idnits 2.17.1 draft-ietf-spring-segment-routing-ldp-interop-13.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- -- The document has examples using IPv4 documentation addresses according to RFC6890, but does not use any IPv6 documentation addresses. Maybe there should be IPv6 examples, too? Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (June 11, 2018) is 2146 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) -- Looks like a reference, but probably isn't: '100' on line 972 -- Looks like a reference, but probably isn't: '300' on line 972 == Outdated reference: A later version (-22) exists of draft-ietf-spring-segment-routing-mpls-13 == Outdated reference: A later version (-25) exists of draft-ietf-isis-segment-routing-extensions-16 == Outdated reference: A later version (-23) exists of draft-ietf-ospf-ospfv3-segment-routing-extensions-11 == Outdated reference: A later version (-27) exists of draft-ietf-ospf-segment-routing-extensions-24 Summary: 0 errors (**), 0 flaws (~~), 5 warnings (==), 4 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group A. Bashandy, Ed. 3 Internet-Draft Individual 4 Intended status: Standards Track C. Filsfils, Ed. 5 Expires: December 13, 2018 S. Previdi 6 Cisco Systems, Inc. 7 B. Decraene 8 S. Litkowski 9 Orange 10 June 11, 2018 12 Segment Routing interworking with LDP 13 draft-ietf-spring-segment-routing-ldp-interop-13 15 Abstract 17 A Segment Routing (SR) node steers a packet through a controlled set 18 of instructions, called segments, by prepending the packet with an SR 19 header. A segment can represent any instruction, topological or 20 service-based. SR allows to enforce a flow through any topological 21 path while maintaining per-flow state only at the ingress node to the 22 SR domain. 24 The Segment Routing architecture can be directly applied to the MPLS 25 data plane with no change in the forwarding plane. This document 26 describes how Segment Routing operates in a network where LDP is 27 deployed and in the case where SR-capable and non-SR-capable nodes 28 coexist. 30 Requirements Language 32 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 33 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 34 document are to be interpreted as described in RFC 2119 [RFC2119]. 36 Status of This Memo 38 This Internet-Draft is submitted in full conformance with the 39 provisions of BCP 78 and BCP 79. 41 Internet-Drafts are working documents of the Internet Engineering 42 Task Force (IETF). Note that other groups may also distribute 43 working documents as Internet-Drafts. The list of current Internet- 44 Drafts is at https://datatracker.ietf.org/drafts/current/. 46 Internet-Drafts are draft documents valid for a maximum of six months 47 and may be updated, replaced, or obsoleted by other documents at any 48 time. It is inappropriate to use Internet-Drafts as reference 49 material or to cite them other than as "work in progress." 51 This Internet-Draft will expire on December 13, 2018. 53 Copyright Notice 55 Copyright (c) 2018 IETF Trust and the persons identified as the 56 document authors. All rights reserved. 58 This document is subject to BCP 78 and the IETF Trust's Legal 59 Provisions Relating to IETF Documents 60 (https://trustee.ietf.org/license-info) in effect on the date of 61 publication of this document. Please review these documents 62 carefully, as they describe your rights and restrictions with respect 63 to this document. Code Components extracted from this document must 64 include Simplified BSD License text as described in Section 4.e of 65 the Trust Legal Provisions and are provided without warranty as 66 described in the Simplified BSD License. 68 Table of Contents 70 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 71 2. SR/LDP Ships-in-the-night coexistence . . . . . . . . . . . . 3 72 2.1. MPLS2MPLS, MPLS2IP and IP2MPLS co-existence . . . . . . . 5 73 3. SR and LDP Interworking . . . . . . . . . . . . . . . . . . . 6 74 3.1. LDP to SR . . . . . . . . . . . . . . . . . . . . . . . . 6 75 3.1.1. LDP to SR Behavior . . . . . . . . . . . . . . . . . 7 76 3.2. SR to LDP . . . . . . . . . . . . . . . . . . . . . . . . 7 77 3.2.1. Segment Routing Mapping Server (SRMS) . . . . . . . . 9 78 3.2.2. SR to LDP Behavior . . . . . . . . . . . . . . . . . 10 79 3.2.3. Interoperability of Multiple SRMSes and Prefix-SID 80 advertisements . . . . . . . . . . . . . . . . . . . 11 81 4. SR/LDP Interworking Use Cases . . . . . . . . . . . . . . . . 12 82 4.1. SR Protection of LDP-based Traffic . . . . . . . . . . . 12 83 4.2. Eliminating Targeted LDP Session . . . . . . . . . . . . 14 84 4.3. Guaranteed FRR coverage . . . . . . . . . . . . . . . . . 15 85 4.4. Inter-AS Option C, Carrier's Carrier . . . . . . . . . . 16 86 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 17 87 6. Manageability Considerations . . . . . . . . . . . . . . . . 17 88 6.1. SR and LDP co-existence . . . . . . . . . . . . . . . . . 17 89 6.2. Dataplane Verification . . . . . . . . . . . . . . . . . 17 90 7. Security Considerations . . . . . . . . . . . . . . . . . . . 17 91 8. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 18 92 9. Contributors' Addresses . . . . . . . . . . . . . . . . . . . 18 93 10. References . . . . . . . . . . . . . . . . . . . . . . . . . 19 94 10.1. Normative References . . . . . . . . . . . . . . . . . . 