idnits 2.17.1 draft-dunbar-sr-sdwan-over-hybrid-networks-04.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 seems to lack a Security Considerations section. (A line matching the expected section header was found, but with an unexpected indentation: ' 6. Security Considerations' ) ** The document seems to lack an IANA Considerations section. (See Section 2.2 of https://www.ietf.org/id-info/checklist for how to handle the case when there are no actions for IANA.) (A line matching the expected section header was found, but with an unexpected indentation: ' 7. IANA Considerations' ) ** There is 1 instance of too long lines in the document, the longest one being 95 characters in excess of 72. ** The document seems to lack a both a reference to RFC 2119 and the recommended RFC 2119 boilerplate, even if it appears to use RFC 2119 keywords. RFC 2119 keyword, line 466: '...appropriate SIDs SHOULD be chosen from...' Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (June 13, 2019) is 1772 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Missing Reference: 'SDWAN-Framework' is mentioned on line 116, but not defined == Missing Reference: 'RFC8085' is mentioned on line 467, but not defined == Missing Reference: 'RFC768' is mentioned on line 469, but not defined == Unused Reference: 'RFC2735' is defined on line 785, but no explicit reference was found in the text == Unused Reference: 'RFC8192' is defined on line 788, but no explicit reference was found in the text == Unused Reference: 'ITU-T-X1036' is defined on line 791, but no explicit reference was found in the text == Unused Reference: 'RFC6071' is defined on line 795, but no explicit reference was found in the text == Unused Reference: 'RFC4364' is defined on line 798, but no explicit reference was found in the text == Unused Reference: 'RFC4664' is defined on line 801, but no explicit reference was found in the text == Unused Reference: 'RFC7510' is defined on line 815, but no explicit reference was found in the text == Outdated reference: A later version (-01) exists of draft-dukes-sr-for-sdwan-00 == Outdated reference: A later version (-26) exists of draft-ietf-6man-segment-routing-header-13 == Outdated reference: A later version (-22) exists of draft-ietf-spring-segment-routing-mpls-13 Summary: 4 errors (**), 0 flaws (~~), 14 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 1 Network Working Group L. Dunbar 2 Internet Draft Futurewei 3 Intended status: Informational Mehmet Toy 4 Expires: Aug 2019 Verizon 5 June 13, 2019 7 Segment routing for SD-WAN paths over hybrid networks 8 draft-dunbar-sr-sdwan-over-hybrid-networks-04 10 Abstract 12 This document describes a method for end-to-end (E2E) SD-WAN paths 13 to traverse specific list of network segments, some of which are 14 private networks which include SR enabled segments, some of which 15 may be the public IP networks that do not support SR, to achieve the 16 desired optimal E2E quality. 18 The method described in this draft uses the principle of segment 19 routing to enforce a SD-WAN path' head-end selected route traversing 20 through a list of specific nodes of multiple network segments 21 without requiring the nodes in each network segment to have the 22 intelligence (or maintaining states) of selecting next hop or next 23 domain. 25 Status of this Memo 27 This Internet-Draft is submitted in full conformance with the 28 provisions of BCP 78 and BCP 79. 30 This Internet-Draft is submitted in full conformance with the 31 provisions of BCP 78 and BCP 79. This document may not be modified, 32 and derivative works of it may not be created, except to publish it 33 as an RFC and to translate it into languages other than English. 35 Internet-Drafts are working documents of the Internet Engineering 36 Task Force (IETF), its areas, and its working groups. Note that 37 other groups may also distribute working documents as Internet- 38 Drafts. 40 Internet-Drafts are draft documents valid for a maximum of six 41 months and may be updated, replaced, or obsoleted by other documents 42 at any time. It is inappropriate to use Internet-Drafts as 43 reference material or to cite them other than as "work in progress." 45 The list of current Internet-Drafts can be accessed at 46 http://www.ietf.org/ietf/1id-abstracts.txt 48 The list of Internet-Draft Shadow Directories can be accessed at 49 http://www.ietf.org/shadow.html 51 This Internet-Draft will expire on December 13, 2019. 