idnits 2.17.1 draft-ietf-i2nsf-framework-10.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: ---------------------------------------------------------------------------- == Mismatching filename: the document gives the document name as 'draft-ietf-i2nsf-framework-09', but the file name used is 'draft-ietf-i2nsf-framework-10' 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 (November 19, 2017) is 2348 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- ** Obsolete normative reference: RFC 5575 (Obsoleted by RFC 8955) == Outdated reference: A later version (-21) exists of draft-ietf-netmod-acl-model-14 == Outdated reference: A later version (-08) exists of draft-ietf-i2nsf-terminology-04 == Outdated reference: A later version (-07) exists of draft-pastor-i2nsf-nsf-remote-attestation-02 Summary: 1 error (**), 0 flaws (~~), 5 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 1 I2NSF D. Lopez 2 Internet-Draft Telefonica I+D 3 Intended status: Informational E. Lopez 4 Expires: May 19, 2018 Curveball Networks 5 L. Dunbar 6 J. Strassner 7 Huawei 8 R. Kumar 9 Juniper Networks 10 November 19, 2017 12 Framework for Interface to Network Security Functions 13 draft-ietf-i2nsf-framework-09 15 Abstract 17 This document describes the framework for the Interface to Network 18 Security Functions (I2NSF), and defines a reference model (including 19 major functional components) for I2NSF. Network security functions 20 (NSFs) are packet-processing engines that inspect and optionally 21 modify packets traversing networks, either directly or in the context 22 of sessions to which the packet is associated. 24 Status of this Memo 26 This Internet-Draft is submitted in full conformance with the 27 provisions of BCP 78 and BCP 79. 29 Internet-Drafts are working documents of the Internet Engineering 30 Task Force (IETF). Note that other groups may also distribute 31 working documents as Internet-Drafts. The list of current Internet- 32 Drafts is at http://datatracker.ietf.org/drafts/current/. 34 Internet-Drafts are draft documents valid for a maximum of six months 35 and may be updated, replaced, or obsoleted by other documents at any 36 time. It is inappropriate to use Internet-Drafts as reference 37 material or to cite them other than as "work in progress." 39 This Internet-Draft will expire on May 19, 2018. 41 Copyright Notice 43 Copyright (c) 2017 IETF Trust and the persons identified as the 44 document authors. All rights reserved. 46 This document is subject to BCP 78 and the IETF Trust's Legal 47 Provisions Relating to IETF Documents 48 (http://trustee.ietf.org/license-info) in effect on the date of 49 publication of this document. Please review these documents 50 carefully, as they describe your rights and restrictions with respect 51 to this document. Code Components extracted from this document must 52 include Simplified BSD License text as described in Section 4.e of 53 the Trust Legal Provisions and are provided without warranty as 54 described in the Simplified BSD License. 56 Table of Contents 58 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 59 2. Conventions used in this document . . . . . . . . . . . . . . 3 60 2.1. Acronyms . . . . . . . . . . . . . . . . . . . . . . . . . 3 61 2.2. Definitions . . . . . . . . . . . . . . . . . . . . . . . 4 62 3. I2NSF Reference Model . . . . . . . . . . . . . . . . . . . . 4 63 3.1. I2NSF Consumer-Facing Interface . . . . . . . . . . . . . 6 64 3.2. I2NSF NSF-Facing Interface . . . . . . . . . . . . . . . . 6 65 3.3. I2NSF Registration Interface . . . . . . . . . . . . . . . 7 66 4. Threats Associated with Externally Provided NSFs . . . . . . . 7 67 5. Avoiding NSF Ossification . . . . . . . . . . . . . . . . . . 8 68 6. The Network Connecting I2NSF Components . . . . . . . . . . . 9 69 6.1. Network Connecting I2NSF Users and I2NSF Controller . . . 9 70 6.2. Network Connecting the Controller and NSFs . . . . . . . 10 71 6.3. Interface to vNSFs . . . . . . . . . . . . . . . . . . . . 11 72 6.4. Consistency . . . . . . . . . . . . . . . . . . . . . . . 12 73 7. I2NSF Flow Security Policy Structure . . . . . . . . . . . . . 12 74 7.1. Customer-Facing Flow Security Policy Structure . . . . . . 12 75 7.2. NSF-Facing Flow Security Policy Structure . . . . . . . . 