19 95 10.2. Informative References . . . . . . . . . . . . . . . . . 20 97 Appendix A. Migration from LDP to SR . . . . . . . . . . . . . . 21 98 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 22 100 1. Introduction 102 Segment Routing, as described in [I-D.ietf-spring-segment-routing], 103 can be used on top of the MPLS data plane without any modification as 104 described in [I-D.ietf-spring-segment-routing-mpls]. 106 Segment Routing control plane can co-exist with current label 107 distribution protocols such as LDP ([RFC5036]). 109 This document outlines the mechanisms through which SR interworks 110 with LDP in cases where a mix of SR-capable and non-SR-capable 111 routers co-exist within the same network and more precisely in the 112 same routing domain. 114 Section 2 describes the co-existence of SR with other MPLS Control 115 Plane. Section 3 documents the interworking between SR and LDP in 116 the case of non-homogeneous deployment. Section 4 describes how a 117 partial SR deployment can be used to provide SR benefits to LDP-based 118 traffic including a possible application of SR in the context of 119 inter-domain MPLS use-cases. Appendix A documents a method to 120 migrate from LDP to SR-based MPLS tunneling. 122 Typically, an implementation will allow an operator to select 123 (through configuration) which of the described modes of SR and LDP 124 co-existence to use. 126 2. SR/LDP Ships-in-the-night coexistence 128 "MPLS Control Plane Client (MCC)" refers to any control plane 129 protocol installing forwarding entries in the MPLS data plane. SR, 130 LDP [RFC5036], RSVP-TE [RFC3209], BGP [RFC8277], etc are examples of 131 MCCs. 133 An MCC, operating at node N, must ensure that the incoming label it 134 installs in the MPLS data plane of Node N has been uniquely allocated 135 to himself. 137 Segment Routing makes use of the Segment Routing Global Block (SRGB, 138 as defined in [I-D.ietf-spring-segment-routing]) for the label 139 allocation. The use of the SRGB allows SR to co-exist with any other 140 MCC. 142 This is clearly the case for the adjacency segment: it is a local 143 label allocated by the label manager, as for any MCC. 145 This is clearly the case for the prefix segment: the label manager 146 allocates the SRGB set of labels to the SR MCC client and the 147 operator ensures the unique allocation of each global prefix segment/ 148 label within the allocated SRGB set. 150 Note that this static label allocation capability of the label 151 manager exists for many years across several vendors and hence is not 152 new. Furthermore, note that the label-manager ability to statically 153 allocate a range of labels to a specific application is not new 154 either. This is required for MPLS-TP operation. In this case, the 155 range is reserved by the label manager and it is the MPLS-TP 156 ([RFC5960]) NMS (acting as an MCC) that ensures the unique allocation 157 of any label within the allocated range and the creation of the 158 related MPLS forwarding entry. 160 Let us illustrate an example of ship-in-the-night (SIN) coexistence. 162 PE2 PE4 163 \ / 164 PE1----A----B---C---PE3 166 Figure 1: SIN coexistence 168 The EVEN VPN service is supported by PE2 and PE4 while the ODD VPN 169 service is supported by PE1 and PE3. The operator wants to tunnel 170 the ODD service via LDP and the EVEN service via SR. 172 This can be achieved in the following manner: 174 The operator configures PE1, PE2, PE3, PE4 with respective 175 loopbacks 192.0.2.201/32, 192.0.2.202/32, 192.0.2.203/32, 176 192.0.2.204/32. These PE's advertised their VPN routes with next- 177 hop set on their respective loopback address. 179 The operator configures A, B, C with respective loopbacks 180 192.0.2.1/32, 192.0.2.2/32, 192.0.2.3/32. 182 The operator configures PE2, A, B, C and PE4 with SRGB [100, 300]. 184 The operator attaches the respective Node Segment Identifiers 185 (Node-SID's, as defined in [I-D.ietf-spring-segment-routing]): 186 202, 101, 102, 103 and 204 to the loopbacks of nodes PE2, A, B, C 187 and PE4. The Node-SID's are configured to request penultimate- 188 hop-popping. 190 PE1, A, B, C and PE3 are LDP capable. 192 PE1 and PE3 are not SR capable. 194 PE3 sends an ODD VPN route to PE1 with next-hop 192.0.2.203 and VPN 195 label 10001. 197 From an LDP viewpoint: PE1 received an LDP label binding (1037) for 198 FEC 192.0.2.203/32 from its next-hop A. A received an LDP label 199 binding (2048) for that FEC from its next-hop B. B received an LDP 200 label binding (3059) for that FEC from its next-hop C. C received 201 implicit-null LDP binding from its next-hop PE3. 203 As a result, PE1 sends its traffic to the ODD service route 204 advertised by PE3 to next-hop A with two labels: the top label is 205 1037 and the bottom label is 10001. Node A swaps 1037 with 2048 and 206 forwards to B. B swaps 2048 with 3059 and forwards to C. C pops 207 3059 and forwards to PE3. 209 PE4 sends an EVEN VPN route to PE2 with next-hop 192.0.2.204 and VPN 210 label 10002. 212 From an SR viewpoint: PE2 maps the IGP route 192.0.2.204/32 onto 213 Node-SID 204; node A swaps 204 with 204 and forwards to B; B swaps 214 204 with 204 and forwards to C; C pops 204 and forwards to PE4. 