53 Copyright Notice 55 Copyright (c) 2019 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 (http://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 63 respect to this document. Code Components extracted from this 64 document must include Simplified BSD License text as described in 65 Section 4.e of the Trust Legal Provisions and are provided without 66 warranty as described in the Simplified BSD License. 68 Table of Contents 70 1. Introduction...................................................3 71 2. Definition of terms............................................4 72 3. Key Use Cases..................................................5 73 3.1. SD-WAN Path over LTE network and SR Domain................5 74 3.2. SD-WAN as Last Mile for Cloud DCs Access..................6 75 3.3. How & Why SR is useful for those use cases................7 76 4. Mechanism for SD-WAN path over one SR Domain and existing access 77 ..................................................................8 78 4.1. Controller Delivers SID Stack to SD-WAN Head-end..........9 79 4.2. Using GRE Key or VXLAN ID to Differentiate Flows.........11 80 4.3. Using UDP Source Port Number to Differentiate Flows......12 81 4.4. GRE Header Extension.....................................15 82 5. SD-WAN path over multiple SP managed domains..................16 83 5.1. When Both SP domains support SR..........................18 84 5.2. When SP-2 does not support SR............................18 85 5.3. When SP-1 and SP-2 don't want to share network information19 86 5.4. TLV to pass Metadata through SRv6 Domain.................19 87 6. Security Considerations.......................................20 88 7. IANA Considerations...........................................21 89 8. References....................................................21 90 8.1. Normative References.....................................21 91 8.2. Informative References...................................21 92 9. Acknowledgments...............................................22 94 1. Introduction 96 This document describes a method to enforce a SD-WAN path's head-end 97 selected route traversing through a list of specific nodes of 98 multiple network segments without requiring the nodes in each 99 network segments to have the intelligence (or maintaining states) of 100 selecting next hop or next segments. Those networks over which the 101 SD-WAN path traverse have at least one SR enabled network, and some 102 network segments (especially the last mile access portion) being 103 existing IP networks (such as existing IPv4, IPv6 or others). 105 SD-WAN, as described by ONUG (Open Network User Group), is about 106 pooling WAN bandwidth from multiple service providers to get better 107 WAN bandwidth management, visibility & control. 109 Throughout this document, the term "Classic SD-WAN" refers to a pair 110 of CPEs in two locations aggregating N Service Providers' paths, 111 such as MPLS Paths and public internet paths. [SR-SD-WAN] describes 112 using explicit routes within the SRv6 or SR-MPLS enabled networks to 113 reach the desired quality for SD-WAN paths over the SRv6 or SR-MPLS 114 enabled networks respectively. 116 [SDWAN-Framework] describes three distinct SD-WAN scenarios from an 117 edge node's perspective: 119 1) All traffic is encrypted over WAN ports, so that network 120 service providers can extend its existing VPN to reach sites only 121 reachable via third party untrusted networks (such as public 122 internet); 124 2) Edge node has some ports connected to VPNs over which traffic 125 can go natively without encryption and other ports to the public 126 Internet over which traffic are encrypted; or 128 3) VPN PEs adding Internet facing WAN ports to offload low 129 priority traffic when the VPN backbone paths/links are congested. 131 This document focuses on the scenario 1) above, where a SD-WAN path 132 is spanning over SR enabled networks and over public IP networks, 133 such as existing IPv4, LTE, etc. Under this scenario, the endpoints 134 of the SD-WAN path (e.g. the CPE devices, one or both) are not 135 directly attached to PEs of a SR domain. 137 The goal is to place as large portion as possible of the SD-WAN path 138 over a provider VPN to achive more optimal transport quality, or 139 steering the SD-WAN path traversing specific ingress/egress PEs to 140 reach optimal cost, quality, regulatory or other reasons. 142 2. Definition of terms 144 Cloud DC: Off-Premises Data Centers that usually host applications 145 and workload owned by different organizations or 146 tenants. 