idnits 2.17.1 draft-bernardos-sfc-discovery-07.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 (October 21, 2021) is 918 days in the past. Is this intentional? Checking references for intended status: Experimental ---------------------------------------------------------------------------- == Outdated reference: A later version (-10) exists of draft-bernardos-sfc-fog-ran-09 Summary: 0 errors (**), 0 flaws (~~), 2 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 SFC WG CJ. Bernardos 3 Internet-Draft UC3M 4 Intended status: Experimental A. Mourad 5 Expires: April 24, 2022 InterDigital 6 October 21, 2021 8 Service Function discovery in fog environments 9 draft-bernardos-sfc-discovery-07 11 Abstract 13 Service function chaining (SFC) allows the instantiation of an 14 ordered set of service functions and subsequent "steering" of traffic 15 through them. Service functions provide an specific treatment of 16 received packets, therefore they need to be known so they can be used 17 in a given service composition via SFC. This document discusses the 18 need for service function discovery mechanisms and propose some 19 solutions for sfc-aware nodes to discover available service functions 20 in fog environments. 22 Status of This Memo 24 This Internet-Draft is submitted in full conformance with the 25 provisions of BCP 78 and BCP 79. 27 Internet-Drafts are working documents of the Internet Engineering 28 Task Force (IETF). Note that other groups may also distribute 29 working documents as Internet-Drafts. The list of current Internet- 30 Drafts is at https://datatracker.ietf.org/drafts/current/. 32 Internet-Drafts are draft documents valid for a maximum of six months 33 and may be updated, replaced, or obsoleted by other documents at any 34 time. It is inappropriate to use Internet-Drafts as reference 35 material or to cite them other than as "work in progress." 37 This Internet-Draft will expire on April 24, 2022. 39 Copyright Notice 41 Copyright (c) 2021 IETF Trust and the persons identified as the 42 document authors. All rights reserved. 44 This document is subject to BCP 78 and the IETF Trust's Legal 45 Provisions Relating to IETF Documents 46 (https://trustee.ietf.org/license-info) in effect on the date of 47 publication of this document. Please review these documents 48 carefully, as they describe your rights and restrictions with respect 49 to this document. Code Components extracted from this document must 50 include Simplified BSD License text as described in Section 4.e of 51 the Trust Legal Provisions and are provided without warranty as 52 described in the Simplified BSD License. 54 Table of Contents 56 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 57 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 3 58 3. Problem statement . . . . . . . . . . . . . . . . . . . . . . 4 59 3.1. Discovery of SF in a multi-provider fog/edge environment 4 60 4. Network-based SF discovery . . . . . . . . . . . . . . . . . 5 61 4.1. ICMPv6-based SF discovery . . . . . . . . . . . . . . . . 8 62 4.2. DHCPv6-based SF discovery . . . . . . . . . . . . . . . . 8 63 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 8 64 6. Security Considerations . . . . . . . . . . . . . . . . . . . 8 65 7. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 8 66 8. Informative References . . . . . . . . . . . . . . . . . . . 8 67 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 8 69 1. Introduction 71 Virtualization of functions provides operators with tools to deploy 72 new services much faster, as compared to the traditional use of 73 monolithic and tightly integrated dedicated machinery. As a natural 74 next step, mobile network operators need to re-think how to evolve 75 their existing network infrastructures and how to deploy new ones to 76 address the challenges posed by the increasing customers' demands, as 77 well as by the huge competition among operators. All these changes 78 are triggering the need for a modification in the way operators and 79 infrastructure providers operate their networks, as they need to 80 significantly reduce the costs incurred in deploying a new service 81 and operating it. Some of the mechanisms that are being considered 82 and already adopted by operators include: sharing of network 83 infrastructure to reduce costs, virtualization of core servers 84 running in data centers as a way of supporting their load-aware 85 elastic dimensioning, and dynamic energy policies to reduce the 86 monthly electricity bill. However, this has proved to be tough to 87 put in practice, and not enough. Indeed, it is not easy to deploy 88 new mechanisms in a running operational network due to the high 89 dependency on proprietary (and sometime obscure) protocols and 90 interfaces, which are complex to manage and often require configuring 91 multiple devices in a decentralized way. 93 Service Functions are widely deployed and essential in many networks. 94 These Service Functions provide a range of features such as security, 95 WAN acceleration, and server load balancing. Service Functions may 96 be instantiated at different points in the network infrastructure 97 such as data center, the WAN, the RAN, and even on mobile nodes. 99 Service functions (SFs), also referred to as VNFs, or just functions, 100 are hosted on compute, storage and networking resources. The hosting 101 environment of a function is called Service Function Provider or 102 NFVI-PoP (using ETSI NFV terminology). 