216 As a result, PE2 sends its traffic to the VPN service route 217 advertised by PE4 to next-hop A with two labels: the top label is 204 218 and the bottom label is 10002. Node A swaps 204 with 204 and 219 forwards to B. B swaps 204 with 204 and forwards to C. C pops 204 220 and forwards to PE4. 222 The two modes of MPLS tunneling co-exist. 224 The ODD service is tunneled from PE1 to PE3 through a continuous 225 LDP LSP traversing A, B and C. 227 The EVEN service is tunneled from PE2 to PE4 through a continuous 228 SR node segment traversing A, B and C. 230 2.1. MPLS2MPLS, MPLS2IP and IP2MPLS co-existence 232 MPLS2MPLS refers the forwarding behavior where a router receives an 233 labeled packet and switches it out as a labeled packet. Several 234 MPLS2MPLS entries may be installed in the data plane for the same 235 prefix. 237 Let us examine A's MPLS forwarding table as an example: 239 Incoming label: 1037 240 - outgoing label: 2048 241 - outgoing next-hop: B 242 Note: this entry is programmed by LDP for 192.0.2.203/32 244 Incoming label: 203 246 - outgoing label: 203 247 - outgoing next-hop: B 248 Note: this entry is programmed by SR for 192.0.2.203/32 250 These two entries can co-exist because their incoming label is 251 unique. The uniqueness is guaranteed by the label manager allocation 252 rules. 254 The same applies for the MPLS2IP forwarding entries. MPLS2IP is the 255 forwarding behavior where a router receives a label IPv4/IPv6 packet 256 with one label only, pops the label, and switches the packet out as 257 IPv4/IPv6. For IP2MPLS coexistence, refer to Section 6.1. 259 3. SR and LDP Interworking 261 This section analyzes the case where SR is available in one part of 262 the network and LDP is available in another part. It describes how a 263 continuous MPLS tunnel can be built throughout the network. 265 PE2 PE4 266 \ / 267 PE1----P5--P6--P7--P8---PE3 269 Figure 2: SR and LDP Interworking 271 Let us analyze the following example: 273 P6, P7, P8, PE4 and PE3 are LDP capable. 275 PE1, PE2, P5 and P6 are SR capable. PE1, PE2, P5 and P6 are 276 configured with SRGB (100, 200) and respectively with node 277 segments 101, 102, 105 and 106. 279 A service flow must be tunneled from PE1 to PE3 over a continuous 280 MPLS tunnel encapsulation and hence SR and LDP need to interwork. 282 3.1. LDP to SR 284 In this section, a right-to-left traffic flow is analyzed. 286 PE3 has learned a service route whose next-hop is PE1. PE3 has an 287 LDP label binding from the next-hop P8 for the FEC "PE1". Hence PE3 288 sends its service packet to P8 as per classic LDP behavior. 290 P8 has an LDP label binding from its next-hop P7 for the FEC "PE1" 291 and hence P8 forwards to P7 as per classic LDP behavior. 293 P7 has an LDP label binding from its next-hop P6 for the FEC "PE1" 294 and hence P7 forwards to P6 as per classic LDP behavior. 296 P6 does not have an LDP binding from its next-hop P5 for the FEC 297 "PE1". However P6 has an SR node segment to the IGP route "PE1". 298 Hence, P6 forwards the packet to P5 and swaps its local LDP-label for 299 FEC "PE1" by the equivalent node segment (i.e. 101). 301 P5 pops 101 (assuming PE1 advertised its node segment 101 with the 302 penultimate-pop flag set) and forwards to PE1. 304 PE1 receives the tunneled packet and processes the service label. 306 The end-to-end MPLS tunnel is built from an LDP LSP from PE3 to P6 307 and the related node segment from P6 to PE1. 309 3.1.1. LDP to SR Behavior 311 It has to be noted that no additional signaling or state is required 312 in order to provide interworking in the direction LDP to SR. 314 A SR node having LDP neighbors MUST create LDP bindings for each 315 Prefix-SID learned in the SR domain by treating SR learned labels as 316 if they were learned through an LDP neighbot. In addition for each 317 FEC, the SR node stitches the incoming LDP label to the outgoing SR 318 label. This has to be done in both LDP independent and ordered label 319 distribution control modes as defined in [RFC5036]. 321 3.2. SR to LDP 323 In this section, the left-to-right traffic flow is analyzed. 325 This section defines the Segment Routing Mapping Server (SRMS). The 326 SRMS is a IGP node advertising mapping between Segment Identifiers 327 (SID) and prefixes advertised by other IGP nodes. The SRMS uses a 328 dedicated IGP extension (IS-IS, OSPF and OSPFv3) which is protocol 329 specific and defined in [I-D.ietf-isis-segment-routing-extensions], 330 [I-D.ietf-ospf-segment-routing-extensions], and 331 [I-D.ietf-ospf-ospfv3-segment-routing-extensions]. 333 The SRMS function of a SR capable router allows distribution of 334 mappings for prefixes not locally attached to the advertising router 335 and therefore allows advertisement of mappings on behalf of non-SR 336 capable routers. 338 The SRMS is a control plane only function which may be located 339 anywhere in the IGP flooding scope. At least one SRMS server MUST 340 exist in a routing domain to advertise prefix-SIDs on behalf non-SR 341 nodes, thereby allowing non-LDP routers to send and receive labeled 342 traffic from LDP-only routers. Multiple SRMSs may be present in the 343 same network (for redundancy). This implies that there are multiple 344 ways a prefix-to-SID mapping can be advertised. Conflicts resulting 345 from inconsistent advertisements are addressed by 346 [I-D.ietf-spring-segment-routing-mpls]. 348 The example diagram depicted in Figure 3 assumes that the operator 349 configures P5 to act as a Segment Routing Mapping Server (SRMS) and 350 advertises the following mappings: (P7, 107), (P8, 108), (PE3, 103) 351 and (PE4, 104). 353 The mappings advertised by one or more SRMSs result from local policy 354 information configured by the operator. 356 If PE3 had been SR capable, the operator would have configured PE3 357 with node segment 103. Instead, as PE3 is not SR capable, the 358 operator configures that policy at the SRMS and it is the latter 359 which advertises the mapping. 361 The mapping server advertisements are only understood by SR capable 362 routers. The SR capable routers install the related node segments in 363 the MPLS data plane exactly like the node segments had been 364 advertised by the nodes themselves. 366 For example, PE1 installs the node segment 103 with next-hop P5 367 exactly as if PE3 had advertised node segment 103. 369 PE1 has a service route whose next-hop is PE3. PE1 has a node 370 segment for that IGP route: 103 with next-hop P5. Hence PE1 sends 371 its service packet to P5 with two labels: the bottom label is the 372 service label and the top label is 103. 374 P5 swaps 103 for 103 and forwards to P6. 376 P6's next-hop for the IGP route "PE3" is not SR capable (P7 does not 377 advertise the SR capability). However, P6 has an LDP label binding 378 from that next-hop for the same FEC (e.g. LDP label 1037). Hence, 379 P6 swaps 103 for 1037 and forwards to P7. 381 P7 swaps this label with the LDP-label received from P8 and forwards 382 to P8. 384 P8 pops the LDP label and forwards to PE3. 386 PE3 receives the tunneled packet and processes the service label. 388 The end-to-end MPLS tunnel is built from an SR node segment from PE1 389 to P6 and an LDP LSP from P6 to PE3. 391 SR mapping advertisement for a given prefix provides no information 392 about the Penultimate Hop Popping. Other mechanisms, such as IGP 393 specific mechanisms ([I-D.ietf-isis-segment-routing-extensions], 394 [I-D.ietf-ospf-segment-routing-extensions] and 395 [I-D.ietf-ospf-ospfv3-segment-routing-extensions]), MAY be used to 396 determine the Penultimate Hop Popping in such case. 398 Note: In the previous example, Penultimate Hop Popping is not 399 performed at the SR/LDP border for segment 103 (PE3), because none of 400 the routers in the SR domain is Penultimate Hop for segment 103. In 401 this case P6 requires the presence of the segment 103 such as to map 402 it to the LDP label 1037. 404 3.2.1. Segment Routing Mapping Server (SRMS) 406 This section specifies the concept and externally visible 407 functionality of a segment routing mapping server (SRMS). 409 The purpose of a SRMS functionality is to support the advertisement 410 of prefix-SIDs to a prefix without the need to explicitly advertise 411 such assignment within a prefix reachability advertisment. Examples 412 of explicit prefix-SID advertisment are the prefix-SID sub-TLVs 413 defined in ([I-D.ietf-isis-segment-routing-extensions], 414 [I-D.ietf-ospf-segment-routing-extensions], and 415 [I-D.ietf-ospf-ospfv3-segment-routing-extensions]). 417 The SRMS functionality allows assigning of prefix-SIDs to prefixes 418 owned by non-SR-capable routers as well as to prefixes owned by SR 419 capable nodes. It is the former capability which is essential to the 420 SR-LDP interworking described later in this section 422 The SRMS functionality consists of two functional blocks: the Mapping 423 Server (MS) and Mapping Client (MC). 425 A MS is a node that advertises an SR mappings. Advertisements sent 426 by an MS define the assignment of a prefix-SID to a prefix 427 independent of the advertisment of reachability to the prefix itself. 428 An MS MAY advertise SR mappings for any prefix whether or not it 429 advertises reachability for the prefix and irrespective of whether 430 that prefix is advertised by or even reachable through any router in 431 the network. 433 An MC is a node that receives and uses the MS mapping advertisments. 434 Note that a node may be both an MS and an MC. An MC interprets the 435 SR mapping advertisment as an assignment of a prefix-SID to a prefix. 436 For a given prefix, if an MC receives an SR mapping advertisement 437 from a mapping server and also has received a prefix-SID 438 advertisement for that same prefix in a prefix reachability 439 advertisement, then the MC MUST prefer the SID advertised in the 440 prefix reachability advertisement over the mapping server 441 advertisement i.e., the mapping server advertisment MUST be ignored 442 for that prefix. Hence assigning a prefix-SID to a prefix using the 443 SRMS functionality does not preclude assigning the same or different 444 prefix-SID(s) to the same prefix using explict prefix-SID 445 advertisement such as the aforementioned prefix-SID sub-TLV. 447 For example consider an IPv4 prefix advertisement received by an IS- 448 IS router in TLV 135. Suppose TLV 135 contained the prefix-SID sub- 449 TLV. If the router that receives TLV 135 with the prefix-SID sub-TLV 450 also received an SR mapping advertisement for the same prefix through 451 the SID/label binding TLV, then the receiving router must prefer the 452 prefix-SID sub-TLV over the SID/label binding TLV for that prefix. 