a> for a brief discussion of
826 BGP-TE.
827
828 In models such as ASON (see <a href="#section-5.2">Section 5.2</a>), it is possible to consider
829 a separate instance of an IGP running within the parent domain where
830 the participating protocol speakers are the nodes directly present in
831 that domain and the PCEs (Routing Controllers) responsible for each
832 of the child domains.
833
834
835
836 <span class="grey">King &amp; Farrel Informational [Page 15]</span></pre>
837 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-16" ></span>
838 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
839
840
841 <span class="h3"><a class="selflink" id="section-4.5" href="#section-4.5">4.5</a>. Determination of Destination Domain</span>
842
843 The PCC asking for an inter-domain path computation is aware of the
844 identity of the destination node by definition. If it knows the
845 egress domain, it can supply this information as part of the path
846 computation request. However, if it does not know the egress domain,
847 this information must be known by the child PCE or known/determined
848 by the parent PCE.
849
850 In some specialist topologies the parent PCE could determine the
851 destination domain based on the destination address, for example,
852 from configuration. However, this is not appropriate for many multi-
853 domain addressing scenarios. In IP-based multi-domain networks, the
854 parent PCE may be able to determine the destination domain by
855 participating in inter-domain routing. Finally, the parent PCE could
856 issue specific requests to the child PCEs to discover if they contain
857 the destination node, but this has scaling implications.
858
859 For the purposes of this document, the precise mechanism of the
860 discovery of the destination domain is left out of scope. Suffice to
861 say that for each multi-domain path computation some mechanism will
862 be required to determine the location of the destination.
863
864 <span class="h3"><a class="selflink" id="section-4.6" href="#section-4.6">4.6</a>. Hierarchical PCE Examples</span>
865
866 The following example describes the generic hierarchical domain
867 topology. Figure 1 demonstrates four interconnected domains within a
868 fifth, parent domain. Each domain contains a single PCE:
869
870 o Domain 1 is the ingress domain and child PCE 1 is able to compute
871 paths within the domain. Its neighbors are Domain 2 and Domain 4.
872 The domain also contains the source LSR (S) and three egress
873 boundary nodes (BN11, BN12, and BN13).
874
875 o Domain 2 is served by child PCE 2. Its neighbors are Domain 1 and
876 Domain 3. The domain also contains four boundary nodes (BN21,
877 BN22, BN23, and BN24).
878
879 o Domain 3 is the egress domain and is served by child PCE 3. Its
880 neighbors are Domain 2 and Domain 4. The domain also contains the
881 destination LSR (D) and three ingress boundary nodes (BN31, BN32,
882 and BN33).
883
884 o Domain 4 is served by child PCE 4. Its neighbors are Domain 2 and
885 Domain 3. The domain also contains two boundary nodes (BN41 and
886 BN42).
887
888
889
890
891
892 <span class="grey">King &amp; Farrel Informational [Page 16]</span></pre>
893 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-17" ></span>
894 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
895
896
897 All of these domains are contained within Domain 5, which is served
898 by the parent PCE (PCE 5).
899
900 -----------------------------------------------------------------
901 | Domain 5 |
902 | ----- |
903 | |PCE 5| |
904 | ----- |
905 | |
906 | ---------------- ---------------- ---------------- |
907 | | Domain 1 | | Domain 2 | | Domain 3 | |
908 | | | | | | | |
909 | | ----- | | ----- | | ----- | |
910 | | |PCE 1| | | |PCE 2| | | |PCE 3| | |
911 | | ----- | | ----- | | ----- | |
912 | | | | | | | |
913 | | ----| |---- ----| |---- | |
914 | | |BN11+---+BN21| |BN23+---+BN31| | |
915 | | - ----| |---- ----| |---- - | |
916 | | |S| | | | | |D| | |
917 | | - ----| |---- ----| |---- - | |
918 | | |BN12+---+BN22| |BN24+---+BN32| | |
919 | | ----| |---- ----| |---- | |
920 | | | | | | | |
921 | | ---- | | | | ---- | |
922 | | |BN13| | | | | |BN33| | |
923 | -----------+---- ---------------- ----+----------- |
924 | \ / |
925 | \ ---------------- / |
926 | \ | | / |
927 | \ |---- ----| / |
928 | ----+BN41| |BN42+---- |
929 | |---- ----| |
930 | | | |
931 | | ----- | |
932 | | |PCE 4| | |
933 | | ----- | |
934 | | | |
935 | | Domain 4 | |
936 | ---------------- |
937 | |
938 -----------------------------------------------------------------
939
940 Figure 1: Sample Hierarchical Domain Topology
941
942
943
944
945
946
947
948 <span class="grey">King &amp; Farrel Informational [Page 17]</span></pre>
949 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-18" ></span>
950 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
951
952
953 Figure 2 shows the view of the domain topology as seen by the parent
954 PCE (PCE 5). This view is an abstracted topology; PCE 5 is aware of
955 domain connectivity but not of the internal topology within each
956 domain.
957
958 ----------------------------
959 | Domain 5 |
960 | ---- |
961 | |PCE5| |
962 | ---- |
963 | |
964 | ---- ---- ---- |
965 | | |---| |---| | |
966 | | D1 | | D2 | | D3 | |
967 | | |---| |---| | |
968 | ---- ---- ---- |
969 | \ ---- / |
970 | \ | | / |
971 | ----| D4 |---- |
972 | | | |
973 | ---- |
974 | |
975 ----------------------------
976
977 Figure 2: Abstract Domain Topology as Seen by the Parent PCE
978
979 <span class="h4"><a class="selflink" id="section-4.6.1" href="#section-4.6.1">4.6.1</a>. Hierarchical PCE Initial Information Exchange</span>
980
981 Based on the topology in Figure 1, the following is an illustration
982 of the initial hierarchical PCE information exchange.
