idnits 2.17.1 draft-ietf-sfc-architecture-06.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 : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (March 5, 2015) is 3311 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(See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Outdated reference: A later version (-13) exists of draft-ietf-sfc-problem-statement-10 Summary: 0 errors (**), 0 flaws (~~), 2 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group J. Halpern, Ed. 3 Internet-Draft Ericsson 4 Intended status: Standards Track C. Pignataro, Ed. 5 Expires: September 6, 2015 Cisco 6 March 5, 2015 8 Service Function Chaining (SFC) Architecture 9 draft-ietf-sfc-architecture-06 11 Abstract 13 This document describes an architecture for the specification, 14 creation, and ongoing maintenance of Service Function Chains (SFC) in 15 a network. It includes architectural concepts, principles, and 16 components used in the construction of composite services through 17 deployment of SFCs, with a focus on those to be standardized in the 18 IETF. This document does not propose solutions, protocols, or 19 extensions to existing protocols. 21 Status of This Memo 23 This Internet-Draft is submitted in full conformance with the 24 provisions of BCP 78 and BCP 79. 26 Internet-Drafts are working documents of the Internet Engineering 27 Task Force (IETF). Note that other groups may also distribute 28 working documents as Internet-Drafts. The list of current Internet- 29 Drafts is at http://datatracker.ietf.org/drafts/current/. 31 Internet-Drafts are draft documents valid for a maximum of six months 32 and may be updated, replaced, or obsoleted by other documents at any 33 time. It is inappropriate to use Internet-Drafts as reference 34 material or to cite them other than as "work in progress." 36 This Internet-Draft will expire on September 6, 2015. 38 Copyright Notice 40 Copyright (c) 2015 IETF Trust and the persons identified as the 41 document authors. All rights reserved. 43 This document is subject to BCP 78 and the IETF Trust's Legal 44 Provisions Relating to IETF Documents 45 (http://trustee.ietf.org/license-info) in effect on the date of 46 publication of this document. Please review these documents 47 carefully, as they describe your rights and restrictions with respect 48 to this document. Code Components extracted from this document must 49 include Simplified BSD License text as described in Section 4.e of 50 the Trust Legal Provisions and are provided without warranty as 51 described in the Simplified BSD License. 53 Table of Contents 55 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 56 1.1. Scope . . . . . . . . . . . . . . . . . . . . . . . . . . 3 57 1.2. Assumptions . . . . . . . . . . . . . . . . . . . . . . . 4 58 1.3. Definition of Terms . . . . . . . . . . . . . . . . . . . 4 59 2. Architectural Concepts . . . . . . . . . . . . . . . . . . . 6 60 2.1. Service Function Chains . . . . . . . . . . . . . . . . . 7 61 2.2. Service Function Chain Symmetry . . . . . . . . . . . . . 8 62 2.3. Service Function Paths . . . . . . . . . . . . . . . . . 9 63 3. Architecture Principles . . . . . . . . . . . . . . . . . . . 10 64 4. Core SFC Architecture Components . . . . . . . . . . . . . . 11 65 4.1. SFC Encapsulation . . . . . . . . . . . . . . . . . . . . 12 66 4.2. Service Function (SF) . . . . . . . . . . . . . . . . . . 13 67 4.3. Service Function Forwarder (SFF) . . . . . . . . . . . . 13 68 4.3.1. Transport Derived SFF . . . . . . . . . . . . . . . . 15 69 4.4. SFC-Enabled Domain . . . . . . . . . . . . . . . . . . . 15 70 4.5. Network Overlay and Network Components . . . . . . . . . 15 71 4.6. SFC Proxy . . . . . . . . . . . . . . . . . . . . . . . . 16 72 4.7. Classification . . . . . . . . . . . . . . . . . . . . . 17 73 4.8. Re-Classification and Branching . . . . . . . . . . . . . 17 74 4.9. Shared Metadata . . . . . . . . . . . . . . . . . . . . . 18 75 5. Additional Architectural Concepts . . . . . . . . . . . . . . 18 76 5.1. The Role of Policy . . . . . . . . . . . . . . . . . . . 18 77 5.2. SFC Control Plane . . . . . . . . . . . . . . . . . . . . 19 78 5.3. Resource Control . . . . . . . . . . . . . . . . . . . . 20 79 5.4. Infinite Loop Detection and Avoidance . . . . . . . . . . 20 80 5.5. Load Balancing Considerations . . . . . . . . . . . . . . 21 81 5.6. MTU and Fragmentation Considerations . . . . . . . . . . 22 82 5.7. SFC OAM . . . . . . . . . . . . . . . . . . . . . . . . . 23 83 5.8. Resilience and Redundancy . . . . . . . . . . . . . . . . 24 84 6. Security Considerations . . . . . . . . . . . . . . . . . . . 24 85 7. Contributors and Acknowledgments . . . . . . . . . . . . . . 25 86 8. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 27 87 9. Informative References . . . . . . . . . . . . . . . . . . . 27 88 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 28 90 1. Introduction 92 The delivery of end-to-end services often requires various service 93 functions. These include traditional network service functions such 94 as firewalls and traditional IP Network Address Translators (NATs), 95 as well as application-specific functions. The definition and 96 instantiation of an ordered set of service functions and subsequent 97 'steering' of traffic through them is termed Service Function 98 Chaining (SFC). 100 This document describes an architecture used for the creation and 101 ongoing maintenance of Service Function Chains (SFC) in a network. 102 It includes architectural concepts, principles, and components, with 103 a focus on those to be standardized in the IETF. Service function 104 chains enable composite services that are constructed from one or 105 more service functions. 107 An overview of the issues associated with the deployment of end-to- 108 end service function chains, abstract sets of service functions and 109 their ordering constraints that create a composite service and the 110 subsequent "steering" of traffic flows through said service 111 functions, is described in [I-D.ietf-sfc-problem-statement]. 113 The current service function deployment models are relatively static, 114 coupled to network topology and physical resources, greatly reducing 115 or eliminating the ability of an operator to introduce new services 116 or dynamically create service function chains. This architecture 117 presents a model addressing the problematic aspects of existing 118 service deployments, including topological independence and 119 configuration complexity. 121 1.1. Scope 123 This document defines the architecture for Service Function Chaining 124 (SFC) with minimum requirements on the physical topology of the 125 network, as standardized in the IETF. In this architecture packets 126 are classified on ingress for handling by the required set of Service 127 Functions (SFs) in the SFC-enabled domain and are then forwarded 128 through that set of functions for processing by each function in 129 turn. Packets may be re-classified as a result of this processing. 131 The architecture described in this document is independent of the 132 planned usage of the network and deployment context and thus, for 133 example, is applicable to both fixed and mobile networks as well as 134 being useful in many Data Center applications. 136 The architecture described herein is assumed to be applicable to a 137 single network administrative domain. While it is possible for the 138 architectural principles and components to be applied to inter-domain 139 SFCs, these are left for future study. 