453 Refer to ([I-D.ietf-isis-segment-routing-extensions], for details 454 about the prefix-SID sub-TLV and SID/label binding TLV. 456 3.2.2. SR to LDP Behavior 458 SR to LDP interworking requires a SRMS as defined above. 460 Each SR capable router installs in the MPLS data plane Node-SIDs 461 learned from the SRMS exactly like if these SIDs had been advertised 462 by the nodes themselves. 464 A SR node having LDP neighbors MUST stitch the incoming SR label 465 (whose SID is advertised by the SRMS) to the outgoing LDP label. 467 It has to be noted that the SR to LDP behavior does not propagate the 468 status of the LDP FEC which was signaled if LDP was configured to use 469 the ordered mode. 471 It has to be noted that in the case of SR to LDP, the label binding 472 is equivalent to the independent LDP Label Distribution Control Mode 473 ([RFC5036]) where a label in bound to a FEC independently from the 474 received binding for the same FEC. 476 3.2.3. Interoperability of Multiple SRMSes and Prefix-SID 477 advertisements 479 In the case of SR/LDP interoperability through the use of a SRMS, 480 mappings are advertised by one or more SRMS. 482 SRMS function is implemented in the link-state protocol (such as IS- 483 IS and OSPF). Link-state protocols allow propagation of updates 484 across area boundaries and therefore SRMS advertisements are 485 propagated through the usual inter-area advertisement procedures in 486 link-state protocols. 488 Multiple SRMSs can be provisioned in a network for redundancy. 489 Moreover, a preference mechanism may also be used among SRMSs so to 490 deploy a primary/secondary SRMS scheme allowing controlled 491 modification or migration of SIDs. 493 The content of SRMS advertisement (i.e.: mappings) are a matter of 494 local policy determined by the operator. When multiple SRMSs are 495 active, it is necessary that the information (mappings) advertised by 496 the different SRMSs is aligned and consistent. The following 497 mechanism is applied to determine the preference of SRMS 498 advertisements: 500 If a node acts as an SRMS, it MAY advertise a preference to be 501 associated with all SRMS SID advertisements sent by that node. The 502 means of advertising the preference is defined in the protocol 503 specific drafts e.g.,[I-D.ietf-isis-segment-routing-extensions] , 504 [I-D.ietf-ospf-segment-routing-extensions], and 505 [I-D.ietf-ospf-ospfv3-segment-routing-extensions]. The preference 506 value is an unsigned 8 bit integer with the following properties: 508 0 - Reserved value indicating advertisements from that node MUST 509 NOT be used. 511 1 - 255 Preference value 513 Advertisement of a preference value is optional. Nodes which do not 514 advertise a preference value are assigned a preference value of 128. 516 A MCC on a node receiving one or more SRMS mapping advertisements 517 applies them as follows 519 - For any prefix for which it did not receive a prefix-SID 520 advertisement, the MCC applies the SRMS mapping advertisments with 521 the highest preference. The mechanism by which a prefix-SID is 522 advertised for a given prefix is defined in the protocol 523 specification , [I-D.ietf-isis-segment-routing-extensions], 525 [I-D.ietf-ospf-segment-routing-extensions] and 526 [I-D.ietf-ospf-ospfv3-segment-routing-extensions] 528 - If there is an incoming label collision as specified in 529 [I-D.ietf-spring-segment-routing-mpls] , apply the steps specified 530 in [I-D.ietf-spring-segment-routing-mpls] to resolve the 531 collision. 533 When the SRMS advertise mappings, an implementation should provide a 534 mechanism through which the operator determines which of the IP2MPLS 535 mappings are preferred among the one advertised by the SRMS and the 536 ones advertised by LDP. 538 4. SR/LDP Interworking Use Cases 540 SR can be deployed such as to enhance LDP transport. The SR 541 deployment can be limited to the network region where the SR benefits 542 are most desired. 544 4.1. SR Protection of LDP-based Traffic 546 In Figure 4, let us assume: 548 All link costs are 10 except FG which is 30. 550 All routers are LDP capable. 552 X, Y and Z are PE's participating to an important service S. 554 The operator requires 50msec link-based Fast Reroute (FRR) for 555 service S. 557 A, B, C, D, E, F and G are SR capable. 559 X, Y, Z are not SR capable, e.g. as part of a staged migration 560 from LDP to SR, the operator deploys SR first in a sub-part of the 561 network and then everywhere. 563 X 564 | 565 Y--A---B---E--Z 566 | | \ 567 D---C--F--G 568 30 570 Figure 3: SR/LDP interworking example 572 The operator would like to resolve the following issues: 574 To protect the link BA along the shortest-path of the important 575 flow XY, B requires a Remote LFA (RLFA, [RFC7490]) repair tunnel 576 to D and hence a targeted LDP session from B to D. Typically, 577 network operators prefer avoiding these dynamically established 578 multi-hop LDP sessions in order to reduce the number of protocols 579 running in the network and hence simplify network operations. 581 There is no LFA/RLFA solution to protect the link BE along the 582 shortest path of the important flow XZ. The operator wants a 583 guaranteed link-based FRR solution. 585 The operator can meet these objectives by deploying SR only on A, B, 586 C, D, E, F and G: 588 The operator configures A, B, C, D, E, F and G with SRGB (100, 589 200) and respective node segments 101, 102, 103, 104, 105, 106 and 590 107. 