983
984 1. Child PCE 1, the PCE responsible for Domain 1, is configured with
985 the location of its parent PCE (PCE 5).
986
987 2. Child PCE 1 establishes contact with its parent PCE. The parent
988 applies policy to ensure that communication with PCE 1 is
989 allowed.
990
991 3. Child PCE 1 listens to the IGP in its domain and learns its
992 inter-domain connectivity. That is, it learns about the links
993 BN11-BN21, BN12-BN22, and BN13-BN41.
994
995 4. Child PCE 1 reports its neighbor domain connectivity to its
996 parent PCE.
997
998 5. Child PCE 1 reports any change in the resource availability on
999 its inter-domain links to its parent PCE.
1000
1001
1002
1003
1004 <span class="grey">King &amp; Farrel Informational [Page 18]</span></pre>
1005 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-19" ></span>
1006 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
1007
1008
1009 Each child PCE performs steps 1 through 5 so that the parent PCE can
1010 create a domain topology view as shown in Figure 2.
1011
1012 <span class="h4"><a class="selflink" id="section-4.6.2" href="#section-4.6.2">4.6.2</a>. Hierarchical PCE End-to-End Path Computation Procedure</span>
1013
1014 The procedure below is an example of a source PCC requesting an end-
1015 to-end path in a multi-domain environment. The topology is
1016 represented in Figure 1. It is assumed that the each child PCE has
1017 connected to its parent PCE and exchanged the initial information
1018 required for the parent PCE to create its domain topology view as
1019 described in <a href="#section-4.6.1">Section 4.6.1</a>.
1020
1021 1. The source PCC (the ingress LSR in our example) sends a request
1022 to the PCE responsible for its domain (PCE 1) for a path to the
1023 destination LSR (D).
1024
1025 2. PCE 1 determines the destination is not in domain 1.
1026
1027 3. PCE 1 sends a computation request to its parent PCE (PCE 5).
1028
1029 4. The parent PCE determines that the destination is in Domain 3.
1030 (See <a href="#section-4.5">Section 4.5</a>.)
1031
1032 5. PCE 5 determines the likely domain paths according to the domain
1033 interconnectivity and TE capabilities between the domains. For
1034 example, assuming that the link BN12-BN22 is not suitable for the
1035 requested path, three domain paths are determined:
1036
1037 S-BN11-BN21-D2-BN23-BN31-D
1038 S-BN11-BN21-D2-BN24-BN32-D
1039 S-BN13-BN41-D4-BN42-BN33-D
1040
1041 6. PCE 5 sends edge-to-edge path computation requests to PCE 2,
1042 which is responsible for Domain 2 (i.e., BN21-to-BN23 and
1043 BN21-to-BN24), and to PCE 4 for Domain 4 (i.e., BN41-to-BN42).
1044
1045 7. PCE 5 sends source-to-edge path computation requests to PCE 1,
1046 which is responsible for Domain 1 (i.e., S-to-BN11 and
1047 S-to-BN13).
1048
1049 8. PCE 5 sends edge-to-egress path computation requests to PCE 3,
1050 which is responsible for Domain 3 (i.e., BN31-to-D, BN32-to-D,
1051 and BN33-to-D).
1052
1053 9. PCE 5 correlates all the computation responses from each child
1054 PCE, adds in the information about the inter-domain links, and
1055 applies any requested and locally configured policies.
1056
1057
1058
1059
1060 <span class="grey">King &amp; Farrel Informational [Page 19]</span></pre>
1061 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-20" ></span>
1062 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
1063
1064
1065 10. PCE 5 then selects the optimal end-to-end multi-domain path that
1066 meets the policies and objective functions, and supplies the
1067 resulting path to PCE 1.
1068
1069 11. PCE 1 forwards the path to the PCC (the ingress LSR).
1070
1071 Note that there is no requirement for steps 6, 7, and 8 to be carried
1072 out in parallel or in series. Indeed, they could be overlapped with
1073 step 5. This is an implementation issue.
1074
1075 <span class="h3"><a class="selflink" id="section-4.7" href="#section-4.7">4.7</a>. Hierarchical PCE Error Handling</span>
1076
1077 In the event that a child PCE in a domain cannot find a suitable path
1078 to the egress, the child PCE should return the relevant error to
1079 notify the parent PCE. Depending on the error response, the parent
1080 PCE selects one of the following actions:
1081
1082 o Cancel the request and send the relevant response back to the
1083 initial child PCE that requested an end-to-end path;
1084
1085 o Relax some of the constraints associated with the initial path
1086 request; or
1087
1088 o Select another candidate domain and send the path request to the
1089 child PCE responsible for the domain.
1090
1091 If the parent PCE does not receive a response from a child PCE within
1092 an allotted time period, the parent PCE can elect to:
1093
1094 o Cancel the request and send the relevant response back to the
1095 initial child PCE that requested an end-to-end path; o Send the path
1096 request to another child PCE in the same domain, if a secondary child
1097 PCE exists; o Select another candidate domain and send the path
1098 request to the child PCE responsible for that domain.
1099
1100 The parent PCE may also want to prune any unresponsive child PCE
1101 domain paths from the candidate set.
1102
1103 <span class="h3"><a class="selflink" id="section-4.8" href="#section-4.8">4.8</a>. Requirements for Hierarchical PCEP Protocol Extensions</span>
1104
1105 This section lists the high-level requirements for extensions to the
1106 PCEP to support the hierarchical PCE model. It is provided to offer
1107 guidance to PCEP protocol developers in designing a solution suitable
1108 for use in a hierarchical PCE framework.