104 With the arrival of virtualization, the deployment model for service 105 function is evolving to one where the traffic is steered through the 106 functions wherever they are deployed (functions do not need to be 107 deployed in the traffic path anymore). For a given service, the 108 abstracted view of the required service functions and the order in 109 which they are to be applied is called a Service Function Chain 110 (SFC). An SFC is instantiated through selection of specific service 111 function instances on specific network nodes to form a service graph: 112 this is called a Service Function Path (SFP). The service functions 113 may be applied at any layer within the network protocol stack 114 (network layer, transport layer, application layer, etc.). 116 A mobile terminal can benefit from using service function chaining at 117 the edge/fog to enhance existing applications or to enable new ones. 118 In order to do so, discovery of available service functions is 119 required. This document focuses on this aspect. 121 2. Terminology 123 The following terms used in this document are defined by the IETF in 124 [RFC7665] and [RFC9015]: 126 Service Function (SF): a function that is responsible for specific 127 treatment of received packets (e.g., firewall, load balancer). 129 Service Function Chain (SFC): for a given service, the abstracted 130 view of the required service functions and the order in which they 131 are to be applied. This is somehow equivalent to the Network 132 Function Forwarding Graph (NF-FG) at ETSI. 134 Service Function Forwarder (SFF): A service function forwarder is 135 responsible for forwarding traffic to one or more connected 136 service functions according to information carried in the SFC 137 encapsulation, as well as handling traffic coming back from the 138 SF. 140 SFI: SF instance. 142 Service Function Path (SFP): the selection of specific service 143 function instances on specific network nodes to form a service 144 graph through which an SFC is instantiated. 146 A Service Function Type (SFT) that is the category of Service 147 Function that is provided (such as "firewall"). 149 3. Problem statement 151 [RFC7665] describes an architecture for the specification, creation, 152 and ongoing maintenance of Service Function Chains (SFCs) in a 153 network. It includes architectural concepts, principles, and 154 components used in the construction of composite services through 155 deployment of SFCs. In this architecture, a key element is the 156 service function (SF), which is a function that is responsible for 157 specific treatment of received packets (e.g., a firewall). 159 So far, how the SFs are discovered and composed has been out of the 160 scope of discussions in IETF. There is however a need to define 161 mechanisms that allow SF discovery in fog environments 162 [I-D.bernardos-sfc-fog-ran]. Note that the mechanisms described in 163 this document address fog environments. There are other mechanisms 164 described, like [RFC9015], that cover generic SF discovery in more 165 traditional environments. Some of the solutions described in the 166 present document might be of applicable to other scenarios as well. 168 3.1. Discovery of SF in a multi-provider fog/edge environment 170 The need to provide networking, computing, and storage capabilities 171 closer to the users has recently emerged, due to the demands from 5G 172 applications of very low latency, leading to what is known today as 173 the concept of intelligent edge. ETSI has been the first to address 174 this need recently by developing the framework of mobile edge 175 computing (MEC). Such an intelligent edge could not be envisaged 176 without virtualization. Beyond applications, it raises a clear 177 opportunity for networking functions to execute at the edge 178 benefiting from inherent low latencies. Being in close proximity to 179 the access, the edge becomes an attractive place for hosting 180 different functions, saving bandwidth in their respective domains and 181 offering local breakout options where required. Whilst it is 182 appreciated the particular challenge for the intelligent edge concept 183 in dealing with mobile users, the edge virtualization substrate has 184 been largely assumed to be fixed or stationary. Although little 185 developed, the intelligent edge concept is being extended further to 186 scenarios where for example the edge computing substrate is on the 187 move, e.g., on-board a car or a train, or that it is distributed 188 further down the edge, even integrating resources from different 189 stakeholders, into what is known as the fog. 191 Service composition is a powerful tool which can provide significant 192 benefits when applied in a softwarized network environment. While it 193 is being explored in the core part of networks to compose services 194 using DPIs (Deep Packet Inspections), firewalls, parental control, 195 video accelerators, etc., its applicability to the RAN (Radio Access 196 Network), and in particular to the edge and the fog, has not been 197 explored yet. 199 Running functions (standalone functions or service function chains) 200 at the edge of the network has clear advantages. For example, it 201 enables offloading functions from the end-user terminal so that it 202 can become more efficient in terms of cost and energy consumption. 