592 The operator configures D as an SR Mapping Server with the 593 following policy mapping: (X, 201), (Y, 202), (Z, 203). 595 Each SR node automatically advertises local adjacency segment for 596 its IGP adjacencies. Specifically, F advertises adjacency segment 597 9001 for its adjacency FG. 599 A, B, C, D, E, F and G keep their LDP capability and hence the flows 600 XY and XZ are transported over end-to-end LDP LSP's. 602 For example, LDP at B installs the following MPLS data plane entries: 604 Incoming label: local LDP label bound by B for FEC Y 605 Outgoing label: LDP label bound by A for FEC Y 606 Outgoing next-hop: A 608 Incoming label: local LDP label bound by B for FEC Z 609 Outgoing label: LDP label bound by E for FEC Z 610 Outgoing next-hop: E 612 The novelty comes from how the backup chains are computed for these 613 LDP-based entries. While LDP labels are used for the primary next- 614 hop and outgoing labels, SR information is used for the FRR 615 construction. In steady state, the traffic is transported over LDP 616 LSP. In transient FRR state, the traffic is backup thanks to the SR 617 enhanced capabilities. 619 The RLFA paths are dynamically pre-computed as defined in [RFC7490]. 620 Typically, implementations allow to enable RLFA mechanism through a 621 simple configuration command that triggers both the pre-computation 622 and installation of the repair path. The details on how RLFA 623 mechanisms are implemented and configured is outside the scope of 624 this document and not relevant to the aspects of SR/LDP interwork 625 explained in this document. 627 This helps meet the requirements of the operator: 629 Eliminate targeted LDP session. 631 Guaranteed FRR coverage. 633 Keep the traffic over LDP LSP in steady state. 635 Partial SR deployment only where needed. 637 4.2. Eliminating Targeted LDP Session 639 B's MPLS entry to Y becomes: 641 - Incoming label: local LDP label bound by B for FEC Y 642 Outgoing label: LDP label bound by A for FEC Y 643 Backup outgoing label: SR node segment for Y {202} 644 Outgoing next-hop: A 645 Backup next-hop: repair tunnel: node segment to D {104} 646 with outgoing next-hop: C 648 It has to be noted that D is selected as Remote Loop-Free Alternate 649 (RLFA) as defined in [RFC7490]. 651 In steady-state, X sends its Y-destined traffic to B with a top label 652 which is the LDP label bound by B for FEC Y. B swaps that top label 653 for the LDP label bound by A for FEC Y and forwards to A. A pops the 654 LDP label and forwards to Y. 656 Upon failure of the link BA, B swaps the incoming top-label with the 657 node segment for Y (202) and sends the packet onto a repair tunnel to 658 D (node segment 104). Thus, B sends the packet to C with the label 659 stack {104, 202}. C pops the node segment 104 and forwards to D. D 660 swaps 202 for 202 and forwards to A. A's next-hop to Y is not SR 661 capable and hence node A swaps the incoming node segment 202 to the 662 LDP label announced by its next-hop (in this case, implicit null). 664 After IGP convergence, B's MPLS entry to Y will become: 666 - Incoming label: local LDP label bound by B for FEC Y 667 Outgoing label: LDP label bound by C for FEC Y 668 Outgoing next-hop: C 670 And the traffic XY travels again over the LDP LSP. 672 Conclusion: the operator has eliminated the need for targeted LDP 673 sessions (no longer required) and the steady-state traffic is still 674 transported over LDP. The SR deployment is confined to the area 675 where these benefits are required. 677 Despite that in general, an implementation would not require a manual 678 configuration of LDP Targeted sessions however, it is always a gain 679 if the operator is able to reduce the set of protocol sessions 680 running on the network infrastructure. 682 4.3. Guaranteed FRR coverage 684 As mentioned in Section 4.1 above, in the example topology described 685 in Figure 4, there is no RLFA-based solution for protecting the 686 traffic flow YZ against the failure of link BE because there is no 687 intersection between the extended P-space and Q-space (see [RFC7490] 688 for details). However: 690 - G belongs to the Q space of Z. 692 - G can be reached from B via a "repair SR path" {106, 9001} that is 693 not affected by failure of link BE (The method by which G and the 694 repair tunnel to it from B are identified are out of scope of this 695 document.) 697 B's MPLS entry to Z becomes: 699 - Incoming label: local LDP label bound by B for FEC Z 700 Outgoing label: LDP label bound by E for FEC Z 701 Backup outgoing label: SR node segment for Z {203} 702 Outgoing next-hop: E 703 Backup next-hop: repair tunnel to G: {106, 9001} 705 G is reachable from B via the combination of a 706 node segment to F {106} and an adjacency segment 707 FG {9001} 709 Note that {106, 107} would have equally work. 710 Indeed, in many case, P's shortest path to Q is 711 over the link PQ. The adjacency segment from P to 712 Q is required only in very rare topologies where 713 the shortest-path from P to Q is not via the link 714 PQ. 716 In steady-state, X sends its Z-destined traffic to B with a top label 717 which is the LDP label bound by B for FEC Z. B swaps that top label 718 for the LDP label bound by E for FEC Z and forwards to E. E pops the 719 LDP label and forwards to Z. 721 Upon failure of the link BE, B swaps the incoming top-label with the 722 node segment for Z (203) and sends the packet onto a repair tunnel to 723 G (node segment 106 followed by adjacency segment 9001). Thus, B 724 sends the packet to C with the label stack {106, 9001, 203}. C pops 725 the node segment 106 and forwards to F. F pops the adjacency segment 726 9001 and forwards to G. G swaps 203 for 203 and forwards to E. E's 727 next-hop to Z is not SR capable and hence E swaps the incoming node 728 segment 203 for the LDP label announced by its next-hop (in this 729 case, implicit null). 