1109
1110
1111
1112
1113
1114
1115
1116 <span class="grey">King &amp; Farrel Informational [Page 20]</span></pre>
1117 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-21" ></span>
1118 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
1119
1120
1121 <span class="h4"><a class="selflink" id="section-4.8.1" href="#section-4.8.1">4.8.1</a>. PCEP Request Qualifiers</span>
1122
1123 Path Computation Request (PCReq) messages are used by a PCC or a PCE
1124 to make a computation request or enquiry to a PCE. The requests are
1125 qualified so that the PCE knows what type of action is required.
1126
1127 Support of the hierarchical PCE architecture will introduce two new
1128 qualifications as follows:
1129
1130 o It must be possible for a child PCE to indicate that the response
1131 it receives from the parent PCE should consist of a domain
1132 sequence only (i.e., not a fully specified end-to-end path). This
1133 allows the child PCE to initiate Per-Domain or BRPC.
1134
1135 o A parent PCE may need to be able to ask a child PCE whether a
1136 particular node address (the destination of an end-to-end path) is
1137 present in the domain that the child PCE serves.
1138
1139 In PCEP, such request qualifications are carried as bit flags in the
1140 RP object (Request Parameter object) within the PCReq message.
1141
1142 <span class="h4"><a class="selflink" id="section-4.8.2" href="#section-4.8.2">4.8.2</a>. Indication of Hierarchical PCE Capability</span>
1143
1144 Although parent/child PCE relationships are likely configured, it
1145 will assist network operations if the parent PCE is able to indicate
1146 to the child that it really is capable of acting as a parent PCE.
1147 This will help to trap misconfigurations.
1148
1149 In PCEP, such capabilities are carried in the Open Object within the
1150 Open message.
1151
1152 <span class="h4"><a class="selflink" id="section-4.8.3" href="#section-4.8.3">4.8.3</a>. Intention to Utilize Parent PCE Capabilities</span>
1153
1154 A PCE that is capable of acting as a parent PCE might not be
1155 configured or willing to act as the parent for a specific child PCE.
1156 This fact could be determined when the child sends a PCReq that
1157 requires parental activity (such as querying other child PCEs), and
1158 could result in a negative response in a PCEP Error (PCErr) message.
1159
1160 However, the expense of a poorly targeted PCReq can be avoided if the
1161 child PCE indicates that it might wish to use the parent-capable PCE
1162 as a parent (for example, on the Open message), and if the parent-
1163 capable PCE determines at that time whether it is willing to act as a
1164 parent to this child.
1165
1166
1167
1168
1169
1170
1171
1172 <span class="grey">King &amp; Farrel Informational [Page 21]</span></pre>
1173 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-22" ></span>
1174 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
1175
1176
1177 <span class="h4"><a class="selflink" id="section-4.8.4" href="#section-4.8.4">4.8.4</a>. Communication of Domain Connectivity Information</span>
1178
1179 <a href="#section-4.4">Section 4.4</a> describes how the parent PCE needs a parent TED and
1180 indicates that the information might be supplied from the child PCEs
1181 in each domain. This requires a mechanism whereby information about
1182 inter-domain links can be supplied by a child PCE to a parent PCE,
1183 for example, on a PCEP Notify (PCNtf) message.
1184
1185 The information that would be exchanged includes:
1186
1187 o Identifier of advertising child PCE
1188 o Identifier of PCE's domain
1189 o Identifier of the link
1190 o TE properties of the link (metrics, bandwidth)
1191 o Other properties of the link (technology-specific)
1192 o Identifier of link endpoints
1193 o Identifier of adjacent domain
1194
1195 It may be desirable for this information to be periodically updated,
1196 for example, when available bandwidth changes. In this case, the
1197 parent PCE might be given the ability to configure thresholds in the
1198 child PCE to prevent flapping of information.
1199
1200 <span class="h4"><a class="selflink" id="section-4.8.5" href="#section-4.8.5">4.8.5</a>. Domain Identifiers</span>
1201
1202 Domain identifiers are already present in PCEP to allow a PCE to
1203 indicate which domains it serves, and to allow the representation of
1204 domains as abstract nodes in paths. The wider use of domains in the
1205 context of this work on hierarchical PCE will require that domains
1206 can be identified in more places within objects in PCEP messages.
1207 This should pose no problems.
1208
1209 However, more attention may need to be applied to the precision of
1210 domain identifier definitions to ensure that it is always possible to
1211 unambiguously identify a domain from its identifier. This work will
1212 be necessary in configuration, and also in protocol specifications
1213 (for example, an OSPF area identifier is sufficient within an
1214 Autonomous System, but becomes ambiguous in a path that crosses
1215 multiple Autonomous Systems).
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228 <span class="grey">King &amp; Farrel Informational [Page 22]</span></pre>
1229 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-23" ></span>
1230 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
1231
1232
1233 <span class="h2"><a class="selflink" id="section-5" href="#section-5">5</a>. Hierarchical PCE Applicability</span>
1234
1235 As per [<a href="./rfc4655" title="&quot;A Path Computation Element (PCE)-Based Architecture&quot;">RFC4655</a>], PCE can inherently support inter-domain path
1236 computation for any definition of a domain as set out in <a href="#section-1.2">Section 1.2</a>
1237 of this document.
1238
1239 Hierarchical PCE can be applied to inter-domain environments,
1240 including autonomous Systems and IGP areas. The hierarchical PCE
1241 procedures make no distinction between, autonomous Systems and IGP
1242 area applications, although it should be noted that the TED
1243 maintained by a parent PCE must be able to support the concept of
1244 child domains connected by inter-domain links or directly connected
1245 at boundary nodes (see <a href="#section-3">Section 3</a>).
1246
1247 This section sets out the applicability of hierarchical PCE to three
1248 environments:
1249
1250 o MPLS traffic engineering across multiple Autonomous Systems
1251 o MPLS traffic engineering across multiple IGP areas
1252 o GMPLS traffic engineering in the ASON architecture
1253
1254 <span class="h3"><a class="selflink" id="section-5.1" href="#section-5.1">5.1</a>. Autonomous Systems and Areas</span>
1255
1256 Networks are comprised of domains. A domain can be considered to be
1257 a collection of network elements within an AS or area that has a
1258 common sphere of address management or path computational
1259 responsibility.
1260
1261 As networks increase in size and complexity it may be required to
1262 introduce scaling methods to reduce the amount information flooded
1263 within the network and make the network more manageable. An IGP
1264 hierarchy is designed to improve IGP scalability by dividing the IGP
1265 domain into areas and limiting the flooding scope of topology
1266 information to within area boundaries. This restricts a router's
1267 visibility to information about links and other routers within the
1268 single area. If a router needs to compute a route to destination
1269 located in another area, a method is required to compute a path
1270 across the area boundary.
1271
1272 When an LSR within an AS or area needs to compute a path across an
1273 area or AS boundary, it must also use an inter-AS computation
1274 technique. Hierarchical PCE is equally applicable to computing
1275 inter-area and inter-AS MPLS and GMPLS paths across domain
1276 boundaries.
1277
1278
1279
1280
1281
1282
1283
1284 <span class="grey">King &amp; Farrel Informational [Page 23]</span></pre>
1285 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-24" ></span>
1286 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
1287
1288
1289 <span class="h3"><a class="selflink" id="section-5.2" href="#section-5.2">5.2</a>. ASON Architecture</span>
1290
1291 The International Telecommunication Union (ITU) defines the ASON
1292 architecture in [<a href="#ref-G-8080">G-8080</a>]. [<a href="#ref-G-7715">G-7715</a>] defines the routing architecture
1293 for ASON and introduces a hierarchical architecture. In this
1294 architecture, the Routing Areas (RAs) have a hierarchical
1295 relationship between different routing levels, which means a parent
1296 (or higher-level) RA can contain multiple child RAs. The
1297 interconnectivity of the lower RAs is visible to the higher-level RA.
1298 Note that the RA hierarchy can be recursive.
1299
1300 In the ASON framework, a path computation request is termed a Route
1301 Query. This query is executed before signaling is used to establish
1302 an LSP termed a Switched Connection (SC) or a Soft Permanent
1303 Connection (SPC). [<a href="#ref-G-7715-2">G-7715-2</a>] defines the requirements and
1304 architecture for the functions performed by Routing Controllers (RCs)
1305 during the operation of remote route queries -- an RC is synonymous
1306 with a PCE. For an end-to-end connection, the route may be computed
1307 by a single RC or multiple RCs in a collaborative manner (i.e., RC
1308 federations). In the case of RC federations, [<a href="#ref-G-7715-2">G-7715-2</a>] describes
1309 three styles during remote route query operation:
1310
1311 o step-by-step remote path computation
1312 o hierarchical remote path computation
1313 o a combination of the above.
1314
1315 In a hierarchical ASON routing environment, a child RC may
1316 communicate with its parent RC (at the next higher level of the ASON
1317 routing hierarchy) to request the computation of an end-to-end path
1318 across several RAs. It does this using a route query message (known
1319 as the abstract message RI_QUERY). The corresponding parent RC may
1320 communicate with other child RCs that belong to other child RAs at
1321 the next lower hierarchical level. Thus, a parent RC can act as
1322 either a Route Query Requester or Route Query Responder.
1323
1324 It can be seen that the hierarchical PCE architecture fits the
1325 hierarchical ASON routing architecture well. It can be used to
1326 provide paths across subnetworks and to determine end-to-end paths in
1327 networks constructed from multiple subnetworks or RAs.
1328
1329 When hierarchical PCE is applied to implement hierarchical remote
1330 path computation in [<a href="#ref-G-7715-2">G-7715-2</a>], it is very important for operators to
1331 understand the different terminology and implicit consistency between
1332 hierarchical PCE and [<a href="#ref-G-7715-2">G-7715-2</a>].
1333
1334
1335
1336
1337
1338
1339
1340 <span class="grey">King &amp; Farrel Informational [Page 24]</span></pre>
1341 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-25" ></span>
1342 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
1343
1344
1345 <span class="h4"><a class="selflink" id="section-5.2.1" href="#section-5.2.1">5.2.1</a>. Implicit Consistency between Hierarchical PCE and G.7715.2</span>
1346
1347 This section highlights the correspondence between features of the
1348 hierarchical PCE architecture and the ASON routing architecture.
1349
1350 (1) RC (Routing Controller) and PCE (Path Computation Element)
1351
1352 [<a id="ref-G-8080">G-8080</a>] describes the Routing Controller component as an
1353 abstract entity, which is responsible for responding to requests
1354 for path (route) information and topology information. It can be
1355 implemented as a single entity, or as a distributed set of
1356 entities that make up a cooperative federation.
1357
1358 [<a id="ref-RFC4655">RFC4655</a>] describes PCE (Path Computation Element) is an entity
1359 (component, application, or network node) that is capable of
1360 computing a network path or route based on a network graph and
1361 applying computational constraints.
1362
1363 Therefore, in the ASON architecture, a PCE can be regarded as a
1364 realization of the RC.
1365
1366 (2) Route Query Requester/Route Query Responder and PCC/PCE
1367
1368 [<a id="ref-G-7715-2">G-7715-2</a>] describes the Route Query Requester as a Connection
1369 Controller or Routing Controller that sends a route query message
1370 to a Routing Controller requesting one or more paths that satisfy
1371 a set of routing constraints. The Route Query Responder is a
1372 Routing Controller that performs path computation upon receipt of
1373 a route query message from a Route Query Requester, sending a
1374 response back at the end of the path computation.