141 1.2. Assumptions 143 The following assumptions are made: 145 o There is no standard definition or characterization applicable to 146 all SFs, and thus the architecture considers each SF as an opaque 147 processing element. 149 o There is no global or standard list of SFs enabled in a given 150 administrative domain. The set of SFs enabled in a given domain 151 is a function of the currently active services which may vary with 152 time and according to the networking environment. 154 o There is no global or standard SF chaining logic. The ordered set 155 of SFs that needs to be applied to deliver a given service is 156 specific to each administrative entity. 158 o The chaining of SFs and the criteria to invoke them are specific 159 to each administrative entity that operates an SF-enabled domain. 161 o Several SF chaining policies can be simultaneously applied within 162 an administrative domain to meet various business requirements. 164 o The underlay is assumed to provide the necessary connectivity to 165 interconnect the SFFs, but the architecture places no constraints 166 on how that connectivity is realized other than it have the 167 required bandwidth, latency, and jitter to support the SFC. 169 o No assumption is made on how FIBs and RIBs of involved nodes are 170 populated. 172 o How to bind traffic to a given SF chain is policy-based. 174 1.3. Definition of Terms 176 Network Service: An offering provided by an operator that is 177 delivered using one or more service functions. This may also be 178 referred to as a composite service. The term "service" is used 179 to denote a "network service" in the context of this document. 181 Note: Beyond this document, the term "service" is overloaded 182 with varying definitions. For example, to some a service is an 183 offering composed of several elements within the operator's 184 network, whereas for others a service, or more specifically a 185 network service, is a discrete element such as a "firewall". 186 Traditionally, such services (in the latter sense) host a set of 187 service functions and have a network locator where the service 188 is hosted. 190 Classification: Locally instantiated matching of traffic flows 191 against policy for subsequent application of the required set of 192 network service functions. The policy may be customer/network/ 193 service specific. 195 Classifier: An element that performs Classification. 197 Service Function Chain (SFC): A service function chain defines an 198 ordered set of abstract service functions (SFs) and ordering 199 constraints that must be applied to packets and/or frames and/or 200 flows selected as a result of classification. An example of an 201 abstract service function is "a firewall". The implied order 202 may not be a linear progression as the architecture allows for 203 SFCs that copy to more than one branch, and also allows for 204 cases where there is flexibility in the order in which service 205 functions need to be applied. The term service chain is often 206 used as shorthand for service function chain. 208 Service Function (SF): A function that is responsible for specific 209 treatment of received packets. A Service Function can act at 210 various layers of a protocol stack (e.g., at the network layer 211 or other OSI layers). As a logical component, a Service 212 Function can be realized as a virtual element or be embedded in 213 a physical network element. One or more Service Functions can 214 be embedded in the same network element. Multiple occurrences 215 of the Service Function can exist in the same administrative 216 domain. 218 One or more Service Functions can be involved in the delivery of 219 added-value services. A non-exhaustive list of abstract Service 220 Functions includes: firewalls, WAN and application acceleration, 221 Deep Packet Inspection (DPI), LI (Lawful Intercept), server load 222 balancing, NAT44 [RFC3022], NAT64 [RFC6146], NPTv6 [RFC6296], 223 HOST_ID injection, HTTP Header Enrichment functions, TCP 224 optimizer. 226 An SF may be SFC encapsulation aware, that is it receives and 227 acts on information in the SFC encapsulation, or unaware, in 228 which case data forwarded to the SF does not contain the SFC 229 encapsulation. 231 Service Function Forwarder (SFF): A service function forwarder is 232 responsible for forwarding traffic to one or more connected 233 service functions according to information carried in the SFC 234 encapsulation, as well as handling traffic coming back from the 235 SF. Additionally, a service function forwarder is responsible 236 for delivering traffic to a classifier when needed and 237 supported, transporting traffic to another SFF (in the same or 238 different type of overlay), and terminating the SFP. 240 Metadata: provides the ability to exchange context information 241 between classifiers and SFs and among SFs. 243 Service Function Path (SFP): The SFP provides a level of indirection 244 between the fully abstract notion of service chain as a sequence 245 of abstract service functions to be delivered, and the fully 246 specified notion of exactly which SFF/SFs the packet will visit 247 when it actually traverses the network. By allowing the control 248 components to specify this level of indirection, the operator 249 may control the degree of SFF/SF selection authority that is 250 delegated to the network. 252 SFC Encapsulation: The SFC Encapsulation provides at a minimum SFP 253 identification, and is used by the SFC-aware functions, such as 254 the SFF and SFC-aware SFs. The SFC Encapsulation is not used 255 for network packet forwarding. In addition to SFP 256 identification, the SFC encapsulation carries metadata including 257 data plane context information. 259 Rendered Service Path (RSP): The Service Function Path is a 260 constrained specification of where packets assigned to a certain 261 service function path must go. While it may be so constrained 262 as to identify the exact locations, it can also be less 263 specific. Packets themselves are of course transmitted from and 264 to specific places in the network, visiting a specific sequence 265 of SFFs and SFs. This sequence of actual visits by a packet to 266 specific SFFs and SFs in the network is known as the Rendered 267 Service Path (RSP). This definition is included here for use by 268 later documents, such as when solutions may need to discuss the 269 actual sequence of locations the packets visit. 271 SFC-enabled Domain: A network or region of a network that implements 272 SFC. An SFC-enabled Domain is limited to a single network 273 administrative domain. 275 SFC Proxy: Removes and inserts SFC encapsulation on behalf of an 276 SFC-unaware service function. SFC proxies are logical elements. 278 2. Architectural Concepts 280 The following sections describe the foundational concepts of service 281 function chaining and the SFC architecture. 283 Service Function Chaining enables the creation of composite (network) 284 services that consist of an ordered set of Service Functions (SF) 285 that must be applied to packets and/or frames and/or flows selected 286 as a result of classification. Each SF is referenced using an 287 identifier that is unique within an SF-enabled domain. 289 Service Function Chaining is a concept that provides for more than 290 just the application of an ordered set of SFs to selected traffic; 291 rather, it describes a method for deploying SFs in a way that enables 292 dynamic ordering and topological independence of those SFs as well as 293 the exchange of metadata between participating entities. 295 2.1. Service Function Chains 297 In most networks, services are constructed as abstract sequences of 298 SFs that represent SFCs. At a high level, an SFC is an abstracted 299 view of a service that specifies the set of required SFs as well as 300 the order in which they must be executed. Graphs, as illustrated in 301 Figure 1, define an SFC, where each graph node represents the 302 required existence of at least one abstract SF. Such graph nodes 303 (SFs) can be part of zero, one, or many SFCs. A given graph node 304 (SF) can appear one time or multiple times in a given SFC. 306 SFCs can start from the origination point of the service function 307 graph (i.e., node 1 in Figure 1), or from any subsequent node in the 308 graph. SFs may therefore become branching nodes in the graph, with 309 those SFs selecting edges that move traffic to one or more branches. 310 An SFC can have more than one terminus. 312 ,-+-. ,---. ,---. ,---. 313 / \ / \ / \ / \ 314 ( 1 )+--->( 2 )+---->( 6 )+---->( 8 ) 315 \ / \ / \ / \ / 316 `---' `---' `---' `---' 318 ,-+-. ,---. ,---. ,---. ,---. 319 / \ / \ / \ / \ / \ 320 ( 1 )+--->( 2 )+---->( 3 )+---->( 7 )+---->( 9 ) 321 \ / \ / \ / \ / \ / 322 `---' `---' `---' `---' `---' 324 ,-+-. ,---. ,---. ,---. ,---. 325 / \ / \ / \ / \ / \ 326 ( 1 )+--->( 7 )+---->( 8 )+---->( 4 )+---->( 7 ) 327 \ / \ / \ / \ / \ / 328 `---' `---' `---' `---' `---' 330 Figure 1: Service Function Chain Graphs 332 The concepts of classification, re-classification, and branching are 333 covered in subsequent sections of this architecture (see Section 4.7 334 and Section 4.8). 336 2.2. Service Function Chain Symmetry 338 SFCs may be unidirectional or bidirectional. A unidirectional SFC 339 requires that traffic be forwarded through the ordered SFs in one 340 direction (SF1 -> SF2 -> SF3), whereas a bidirectional SFC requires a 341 symmetric path (SF1 -> SF2 -> SF3 and SF3 -> SF2 -> SF1), and in 342 which the SF instances are the same in opposite directions. A hybrid 343 SFC has attributes of both unidirectional and bidirectional SFCs; 344 that is to say some SFs require symmetric traffic, whereas other SFs 345 do not process reverse traffic or are independent of the 346 corresponding forward traffic. 348 SFCs may contain cycles; that is traffic may need to traverse one or 349 more SFs within an SFC more than once. Solutions will need to ensure 350 suitable disambiguation for such situations. 352 The architectural allowance that is made for SFPs that delegate 353 choice to the network for which SFs and/or SFFs a packet will visit 354 creates potential issues here. A solution that allows such 355 delegation needs to also describe how the solution ensures that those 356 service chains that require service function chain symmetry can 357 achieve that. 359 Further, there are state tradeoffs in symmetry. Symmetry may be 360 realized in several ways depending on the SFF and classifier 361 functionality. In some cases, "mirrored" classification (i.e., from 362 Source to Destination and from Destination to Source) policy may be 363 deployed, whereas in others shared state between classifiers may be 364 used to ensure that symmetric flows are correctly identified, then 365 steered along the required SFP. At a high level, there are various 366 common cases. In a non-exhaustive way, there can be for example: 368 o A single classifier (or a small number of classifiers), in which 369 case both incoming and outgoing flows could be recognized at the 370 same classifier, so the synchronization would be feasible by 371 internal mechanisms internal to the classifier. 373 o Stateful classifiers where several classifiers may be clustered 374 and share state. 376 o Fully distributed classifiers, where synchronization needs to be 377 provided through unspecified means. 379 o A classifier that learns state from the egress packets/flows that 380 is then used to provide state for the return packets/flow. 382 o Symmetry may also be provided by stateful forwarding logic in the 383 SFF in some implementations. 385 This is a non-comprehensive list of common cases. 387 2.3. Service Function Paths 389 A service function path (SFP) is a mechanism used by service chaining 390 to express the result of applying more granular policy and 391 operational constraints to the abstract requirements of a service 392 chain (SFC). This architecture does not mandate the degree of 393 specificity of the SFP. Architecturally, within the same SFC-enabled 394 domain, some SFPs may be fully specified, selecting exactly which SFF 395 and which SF are to be visited by packets using that SFP, while other 396 SFPs may be quite vague, deferring to the SFF the decisions about the 397 exact sequence of steps to be used to realize the SFC. The 398 specificity may be anywhere in between these extremes. 400 As an example of such an intermediate specificity, there may be two 401 SFPs associated with a given SFC, where one SFP specifies that any 402 order of SFF and SF may be used as long as it is within data center 403 1, and where the second SFP allows the same latitude, but only within 404 data center 2. 406 Thus, the policies and logic of SFP selection or creation (depending 407 upon the solution) produce what may be thought of as a constrained 408 version of the original SFC. Since multiple policies may apply to 409 different traffic that uses the same SFC, it also follows that there 410 may be multiple SFPs may be associated with a single SFC. 412 The architecture allows for the same SF to be reachable through 413 multiple SFFs. In these cases, some SFPs may constrain which SFF is 414 used to reach which SF, while some SFPs may leave that decision to 415 the SFF itself. 417 Further, the architecture allows for two or more SFs to be attached 418 to the same SFF, and possibly connected via internal means allowing 419 more effective communication. In these cases, some solutions or 420 deployments may choose to use some form of internal inter-process or 421 inter-VM messaging (communication behind the virtual switching 422 element) that is optimized for such an environment. This must be 423 coordinated with the SFF so that the service function forwarding can 424 properly perform its job. Implementation details of such mechanisms 425 are considered out of scope for this document, and can include a 426 spectrum of methods: for example situations including all next-hops 427 explicitly, others where a list of possible next-hops is provided and 428 the selection is local, or cases with just an identifier, where all 429 resolution is local. 431 This architecture also allows the same SF to be part of multiple 432 SFPs. 434 3. Architecture Principles 436 Service function chaining is predicated on several key architectural 437 principles: 439 1. Topological independence: no changes to the underlay network 440 forwarding topology - implicit, or explicit - are needed to 441 deploy and invoke SFs or SFCs. 443 2. Plane separation: dynamic realization of SFPs is separated from 444 packet handling operations (e.g., packet forwarding). 446 3. Classification: traffic that satisfies classification rules is 447 forwarded according to a specific SFP. For example, 448 classification can be as simple as an explicit forwarding entry 449 that forwards all traffic from one address into the SFP. 450 Multiple classification points are possible within an SFC (i.e., 451 forming a service graph) thus enabling changes/updates to the SFC 452 by SFs. 454 Classification can occur at varying degrees of granularity; for 455 example, classification can use a 5-tuple, a transport port or 456 set of ports, part of the packet payload, it can be the result of 457 high-level inspections, or it can come from external systems. 459 4. Shared Metadata: Metadata/context data can be shared amongst SFs 460 and classifiers, between SFs, and between external systems and 461 SFs (e.g., orchestration). 463 One use of metadata is to provide and share the result of 464 classification (that occurs within the SFC-enabled domain, or 465 external to it) along an SFP. For example, an external 466 repository might provide user/subscriber information to a service 467 chain classifier. This classifier could in turn impose that 468 information in the SFC encapsulation for delivery to the 469 requisite SFs. The SFs could in turn utilize the user/subscriber 470 information for local policy decisions. Metadata can also share 471 SF output along the SFP. 473 5. Service definition independence: The SFC architecture does not 474 depend on the details of SFs themselves. 476 6. Service function chain independence: The creation, modification, 477 or deletion of an SFC has no impact on other SFCs. The same is 478 true for SFPs. 480 7. Heterogeneous control/policy points: The architecture allows SFs 481 to use independent mechanisms (out of scope for this document) to 482 populate and resolve local policy and (if needed) local 483 classification criteria. 485 4. Core SFC Architecture Components 487 At a very high level, the logical architecture of an SFC-enabled 488 Domain comprises: 490 o . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491 . +--------------+ +------------------~~~ 492 . | Service | SFC | Service +---+ +---+ 493 . |Classification| Encapsulation | Function |sf1|...|sfn| 494 +---->| Function |+---------------->| Path +---+ +---+ 495 . +--------------+ +------------------~~~ 496 . SFC-enabled Domain 497 o . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499 Figure 2: Service Function Chain Architecture 501 The following sub-sections provide details on each logical component 502 that form the basis of the SFC architecture. A detailed overview of 503 how each of these architectural components interact is provided in 504 Figure 3: 506 +----------------+ +----------------+ 507 | SFC-aware | | SFC-unaware | 508 |Service Function| |Service Function| 509 +-------+--------+ +-------+--------+ 510 | | 511 SFC Encapsulation No SFC Encapsulation 512 | SFC | 513 +---------+ +----------------+ Encapsulation +---------+ 514 |SFC-Aware|-----------------+ \ +------------|SFC Proxy| 515 | SF | ... ----------+ \ \ / +---------+ 516 +---------+ \ \ \ / 517 +-------+--------+ 518 | SF Forwarder | 519 | (SFF) | 520 +-------+--------+ 521 | 522 SFC Encapsulation 523 | 524 ... SFC-enabled Domain ... 525 | 526 Network Overlay Transport 527 | 528 _,....._ 529 ,-' `-. 530 / `. 531 | Network | 532 `. / 533 `.__ __,-' 534 `'''' 536 Figure 3: Service Function Chain Architecture Components 538 4.1. SFC Encapsulation 540 The SFC encapsulation enables service function path selection. It 541 also enables the sharing of metadata/context information when such 542 metadata exchange is required. 544 The SFC encapsulation carries explicit information used to identify 545 the SFP. However, the SFC encapsulation is not a transport 546 encapsulation itself: it is not used to forward packets within the 547 network fabric. If packets need to flow between separate physical 548 platforms, the SFC encapsulation therefore relies on an outer network 549 transport. Transit forwarders -- such as router and switches -- 550 forward SFC encapsulated packets based on the outer (non-SFC) 551 encapsulation. 553 One of the key architecture principles of SFC is that the SFC 554 encapsulation remain transport independent. As such any network 555 transport protocol may be used to carry the SFC encapsulated traffic. 557 4.2. Service Function (SF) 559 The concept of an SF evolves; rather than being viewed as a bump in 560 the wire, an SF becomes a resource within a specified administrative 561 domain that is available for consumption as part of a composite 562 service. SFs send/receive data to/from one or more SFFs. SFC-aware 563 SFs receive this traffic with the SFC encapsulation. 565 While the SFC architecture defines the concept and specifies some 566 characteristics of a new encapsulation - the SFC encapsulation - and 567 several logical components for the construction of SFCs, existing SF 568 implementations may not have the capabilities to act upon or fully 569 integrate with the new SFC encapsulation. In order to provide a 570 mechanism for such SFs to participate in the architecture, an SFC 571 proxy function is defined (see Section 4.6). The SFC proxy acts as a 572 gateway between the SFC encapsulation and SFC-unaware SFs. The 573 integration of SFC-unaware service functions is discussed in more 574 detail in the SFC proxy section. 576 This architecture allows an SF to be part of multiple SFPs and SFCs. 578 4.3. Service Function Forwarder (SFF) 580 The SFF is responsible for forwarding packets and/or frames received 581 from the network to one or more SFs associated with a given SFF using 582 information conveyed in the SFC encapsulation. Traffic from SFs 583 eventually returns to the same SFF, which is responsible for 584 injecting traffic back onto the network. Some SFs, such as 585 firewalls, could also consume a packet. 587 The collection of SFFs and associated SFs creates a service plane 588 overlay in which SFC-aware SFs, as well as SFC-unaware SFs reside. 589 Within this service plane, the SFF component connects different SFs 590 that form a service function path. 592 SFFs maintain the requisite SFP forwarding information. SFP 593 forwarding information is associated with a service path identifier 594 that is used to uniquely identify an SFP. The service forwarding 595 state enables an SFF to identify which SFs of a given SFP should be 596 applied, and in what order, as traffic flows through the associated 597 SFP. While there may appear to the SFF to be only one available way 598 to deliver the given SF, there may also be multiple choices allowed 599 by the constraints of the SFP. 601 If there are multiple choices, the SFF needs to preserve the property 602 that all packets of a given flow are handled the same way, since the 603 SF may well be stateful. Additionally, the SFF may preserve the 604 handling of packets based on other properties on top of a flow, such 605 as a subscriber, session, or application instance identification. 607 The SFF also has the information to allow it to forward packets to 608 the next SFF after applying local service functions. Again, while 609 there may be only a single choice available, the architecture allows 610 for multiple choices for the next SFF. As with SFs, the solution 611 needs to operate such that the behavior with regard to specific flows 612 (see the Rendered Service Path) is stable. The selection of 613 available SFs and next SFFs may be interwoven when an SFF supports 614 multiple distinct service functions and the same service function is 615 available at multiple SFFs. Solutions need to be clear about what is 616 allowed in these cases. 618 Even when the SFF supports and utilizes multiple choices, the 619 decision as to whether to use flow-specific mechanisms or coarser 620 grained means to ensure that the behavior of specific flows is stable 621 is a matter for specific solutions and specific implementations. 623 The SFF component has the following primary responsibilities: 625 1. SFP forwarding : Traffic arrives at an SFF from the network. The 626 SFF determines the appropriate SF the traffic should be forwarded 627 to via information contained in the SFC encapsulation. Post-SF, 628 the traffic is returned to the SFF, and if needed forwarded to 629 another SF associated with that SFF. If there is another non- 630 local (i.e., different SFF) hop in the SFP, the SFF further 631 encapsulates the traffic in the appropriate network transport 632 protocol and delivers it to the network for delivery to the next 633 SFF along the path. Related to this forwarding responsibility, 634 an SFF should be able to interact with metadata. 636 2. Terminating SFPs : An SFC is completely executed when traffic has 637 traversed all required SFs in a chain. When traffic arrives at 638 the SFF after the last SF has finished processing it, the final 639 SFF knows from the service forwarding state that the SFC is 640 complete. The SFF removes the SFC encapsulation and delivers the 641 packet back to the network for forwarding. 643 3. Maintaining flow state: In some cases, the SFF may be stateful. 644 It creates flows and stores flow-centric information. This state 645 information may be used for a range of SFP-related tasks such as 646 ensuring consistent treatment of all packets in a given flow, 647 ensuring symmetry or for state-aware SFC Proxy functionality (see 648 Section 4.8). 650 4.3.1. Transport Derived SFF 652 Service function forwarding, as described above, directly depends 653 upon the use of the service path information contained in the SFC 654 encapsulation. However, existing implementations may not be able to 655 act on the SFC encapsulation. These platforms may opt to use 656 existing transport information if it can be arranged to provide 657 explicit service path information. 659 This results in the same architectural behavior and meaning for 660 service function forwarding and service function paths. It is the 661 responsibility of the control components to ensure that the transport 662 path executed in such a case is fully aligned with the path 663 identified by the information in the service chaining encapsulation. 665 4.4. SFC-Enabled Domain 667 Specific features may need to be enforced at the boundaries of an 668 SFC-enabled domain, for example to avoid leaking SFC information. 669 Using the term node to refer generically to an entity that is 670 performing a set of functions, in this context, an SFC Boundary Node 671 denotes a node that connects one SFC-enabled domain to a node either 672 located in another SFC-enabled domain or in a domain that is SFC- 673 unaware. 675 An SFC Boundary node can act as egress or ingress. An SFC Egress 676 Node denotes a SFC Boundary Node that handles traffic leaving the 677 SFC-enabled domain the Egress Node belongs to. Such a node is 678 required to remove any information specific to the SFC Domain, 679 typically the SFC Encapsulation. Further, from a privacy 680 perspective, an SFC Egress Node is required to ensure that any 681 sensitive information added as part of SFC gets removed. An SFC 682 Ingress Node denotes an SFC Boundary Node that handles traffic 683 entering the SFC-enabled domain. In most solutions and deployments 684 this will need to include a classifier, and will be responsible for 685 adding the SFC encapsulation to the packet. 687 4.5. Network Overlay and Network Components 689 Underneath the SFF there are components responsible for performing 690 the transport (overlay) forwarding. They do not consult the SFC 691 encapsulation or inner payload for performing this forwarding. They 692 only consult the outer-transport encapsulation for the transport 693 (overlay) forwarding. 695 4.6. SFC Proxy 697 In order for the SFC architecture to support SFC-unaware SFs (e.g., 698 legacy service functions) a logical SFC proxy function may be used. 699 This function sits between an SFF and one or more SFs to which the 700 SFF is directing traffic (see Figure 3). 702 The proxy accepts packets from the SFF on behalf of the SF. It 703 removes the SFC encapsulation, and then uses a local attachment 704 circuit to deliver packets to SFC unaware SFs. It also receives 705 packets back from the SF, reapplies the SFC encapsulation, and 706 returns them to the SFF for processing along the service function 707 path. 709 Thus, from the point of view of the SFF, the SFC proxy appears to be 710 part of an SFC aware SF. 712 Communication details between the SFF and the SFC Proxy are the same 713 as those between the SFF and an SFC aware SF. The details of that 714 are not part of this architecture. The details of the communication 715 methods over the local attachment circuit between the SFC proxy and 716 the SFC-unaware SF are dependent upon the specific behaviors and 717 capabilities of that SFC-unaware SF, and thus are also out of scope 718 for this architecture. 720 Specifically, for traffic received from the SFF intended for the SF 721 the proxy is representing, the SFC proxy: 723 o Removes the SFC encapsulation from SFC encapsulated packets. 725 o Identifies the required SF to be applied based on available 726 information including that carried in the SFC encapsulation. 728 o Selects the appropriate outbound local attachment circuit through 729 which the next SF for this SFP is reachable. This is derived from 730 the identification of the SF carried in the SFC encapsulation, and 731 may include local techniques. Examples of a local attachment 732 circuit include, but are not limited to, VLAN, IP-in-IP, L2TPv3, 733 GRE, VXLAN. 735 o Forwards the original payload via the selected local attachment 736 circuit to the appropriate SF. 738 When traffic is returned from the SF: 740 o Applies the required SFC encapsulation. The determination of the 741 encapsulation details may be inferred by the local attachment 742 circuit through which the packet and/or frame was received, or via 743 packet classification, or other local policy. In some cases, 744 packet ordering or modification by the SF may necessitate 745 additional classification in order to re-apply the correct SFC 746 encapsulation. 748 o Delivers the packet with the SFC Encapsulation to the SFF, as 749 would happen with packets returned from an SFC-aware SF. 751 4.7. Classification 753 Traffic from the network that satisfies classification criteria is 754 directed into an SFP and forwarded to the requisite service 755 function(s). Classification is handled by a service classification 756 function; initial classification occurs at the ingress to the SFC 757 domain. The granularity of the initial classification is determined 758 by the capabilities of the classifier and the requirements of the SFC 759 policy. For instance, classification might be relatively coarse: all 760 packets from this port are subject to SFC policy X and directed into 761 SFP A, or quite granular: all packets matching this 5-tuple are 762 subject to SFC policy Y and directed into SFP B. 764 As a consequence of the classification decision, the appropriate SFC 765 encapsulation is imposed on the data, and a suitable SFP is selected 766 or created. Classification results in attaching the traffic to a 767 specific SFP. 769 4.8. Re-Classification and Branching 771 The SFC architecture supports re-classification (or non-initial 772 classification) as well. As packets traverse an SFP, re- 773 classification may occur - typically performed by a classification 774 function co-resident with a service function. Reclassification may 775 result in the selection of a new SFP, an update of the associated 776 metadata, or both. This is referred to as "branching". 778 For example, an initial classification results in the selection of 779 SFP A: DPI_1 --> SLB_8. However, when the DPI service function is 780 executed, attack traffic is detected at the application layer. DPI_1 781 re-classifies the traffic as attack and alters the service path to 782 SFP B, to include a firewall for policy enforcement: dropping the 783 traffic: DPI_1 --> FW_4. Subsequent to FW_4, surviving traffic would 784 be returned to the original SFF. In this simple example, the DPI 785 service function re-classifies the traffic based on local application 786 layer classification capabilities (that were not available during the 787 initial classification step). 789 When traffic arrives after being steered through an SFC-unaware SF, 790 the SFC Proxy must perform re-classification of traffic to determine 791 the SFP. The SFC Proxy is concerned with re-attaching information 792 for SFC-unaware SFs, and a stateful SFC Proxy simplifies such 793 classification to a flow lookup. 795 4.9. Shared Metadata 797 Sharing metadata allows the network to provide network-derived 798 information to the SFs, SF-to-SF information exchange and the sharing 799 of service-derived information to the network. Some SFCs may not 800 require metadata exchange. SFC infrastructure enables the exchange 801 of this shared data along the SFP. The shared metadata serves 802 several possible roles within the SFC architecture: 804 o Allows elements that typically operate as ships in the night to 805 exchange information. 807 o Encodes information about the network and/or data for post- 808 service forwarding. 810 o Creates an identifier used for policy binding by SFs. 812 Context information can be derived in several ways: 814 o External sources 816 o Network node classification 818 o Service function classification 820 5. Additional Architectural Concepts 822 There are a number of issues which solutions need to address, and 823 which the architecture informs but does not determine. This section 824 lays out some of those concepts. 826 5.1. The Role of Policy 828 Much of the behavior of service chains is driven by operator and per- 829 customer policy. This architecture is structured to isolate the 830 policy interactions from the data plane and control logic. 832 Specifically, it is assumed that the service chaining control plane 833 creates the service paths. The service chaining data plane is used 834 to deliver the classified packets along the service chains to the 835 intended service functions. 837 Policy, in contrast, interacts with the system in other places. 838 Policies and policy engines may monitor service functions to decide 839 if additional (or fewer) instances of services are needed. When 840 applicable, those decisions may in turn result in interactions that 841 direct the control logic to change the SFP placement or packet 842 classification rules. 844 Similarly, operator service policy, often managed by operational or 845 business support systems (OSS or BSS), will frequently determine what 846 service functions are available. Operator service policies also 847 determine which sequences of functions are valid and are to be used 848 or made available. 850 The offering of service chains to customers, and the selection of 851 which service chain a customer wishes to use, are driven by a 852 combination of operator and customer policies using appropriate 853 portals in conjunction with the OSS and BSS tools. These selections 854 then drive the service chaining control logic, which in turn 855 establishes the appropriate packet classification rules. 857 5.2. SFC Control Plane 859 This is part of the overall architecture but outside the scope of 860 this document. 862 The SFC control plane is responsible for constructing SFPs, 863 translating SFCs to forwarding paths and propagating path information 864 to participating nodes to achieve requisite forwarding behavior to 865 construct the service overlay. For instance, an SFC construction may 866 be static; selecting exactly which SFFs and which SFs from those SFFs 867 are to be used, or it may be dynamic, allowing the network to perform 868 some or all of the choices of SFF or SF to use to deliver the 869 selected service chain within the constraints represented by the 870 service path. 872 In the SFC architecture, SFs are resources; the control plane manages 873 and communicates their capabilities, availability and location in 874 fashions suitable for the transport and SFC operations in use. The 875 control plane is also responsible for the creation of the context 876 (see below). The control plane may be distributed (using new or 877 existing control plane protocols), or be centralized, or a 878 combination of the two. 880 The SFC control plane provides the following functionality: 882 1. An SFC-enabled domain wide view of all available service function 883 resources as well as the network locators through which they are 884 reachable. 886 2. Uses SFC policy to construct service function chains, and 887 associated service function paths. 889 3. Selection of specific SFs for a requested SFC, either statically 890 (using specific SFs) or dynamically (using service explicit SFs 891 at the time of delivering traffic to them). 893 4. Provides requisite SFC data plane information to the SFC 894 architecture components, most notably the SFF. 896 5. Provide the metadata and usage information classifiers need so 897 that they in turn can provide this metadata for appropriate 898 packets in the data plane. 900 6. When needed, provide information including policy information to 901 other SFC elements to be able to properly interpret metadata. 903 5.3. Resource Control 905 The SFC system may be responsible for managing all resources 906 necessary for the SFC components to function. This includes network 907 constraints used to plan and choose network path(s) between service 908 function forwarders, network communication paths between service 909 function forwarders and their attached service functions, 910 characteristics of the nodes themselves such as memory, number of 911 virtual interfaces, routes, and instantiation, configuration, and 912 deletion of SFs. 914 The SFC system will also be required to reflect policy decisions 915 about resource control, as expressed by other components in the 916 system. 918 While all of these aspects are part of the overall system, they are 919 beyond the scope of this architecture. 921 5.4. Infinite Loop Detection and Avoidance 923 This SFC architecture is predicated on topological independence from 924 the underlying forwarding topology. Consequently, a service topology 925 is created by Service Function Paths or by the local decisions of the 926 Service Function Forwarders based on the constraints expressed in the 927 SFP. Due to the overlay constraints, the packet-forwarding path may 928 need to visit the same SFF multiple times, and in some less common 929 cases may even need to visit the same SF more than once. The Service 930 Chaining solution needs to permit these limited and policy-compliant 931 loops. At the same time, the solutions must ensure that indefinite 932 and unbounded loops cannot be formed, as such would consume unbounded 933 resources without delivering any value. 935 In other words, this architecture requires the solution to prevent 936 infinite Service Function Loops, even when Service Functions may be 937 invoked multiple times in the same SFP. 939 5.5. Load Balancing Considerations 941 Supporting function elasticity and high-availability should not 942 overly complicate SFC or lead to unnecessary scalability problems. 944 In the simplest case, where there is only a single function in the 945 SFP (the next hop is either the destination address of the flow or 946 the appropriate next hop to that destination), one could argue that 947 there may be no need for SFC. 949 In the cases where the classifier is separate from the single 950 function or a function at the terminal address may need sub-prefix 951 (e.g., finer grained address information) or per-subscriber metadata, 952 a single SFP exists (i.e., the metadata changes but the SFP does 953 not), regardless of the number of potential terminal addresses for 954 the flow. This is the case of the simple load balancer. See 955 Figure 4. 957 +---+ +---++--->web server 958 source+-->|sff|+-->|sf1|+--->web server 959 +---+ +---++--->web server 961 Figure 4: Simple Load Balancing 963 By extrapolation, in the case where intermediary functions within a 964 chain had similar "elastic" behaviors, we do not need separate chains 965 to account for this behavior - as long as the traffic coalesces to a 966 common next-hop after the point of elasticity. 968 In Figure 5, we have a chain of five service functions between the 969 traffic source and its destination. 971 +---+ +---+ +---+ +---+ +---+ +---+ 972 |sf2| |sf2| |sf3| |sf3| |sf4| |sf4| 973 +---+ +---+ +---+ +---+ +---+ +---+ 974 | | | | | | 975 +-----+-----+ +-----+-----+ 976 | | 977 + + 978 +---+ +---+ +---+ +---+ +---+ 979 source+-->|sff|+-->|sff|+--->|sff|+--->|sff|+-->|sff|+-->destination 980 +---+ +---+ +---+ +---+ +---+ 981 + + + 982 | | | 983 +---+ +---+ +---+ 984 |sf1| |sf3| |sf5| 985 +---+ +---+ +---+ 987 Figure 5: Load Balancing 989 This would be represented as one service function path: 990 sf1->sf2->sf3->sf4->sf5. The SFF is a logical element, which may be 991 made up of one or multiple components. In this architecture, the SFF 992 may handle load distribution based on policy. 994 It can also be seen in the above that the same service function may 995 be reachable through multiple SFFs, as discussed earlier. The 996 selection of which SFF to use to reach SF3 may be made by the control 997 logic in defining the SFP, or may be left to the SFFs themselves, 998 depending upon policy, solution, and deployment constraints. In the 999 latter case, it needs to be assured that exactly one SFF takes 1000 responsibility to steer traffic through SF3. 1002 5.6. MTU and Fragmentation Considerations 1004 This architecture prescribes additional information being added to 1005 packets to identify service function paths and often to represent 1006 metadata. It also envisions adding transport information to carry 1007 packets along service function paths, at least between service 1008 function forwarders. This added information increases the size of 1009 the packet to be carried by service chaining. Such additions could 1010 potentially increase the packet size beyond the MTU supported on some 1011 or all of the media used in the service chaining domain. 1013 Such packet size increases can thus cause operational MTU problems. 1014 Requiring fragmentation and reassembly in an SFF would be a major 1015 processing increase, and might be impossible with some transports. 1016 Expecting service functions to deal with packets fragmented by the 1017 SFC function might be onerous even when such fragmentation was 1018 possible. Thus, at the very least, solutions need to pay attention 1019 to the size cost of their approach. There may be alternative or 1020 additional means available, although any solution needs to consider 1021 the tradeoffs. 1023 These considerations apply to any generic architecture that increases 1024 the header size. There are also more specific MTU considerations: 1025 Effects on Path MTU Discovery (PMTUD) as well as deployment 1026 considerations. Deployments within a single administrative control 1027 or even a single Data Center complex can afford more flexibility in 1028 dealing with larger packets, and deploying existing mitigations that 1029 decrease the likelihood of fragmentation or discard. 1031 5.7. SFC OAM 1033 Operations, Administration, and Maintenance (OAM) tools are an 1034 integral part of the architecture. These serve various purposes, 1035 including fault detection and isolation, and performance management. 1036 For example, there are many advantages of SFP liveness detection, 1037 including status reporting, support for resiliency operations and 1038 policies, and an enhanced ability to balance load. 1040 Service Function Paths create a services topology, and OAM performs 1041 various functions within this service layer. Furthermore, SFC OAM 1042 follows the same architectural principles of SFC in general. For 1043 example, topological independence (including the ability to run OAM 1044 over various overlay technologies) and classification-based policy. 1046 We can subdivide the SFC OAM architecture in two parts: 1048 o In-band: OAM packets follow the same path and share fate with user 1049 packets, within the service topology. For this, they also follow 1050 the architectural principle of consistent policy identifiers, and 1051 use the same path IDs as the service chain data packets. Load 1052 balancing and SFC encapsulation with packet forwarding are 1053 particularly important here. 1055 o Out-of-band: reporting beyond the actual data plane. An 1056 additional layer beyond the data-plane OAM allows for additional 1057 alerting and measurements. 1059 This architecture prescribes end-to-end SFP OAM functions, which 1060 implies SFF understanding of whether an in-band packet is an OAM or 1061 user packet. However, service function validation is outside of the 1062 scope of this architecture, and application-level OAM is not what 1063 this architecture prescribes. 1065 Some of the detailed functions performed by SFC OAM include fault 1066 detection and isolation in a Service Function Path or a Service 1067 Function, verification that connectivity using SFPs is both effective 1068 and directing packets to the intended service functions, service path 1069 tracing, diagnostic and fault isolation, alarm reporting, performance 1070 measurement, locking and testing of service functions, validation 1071 with the control plane (see Section 5.2), and also allow for vendor- 1072 specific as well as experimental functions. SFC should leverage, and 1073 if needed extend relevant existing OAM mechanisms. 