731 After IGP convergence, B's MPLS entry to Z will become: 733 - Incoming label: local LDP label bound by B for FEC Z 734 Outgoing label: LDP label bound by C for FEC Z 735 Outgoing next-hop: C 737 And the traffic XZ travels again over the LDP LSP. 739 Conclusions: 741 - the operator has eliminated its second problem: guaranteed FRR 742 coverage is provided. The steady-state traffic is still 743 transported over LDP. The SR deployment is confined to the area 744 where these benefits are required. 746 - FRR coverage has been achieved without any signaling for setting 747 up the repair LSP and without setting up a targeted LDP session 748 between B and G. 750 4.4. Inter-AS Option C, Carrier's Carrier 752 In inter-AS Option C [RFC4364], two interconnected ASes sets up 753 inter-AS MPLS connectivity. SR may be independently deployed in each 754 AS. 756 PE1---R1---B1---B2---R2---PE2 757 <-----------> <-----------> 758 AS1 AS2 760 Figure 4: Inter-AS Option C 762 In Inter-AS Option C, B2 advertises to B1 a labeled BGP route 763 [RFC8277] for PE2 and B1 reflects it to its internal peers, such as 764 PE1. PE1 learns from a service route reflector a service route whose 765 next-hop is PE2. PE1 resolves that service route on the labeled BGP 766 route to PE2. That labeled BGP route to PE2 is itself resolved on 767 the AS1 IGP route to B1. 769 If AS1 operates SR, then the tunnel from PE1 to B1 is provided by the 770 node segment from PE1 to B1. 772 PE1 sends a service packet with three labels: the top one is the node 773 segment to B1, the next-one is the label in the labeled BGP route 774 provided by B1 for the route "PE2" and the bottom one is the service 775 label allocated by PE2. 777 5. IANA Considerations 779 This document does not introduce any new codepoint. 781 6. Manageability Considerations 783 6.1. SR and LDP co-existence 785 When both SR and LDP co-exist, the following applies: 787 - If both SR and LDP propose an IP2MPLS entry for the same IP 788 prefix, then by default the LDP route SHOULD be selected. This is 789 because it is expected that SR is introduced into network that 790 contain routers that do not support SR. Hence by having a 791 behavior that prefers LDP over SR, traffic flow is unlikely to be 792 disrupted 794 - A local policy on a router MUST allow to prefer the SR-provided 795 IP2MPLS entry. 797 - Note that this policy MAY be locally defined. There is no 798 requirement that all routers use the same policy. 800 6.2. Dataplane Verification 802 When Label switch paths (LSPs) are defined by stitching LDP LSPs with 803 SR LSPs, it is necessary to have mechanisms allowing the verification 804 of the LSP connectivity as well as validation of the path. These 805 mechanisms are described in [RFC8287]. 807 7. Security Considerations 809 This document does not introduce any change to the MPLS dataplane 810 [RFC3031] and therefore no additional security of the MPLS dataplane 811 is required. 813 This document introduces another form of label binding 814 advertisements. The security associated with these advertisement is 815 part of the security applied to routing protocols such as IS-IS 816 [RFC5304] and OSPF [RFC5709] which both make use of cryptographic 817 authentication mechanisms. This document also specifies a mechanism 818 by which the ill effects of advertising conflicting label bindings 819 can be mitigated. Because this document recognizes that 820 miscofiguration and/or programming may result in false or conflicting 821 label binding advertisements, thereby compromising traffic 822 forwarding, the document recommends strict configuration/ 823 programmability control as well as montoring the SID advertised and 824 log/error messages by the operator to avoid or at least significantly 825 minimize the possibility of such risk. 827 8. Acknowledgements 829 The authors would like to thank Pierre Francois, Ruediger Geib and 830 Alexander Vainshtein for their contribution to the content of this 831 document. 833 9. Contributors' Addresses 834 Edward Crabbe 835 Individual 836 Email: edward.crabbe@gmail.com 838 Igor Milojevic 839 Email: milojevicigor@gmail.com 841 Saku Ytti 842 TDC 843 Email: saku@ytti.fi 845 Rob Shakir 846 Google 847 Email: robjs@google.com 849 Martin Horneffer 850 Deutsche Telekom 851 Email: Martin.Horneffer@telekom.de 853 Wim Henderickx 854 Nokia 855 Email: wim.henderickx@nokia.com 857 Jeff Tantsura 858 Individual 859 Email: jefftant@gmail.com 861 Les Ginseberg 862 Cisco Systems 863 Email: ginsberg@cisco.com 865 10. References 867 10.1. Normative References 869 [I-D.ietf-spring-segment-routing] 870 Filsfils, C., Previdi, S., Decraene, B., Litkowski, S., 871 and R. Shakir, "Segment Routing Architecture", January 872 2018. 874 [I-D.ietf-spring-segment-routing-mpls] 875 Bashandy, A., Filsfils, C., Previdi, S., Decraene, B., 876 Litkowski, S., and R. Shakir, "Segment Routing with MPLS 877 data plane", draft-ietf-spring-segment-routing-mpls-13 878 (work in progress), April 2018. 880 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 881 Requirement Levels", BCP 14, RFC 2119, 882 DOI 10.17487/RFC2119, March 1997, 883 . 885 [RFC5036] Andersson, L., Ed., Minei, I., Ed., and B. Thomas, Ed., 886 "LDP Specification", RFC 5036, DOI 10.17487/RFC5036, 887 October 2007, . 889 10.2. Informative References 891 [I-D.ietf-isis-segment-routing-extensions] 892 Previdi, S., Ginsberg, L., Filsfils, C., Bashandy, A., 893 Gredler, H., Litkowski, S., Decraene, B., and J. Tantsura, 894 "IS-IS Extensions for Segment Routing", draft-ietf-isis- 895 segment-routing-extensions-16 (work in progress), April 896 2018. 898 [I-D.ietf-ospf-ospfv3-segment-routing-extensions] 899 Psenak, P., Filsfils, C., Previdi, S., Gredler, H., 900 Shakir, R., Henderickx, W., and J. Tantsura, "OSPFv3 901 Extensions for Segment Routing", draft-ietf-ospf-ospfv3- 902 segment-routing-extensions-11 (work in progress), January 903 2018. 905 [I-D.ietf-ospf-segment-routing-extensions] 906 Psenak, P., Previdi, S., Filsfils, C., Gredler, H., 907 Shakir, R., Henderickx, W., and J. Tantsura, "OSPF 908 Extensions for Segment Routing", draft-ietf-ospf-segment- 909 routing-extensions-24 (work in progress), December 2017. 911 [RFC3031] Rosen, E., Viswanathan, A., and R. Callon, "Multiprotocol 912 Label Switching Architecture", RFC 3031, 913 DOI 10.17487/RFC3031, January 2001, 914 . 916 [RFC3209] Awduche, D., Berger, L., Gan, G., Li, T., Srinivasan, V., 917 and G. Srinivasan, "RSVP-TE: Extensions to RSVP for LSP 918 Tunnels", December 2001. 920 [RFC4364] Rosen, E. and Y. Rekhter, "BGP/MPLS IP Virtual Private 921 Networks (VPNs)", RFC 4364, DOI 10.17487/RFC4364, February 922 2006, . 924 [RFC5304] Li, T. and R. Atkinson, "IS-IS Cryptographic 925 Authentication", RFC 5304, DOI 10.17487/RFC5304, October 926 2008, . 928 [RFC5709] Bhatia, M., Manral, V., Fanto, M., White, R., Barnes, M., 929 Li, T., and R. Atkinson, "OSPFv2 HMAC-SHA Cryptographic 930 Authentication", RFC 5709, DOI 10.17487/RFC5709, October 931 2009, . 933 [RFC5960] Frost, D., Ed., Bryant, S., Ed., and M. Bocci, Ed., "MPLS 934 Transport Profile Data Plane Architecture", RFC 5960, 935 DOI 10.17487/RFC5960, August 2010, 936 . 938 [RFC7490] Bryant, S., Filsfils, C., Previdi, S., Shand, M., and N. 939 So, "Remote Loop-Free Alternate (LFA) Fast Reroute (FRR)", 940 RFC 7490, DOI 10.17487/RFC7490, April 2015, 941 . 943 [RFC8277] Rosen, E., "Using BGP to Bind MPLS Labels to Address 944 Prefixes", October 2017. 946 [RFC8287] Kumar, N., Pignataro, C., Swallow, G., Akiya, N., Kini, 947 S., and M. Chen, "Label Switched Path (LSP) Ping/ 948 Traceroute for Segment Routing (SR) IGP-Prefix and IGP- 949 Adjacency Segment Identifiers (SIDs) with MPLS Data 950 Planes", December 2017. 952 [RFC8355] Filsfils, C., Previdi, S., Decraene, B., and R. Shakir, 953 "Resiliency Use Cases in Source Packet Routing in 954 Networking (SPRING) Networks", March 2018. 956 Appendix A. Migration from LDP to SR 958 PE2 PE4 959 \ / 960 PE1----P5--P6--P7---PE3 962 Figure 5: Migration 964 Several migration techniques are possible. The technique described 965 here is inspired by the commonly used method to migrate from one IGP 966 to another. 968 At time T0, all the routers run LDP. Any service is tunneled from an 969 ingress PE to an egress PE over a continuous LDP LSP. 971 At time T1, all the routers are upgraded to SR. They are configured 972 with the SRGB range [100, 300]. PE1, PE2, PE3, PE4, P5, P6 and P7 973 are respectively configured with the node segments 101, 102, 103, 974 104, 105, 106 and 107 (attached to their service-recursing loopback). 976 At this time, the service traffic is still tunneled over LDP LSP. 977 For example, PE1 has an SR node segment to PE3 and an LDP LSP to 978 PE3 but by default, as seen earlier, the LDP IP2MPLS encapsulation 979 is preferred. However, it has to be noted that the SR 980 infrastructure is usable, e.g. for Fast Reroute (FRR) or IGP Loop 981 Free Convergence to protect existing IP and LDP traffic. FRR 982 mechanisms are described in and [RFC8355]. 984 At time T2, the operator enables the local policy at PE1 to prefer SR 985 IP2MPLS encapsulation over LDP IP2MPLS. 987 The service from PE1 to any other PE is now riding over SR. All 988 other service traffic is still transported over LDP LSP. 990 At time T3, gradually, the operator enables the preference for SR 991 IP2MPLS encapsulation across all the edge routers. 993 All the service traffic is now transported over SR. LDP is still 994 operational and services could be reverted to LDP. 996 At time T4, LDP is unconfigured from all routers. 998 Authors' Addresses 1000 Ahmed Bashandy (editor) 1001 Individual 1002 USA 1004 Email: abashandy.ietf@gmail.com 1006 Clarence Filsfils (editor) 1007 Cisco Systems, Inc. 1008 Brussels 1009 BE 1011 Email: cfilsfil@cisco.com 1013 Stefano Previdi 1014 Cisco Systems, Inc. 1015 IT 1017 Email: stefano@previdi.net 1018 Bruno Decraene 1019 Orange 1020 FR 1022 Email: bruno.decraene@orange.com 1024 Stephane Litkowski 1025 Orange 1026 FR 1028 Email: stephane.litkowski@orange.com