1375
1376 In the context of ASON, a Signaling Controller initiates and
1377 processes signaling messages and is closely coupled to a
1378 Signaling Protocol Speaker. A Routing Controller makes routing
1379 decisions and is usually coupled to configuration entities and/or
1380 a Routing Protocol Speaker.
1381
1382 It can be seen that a PCC corresponds to a Route Query Requester,
1383 and a PCE corresponds to a Route Query Responder. A PCE/RC can
1384 also act as a Route Query Requester sending requests to another
1385 Route Query Responder.
1386
1387 The Path Computation Request (PCReq) and Path Computation Reply
1388 (PCRep) messages between PCC and PCE correspond to the RI_QUERY
1389 and RI_UPDATE messages in [<a href="#ref-G-7715-2">G-7715-2</a>].
1390
1391
1392
1393
1394
1395
1396 <span class="grey">King &amp; Farrel Informational [Page 25]</span></pre>
1397 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-26" ></span>
1398 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
1399
1400
1401 (3) Routing Area Hierarchy and Hierarchical Domain
1402
1403 The ASON routing hierarchy model is shown in Figure 6 of [<a href="#ref-G-7715">G-7715</a>]
1404 through an example that illustrates routing area levels. If the
1405 hierarchical remote path computation mechanism of [<a href="#ref-G-7715-2">G-7715-2</a>] is
1406 applied in this scenario, each routing area should have at least
1407 one RC to perform the route query function, and the child RCs
1408 within the area should have a parent RC.
1409
1410 According to [<a href="#ref-G-8080">G-8080</a>], the parent RC has visibility of the
1411 structure of the lower level, so it knows the interconnectivity
1412 of the RAs in the lower level. Each child RC can compute edge-
1413 to-edge paths across its own child RA.
1414
1415 Thus, an RA corresponds to a domain in the PCE architecture, and
1416 the hierarchical relationship between RAs corresponds to the
1417 hierarchical relationship between domains in the hierarchical PCE
1418 architecture. Furthermore, a parent PCE in a parent domain can
1419 be regarded as parent RC in a higher routing level, and a child
1420 PCE in a child domain can be regarded as child RC in a lower
1421 routing level.
1422
1423 <span class="h4"><a class="selflink" id="section-5.2.2" href="#section-5.2.2">5.2.2</a>. Benefits of Hierarchical PCEs in ASON</span>
1424
1425 RCs in an ASON environment can use the hierarchical PCE model to
1426 fully match the ASON hierarchical routing model, so the hierarchical
1427 PCE mechanisms can be applied to fully satisfy the architecture and
1428 requirements of [<a href="#ref-G-7715-2">G-7715-2</a>] without any changes. If the hierarchical
1429 PCE mechanism is applied in ASON, it can be used to determine end-to-
1430 end optimized paths across subnetworks and RAs before initiating
1431 signaling to create the connection. It can also improve the
1432 efficiency of connection setup to avoid crankback.
1433
1434 <span class="h2"><a class="selflink" id="section-6" href="#section-6">6</a>. A Note on BGP-TE</span>
1435
1436 The concept of exchange of TE information between Autonomous Systems
1437 (ASes) is discussed in [<a href="#ref-BGP-TE" title="&quot;North-Bound Distribution of Link-State and TE Information using BGP&quot;">BGP-TE</a>]. The information exchanged in this
1438 way could be the full TE information from the AS, an aggregation of
1439 that information, or a representation of the potential connectivity
1440 across the AS. Furthermore, that information could be updated
1441 frequently (for example, for every new LSP that is set up across the
1442 AS) or only at threshold-crossing events.
1443
1444 There are a number of discussion points associated with the use of
1445 [<a href="#ref-BGP-TE" title="&quot;North-Bound Distribution of Link-State and TE Information using BGP&quot;">BGP-TE</a>] concerning the volume of information, the rate of churn of
1446 information, the confidentiality of information, the accuracy of
1447 aggregated or potential-connectivity information, and the processing
1448 required to generate aggregated information. The PCE architecture
1449
1450
1451
1452 <span class="grey">King &amp; Farrel Informational [Page 26]</span></pre>
1453 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-27" ></span>
1454 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
1455
1456
1457 and the architecture enabled by [<a href="#ref-BGP-TE" title="&quot;North-Bound Distribution of Link-State and TE Information using BGP&quot;">BGP-TE</a>] make different assumptions
1458 about the operational objectives of the networks, and this document
1459 does not attempt to make one of the approaches "right" and the other
1460 "wrong". Instead, this work assumes that a decision has been made to
1461 utilize the PCE architecture.
1462
1463 <span class="h3"><a class="selflink" id="section-6.1" href="#section-6.1">6.1</a>. Use of BGP for TED Synchronization</span>
1464
1465 Indeed, [<a href="#ref-BGP-TE" title="&quot;North-Bound Distribution of Link-State and TE Information using BGP&quot;">BGP-TE</a>] may have some uses within the PCE model. For
1466 example, [<a href="#ref-BGP-TE" title="&quot;North-Bound Distribution of Link-State and TE Information using BGP&quot;">BGP-TE</a>] could be used as a "northbound" TE advertisement
1467 such that a PCE does not need to listen to an IGP in its domain, but
1468 has its TED populated by messages received (for example) from a Route
1469 Reflector. Furthermore, the inter-domain connectivity and
1470 capabilities that are required information for a parent PCE could be
1471 obtained as a filtered subset of the information available in
1472 [<a href="#ref-BGP-TE" title="&quot;North-Bound Distribution of Link-State and TE Information using BGP&quot;">BGP-TE</a>]. This scenario is discussed further in [<a href="#ref-PCE-AREA-AS" title="&quot;Applicability of the Path Computation Element to Inter-Area and Inter-AS MPLS and GMPLS Traffic Engineering&quot;">PCE-AREA-AS</a>].