1075 5.8. Resilience and Redundancy 1077 As a practical operational requirement, any service chaining solution 1078 needs to be able to respond effectively, and usually very quickly, to 1079 failure conditions. These may be failures of connectivity in the 1080 network between SFFs, failures of SFFs, or failures of SFs. Per-SF 1081 state, as for example stateful-firewall state, is the responsibility 1082 of the SF, and not addressed by this architecture. 1084 Multiple techniques are available to address this issue. Solutions 1085 can describe both what they require and what they allow to address 1086 failure. Solutions can make use of flexible specificity of service 1087 function paths, if the SFF can be given enough information in a 1088 timely fashion to do this. Solutions can also make use of MAC or IP 1089 level redundancy mechanisms such as VRRP. Also, particularly for SF 1090 failures, load balancers co-located with the SFF or as part of the 1091 service function delivery mechanism can provide such robustness. 1093 Similarly, operational requirements imply resilience in the face of 1094 load changes. While mechanisms for managing (e.g., monitoring, 1095 instantiating, loading images, providing configuration to service 1096 function chaining control, deleting, etc.) virtual machines are out 1097 of scope for this architecture, solutions can and are aided by 1098 describing how they can make use of scaling mechanisms. 1100 6. Security Considerations 1102 Security considerations apply to the realization of this 1103 architecture, in particular to the documents that will define 1104 protocols. Such realization ought to provide means to protect 1105 against security and privacy attacks in the areas hereby described. 1107 Following the categorization of [I-D.ietf-sfc-problem-statement], we 1108 can largely divide the security consdierations in three areas: 1110 Service Overlay: Underneath the Service Function Forwarders, the 1111 components that are responsible for performing the transport 1112 forwarding consult the outer-transport encapsulation for 1113 underlay forwarding. Used transport mechanisms should satisfy 1114 the security requirements of the specific SFC deployment. These 1115 requirements typically include varying degrees of traffic 1116 separation, protection against different attacks (e.g., 1117 spoofing, man-in-the-middle, brute-force, or insertion attacks), 1118 and can also include authenticity and integrity checking, and/or 1119 confidentiality provisions. 1121 Classification: Specific requirements may need to be enforced at the 1122 boundaries of an SFC-enabled domain. These include, for 1123 example, to avoid leaking SFC information, and to protect its 1124 borders against various forms of attacks, including DDoS 1125 attacks. Classification is used at the ingress edge of an SFC- 1126 enabled domain. Policy for this classification is done using a 1127 plurality of methods. Whatever method is used needs to consider 1128 a range of security issues. These include appropriate 1129 authentication and authorization of classification policy, 1130 potential confidentiality issues of that policy, protection 1131 against corruption, and proper application of policy with needed 1132 segregation of application. This includes proper controls on 1133 the policies which drive the application of the SFC 1134 Encapsulation and associated metadata to packets. Similar 1135 issues need to be addressed if classification is performed 1136 within a service chaining domain, i.e., re-classification. 1138 SFC Encapsulation: The SFC Encapsulation provides at a minimum SFP 1139 identification, and carries metadata. An operator may consider 1140 the SFC Metadata as sensitive. From a privacy perspective, a 1141 user may be concerned about the operator revealing data about 1142 (and not belonging to) the customer. Therefore, solutions 1143 should consider whether there is a risk of sensitive information 1144 slipping out of the operators control. Further, the SFC 1145 Encapsulation includes SFC OAM Functions, which need to not 1146 negatively affect the security considerations of an SFC-enabled 1147 domain. 1149 Finally, all entities (software or hardware) interacting with the 1150 service chaining mechanisms need to provide means of security against 1151 malformed, poorly configured (deliberate or not) protocol constructs 1152 and loops. These considerations are largely the same as those in any 1153 network, particularly an overlay network. 1155 7. Contributors and Acknowledgments 1157 The editors would like to thank Sam Aldrin, Nicolas Bouthors, Stewart 1158 Bryant, Linda Dunbar, Alla Goldner, Ken Gray, Barry Greene, Anil 1159 Gunturu, David Harrington, Shunsuke Homma, Dave Hood, Nagendra Kumar, 1160 Hongyu Li, Andrew Malis, Guy Meador III, Kengo Naito, Thomas Narten, 1161 Ron Parker, Reinaldo Penno, Naiming Shen, Xiaohu Xu, and Lucy Yong 1162 for a thorough review and useful comments. 1164 The initial version of this "Service Function Chaining (SFC) 1165 Architecture" document is the result of merging two previous 1166 documents, and this section lists the aggregate of authors, editors, 1167 contributors and acknowledged participants, all who provided 1168 important ideas and text that fed into this architecture. 1170 [I-D.boucadair-sfc-framework]: 1172 Authors: 1174 Mohamed Boucadair 1175 Christian Jacquenet 1176 Ron Parker 1177 Diego R. Lopez 1178 Jim Guichard 1179 Carlos Pignataro 1181 Contributors: 1183 Parviz Yegani 1184 Paul Quinn 1185 Linda Dunbar 1187 Acknowledgements: 1189 Many thanks to D. Abgrall, D. Minodier, Y. Le Goff, D. 1190 Cheng, R. White, and B. Chatras for their review and 1191 comments. 1193 [I-D.quinn-sfc-arch]: 1195 Authors: 1197 Paul Quinn (editor) 1198 Joel Halpern (editor) 1200 Contributors: 1202 Puneet Agarwal 1203 Andre Beliveau 1204 Kevin Glavin 1205 Ken Gray 1206 Jim Guichard 1207 Surendra Kumar 1208 Darrel Lewis 1209 Nic Leymann 1210 Rajeev Manur 1211 Thomas Nadeau 1212 Carlos Pignataro 1213 Michael Smith 1214 Navindra Yadav 1216 Acknowledgements: 1218 The authors would like to thank David Ward, Abhijit Patra, 1219 Nagaraj Bagepalli, Darrel Lewis, Ron Parker, Lucy Yong and 1220 Christian Jacquenet for their review and comments. 1222 8. IANA Considerations 1224 [RFC Editor: please remove this section prior to publication.] 1226 This document has no IANA actions. 1228 9. Informative References 1230 [I-D.boucadair-sfc-framework] 1231 Boucadair, M., Jacquenet, C., Parker, R., Lopez, D., 1232 Guichard, J., and C. Pignataro, "Service Function 1233 Chaining: Framework & Architecture", draft-boucadair-sfc- 1234 framework-02 (work in progress), February 2014. 1236 [I-D.ietf-sfc-problem-statement] 1237 Quinn, P. and T. Nadeau, "Service Function Chaining 1238 Problem Statement", draft-ietf-sfc-problem-statement-10 1239 (work in progress), August 2014. 1241 [I-D.quinn-sfc-arch] 1242 Quinn, P. and J. Halpern, "Service Function Chaining (SFC) 1243 Architecture", draft-quinn-sfc-arch-05 (work in progress), 1244 May 2014. 1246 [RFC3022] Srisuresh, P. and K. Egevang, "Traditional IP Network 1247 Address Translator (Traditional NAT)", RFC 3022, January 1248 2001. 1250 [RFC6146] Bagnulo, M., Matthews, P., and I. van Beijnum, "Stateful 1251 NAT64: Network Address and Protocol Translation from IPv6 1252 Clients to IPv4 Servers", RFC 6146, April 2011. 1254 [RFC6296] Wasserman, M. and F. Baker, "IPv6-to-IPv6 Network Prefix 1255 Translation", RFC 6296, June 2011. 1257 Authors' Addresses 1259 Joel Halpern (editor) 1260 Ericsson 1262 Email: jmh@joelhalpern.com 1264 Carlos Pignataro (editor) 1265 Cisco Systems, Inc. 1267 Email: cpignata@cisco.com