148 Controller: Used interchangeably with SD-WAN controller to manage 149 SD-WAN overlay path creation/deletion and monitoring the 150 path conditions between two or more sites. 152 DMVPN: Dynamic Multipoint Virtual Private Network. DMVPN is a 153 secure network that exchanges data between sites without 154 needing to pass traffic through an organization's 155 headquarter virtual private network (VPN) server or 156 router. 158 Heterogeneous Cloud: applications & workloads split among Cloud DCs 159 owned & managed by different Cloud Providers. 161 Hybrid Cloud: applications & workloads split between on-premises 162 Data centers and Cloud DCs. In this document Hybrid 163 Cloud also include heterogeneous cloud as well. 165 SD-WAN: Software Defined Wide Area Network, "SDWAN" refers to 166 the solutions of pooling WAN bandwidth from multiple 167 underlay networks to get better WAN bandwidth 168 management, visibility & control. When the underlay 169 networks are private, traffic can traverse without 170 additional encryption; when the underlay networks are 171 public, such as the Internet, some traffic needs to be 172 encrypted when traversing through (depending on user 173 provided policies). 175 SP: Network Service Provider 177 SR: Segment Routing 179 SR Domain: A domain that supports Segment Routing 181 VPC: Virtual Private Cloud. A service offered by many Cloud 182 DC operators to allocate a logically isolated cloud 183 resources, including computing, networking and storage. 185 3. Key Use Cases 187 3.1. SD-WAN Path over LTE network and SR Domain 189 MEF Cloud Service Architecture [MEF-Cloud] describes a use case of 190 network operators needing to use SD-WAN over LTE for the last mile 191 access that they do not have physical infrastructure, as shown 192 below: 194 +********SD-WAN Overlay Path *********+ 195 * * 196 * +------------+ * 197 * |SD-WAN Ctrl | * 198 * +===+------------+====+ * 199 * // \\ * 200 * // <---Overlay Path ---> \\ * 201 * +-+--+ +--+ +--+ +--+-+ * 202 ***+ E1 |==|C1+--------+C4+==+ E2 |****+ 203 A --+ | | | | | | +----Z 204 A2--+--||+ ++-+ ++-+ ++---+\---Z2 205 LTE || | | // 206 || | SR +-+---+ +----+ 207 || | Network | C6 | |E3 | 208 || | | |----| | 209 +===+-+---+ +-+---+ +----+ 210 + C3 |-------+ 211 +---+-+ 212 | 213 +---+-+ 214 | E4 | 215 +-----+ 216 -- Directly attached 217 == || Public Internet or LTE path 218 *** Overlay path 220 Figure 1: SD-WAN end points are attached to VPN via LTE 222 3.2. SD-WAN as Last Mile for Cloud DCs Access 224 Digital Transformation is propelling more and more enterprises to 225 move their workloads/Apps to cloud DCs that are geographically close 226 to their end users to improve end-to-end latency & overall user 227 experience, or to comply with local data protection regulations. 228 Conversely, workloads/Apps in those Cloud DCs can be easily shutdown 229 when their end users' geographic base changes. 231 Because of the ephemeral property of the selected Cloud DCs, an 232 enterprise or its network service provider may not have the direct 233 links to the Cloud DCs that are optimal for hosting the enterprise's 234 specific workloads/Apps. Under those circumstances, SD-WAN is a very 235 flexible choice to interconnect the enterprise on-premises data 236 centers & branch offices to its desired Cloud DCs. 238 However, SD-WAN paths over public internet can have unpredictable 239 performance, especially over long distances and cross state/country 240 boundaries. Therefore, it is highly desirable to place as much as 241 possible the portion of SD-WAN paths over service provider VPN (e.g. 242 enterprise's existing VPN) that have guaranteed SLA and to minimize 243 the distance/segments over public internet. 245 Under this scenario, one or both of the SD-WAN end points may not 246 directly attached to the PEs of a SR Domain. 248 3.3. How & Why SR is useful for those use cases 250 Let us assume that the SD-WAN Controller is capable of computing 251 optimal paths between two end-points (e.g. E1<->E2 in the Figure 2), 252 either by communicating with the SR Domain controller/management- 253 system, or by other methods which is out of the scope of this 254 document. 256 The SR domain must have a set of PEs that have at least one port 257 facing the external networks (such as the public internet or LTE 258 termination). 