1473
1474 <span class="h2"><a class="selflink" id="section-7" href="#section-7">7</a>. Management Considerations</span>
1475
1476 General PCE management considerations are discussed in [<a href="./rfc4655" title="&quot;A Path Computation Element (PCE)-Based Architecture&quot;">RFC4655</a>]. In
1477 the case of the hierarchical PCE architecture, there are additional
1478 management considerations.
1479
1480 The administrative entity responsible for the management of the
1481 parent PCEs must be determined. In the case of multi-domains (e.g.,
1482 IGP areas or multiple ASes) within a single service provider network,
1483 the management responsibility for the parent PCE would most likely be
1484 handled by the service provider. In the case of multiple ASes within
1485 different service provider networks, it may be necessary for a third
1486 party to manage the parent PCEs according to commercial and policy
1487 agreements from each of the participating service providers.
1488
1489 <span class="h3"><a class="selflink" id="section-7.1" href="#section-7.1">7.1</a>. Control of Function and Policy</span>
1490
1491 <span class="h4"><a class="selflink" id="section-7.1.1" href="#section-7.1.1">7.1.1</a>. Child PCE</span>
1492
1493 Support of the hierarchical procedure will be controlled by the
1494 management organization responsible for each child PCE. A child PCE
1495 must be configured with the address of its parent PCE in order for it
1496 to interact with its parent PCE. The child PCE must also be
1497 authorized to peer with the parent PCE.
1498
1499 <span class="h4"><a class="selflink" id="section-7.1.2" href="#section-7.1.2">7.1.2</a>. Parent PCE</span>
1500
1501 The parent PCE must only accept path computation requests from
1502 authorized child PCEs. If a parent PCE receives requests from an
1503 unauthorized child PCE, the request should be dropped.
1504
1505
1506
1507
1508 <span class="grey">King &amp; Farrel Informational [Page 27]</span></pre>
1509 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-28" ></span>
1510 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
1511
1512
1513 This means that a parent PCE must be configured with the identities
1514 and security credentials of all of its child PCEs, or there must be
1515 some form of shared secret that allows an unknown child PCE to be
1516 authorized by the parent PCE.
1517
1518 <span class="h4"><a class="selflink" id="section-7.1.3" href="#section-7.1.3">7.1.3</a>. Policy Control</span>
1519
1520 It may be necessary to maintain a policy module on the parent PCE
1521 [<a href="./rfc5394" title="&quot;Policy-Enabled Path Computation Framework&quot;">RFC5394</a>]. This would allow the parent PCE to apply commercially
1522 relevant constraints such as SLAs, security, peering preferences, and
1523 monetary costs.
1524
1525 It may also be necessary for the parent PCE to limit end-to-end path
1526 selection by including or excluding specific domains based on
1527 commercial relationships, security implications, and reliability.
1528
1529 <span class="h3"><a class="selflink" id="section-7.2" href="#section-7.2">7.2</a>. Information and Data Models</span>
1530
1531 A PCEP MIB module is defined in [<a href="#ref-PCEP-MIB" title="&quot;PCE communication protocol (PCEP) Management Information Base&quot;">PCEP-MIB</a>] that describes managed
1532 objects for modeling of PCEP communication. An additional PCEP MIB
1533 will be required to report parent PCE and child PCE information,
1534 including:
1535
1536 o parent PCE configuration and status,
1537
1538 o child PCE configuration and information,
1539
1540 o notifications to indicate session changes between parent PCEs and
1541 child PCEs, and
1542
1543 o notification of parent PCE TED updates and changes.
1544
1545 <span class="h3"><a class="selflink" id="section-7.3" href="#section-7.3">7.3</a>. Liveness Detection and Monitoring</span>
1546
1547 The hierarchical procedure requires interaction with multiple PCEs.
1548 Once a child PCE requests an end-to-end path, a sequence of events
1549 occurs that requires interaction between the parent PCE and each
1550 child PCE. If a child PCE is not operational, and an alternate
1551 transit domain is not available, then a failure must be reported.
1552
1553 <span class="h3"><a class="selflink" id="section-7.4" href="#section-7.4">7.4</a>. Verifying Correct Operation</span>
1554
1555 Verifying the correct operation of a parent PCE can be performed by
1556 monitoring a set of parameters. The parent PCE implementation should
1557 provide the following parameters monitored by the parent PCE:
1558
1559
1560
1561
1562
1563
1564 <span class="grey">King &amp; Farrel Informational [Page 28]</span></pre>
1565 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-29" ></span>
1566 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
1567
1568
1569 o number of child PCE requests,
1570
1571 o number of successful hierarchical PCE procedures completions on a
1572 per-PCE-peer basis,
1573
1574 o number of hierarchical PCE procedure completion failures on a per-
1575 PCE-peer basis, and
1576
1577 o number of hierarchical PCE procedure requests from unauthorized
1578 child PCEs.
1579
1580 <span class="h3"><a class="selflink" id="section-7.5" href="#section-7.5">7.5</a>. Impact on Network Operation</span>
1581
1582 The hierarchical PCE procedure is a multiple-PCE path computation
1583 scheme. Subsequent requests to and from the child and parent PCEs do
1584 not differ from other path computation requests and should not have
1585 any significant impact on network operations.
1586
1587 <span class="h2"><a class="selflink" id="section-8" href="#section-8">8</a>. Security Considerations</span>
1588
1589 The hierarchical PCE procedure relies on PCEP and inherits the
1590 security requirements defined in [<a href="./rfc5440" title="&quot;Path Computation Element (PCE) Communication Protocol (PCEP)&quot;">RFC5440</a>]. As noted in <a href="#section-7">Section 7</a>,
1591 there is a security relationship between child and parent PCEs. This
1592 relationship, like any PCEP relationship, assumes pre-configuration
1593 of identities, authority, and keys, or can operate through any key
1594 distribution mechanism outside the scope of PCEP. As PCEP operates
1595 over TCP, it may make use of any TCP security mechanism.
1596
1597 The hierarchical PCE architecture makes use of PCE policy [<a href="./rfc5394" title="&quot;Policy-Enabled Path Computation Framework&quot;">RFC5394</a>]
1598 and the security aspects of the PCE Communication Protocol documented
1599 in [<a href="./rfc5440" title="&quot;Path Computation Element (PCE) Communication Protocol (PCEP)&quot;">RFC5440</a>]. It is expected that the parent PCE will require all
1600 child PCEs to use full security when communicating with the parent
1601 and that security will be maintained by not supporting the discovery
1602 by a parent of child PCEs.
1603
1604 PCE operation also relies on information used to build the TED.
1605 Attacks on a PCE system may be achieved by falsifying or impeding
1606 this flow of information. The child PCE TEDs are constructed as
1607 described in [<a href="./rfc4655" title="&quot;A Path Computation Element (PCE)-Based Architecture&quot;">RFC4655</a>] and are unchanged in this document: if the PCE
1608 listens to the IGP for this information, then normal IGP security
1609 measures may be applied, and it should be noted that an IGP routing
1610 system is generally assumed to be a trusted domain such that router
1611 subversion is not a risk. The parent PCE TED is constructed as
1612 described in this document and may involve:
1613
1614
1615
1616
1617
1618
1619
1620 <span class="grey">King &amp; Farrel Informational [Page 29]</span></pre>
1621 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-30" ></span>
1622 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
1623
1624
1625 - multiple parent-child relationships using PCEP (as already
1626 described)
1627
1628 - the parent PCE listening to child domain IGPs (with the same
1629 security features as a child PCE listening to its IGP)
1630
1631 - an external mechanism (such as [<a href="#ref-BGP-TE" title="&quot;North-Bound Distribution of Link-State and TE Information using BGP&quot;">BGP-TE</a>]), which will need to be
1632 authorized and secured.
1633
1634 Any multi-domain operation necessarily involves the exchange of
1635 information across domain boundaries. This is bound to represent a
1636 significant security and confidentiality risk especially when the
1637 child domains are controlled by different commercial concerns. PCEP
1638 allows individual PCEs to maintain confidentiality of their domain
1639 path information using path-keys [<a href="./rfc5520" title="&quot;Preserving Topology Confidentiality in Inter-Domain Path Computation Using a Path-Key-Based Mechanism&quot;">RFC5520</a>], and the hierarchical PCE
1640 architecture is specifically designed to enable as much isolation of
1641 domain topology and capabilities information as is possible.
1642
1643 For further considerations of the security issues related to inter-AS
1644 path computation, see [<a href="./rfc5376" title="&quot;Inter-AS Requirements for the Path Computation Element Communication Protocol (PCECP)&quot;">RFC5376</a>].
1645
1646 <span class="h2"><a class="selflink" id="section-9" href="#section-9">9</a>. Acknowledgements</span>
1647
1648 The authors would like to thank David Amzallag, Oscar Gonzalez de
1649 Dios, Franz Rambach, Ramon Casellas, Olivier Dugeon, Filippo Cugini,
1650 Dhruv Dhody, and Julien Meuric for their comments and suggestions.
1651
1652 <span class="h2"><a class="selflink" id="section-10" href="#section-10">10</a>. References</span>
1653
1654 <span class="h3"><a class="selflink" id="section-10.1" href="#section-10.1">10.1</a>. Normative References</span>
1655
1656 [<a id="ref-RFC4655">RFC4655</a>] Farrel, A., Vasseur, J.-P., and J. Ash, "A Path
1657 Computation Element (PCE)-Based Architecture", <a href="./rfc4655">RFC</a>
1658 <a href="./rfc4655">4655</a>, August 2006.
1659
1660 [<a id="ref-RFC5152">RFC5152</a>] Vasseur, JP., Ed., Ayyangar, A., Ed., and R. Zhang, "A
1661 Per-Domain Path Computation Method for Establishing
1662 Inter-Domain Traffic Engineering (TE) Label Switched
1663 Paths (LSPs)", <a href="./rfc5152">RFC 5152</a>, February 2008.
1664
1665 [<a id="ref-RFC5394">RFC5394</a>] Bryskin, I., Papadimitriou, D., Berger, L., and J. Ash,
1666 "Policy-Enabled Path Computation Framework", <a href="./rfc5394">RFC 5394</a>,
1667 December 2008.
1668
1669 [<a id="ref-RFC5440">RFC5440</a>] Vasseur, JP., Ed., and JL. Le Roux, Ed., "Path
1670 Computation Element (PCE) Communication Protocol
1671 (PCEP)", <a href="./rfc5440">RFC 5440</a>, March 2009.
1672
1673
1674
1675
1676 <span class="grey">King &amp; Farrel Informational [Page 30]</span></pre>
1677 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-31" ></span>
1678 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
1679
1680
1681 [<a id="ref-RFC5441">RFC5441</a>] Vasseur, JP., Ed., Zhang, R., Bitar, N., and JL. Le
1682 Roux, "A Backward-Recursive PCE-Based Computation
1683 (BRPC) Procedure to Compute Shortest Constrained Inter-
1684 Domain Traffic Engineering Label Switched Paths", <a href="./rfc5441">RFC</a>
1685 <a href="./rfc5441">5441</a>, April 2009.
1686
1687 [<a id="ref-RFC5520">RFC5520</a>] Bradford, R., Ed., Vasseur, JP., and A. Farrel,
1688 "Preserving Topology Confidentiality in Inter-Domain
1689 Path Computation Using a Path-Key-Based Mechanism", <a href="./rfc5520">RFC</a>
1690 <a href="./rfc5520">5520</a>, April 2009.
1691
1692 <span class="h3"><a class="selflink" id="section-10.2" href="#section-10.2">10.2</a>. Informative References</span>
1693
1694 [<a id="ref-RFC4105">RFC4105</a>] Le Roux, J.-L., Ed., Vasseur, J.-P., Ed., and J. Boyle,
1695 Ed., "Requirements for Inter-Area MPLS Traffic
1696 Engineering", <a href="./rfc4105">RFC 4105</a>, June 2005.
1697
1698 [<a id="ref-RFC4216">RFC4216</a>] Zhang, R., Ed., and J.-P. Vasseur, Ed., "MPLS Inter-
1699 Autonomous System (AS) Traffic Engineering (TE)
1700 Requirements", <a href="./rfc4216">RFC 4216</a>, November 2005.
1701
1702 [<a id="ref-RFC4726">RFC4726</a>] Farrel, A., Vasseur, J.-P., and A. Ayyangar, "A
1703 Framework for Inter-Domain Multiprotocol Label
1704 Switching Traffic Engineering", <a href="./rfc4726">RFC 4726</a>, November
1705 2006.
1706
1707 [<a id="ref-RFC5316">RFC5316</a>] Chen, M., Zhang, R., and X. Duan, "ISIS Extensions in
1708 Support of Inter-Autonomous System (AS) MPLS and GMPLS
1709 Traffic Engineering", <a href="./rfc5316">RFC 5316</a>, December 2008.
1710
1711 [<a id="ref-RFC5376">RFC5376</a>] Bitar, N., Zhang, R., and K. Kumaki, "Inter-AS
1712 Requirements for the Path Computation Element
1713 Communication Protocol (PCECP)", <a href="./rfc5376">RFC 5376</a>, November
1714 2008.
1715
1716 [<a id="ref-RFC5392">RFC5392</a>] Chen, M., Zhang, R., and X. Duan, "OSPF Extensions in
1717 Support of Inter-Autonomous System (AS) MPLS and GMPLS
1718 Traffic Engineering", <a href="./rfc5392">RFC 5392</a>, January 2009.
1719
1720 [<a id="ref-RFC5541">RFC5541</a>] Le Roux, JL., Vasseur, JP., and Y. Lee, "Encoding of
1721 Objective Functions in the Path Computation Element
1722 Communication Protocol (PCEP)", <a href="./rfc5541">RFC 5541</a>, June 2009.
1723
1724 [<a id="ref-G-8080">G-8080</a>] ITU-T Recommendation G.8080/Y.1304, Architecture for
1725 the automatically switched optical network (ASON).
1726
1727
1728
1729
1730
1731
1732 <span class="grey">King &amp; Farrel Informational [Page 31]</span></pre>
1733 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-32" ></span>
1734 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
1735
1736
1737 [<a id="ref-G-7715">G-7715</a>] ITU-T Recommendation G.7715 (2002), Architecture and
1738 Requirements for the Automatically Switched Optical
1739 Network (ASON).
1740
1741 [<a id="ref-G-7715-2">G-7715-2</a>] ITU-T Recommendation G.7715.2 (2007), ASON routing
1742 architecture and requirements for remote route query.
1743
1744 [<a id="ref-BGP-TE">BGP-TE</a>] Gredler, H., Medved, J., Previdi, S., Farrel, A., and
1745 S. Ray, "North-Bound Distribution of Link-State and TE
1746 Information using BGP", Work in Progress, October 2012.
1747
1748 [<a id="ref-PCE-AREA-AS">PCE-AREA-AS</a>] King, D., Meuric, J., Dugeon, O., Zhao, Q., Gonzalez de
1749 Dios, O., and F. Chico, "Applicability of the Path
1750 Computation Element to Inter-Area and Inter-AS MPLS and
1751 GMPLS Traffic Engineering", Work in Progress, January
1752 2012.
1753
1754 [<a id="ref-PCEP-MIB">PCEP-MIB</a>] Koushik, A., Emile, S., Zhao, Q., King, D., and J.
1755 Hardwick, "PCE communication protocol (PCEP) Management
1756 Information Base", Work in Progress, July 2012.
1757
1758 <span class="h2"><a class="selflink" id="section-11" href="#section-11">11</a>. Contributors</span>
1759
1760 Quintin Zhao
1761 Huawei Technology
1762 125 Nagog Technology Park
1763 Acton, MA 01719
1764 US
1765
1766 EMail: [email protected]
1767
1768
1769 Fatai Zhang
1770 Huawei Technologies
1771 F3-5-B R&amp;D Center, Huawei Base
1772 Bantian, Longgang District
1773 Shenzhen 518129
1774 P.R. China
1775
1776 EMail: [email protected]
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788 <span class="grey">King &amp; Farrel Informational [Page 32]</span></pre>
1789 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-33" ></span>
1790 <span class="grey"><a href="./rfc6805">RFC 6805</a> PCE Hierarchy Framework November 2012</span>
1791
1792
1793 Authors' Addresses
1794
1795 Daniel King
1796 Old Dog Consulting
1797 UK
1798
1799 EMail: [email protected]
1800
1801
1802 Adrian Farrel
1803 Old Dog Consulting
1804 UK
1805
1806 EMail: [email protected]
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844 King &amp; Farrel Informational [Page 33]
1845 </pre>