204 A mobile terminal can benefit from using service function chaining at 205 the edge/fog to enhance existing applications or to enable new ones. 206 Some examples of such applications are: privacy enhancement by local 207 anchoring, opportunistic local breakout, assisted encryption, video 208 transcoding, personal firewalling, etc. The mobile terminal might 209 look for function hosting opportunities at the edge for various 210 reasons such as: 212 o to increase battery life in critical situations by offloading 213 energy demanding operations (e.g., video transcoding, augmented 214 reality) to the edge/cloud; 216 o to reduce communications latency (e.g., by using local breakout at 217 the edge for selected applications demanding low latency); 219 o to enable new functions (e.g., privacy improvements, personal 220 firewalling) which demand additional intelligence/resources at the 221 network; 223 o to benefit from context information available at the edge (e.g., 224 enrich networking decisions by executing functions at the edge 225 using RAN information); 227 Several key challenges need to be addressed to enable controlled 228 service function chaining for a mobile terminal, and one of them is 229 the discovery of the functions available for use at the Fog/Edge/ 230 Cloud. 232 4. Network-based SF discovery 234 In this section we describe several mechanisms for a mobile SFC-aware 235 node to discover what SFs are available in the network. Different 236 alternatives (protocol containers) are considered to enable the 237 mobile node to obtain the following information per SF available: 239 o Service Function Type, identifying the category of SF provided. 241 o SFC-aware: Yes/No. Indicates if the SF is SFC-aware. 243 o Route Distinguisher (RD): IP address indicating the location of 244 the SF(I). 246 o Pricing/costs details. 248 o Migration capabilities of the SF: whether a given function can be 249 moved to another provider (potentially including information about 250 compatible providers topologically close). 252 o Mobility of the device hosting the SF, with e.g. the following 253 sub-options: 255 Level: no, low, high; or a corresponding scale (e.g., 1 to 10). 257 Current geographical area (e.g., GPS coordinates, post code). 259 Target moving area (e.g., GPS coordinates, post code). 261 o Power source of the device hosting the SF, with e.g. the following 262 sub-options: 264 Battery: Yes/No. If Yes, the following sub-options could be 265 defined: 267 Capacity of the battery (e.g., mmWh). 269 Charge status (e.g., %). 271 Lifetime (e.g., minutes). 273 Figure 1 shows the generic mechanism for SF discovery, with network 274 support. In this scenario, SFs (which might belong to different 275 administrative domains) are previously registered at the network, 276 which can then reply to requests sent from mobile nodes that have 277 just attached to the network. A request might optionally include the 278 SFs of interest for the terminal, instead of a request for all known 279 SFs. 281 The network might also send periodic advertisements in addition to 282 responses to solicited requests. These responses/advertisements 283 include the information about known SFs (or only about the ones 284 queried by the terminal), which can then be used by the terminal to 285 decide whether to use (some of) them in a certain SFC. How the 286 mobile terminal then configures this SFC is not covered in this 287 document. 289 ___________ 290 _( )_ 291 _( SF1 SF2 )_ 292 ------------ ----------- _( SF3 )_ 293 | terminal | | network |-(_ SF4 SF5 _) 294 ------------ ----------- (_ SF6 SF7 _) 295 | | (_ SF8 _) 296 XXX (1. attachment) | (___________) 297 | | 298 +---2. Request (optional)------>| 299 | | 300 |<--------3. Response/Advert.---| 301 | (SF1,SF2...,SF8) | 302 | | 304 Figure 1: SF (network) discovery 306 In addition to the discovery of SFs at the infrastructure, mobile 307 terminals can also host SF(I)s, and therefore they also need to be 308 discovered. A similar approach can be followed, as showin in 309 Figure 2. 311 ------------ 312 | SF3 | 313 ------------ | terminal | ------------ 314 | SF1 SF2 | ------------ | SF4 SF5 | 315 | terminal | | | terminal | 316 ------------ | ------------ 317 | | | 318 +--1. Request->+-------------->| 319 | (SF1,SF2) | | 320 |<----------------2. Response--+ 321 | (SF4,SF5) | 322 |<-2. Response-+ | 323 | (SF3) | | 324 | | | 326 Figure 2: SF (mobiles) discovery 328 SFs might belong to different administrative domains. This might 329 require the use of additional security and authentication mechanisms. 330 Policies can be used (both in single and multi-domain scenarios) to 331 adapt/limit the type and number of SFs that are advertised, depending 332 on the relationship of the requester and the advertiser. 334 Next sections describe different protocol alternatives for this SF 335 discovery in fog environments. 337 4.1. ICMPv6-based SF discovery 339 TBD. 341 4.2. DHCPv6-based SF discovery 343 TBD. 345 5. IANA Considerations 347 N/A. 349 6. Security Considerations 351 TBD. 353 7. Acknowledgments 355 The work in this draft has been explored under the framework of the 356 H2020 5G-DIVE project (Grant 859881). 358 8. Informative References 360 [I-D.bernardos-sfc-fog-ran] 361 Bernardos, C. J., Rahman, A., and A. Mourad, "Service 362 Function Chaining Use Cases in Fog RAN", draft-bernardos- 363 sfc-fog-ran-09 (work in progress), March 2021. 365 [RFC7665] Halpern, J., Ed. and C. Pignataro, Ed., "Service Function 366 Chaining (SFC) Architecture", RFC 7665, 367 DOI 10.17487/RFC7665, October 2015, 368 . 370 [RFC9015] Farrel, A., Drake, J., Rosen, E., Uttaro, J., and L. 371 Jalil, "BGP Control Plane for the Network Service Header 372 in Service Function Chaining", RFC 9015, 373 DOI 10.17487/RFC9015, June 2021, 374 . 376 Authors' Addresses 377 Carlos J. Bernardos 378 Universidad Carlos III de Madrid 379 Av. Universidad, 30 380 Leganes, Madrid 28911 381 Spain 383 Phone: +34 91624 6236 384 Email: cjbc@it.uc3m.es 385 URI: http://www.it.uc3m.es/cjbc/ 387 Alain Mourad 388 InterDigital Europe 390 Email: Alain.Mourad@InterDigital.com 391 URI: http://www.InterDigital.com/