13 76 7.3. Differences from ACL Data Models . . . . . . . . . . . . . 15 77 8. Capability Negotiation . . . . . . . . . . . . . . . . . . . . 16 78 9. Registration Considerations . . . . . . . . . . . . . . . . . 16 79 9.1. Flow-Based NSF Capability Characterization . . . . . . . . 16 80 9.2. Registration Categories . . . . . . . . . . . . . . . . . 17 81 10. Manageability Considerations . . . . . . . . . . . . . . . . 20 82 11. Security Considerations . . . . . . . . . . . . . . . . . . 21 83 12. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 21 84 13. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 21 85 14. References . . . . . . . . . . . . . . . . . . . . . . . . . . 21 86 14.1. Normative References . . . . . . . . . . . . . . . . . . . 21 87 14.2. Informative References . . . . . . . . . . . . . . . . . . 22 88 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 23 90 1. Introduction 92 This document describes the framework for the Interface to Network 93 Security Functions (I2NSF), and defines a reference model (including 94 major functional components) for I2NSF. This includes an analysis of 95 the threats implied by the deployment of Network Security Functions 96 (NSFs) that are externally provided. It also describes how I2NSF 97 facilitates implementing security functions in a technology- and 98 vendor-independent manner in Software-Defined Networking (SDN) and 99 Network Function Virtualization (NFV) environments, while avoiding 100 potential constraints that could limit the capabilities of NSFs. 102 The I2NSF use cases [RFC8192] call for standard interfaces for users 103 of an I2NSF system (e.g., applications, overlay or cloud network 104 management system, or enterprise network administrator or management 105 system), to inform the I2NSF system which I2NSF functions should be 106 applied to which traffic (or traffic patterns). The I2NSF system 107 realizes this as a set of security rules for monitoring and 108 controlling the behavior of different traffic. It also provides 109 standard interfaces for users to monitor flow-based security 110 functions hosted and managed by different administrative domains. 112 [RFC8192] also describes the motivation and the problem space for 113 an Interface to Network Security Functions system. 115 2. Conventions used in this document 117 This memo does not propose a protocol standard, and the use of words 118 such as "should" follow their ordinary English meaning, and not that 119 for normative languages defined in [RFC2119][RFC8174]. 121 2.1. Acronyms 123 The following acronyms are used in this document: 125 DOTS Distributed Denial-of-Service Open Threat Signaling 126 IDS Intrusion Detection System 127 IoT Internet of Things 128 IPS Intrusion Protection System 129 NSF Network Security Function 131 2.2. Definitions 133 The following terms, which are used in this document, are defined in 134 the I2NSF terminology document [I-D.ietf-i2nsf-terminology]: 136 Capability 138 Controller 140 Firewall 142 I2NSF Consumer 144 I2NSF NSF-Facing Interface 146 I2NSF Policy Rule 148 I2NSF Producer 150 I2NSF Registration Interface 152 I2NSF Registry 154 Interface 156 Interface Group 158 Intrusion Detection System 160 Intrusion Protection System 162 Network Security Function 164 Role 166 3. I2NSF Reference Model 168 Figure 1 shows a reference model (including major functional 169 components and interfaces) for an I2NSF system. This figure is drawn 170 from the point-of-view of the Network Operator Management System; 171 hence, this view does not assume any particular management 172 architecture for either the NSFs or for how NSFs are managed (on the 173 developer's side). In particular, the Network Operator Management 174 System does not participate in NSF data plane activities. 176 Note that the term "Controller" is defined in 177 [I-D.ietf-i2nsf-terminology]. 179 +-------------------------------------------------------+ 180 | I2NSF User (e.g., Overlay Network Mgmt, Enterprise | 181 | Network Mgmt, another network domain's mgmt, etc.) | 182 +--------------------+----------------------------------+ 183 | 184 | I2NSF Consumer-Facing Interface 185 | 186 | I2NSF 187 +------------+---------+ Registration +-------------+ 188 | Network Operator Mgmt| