idnits 2.17.1 draft-galis-precision-netslices-problem-statement-00.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 == Line 163 has weird spacing: '...virtual infra...' == Line 164 has weird spacing: '...diverse resou...' -- The document date (November 4, 2018) is 1972 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Missing Reference: 'RFC1958' is mentioned on line 704, but not defined == Missing Reference: 'RFC3439' is mentioned on line 707, but not defined == Missing Reference: 'RFC3234' is mentioned on line 711, but not defined == Missing Reference: 'RFC7665' is mentioned on line 654, but not defined == Missing Reference: 'SFC WG' is mentioned on line 102, but not defined == Missing Reference: 'RFC7149' is mentioned on line 714, but not defined == Missing Reference: 'SC6' is mentioned on line 174, but not defined == Missing Reference: 'I-D.leeking-actn-problem-statement 03' is mentioned on line 658, but not defined == Missing Reference: 'I-D.teas-actn-requirements-04' is mentioned on line 663, but not defined == Missing Reference: 'IETF-Slicing1' is mentioned on line 668, but not defined == Missing Reference: 'IETF-Slicing2' is mentioned on line 674, but not defined == Missing Reference: 'IETF-Slicing3' is mentioned on line 679, but not defined == Missing Reference: 'IETF-Slicing4' is mentioned on line 684, but not defined == Missing Reference: 'IETF-Slicing5' is mentioned on line 690, but not defined == Missing Reference: 'IETF-Mobility' is mentioned on line 726, but not defined == Missing Reference: 'IETF-Coding' is mentioned on line 736, but not defined == Missing Reference: 'IETF-Anchoring' is mentioned on line 741, but not defined == Missing Reference: 'RFC2119' is mentioned on line 695, but not defined == Missing Reference: 'RFC4364' is mentioned on line 700, but not defined == Missing Reference: 'I-D.dong-network-slicing-problem-statement' is mentioned on line 644, but not defined == Missing Reference: 'I-D.galis-anima-autonomic-slice-networking' is mentioned on line 649, but not defined == Missing Reference: 'SFG WG' is mentioned on line 719, but not defined == Missing Reference: 'CPP' is mentioned on line 722, but not defined == Missing Reference: 'IETF-Virtualization' is mentioned on line 731, but not defined == Missing Reference: 'RFC6291' is mentioned on line 746, but not defined == Unused Reference: 'ITU-T 2011' is defined on line 772, but no explicit reference was found in the text == Unused Reference: 'GUERZONI-2016' is defined on line 783, but no explicit reference was found in the text == Unused Reference: 'IMT2020-2016bis' is defined on line 798, but no explicit reference was found in the text == Unused Reference: 'KARL-2016' is defined on line 803, but no explicit reference was found in the text == Unused Reference: 'NGMN-2016' is defined on line 814, but no explicit reference was found in the text == Unused Reference: 'NGS-3GPP-2016' is defined on line 819, but no explicit reference was found in the text == Unused Reference: 'ONF-2016' is defined on line 824, but no explicit reference was found in the text == Unused Reference: 'GRAMMATIKOU-2012' is defined on line 837, but no explicit reference was found in the text == Unused Reference: 'GAL' is defined on line 845, but no explicit reference was found in the text == Unused Reference: 'GAPS' is defined on line 850, but no explicit reference was found in the text == Unused Reference: 'NS ARCH' is defined on line 857, but no explicit reference was found in the text Summary: 0 errors (**), 0 flaws (~~), 39 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 NMRG A. Galis 3 Internet-Draft University College London 4 Intended Status: Informational et al. 5 Expires: May 4, 2019 November 4, 2018 7 Management of Precision Network Slicing - Problem Statement 8 draft-galis-precision-netslices-problem-statement-00 10 Abstract 12 This document introduces Precision Network Slicing Management 13 problems and their context. It represents an initial review of the 14 Management of Network Slicing problem statement derived from the 15 analysis of the technical gaps in IETF protocols ecosystem. It 16 complements and brings together the efforts being carried out in 17 several other IETF working groups covering certain aspects of Network 18 Slicing management functions and operations. 20 Status of this Memo 22 This Internet-Draft is submitted in full conformance with the 23 provisions of BCP 78 and BCP 79. 25 Internet-Drafts are working documents of the Internet Engineering 26 Task Force (IETF). Note that other groups may also distribute 27 working documents as Internet-Drafts. The list of current Internet- 28 Drafts is at https://datatracker.ietf.org/drafts/current/. 30 Internet-Drafts are draft documents valid for a maximum of six months 31 and may be updated, replaced, or obsoleted by other documents at any 32 time. It is inappropriate to use Internet-Drafts as reference 33 material or to cite them other than as "work in progress." 35 This Internet-Draft will expire on May 4, 2019. 37 Copyright Notice 39 Copyright (c) 2018 IETF Trust and the persons identified as the 40 document authors. All rights reserved. 42 This document is subject to BCP 78 and the IETF Trust's Legal 43 Provisions Relating to IETF Documents 44 (https://trustee.ietf.org/license-info) in effect on the date of 45 publication of this document. Please review these documents 46 carefully, as they describe your rights and restrictions with respect 47 to this document. Code Components extracted from this document must 48 include Simplified BSD License text as described in Section 4.e of 49 the Trust Legal Provisions and are provided without warranty as 50 described in the Simplified BSD License. 52 Table of Contents 54 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 2 55 1.1 Early Definitions of Slicing . . . . . . . . . . . . . . . . 3 56 1.2 Definition of Terms . . . . . . . . . . . . . . . . . . . . 5 57 1.2.1 Roles . . . . . . . . . . . . . . . . . . . . . . . . . 5 58 1.2.2 Key Terms . . . . . . . . . . . . . . . . . . . . . . . 5 59 1.2.3 Slicing and Sharing of Resources . . . . . . . . . . . . 6 60 1.3 Precision Network Slicing Value Characteristics . . . . . . 7 61 1.4 Precision Network Slicing Work Scope . . . . . . . . . . . . 9 62 2. Management of Precision Network Slicing - Selected Problems 63 and Work Areas . . . . . . . . . . . . . . . . . . . . . . . . 11 64 2.1 Overall management aspects, APIs and functionality of 65 network slices. . . . . . . . . . . . . . . . . . . . . . . 11 66 2.2 Slice Management Characteristics, Capabilities and 67 Assurances. . . . . . . . . . . . . . . . . . . . . . . . . 12 68 2.3 E2E Network Slicing. . . . . . . . . . . . . . . . . . . . . 13 69 3 Security Considerations . . . . . . . . . . . . . . . . . . . . 14 70 4 IANA Considerations . . . . . . . . . . . . . . . . . . . . . . 14 71 5 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . 14 72 6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 73 6.1 IETF References . . . . . . . . . . . . . . . . . . . . . . 14 74 6.2 Informative References . . . . . . . . . . . . . . . . . . 16 75 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 18 77 1 Introduction 79 Network slicing (NS) is an approach to flexible isolation and 80 allocation of network resources and network functions for a logical 81 network instance, providing a high level of such network 82 customization and quality service guarantee that includes also such 83 customized reliability and security levels. NS transforms the 84 networking perspective by abstracting, isolating, orchestrating, 85 softwarizing, and separating logical network components from the 86 underlying physical network supporting the introduction of new 87 network architectures ([RFC1958], [RFC3439], [RFC3234]) and new 88 service delivery [5G-ICN]. In general, a particular network slice 89 consists of a union of subsets of (connectivity, storage, computing) 90 resources & (Virtual) Network Functions & Service Functions [RFC7665] 91 at the data & control & management planes at a given time that are 92 managed together to provide a logical networking infrastructure to 93 support a single service or a set of services. 95 NS enables the dynamic and concurrent deployment of multiple logical, 96 self-contained and independent, logical networks on a common 97 infrastructure. 99 The management plane allocates a group of network resources network 100 resources can be physical, virtual or a combination thereof), it 101 connects with the physical and virtual network and service functions 102 ([SFC WG]) as appropriate, and it instantiates all of the network and 103 service functions assigned to a slice. On the other hand, for slice 104 operations, the slice management plane functionality that may be 105 operated by slice tenant takes over the control and governing of all 106 the network resources, network functions, and service functions 107 assigned to the slice. It (re-) configures them as appropriate and as 108 per elasticity needs, in order to provide an end-to-end service. In 109 particular, slice ingress routers are configured, so that appropriate 110 traffic is bound to the relevant slice. 112 Network operators can use NS to enable different services to receive 113 different treatment and to allow the allocation and release of 114 network resources according to the context and contention policy of 115 the operators. Such an approach using NS would allow a significant 116 reduction of the operations expenditure. In addition, there is an 117 enabling synergy between NS and softwarization. On the one hand, NS 118 makes possible softwarization, programmability ([RFC7149]), and the 119 innovation necessary to enrich the offered services. On the other 120 hand, Network softwarization techniques [IMT2020-2015], [IMT2020- 121 2016] may be used to realize and manage [MANO-2014] network slicing. 122 NS provides the means for the network operators to provide network 123 programmable capabilities to both service providers and other market 124 players without changing their physical infrastructure. 126 Slices may support dynamic multiple services, multi- tenancy and the 127 integration means for vertical market players (e.g. the automotive 128 industry, energy industry, healthcare industry, media and the 129 entertainment industry, etc.) 131 1.1 Early Definitions of Slicing 133 The followings are early definitions of slicing: 135 (i) Active / Programmable Networks research Node operating systems & 136 resource control frameworks (1995 - 2005) [Programmable 137 Networks] 139 (ii) Federated Testbed research: Planet Lab USA (2002), PlanetLab EU 140 (2005), OneLab EU (2007), PlanetLab Japan (2005), OpenLab EU 141 (2012). 143 (iii) GENI Slice (2009): "GENI [GENI-2009] is a shared network 144 testbed, i.e., multiple experimenters may be running multiple 145 experiments at the same time. A GENI slice is: 147 o The unit of isolation for experiments. 149 o A container for resources used in an experiment. GENI 150 experimenters add GENI resources (compute resources, network 151 links, etc..) to slices, and run experiments that use these 152 resources. 154 o A unit of access control. The experimenter that creates a 155 slice can determine which project members have access to the 156 slice, i.e., are members of the slice. 158 (iv) Slice capabilities (2009) [ChinaCom-2009] 160 o 3 Slices Capabilities: "Resource allocation to virtual 161 infrastructures or slices of virtual infrastructure."; 162 "Dynamic creation and management of virtual 163 infrastructures/slices of virtual infrastructure across 164 diverse resources."; "Dynamic mapping and deployment of a 165 service on a virtual infrastructure/slices of virtual 166 infrastructure." 168 o 17 Orchestration capabilities. 170 o 19 Self-functionality mechanisms. 172 o 14 Self-functionality infrastructure capabilities. 174 (v) ITU-T Slicing (2011) as defined in [SC6], it is the basic concept 175 of the Network Softwarization. Slicing allows logically isolated 176 network partitions (LINP) with a slice being considered as a 177 unit of programmable resources such as network, computation, and 178 storage. 180 (vi) NGMN Slice capabilities (2016) [NGMN 2016] consist of 3 layers: 182 1) Service Instance Layer, 2) Network Slice Instance Layer, and 183 3) Resource layer. 185 o The Service Instance Layer represents the services (end-user 186 service or business services), which are to be supported. Each 187 service is represented by a Service Instance. Typically, 188 services can be provided by the network operator or by 3rd 189 parties. 191 o A Network Slice Instance provides the network characteristics, 192 which are required by a Service Instance. A Network Slice 193 Instance may also be shared across multiple Service Instances 194 provided by the network operator. 196 (vii) 3GPP - GPP TR23.799 Study Item "Network Slicing' 2016. 198 (viii) ONF Recommendation TR-526 "Applying SDN architecture to 199 Network Slicing", 2016. 201 Additional characteristics, standard and research activities on 202 Infrastructure slicing and references are presented in [NS Tutorial 203 2018]. 205 1.2 Definition of Terms 207 1.2.1 Roles 209 Resource Provider - It owns the physical resources and infrastructure 210 (network/ cloud/ datacenter) and provides / leases them to 211 operators. 213 Slice Provider - A slice provider is an entity that has appropriate 214 tools for the lifecycle management of network slices. Typically, this 215 a telecommunication service provider that in most cases can also play 216 a role of the Resource Provider. 218 Slice Tenant - A slice tenant is the business owner of a specific 219 network/cloud/datacenter slice, in which customized services are 220 hosted. 222 Infrastructure slice tenants can make requests for the creation of 223 new slice through a service model. 225 1.2.2 Key Terms 227 Network Slice - A set of infrastructures (network, cloud, data 228 center) run-time network functions, infrastructure resources (i.e., 229 managed connectivity, compute, storage resources) and service 230 functions that have attributes specifically designed to meet the 231 needs of an industry vertical or a service. 233 As such a Network Slice is a managed group of subsets of resources, 234 run-time network functions/network virtual functions at the data, 235 control, management/orchestration, and service planes at any given 236 time. The behavior of the Network Slice is realized via network slice 237 instances (i.e., activated slices, dynamically and non-disruptively 238 re-provisioned). The Network Slice key characteristics are provided 239 below: 241 o A Network Slice supports at least one type of service. 243 o A Network Slice may consist of cross-domain components from 244 separate domains in the same or different administrations, or 245 components applicable to the infrastructure. 247 o A resource-only partition is one of the components of a Cloud 248 Network Slice, however on its own does not fully represent a 249 Network Slice. It can be seen as a slice substrate. 251 o A collection of partitions from separate domains is combined and 252 aggregated to form a cloud / network slice. 254 o Underlays / overlays supporting all services equally (with 'best 255 effort" support) are not fully representing a Network Slice. 257 Precision Network Slices - a network slice which guarantees QoS 258 characteristics (e.g. low latency) and/or KPIs (Key Performance 259 Indicators). 261 Network Slicing: Network slicing is a technology or an approach to 262 create separate network slices in support of services, depending on 263 several requirements, on the same physical resources. This is 264 possible by combinations of several network technologies. 266 End-to-End Network Slice (E2E-NS): An E2E-NS is a virtual network 267 connecting between end points of a number of NS subnets (i.e. single 268 domain slices). E2E slices are composed of a single NS subnet or 269 multiple NS subnets. 271 Network Slice as a Service (NSaaS): An NSaaS is a NS distribution 272 model in which a third-party provider can manage the lifecycle of NSs 273 and makes them available to customers. 275 1.2.3 Slicing and Sharing of Resources 277 o From a business point of view, a Network Slice includes a 278 combination of all relevant network and compute resources, 279 functions, and assets required to fulfil a specific business 280 case or service. 282 o From the infrastructure point of view, the infrastructure slice 283 instances require the partitioning and assignment of a set of 284 resources that can be used in an isolated, disjunctive or non- 285 disjunctive manner for that slice. 287 o From the tenant point of view, the infrastructure slice instance 288 provides different capabilities, specifically in terms of their 289 management and control capabilities, and how much of them the 290 network service provider hands over to the slice tenant. As 291 such, there are two types of slices: 293 (i) Internal slices, understood as the partitions used for 294 internal services of the provider, retaining full control 295 and management of them. 297 (ii) External slices, being those partitions hosting customer 298 services, appearing to the customer as dedicated 299 networks/clouds/data centers. 301 o From the management plane point of view, infrastructure slices 302 refer to the managed fully functional dynamically created 303 partitions of physical and/or virtual network resources, network 304 physical/virtual and service functions that can act as an 305 independent instance of a connectivity network and/or as a 306 network / cloud. 308 o From the date plane point of view, infrastructure slices refer 309 to dynamically created partitions of network forwarding devices 310 with guarantees for isolation, customization and security. 312 1.3 Precision Network Slicing Value Characteristics 314 As a differentiation from non-partition networks and those with 315 simple partitions of connectivity resources (e.g. VPNs)/ Virtual 316 Networks/Other abstractions of the data traffic layer, the following 317 Motivation and key value-added characteristics of Network Slicing and 318 the corresponding usage is identified: 320 o Precision Network slicing considerably transforms the networking 321 perspective by abstracting, isolating, orchestrating and 322 separating logical network behaviors from the underlying 323 physical network resources. 325 o Precision Network Slice is a dedicated network that is built on 326 an infrastructure mainly composed of, but not limited to, 327 connectivity, storage and computing. 329 o Each Precision Network Slice has the ability to dynamically 330 expose and possibly negotiate the parameters that characterize 331 an NS. 333 o Each Precision Network Slice will have its own operator/tenant 334 that sees it as a complete network infrastructure (i.e. router 335 instances, programmability, using any appropriate communication 336 protocol, caches, provide dynamic placement of virtual network 337 functions according to traffic patterns, to use its own 338 controller, finally, it can manage its network as its own). 340 o Provision Network slicing supports tenants that are strongly 341 independent on infrastructure. 343 o A Precision Network Slicing aware infrastructures allows 344 operators to use part of the resources to meet stringent 345 resource requirements. 347 o Precision Network slicing introduces an additional layer of 348 abstraction by the creation of logically or physically isolated 349 groups of network resources and network function/virtual network 350 functions configurations separating its behavior from the 351 underlying physical network. 353 o Precision Network slicing covers the full life cycle of slices 354 that are managed groups of infrastructure resources, network 355 functions and services (e.g. the network slice components are: 356 service instance, a network functions instance, resources, slice 357 manager, capability exposure and guarantees for QoS 358 characteristics and/or KPIs). 360 o Precision Network slices are dynamically and non-disruptively 361 reprovisioned. 363 o Precision Network slices will need to be as far as possible 364 self-managed by automated, autonomic and autonomous, systems in 365 order to cope with dynamic requirements, such as scalability or 366 extensibility of an infrastructure (organically 367 growing/shrinking of resources to meet the size of their 368 organizations). 370 o Precision Network slices are configurable and programmable, and 371 they have the ability to expose their capabilities and 372 characteristics. The slice protocols and functions are selected 373 according to slice required features. The behavior of the 374 network slice realized via network slice instance(s). 376 o Precision Network slices are concurrently deployed as multiple 377 logical, self-contained and independent, partitioned network 378 functions and resources on common physical infrastructure. 380 o Network slicing supports dynamic multi-services, multi-tenancy 381 and the means for backing vertical market players. 383 o Network slicing simplifies the provisioning of services 384 manageability of networks and integration and operational 385 challenges especially for supporting communication services. 387 o Precision Network slicing offers native service customization 388 enabled by the selection and configuration of network functions 389 for coordinating/orchestration and control of network 390 resources. 392 o Precision Network Slicing Capability exposure: providers can 393 offer Application Programming Interfaces (APIs) to the vertical 394 business customers for granting the capability of managing their 395 own slices and for the purpose of building of advanced services 396 on top of services offered by the network slice. Such management 397 actions can include dimensioning, configuration, etc. 399 o Hosting applications: providers offer the capability of hosting 400 virtualized versions of network functions or applications, 401 including the activation of the necessary monitoring information 402 for those functions. 404 o Hosting on-demand 3rd parties: empower partners (3rd parties / 405 OTTs) to directly make offers to the end customers augmenting 406 Operator network or other value creation capabilities. 408 1.4 Precision Network Slicing Work Scope 410 The purpose of the NS work in IETF is to develop a set of protocols 411 and/or protocol extensions that enable efficient slice lifecycle 412 management (creation, activation / update /deactivation), slice 413 composition, inter-slice operations (for subslices concatenation, 414 that includes slice discovery and description) slice orchestration, 415 overall network slicing system management, providing slice isolation= 416 as well as management of slice related KPIs (according to SLA), and 417 safe and secure operations within a connectivity network or network 418 cloud / data center environments [NECOS]. 420 While there are, isolated efforts being carried out in several IETF 421 working groups Network WG [I-D.leeking-actn-problem-statement 03], 422 TEAS WG [I-D.teas-actn-requirements-04], [I-D.dong-network-slicing- 423 problem-statement], ANIMA WG [I-D.galis-anima-autonomic-slice- 424 networking], [IETF-Slicing1], [IETF-Slicing2], [IETF-Slicing3], 425 [IETF-Slicing4], [IETF-Slicing5], [IETF-Mobility], [IETF- 426 Virtualization], [IETF-Coding], [IETF-Anchoring] to achieve certain 427 aspects of network slice functions and operations, there is a clear 428 need to look at the complete life-cycle management characteristics of 429 Network Slicing solutions though the discussions based on the 430 following architectural tenets: 432 o Underlay tenet: support for an IP/MPLS-based underlay data Plane 433 (including segment routing). 435 o Governance tenet: a logically centralized authority for network 436 slices in a domain. 438 o Separation tenet: slices may be virtually or physically 439 independent of each other and have an appropriate degree of 440 isolation (note 1) from each other what includes isolation of 441 each slice management systems. 443 o Capability exposure tenet: each slice allows third parties to 444 access via dedicated interfaces and /or APIs and /or programming 445 methods information regarding services provided by the slice 446 (e.g., connectivity information, mobility, autonomicity, etc.) 447 within limits set by the operator or the slice owner. 449 NS approaches that do not adhere to these tenets are explicitly 450 outside of the scope of the proposed work at IETF. 452 In pursuit of the solutions described above, there is a need to 453 document architecture for network slicing within both wide area 454 network and edge/central data center environments. 456 Elicitation of requirements (examples are [RFC2119], [RFC4364]) for 457 both Network Slice control and management planes will be needed, 458 Facilitating the selection, extension, and/or development of the 459 protocols for each of the functional interfaces identified to support 460 the architecture. 462 Additionally, documentation on the common use-cases for slice 463 validation for 5G is needed, such as mission-critical ultra-low 464 latency communication services; massive-connectivity machine 465 communication services (e.g. smart metering, smart grid and sensor 466 networks); extreme QoS; independent operations and management; 467 independent cost and/or energy optimization; independent multi- 468 topology routing; multi-tenant operations; multiple infrastructure 469 providers; new network architecture enablement, etc. 471 The proposed NS work in NMRG would be coordinated with other IETF WGs 472 (e.g. TEAS WG, DETNET WG, ANIMA WG, SFC WG, NETCONF WG, SUPA WG, NVO3 473 WG, DMM WG, Routing Area WG (RTGWG) to ensure that the commonalities 474 and differences in solutions are properly considered. Where suitable 475 protocols, models or methods exist, they will be preferred over 476 creating new ones. 478 2. Management of Precision Network Slicing - Selected Problems and Work 479 Areas 481 The goal of this proposed work is to develop one or more protocol 482 specifications (or extensions to existing protocols) to address 483 specific slicing problems that are not met by the existing tools. The 484 following problems were selected according to the analysis of the 485 technical gaps in the IETF protocols ecosystem. 487 2.1 Overall management aspects, APIs and functionality of network 488 slices. 490 These problems include: 492 Precision NS Life Cycle Management: (1) The management plane creates 493 the grouping of network resources (physical, virtual, or a 494 combination thereof), it connects with the physical and virtual 495 network and service functions, it instantiates all of the network and 496 service functions assigned to the slice and it activates assurance 497 loops for precision network functions (i.e. guaranties for QoS 498 characteristics and/or KPIs). (2) Template/NS repository assists 499 lifecycle management; (3) Resource Registrar manages exposed network 500 infrastructure capabilities; (4) NS Manager oversees individual slice 501 (with capability exposure to the NS Tenant); (5) Uniform Slice 502 lifecycle management: Slice lifecycle management including creation, 503 activation / deactivation, protection, elasticity, extensibility, 504 safety, and sizing of the slicing model per network and per network 505 cloud for slices in access, core and transport networks; for slices 506 in data centers/clouds/; (6) Automated instantiation, scaling and 507 resource reconfiguration of slices during slice lifetime. 509 E2E multi-domain Precision Orchestration (1) Coordination of any 510 number of inter-related resources in a number of subordinate domains, 511 and assurance of transactional integrity as part of the triggering 512 process and assurance of QoS characteristics and or KPIs; (2) 513 Automated control of slice lifecycle management, including discovery 514 and concatenation of slices in each segment of the infrastructure (in 515 data, control, and management planes); (3) Autonomic coordination and 516 triggering of slice elasticity and placement; (4) Coordination and 517 (re)-configuration of resources by taking over the control of all the 518 network functions; (5) reconfiguration of resources taking into 519 account e2e guarantees for QoS characteristics and/or KPIs 521 Full NS FCAPS: (1) Fault, Configuration, Accounting, Performance, 522 Security; (2) Monitoring Subsystem is responsible for monitoring 523 continuously the state all components of a NS; Monitoring Subsystem 524 receives the detailed service monitoring requests with references to 525 resource allocation and Network functions instances in a NS. (3) 526 Discovery and monitoring probes are needed of all NS components and 527 NS itself and for dynamic discovery of service with function 528 instances and their capability. 530 2.2 Slice Management Characteristics, Capabilities and Assurances. 532 These problems include: 534 Programmability and control of Network Slices; Capability exposure 535 for Network Slicing (allowing openness); with APIs for dynamic slice 536 management and interaction. 538 Autonomic slice management: (1) Network slice is a dynamic entity 539 with autonomic characteristics of its lifecycle and operations. (2) 540 The problem of efficient allocation of resources between slices 541 combined with real-time optimization of slice operations can only be 542 solved by continuous autonomic monitoring of slice performance and 543 making continuous autonomic adaptations. (3) Autonomic control of 544 slice life cycle management, including a concatenation of slices in 545 each segment or domain of the infrastructure (in data, control, and 546 management planes); 548 Slice Element Manager & Capability exposure / Key APIs: (1) 549 Description of exclusive control and/or management interfaces and 550 capabilities exposed for a network slice, enabling the deployment of 551 different logical network slices over shared resources; (2) 552 Description of the Slice Element Manager which guarantees a level of 553 service, according to a negotiated SLA between the customer and the 554 slice provider. 556 Guaranteed Isolation - (1) slice creation and deployment with 557 guarantees for separation in each of the Data / Control / Management 558 / Service planes. (2) Methods to enable diverse requirements for 559 slicing, including guarantees for the end-to-end QoS of a service 560 within a slice. 562 Guaranteed QoS characteristics and/or KPIs - (1) slice creation and 563 deployment with guarantees for QoS characteristics and/or KPIs in 564 each of the Data / Control / Management / Service planes. (2) Methods 565 to enable diverse requirements for slicing, including guarantees for 566 the end-to-end QoS characteristics and/or KPIs of a service within a 567 slice. 569 Service / data model & mapping (1) service mapping enables on-demand 570 processing anywhere in the physically distributed network, with 571 dynamic and fine granular service (re)-provisioning; (2) It includes 572 a slice-aware information model based on necessary connectivity, 573 storage, compute resources, network functions, capabilities exposed 574 and service elements. (3) Network Function as a Service; (4) Network 575 Slice as a Service; (5) Slice Network Functions as a Service. (6) 576 Slice Templates & Methods for the design of slices to different 577 scenarios in Vertical market players (such as the automotive 578 industry, energy industry, healthcare industry, media and 579 entertainment industry, holograms, etc.). This outlines an 580 appropriate slice template definition that may include capability 581 exposure of managed partitions of network resources (i.e. 582 connectivity compute and storage resources), physical and/or virtual 583 network and service functions that can act as an independent 584 connectivity network and/or as a network cloud. (7) The Economy of 585 Scale in Slicing: The benefits of slicing grow as the number of 586 service types that you are trying to launch grows. In addition 587 significant automation is needed to be able to do this at scale. 588 Otherwise the complexity and operational challenges are likely to 589 mount up. It's key objective that the provider gears up to support 590 this ambition in development, delivery and operations. 592 High level of recursion, namely methods for network slicing 593 segmentation allowing a slicing hierarchy with parent-child 594 relationships. 596 2.3 E2E Network Slicing. 598 These problems include: 600 E2E Network Slicing Scalability: Scalability: In order to partition 601 network resources in a scalable manner, it is required to clearly 602 define to what extent slice customers can be accommodated or not on a 603 given slice. The application of different SLAs on the offered 604 capabilities of management, control and customization of slices will 605 directly impact the management scalability issue. 607 E2E Precision Slicing (E2E Network Slices with guaranteed QoS / 608 KPIs)- E2E multiple logical, self-contained and independent, shared 609 or partitioned networks on a common infrastructure with guaranties 610 for QoS characteristics and /or KPIs (Key Performance Indicators). 612 E2E Network Slices Reliability - Maintaining the reliability of an 613 E2E network slice instance, which is being terminated, or after 614 resource changes in a subnet. 616 E2E Slice composition / decomposition: The stitching of slices is an 617 operation that modifies the functionality of an existing slice by 618 adding and merging functions of another slice (i.e. enhancing control 619 plane properties be functions defined in another slice template). 620 Stitching of slices is used to enrich slice services: (1) Slice 621 stitching operations are supported by uniform slice descriptors; (2) 622 Efficient stitching/ decomposition (vertically, horizontally, 623 vertically + horizontally). 625 3 Security Considerations 627 Security will be a major part of the design of network slicing. 629 4 IANA Considerations 631 This document requests no IANA actions. 633 5 Acknowledgements 635 Thanks to Kevin Smith (Vodafone), Satoru Matsushima (SoftBank), 636 Christian Jacquenet (Orange), Mohamed Boucadair (Orange) for their 637 contributions to this draft. This work was partially supported by the 638 EU project NECOS - "Novel Enablers for Cloud Slicing" [NECOS]. 640 6 References 642 6.1 IETF References 644 [I-D.dong-network-slicing-problem-statement] Dong, J. and S. Bryant, 645 "Problem Statement of Network Slicing in IP/MPLS 646 Networks", draft-dong-network-slicing-problem-statement-00 647 (work in progress), October 2016. 649 [I-D.galis-anima-autonomic-slice-networking] Galis, A., Makhijani, 650 K., and D. Yu, "Autonomic Slice Networking-Requirements 651 and Reference Model", draft-galis-anima-autonomic-slice- 652 networking-01 (work in progress), October 2016. 654 [RFC7665] Halpern, J., Pignataro, C., "Service Function Chaining 655 (SFC) Architecture", https://tools.ietf.org/html/rfc7665, 656 October 2015. 658 [I-D.leeking-actn-problem-statement 03] Ceccarelli, D., Lee, Y., 659 "Framework for Abstraction and Control of Traffic 660 Engineered Networks", draft-leeking-actn-problem- 661 statement-03 (work in progress), September 2014. 663 [I-D.teas-actn-requirements-04] Lee, Y., Dhody, D., Belotti, S., 664 Pithewan, K., Ceccarelli, D., "Requirements for 665 Abstraction and Control of TE Networks", draft-ietf-teas- 666 actn-requirements-04.txt, January 2017. 668 [IETF-Slicing1] "Presentations - Network Slicing meeting at IETF 97 669 of 15th November 2016", n.d., 670 . 674 [IETF-Slicing2] "Presentations - Network Slicing meeting at IETF 97 675 of 15th November 2016", n.d., 676 . 679 [IETF-Slicing3] "Presentations - Network Slicing meeting at IETF 97 680 of 15th November 2016", n.d., 681 . 684 [IETF-Slicing4] "Presentations - Network Slicing meeting at IETF 97 685 of 15th November 2016", n.d., 686 . 690 [IETF-Slicing5] "Presentations - Network Slicing meeting at IETF 97 691 of 15th November 2016", n.d., 692 . 695 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 696 Requirement Levels", BCP 14, RFC 2119, DOI 697 10.17487/RFC2119, March 1997, . 700 [RFC4364] Rosen, E. and Y. Rekhter, "BGP/MPLS IP Virtual Private 701 Networks (VPNs)", RFC 4364, DOI 10.17487/RFC4364, February 702 2006, . 704 [RFC1958] Carpenter, B., "Architectural Principles of the Internet", 705 RFC 1958, . 707 [RFC3439] Bush, R., Meyer, D., "Some Internet Architectural 708 Guidelines and Philosophy", RFC3439, 709 . 711 [RFC3234] Carpenter, B., Brim S., "Middleboxes: Taxonomy and Issues", 712 RFC3439, . 714 [RFC7149] Boucadair, M., Jacquenet, C. , " Software-Defined 715 Networking: A Perspective from within a Service Provider 716 Environment", RFC 7149, March 2014 717 . 719 [SFG WG] "Service Function Chaining WG" 720 . 722 [CPP] Boucadair M., Jacquenet, C., Wang, N., "IP Connectivity 723 Provisioning Profile (CPP)" 724 https://tools.ietf.org/html/rfc7297 726 [IETF-Mobility] Truong-Xuan Do, Young-Han Kim, "Architecture for 727 delivering multicast mobility services using network 728 slicing" 2016-10-31 731 [IETF-Virtualization] Carlos Bernardos, Akbar Rahman, Juan Zuniga, 732 Luis Contreras, Pedro Aranda, " Network Virtualization 733 Research Challenges" 2016-10-31 736 [IETF-Coding] M.A. Vazquez-Castro, Tan Do-Duy, Paresh Saxena, Magnus 737 Vikstrom, "Network Coding Function Virtualization" 2016- 738 11-14 741 [IETF-Anchoring] Anthony Chan, Xinpeng Wei, Jong-Hyouk Lee, Seil 742 Jeon, Alexandre Petrescu, Fred Templin "Distributed 743 Mobility Anchoring" 2016-12-15 746 [RFC6291] L. Andersson, H. van Helvoort, R. Bonica, D. Romascanu, S. 747 Mansfield "Guidelines for the Use of the "OAM" Acronym in 748 the IETF" - June 2011 https://tools.ietf.org/html/rfc6291 750 6.2 Informative References 752 [NECOS] Novel Enablers for Cloud Slicing - http://www.h2020- 753 necos.eu 755 [Programmable Networks] "Programmable Networks for IP Service 756 Deployment", Galis, A., Denazis, S., Brou, C., Klein, C. - 757 ISBN 1-58053-745-6, pp 450, June 2004, Artech House Books, 758 Online: http://www.artechhouse.com/International/Books/ 759 Programmable-Networks-for-IP-Service-Deployment-1017.aspx 761 [ChinaCom-2009] A. Galis et al - "Management and Service-aware 762 Networking Architectures (MANA) for Future Internet" - 763 Invited paper IEEE 2009 Fourth International Conference on 764 Communications and Networking in China (ChinaCom09) 26-28 765 August 2009, Xi'an, China, n.d., 766 . 768 [GENI-2009] "GENI Key Concepts - Global Environment for Network 769 Innovations (GENI)", n.d., 770 . 772 [ITU-T 2011] ITU-T Y.3011- http://www.itu.int/rec/T-REC-Y.3001- 773 201105-I 775 [NGMN 2016] Network Slicing Framework 776 https://www.ngmn.org/fileadmin/user_upload/ 777 161010_NGMN_Network_Slicing_framework_v1.0.8.pdf 779 [NS Tutorial 2018] "Infrastructure Slicing Landscape": Galis. A, 780 Makhijani, K Tutorial at IEEE NetSoft 2018, Montreal 19 781 July 2018; http://discovery.ucl.ac.uk/10051374/ 783 [GUERZONI-2016] Guerzoni, R., Vaishnavi, I., Perez-Caparros, D., 784 Galis, A., et al "Analysis of End-to-End Multi Domain 785 Management and Orchestration Frameworks for Software 786 Defined Infrastructures - an Architectural Survey", June 787 2016, . 789 [IMT2020-2015] "Report on Gap Analysis", ITU-T FG IMT2020, December 790 2015, . 793 [IMT2020-2016] "Draft Technical Report Application of network 794 softwarization to IMT-2020 (O-041)", ITU-T FG IMT2020, 795 December 2016, . 798 [IMT2020-2016bis] "Draft Terms and definitions for IMT-2020 in ITU-T 799 (O-040)", ITU-T FG IMT2020, December 2016, 800 . 803 [KARL-2016] Karl, H., Peuster, M, Galis, A., et al "DevOps for 804 Network Function Virtualization - An Architectural 805 Approach", July 2016, . 808 [MANO-2014] "Network Functions Virtualisation (NFV); Management and 809 Orchestration v1.1.1.", ETSI European Telecommunications 810 Standards Institute., December 2014, 811 . 814 [NGMN-2016] Hedmar,P., Mschner, K., et al - "Description of Network 815 Slicing Concept", NGMN Alliance NGS-3GPP-2016, January 816 2016, . 819 [NGS-3GPP-2016] "Study on Architecture for Next Generation System - 820 latest version v1.0.2", September 2016, 821 . 824 [ONF-2016] Paul, M, Schallen, S., Betts, M., Hood, D., Shirazipor, 825 M., Lopes, D., Kaippallimalit, J., - Open Network 826 Fundation document "Applying SDN Architecture to 5G 827 Slicing", Open Network Fundation, April 2016, 828 . 832 [5G-ICN] Ravi Ravindran, Asit Chakraborti, Syed Obaid Amin, Aytac 833 Azgin, G.Q.Wang, "5G-ICN: Delivering ICN Services in 5G 834 using Network Slicing", IEEE Communication Magazine, May, 835 2017. 837 [GRAMMATIKOU-2012] Grammatikou, M; Marinos, C; Martinez-Julia, P; 838 Jofre, J; Gheorghiu, S; et al. Proceedings of the 839 International Conference on Parallel and Distributed 840 Processing Techniques and Applications (PDPTA); Athens: 1- 841 5. Athens: The Steering Committee of The World Congress in 842 Computer Science, Computer Engineering and Applied 843 Computing (WorldComp). (2012) 845 [GAL] A. Galis, Chih-Lin I" Towards 5G Network Slicing - Motivation 846 and Challenges" IEEE 5G Tech Focus, Volume 1, Number 1, 847 March 2017 - http://5g.ieee.org/tech-focus/march- 848 2017#networkslicing 850 [GAPS] "Gap Analysis for Network Slicing" draft-qiang-netslices-gap- 851 analysis-01 853 [NS UseCases] "Network Slicing Use Cases: Network Customization for 854 different services" draft-makhijani-netslices-usecase- 855 customization-03 857 [NS ARCH] "Network Slicing Architecture" draft-geng-netslices- 858 architecture-02 860 Authors' Addresses 862 Alex Galis 863 University College London 864 Email: a.galis@ucl.ac.uk 866 Luis Miguel Contreras Murillo 867 Telefonica 868 Email: luismiguel.contrerasmurillo@telefonica.com 870 Liang Geng 871 China Mobile 872 Email: gengliang@chinamobile.com 874 Slawomir Kuklinski 875 Orange Polska 876 Email: slawomir.kuklinski@orange.com 878 Kiran Makhijani 879 Huawei Technologies 880 Email: kiran.makhijani@huawei.com 882 Li Qiang 883 Huawei Technologies 884 Email: qiangli3@huawei.com 886 Hannu Flinck 887 Nokia 888 Email: hannu.flinck@nokia-bell-labs.com 890 Reza Rokui 891 Nokia 892 Email: reza.rokui@nokia.com 894 Pedro Martinez-Julia 895 National Institute of Information and Communications Technology 896 (NICT) 897 Email: pedro@nict.go.jp 899 Christian Rothenberg 900 University of Campinas (Unicamp) 901 Email: chesteve@dca.fee.unicamp.br 903 Joan Serrat 904 Universitat Politecnica de Catalunya (UPC) 905 Email: serrat@tsc.upc.edu 907 Stuart Clayman 908 University College London (UCL) 909 Email: s.clayman@ucl.ac.uk 910 Francesco Tusa 911 University College London (UCL) 912 Email: francesco.tusa@ucl.ac.uk