260 Under this circumstance, SD-WAN end-points usually can reach 261 multiple PEs. 263 In the diagram below, E1 <-> E2 SD-WAN (most likely IPsec encrypted 264 tunnel) path can traverse C1 <-> C4, C1<->C6, C3<->C6, or C3<->C4 265 within the VPN. There are many flows (by different Apps) between E1 266 <-> E2. Some flows may need to traverse C1<->C4, others may need to 267 traverse C3<->C6 or other segments within the VPN, which are 268 determined by the SD-WAN controller based on the characteristics & 269 need of the Apps, such as cost, available bandwidth, latency, or 270 special functions only available at specific locations, etc. 272 Even with the same ingress/egress, some flows may need different 273 segments across the SR Domain. It is not practical, or even 274 possible, for PEs (e.g. C1, C2, C3 in this example) to determine 275 which Apps' flows should egress C4 or C6 where both C4&C6 can reach 276 E2. 278 Segment Routing can be used to steer packets (or path) to traverse 279 the explicit egress node (C4 or C6), or explicit segments through 280 the SR Domain based on the SLA requested by the SD-WAN head-end 281 nodes. 283 +********SD-WAN Overlay Path *********+ 284 * * 285 * +------------+ * 286 * |SD-WAN Ctrl | * 287 * +===+------------+====+ * 288 * // \\ * 289 * // <-- Overlay path----> \\ * 290 * +-+--+ ++-+ ++-+ +--+-+ * 291 ***+ E1 |==|C1|--------|C4+==+ E2 |****+ 292 A --+ | | | | | | +----Z 293 A2--+--+-+ ++-+ ++-+ ++---+\---Z2 294 LTE || | | // 295 || | SR +-+---+ +----+ 296 || | Network | C6 | |E3 | 297 || | | |- --| | 298 || +-+---+ +-+---+ +----+ 299 +==+ C3 |-------+ 300 +---+-+ 301 | 302 +---+-+ 303 | E4 | 304 +-----+ 306 -- Directly attached 307 == || Public Internet or LTE path 308 ** Overlay path 310 Figure 2: SDWAN end points not directly attached to PEs of SR Domain 312 4. Mechanism for SD-WAN path over one SR Domain and existing access 314 This section describes the mechanism to enforce a SD-WAN path' head- 315 end selected route traversing through a list of specific nodes of 316 multiple network segments without requiring the nodes in each 317 network segment to have the intelligence (or maintaining states) of 318 selecting next hop or next domain. 320 There may be two approaches here: 322 1) Controller installs the entire SID stack at E1. 323 2) Controller delivers to E1 a "Key" that the SR ingress PE can use 324 to map to the SID stack for the packets arriving at the SR Ingress 325 PE. Section 4.2 & 4.3 will describe how the "Key" is carried by the 326 packets. 328 The Approach 1) requires less processing at the SR Ingress PE nodes, 329 but only works if the remote CPEs are in the same administrative 330 domain as the SR domain. SR domain usually is not willing to expose 331 its internal binding SIDs to devices in different administration 332 domains. This approach also requires more changes to SD-WAN end 333 nodes and need more header bytes added to the packets when rd traversing through 3 party internet. Some SD-WAN nodes might not be 334 capable of supporting encapsulating packets with the SID stack. 336 The Approach 2) above requires SR Ingress PE nodes to map the "Key" 337 to the SID Stack and prepend the SID stack to the packets (Same 338 processing for other traffic except the mapping is from the received 339 "Key" carried in the payload). 341 4.1. Controller Delivers SID Stack to SD-WAN Head-end 343 This approach is straightforward. 345 E1 -------------------------- > SD-WAN controller 346 request for a SD-WAN path E1<->E2 with a specific SLA 348 E1 <-------------------------- SD-WAN controller 349 Reply with the Ingress PE Node ID or address 350 & the Binding SID. 352 Here is the packet header for SD-WAN Source Node to prepend to the 353 payload: 355 0 1 2 3 356 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 358 IPv4 Header: 359 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 360 |Version| IHL |Type of Service| Total Length | 361 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 362 | Identification |Flags| Fragment Offset | 363 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 364 | Time to Live | Prot.=17(UDP) | Header Checksum | 365 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 366 | SD-WAN Source IPv4 Address | 367 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 368 | SR Ingress PE IPv4 Address | 369 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 371 UDP Header: 372 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 373 | Source Port = | Dest. Port = 4754/4755 | 374 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 375 | UDP Length | UDP Checksum | 376 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 378 GRE Header: 379 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 380 |C| |K|S| Reserved0 | Ver | Protocol Type | 381 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 382 | Checksum (optional) | Reserved1 (Optional) | 383 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 384 | Key (For SR Ingress to map to its SID) | 385 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 386 | Sequence Number (optional) | 387 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 389 To traverse SRv6 domain, SRv6 Header is appended after the GRE 390 header [SRv6-SRH]: 392 To traverse MPLS-SR domain, a stack of MPLS labels is appended after 393 GRE Header [MPLS-SR]. 395 4.2. Using GRE Key or VXLAN ID to Differentiate Flows 397 This section describes a method of SD-WAN head-end node using GRE 398 Key or VXLAN Network ID (VNI) to indicate the desired property for 399 specific flows between SD-WAN end-points (E1<->E2 in the figure 400 above): such as different desired routes through the SR Domain, 401 different egress PEs based on cost, performance or other factors. 402 It might be difficult or impossible for DiffServ bits carried by the 403 packets to describe those flow properties because there can be more 404 than what DiffServ bits can represent. 406 The SR Domain ingress can map the GRE key to different SID through 407 the SR Domain. 409 We assume that the SD-WAN Controller can determine which ingress PE 410 can lead to the optimal path between E1<->E2. It is beyond the scope 411 of this document on how SD-WAN controller computes the paths and how 412 & what SD-WAN controller communicates with the SR Domain controller. 414 Here is the sequence of the flow: 416 E1 -------------------------- > SD-WAN controller 417 request for a SD-WAN path E1<->E2 with a specific SLA 419 E1 <-------------------------- SD-WAN controller 420 Reply with the Ingress PE Node ID or address 421 & (the GRE Key or VXLAN ID). 423 Note: the GRE key (or VXLAN ID) from the SD-WAN controller is for 424 the ingress PEs to correlate desired Path with the list of SIDs to 425 prepend the packet across the SR domain. 427 When SD-WAN Controller get the E1<->E2 path request, it will 428 communicate with the VPN Controller to get the optimal Ingress PE 429 Node ID (or IP address) and the GRE key (or VXLAN ID) to encapsulate 430 the original packets between E1 <-> E2 (assuming IPsec Tunnel mode 431 is used). 433 Upon receiving the GRE encapsulated packets, the provider ingress 434 Edge C1/C3 uses the GRE key (or VXLAN ID) to map to the pre-defined 435 (by the network controller) Binding SIDs, prepend the Binding SIDs 436 to the packets, and forward its desired paths across the provider 437 VPN. 439 Depending on how the SD-WAN path destination can be reached by the 440 egress PE, the egress PE has different processing procedure: 442 - If the destination of the SD-WAN path is directly attached to 443 the egress VPN PE node, the egress VPN PE decapsulates SR 444 header and forward the packets to SD-WAN path destination node, 445 such as the E2 in the figure above. 446 - If the destination of the SD-WAN path is IP reachable via IPv4 447 network from the egress VPN PE node, the egress VPN PE node 448 decapsulates SR header and forward the packets to SD-WAN path 449 destination node via its internet facing port to the SD-WAN 450 path destination (i.e. the E2 node in the figure above). 451 - If the SD-WAN path is traversing multiple domains owned by 452 different network operators, the egress PE processing is 453 described in the next session. 455 4.3. Using UDP Source Port Number to Differentiate Flows 457 [RFC8086] describes how to use GRE-in-UDP source port number as 458 entropy for better ECMP performance. When the remotely attached CPEs 459 is within very close proximity to the PEs, e.g. only one or two 460 hopes away like in LTE access, there is less issue if ECMP put all 461 flows with same traffic classifier into one path. Then, those UDP 462 numbers can also be used as a key to SR PE nodes to map to the 463 appropriate SID to the packets. 465 Same as RFC8086, UDP source port values used as a key for SR PEs to 466 map to appropriate SIDs SHOULD be chosen from the ephemeral port 467 range (49152-65535) [RFC8085]. 469 The GRE-in-UDP encapsulation format contains a UDP header [RFC768] 470 and a GRE header [RFC2890]. The format is shown as follow 471 (presented in bit order): 473 0 1 2 3 474 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 476 IPv4 Header: 477 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 478 |Version| IHL |Type of Service| Total Length | 479 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 480 | Identification |Flags| Fragment Offset | 481 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 482 | Time to Live | Prot.=17(UDP) | Header Checksum | 483 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 484 | SD-WAN Source IPv4 Address | 485 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 486 | SR Ingress PE IPv4 Address | 487 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 489 UDP Header: 490 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 491 | Source Port = SIDs key Value | Dest. Port = 4754/4755 | 492 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 493 | UDP Length | UDP Checksum | 494 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 496 GRE Header: 497 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 498 |C| |K|S| Reserved0 | Ver | Protocol Type | 499 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 500 | Checksum (optional) | Reserved1 (Optional) | 501 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 502 | Key (optional) | 503 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 504 | Sequence Number (optional) | 505 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 507 Figure 3: UDP + GRE Headers in IPv4 509 Here is the GRE Header for IPv6 network, i.e. the SD-WAN Source SD- 510 WAN Destination, and SR PEs are all in IPv6 domain: 512 0 1 2 3 513 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 515 IPv6 Header: 516 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 517 |Version| Traffic Class | Flow Label | 518 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 519 | Payload Length | NxtHdr=17(UDP)| Hop Limit | 520 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 521 | | 522 + + 523 | | 524 + SD-WAN Source IPv6 Address + 525 | | 526 + + 527 | | 528 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 529 | | 530 + + 531 | | 532 + SR Domain Ingress PE IPv6 Address + 533 | | 534 + + 535 | | 536 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 538 UDP Header: 539 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 540 | Source Port = SIDs key value | Dest. Port = 4754/4755 | 541 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 542 | UDP Length | UDP Checksum | 543 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 544 GRE Header: 545 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 546 |C| |K|S| Reserved0 | Ver | Protocol Type | 547 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 548 | Checksum (optional) | Reserved1 (Optional) | 549 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 550 | Key (optional) | 551 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 552 | Sequence Number (optional) | 553 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 555 Figure 4: GRE+UDP for IPv6 557 4.4. GRE Header Extension 559 A new protocol type can be added to the GRE header [RFC2890] to make 560 it easier for the SR PE to do the proper actions: 562 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 563 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 564 |C| Reserved0 | Ver | Protocol Type | 565 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 566 | Checksum (optional) | Reserved1 (Optional) | 567 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 569 The proposed GRE header will have the following format: 571 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 572 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 573 |C| |K|S| Reserved0 | Ver | Protocol Type | 574 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 575 | Checksum (optional) | Reserved1 (Optional) | 576 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 577 | Key (optional) | 578 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 579 | Sequence Number (Optional) | 580 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 582 New protocol type (value to be assigned by IANA): 584 UDP-Key: Using UDP source port value as a Key for SR Ingress PE to 585 map to the appropriate SIDs. 587 GRE-KEY: Using GRE Key value as a key for SR ingress PE to map to 588 the appropriate SIDs 590 5. SD-WAN path over multiple SP managed domains 591 The following figure shows a SD-WAN Path E1<->E2 over two SP domains 592 which are interconnected by public internet. 594 +********SD-WAN Overlay Path *************+ 595 * * 596 * +------------+ * 597 * |SD-WAN Ctrl | * 598 * +===+------------+====E2/E3/E4.. * 599 * // * 600 * // * 601 * +-+--+ ++-+ ++-+ +--+-+ * 602 ***+ E1 |==|C1+--------+C7+--+ E7 | * 603 A --+ | | | | | | + * 604 A2--+----+ ++-+ ++-+ ++---+ * 605 LTE || | SP1 | // * 606 +---+ SR +-+---+ +----+ * 607 +--------+ |C3 | Network | C4 | |E3 | * 608 | E4 | | | | |- --| | * 609 | | +---+-----+---+---+-+ +----+ * 610 +--------+=======+ C2 | || * 611 +---+-+ // * 612 // // * 613 +-+-+--------+--+-+ * 614 |D1 | |D4 | * 615 | | | | * 616 ++--+ ++---+ * 617 | SP2 | * 618 | SR +--+--+ +-+--+ 619 +--------+ | Network | D2 | |E2 +----Z 620 | E6 | | | +====+ +---Z2 621 | | +--+-----+ +-+---+LTE +----+ 622 +-+--+-+-+-------+ D3 |-----+ 623 +---+-+ 625 -- Directly attached 626 == || Public Internet or LTE path 627 ** Overlay path 628 Figure 5: SD-WAN path over two different SP domains 630 Let's assume that the SP-1 domain's egress node for the SD-WAN path 631 E1<->E2 is C2, which can reach D1 or D4 of SP-2 via public IP 632 network (say IPv4 network). 634 Let's also assume that the optimal route for some flows over SD-WAN 635 path E1<->E2 are C1->C2->D1 and other flows are over C1->C2->D4 (out 636 of the scope of this document on how the path is calculated). 638 If SP-1 is SR enabled, the mechanism described in Section 4 is 639 applicable to the SD-WAN path source node E1 and the SP-1's ingress 640 PE (e.g. C1 or C3 in the figure). 641 However, the processing at egress node might be different depending 642 on how the SP-1's egress edges are connected to the SP-2's ingress 643 edge nodes. 645 5.1. When Both SP domains support SR 647 There may be three approaches here: 649 1) Controller installs the entire SID stack at E1, and the SID list 650 contains SID entries belong to both domains. 652 2) Controller delivers to E1 the SID stack that only for the first 653 domain, but delivers to C6 (egress node of first domain) the binding 654 SID of the second domain. 656 3) Controller delivers a "Key" to E1, which can be encoded as GRE 657 KEY or represented by the Source UDP port of the GRE encapsulation, 658 for Ingress PE of the first SR Domain to map to its own SID stack as 659 described in Section 4. The first SR Domain will reserve the "Key" 660 through its domain and pass the "Key" to the second SR domain. The 661 second SR Domain Ingress node will use the same method to map the 662 "Key" to its SID stack. 664 5.2. When SP-2 does not support SR 666 Under this circumstance (which can be caused by SP-2 not supporting 667 SR or not willing to share Binding SIDs to SP-1), if the packets 668 arriving at SP-1 egress node C6 do not have any metadata indicating 669 the types of encrypted payload, C6 does not really have much choice 670 other than simply forwarding the packets to E2 via public IP 671 network. This way, the packets may or may not traverse through the 672 SP-2 domain. If the distance between C6 and E2 is far, the quality 673 of service can be unpredictable. 675 5.3. When SP-1 and SP-2 don't want to share network information 677 If SP-1's ingress node C1 can include the GRE KEY it receives from 678 E1 in the data packets' SR header, the SP-1's egress node can map 679 the Key to the SP-2's Ingress node and encapsulate the data packet 680 in a new GRE header destined towards the SP-2's Ingress node. Then 681 the SP-2's Ingress node can follow the procedure described in the 682 Section 4 to forward the data packets across its domain. 684 If the first SR Domain does not support adding metadata to carry the 685 "key" through its domain, the controller can deliver the "key" to 686 SP-1's egress node the same time as it delivers the key to E1, 687 knowing the SD-WAN path will need to traverse two domains with the 688 second one does support SR but the two SPs don't want to exchange 689 network information. 691 5.4. TLV to pass Metadata through SRv6 Domain 693 If SP-1 is SRv6 based, the ingress node C1 can append a TLV to the 694 end of the SR Header [SRv6-SRH] to carry the KEY it receives from 695 E1. 697 The SP-1 egress node C6 can get the mapping between the KEYs and the 698 Node-IDs (or Addresses) of the next domain's ingress edge node (i.e. 699 D1 or D4 in the figure 3 above) from its network controller ahead of 700 time. 702 0 1 2 3 703 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 704 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 705 | Type | Length | RESERVED | 706 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 707 | Key ID (4 octets) from the GRE tunnel remote ingress node | 708 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 709 | Optional // 710 | Node ID or address for the ingress node Next domain // 711 | Variable length (0~32 octets) // 712 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 714 TYPE: (to be assigned by IANA) is to indicate the TLV is for 715 carrying the flow identifier of the packet encoded by the SD-WAN 716 source node. 718 Upon receiving the packet, the egress node (C6) can 720 - find the Node-ID (or the address) for the next domain's ingress 721 node, 722 - construct a GRE header with the Key received from the TLV above 723 and the destination address from the mapping given by the 724 controller, 725 - encapsulate the GRE header to the data packet (which has 726 decapsulated SR header), 727 - and forward the packet to the public internet. 729 6. Security Considerations 731 Remotely attached CPEs might brought the following security risks: 733 1) Potential DDoS attack to the PEs with ports facing internet. 734 I.e. the PE resource being attacked by unwanted traffic. 735 2) Potential risk of provider VPN network bandwidth being stolen 736 by the entities who spoofed the addresses of SD-WAN end nodes. 738 To mitigate security risk of 1) above, it is absolutely necessary 739 for PEs which accept remotely attached CPEs or simply have ports 740 facing internet to enable Anti-DDoS feature to prevent major DDoS 741 attack to those PEs. 743 To mitigate the security risk of 2) above, RFC7510 defines the use 744 of DTLS to authenticate and encrypt the RFC7510 encapsulation. 746 However, for the scenario of SD-WAN source node being remotely 747 attached to PEs, using the method recommended by RFC7510 means the 748 source node has to perform DTLS on top of the IPSec encryption 749 between SD-WAN end points E1<->E2. This can be too processing heavy 750 for the SD-WAN end nodes. In addition, if there are many SD-WAN 751 flows to traverse through the ingress PE (e.g. C1, C2, C4 in the 752 figure 1 above), heavy processing is required on the ingress PEs. 754 Since the payload between E2<->E2 is already encrypted, the 755 confidentiality of the payload is already ensured. The network 756 operators need to balance between how much they can tolerant some 757 percentage of bandwidth being stolen and how much extra cost they 758 are willing to pay for completely prevent any unpaid traffic 759 traversing through its VPN networks. For operators who opt for lower 760 cost ingress PEs and CPEs, but can tolerant some percentage of 761 bandwidth being used by unpaid subscribers, a simple approach can be 762 considered: 764 - Embed a key in the packets, which can be changed periodically, 765 like the digital signature used by a certificate authority or 766 certification authority (CA). 767 - The key can be encoded in the GRE Key field between SD-WAN end 768 node and Ingress PE. Since GRE has 24 bits, some fixed bits 769 can be used to represent the signature of paid subscribers. 771 7. IANA Considerations 773 This document requires new protocol type: 775 Protocol type to be added to GRE header: SR_Route 777 8. References 779 8.1. Normative References 780 [RFC2890] G. Dommety "Key and Sequence Number Extensions to GRE". 781 Sep. 2000. 783 8.2. Informative References 785 [RFC2735] B. Fox, et al "NHRP Support for Virtual Private 786 networks". Dec. 1999. 788 [RFC8192] S. Hares, et al "Interface to Network Security Functions 789 (I2NSF) Problem Statement and Use Cases", July 2017 791 [ITU-T-X1036] ITU-T Recommendation X.1036, "Framework for creation, 792 storage, distribution and enforcement of policies for 793 network security", Nov 2007. 795 [RFC6071] S. Frankel and S. Krishnan, "IP Security (IPsec) and 796 Internet Key Exchange (IKE) Document Roadmap", Feb 2011. 798 [RFC4364] E. Rosen and Y. Rekhter, "BGP/MPLS IP Virtual Private 799 Networks (VPNs)", Feb 2006 801 [RFC4664] L. Andersson and E. Rosen, "Framework for Layer 2 Virtual 802 Private Networks (L2VPNs)", Sept 2006. 804 [SR-SD-WAN] D. Dukes, et al, "SR for SDWAN: VPN with Underlay SLA", 805 draft-dukes-sr-for-sdwan-00, in progress, Oct 2017 807 [SRv6-SRH] S. Previdi, et al, "IPv6 Segment Routing Header (SRH)", 808 draft-ietf-6man-segment-routing-header-13, in progress, 809 April 2018. 811 [MPLS-SR] A. Bashandy, et al, "Segment Routing with MPLS data 812 plane", draft-ietf-spring-segment-routing-mpls-13, in 813 progress, April 2018. 815 [RFC7510] X. Xu, et al, "Encapsulating MPLS in UDP", April 2015. 817 [RFC8086] L. Yong, et al, "GRE-in-UDP Encapsulation", March 2017. 819 [MEF-Cloud] "Cloud Services Architecture Technical Specification", 820 Work in progress, April 2018 822 9. Acknowledgments 824 Many thanks to Dean Cheng and Jim Guichard for the discussion and 825 contributions. 827 Authors' Addresses 829 Linda Dunbar 830 Futurewei 831 Email: Linda.Dunbar@futurewei.com 833 Mehmet Toy 834 Verizon 835 One Verizon Way 836 Basking Ridge, NJ 07920 837 Email: mehmet.toy@verizon.com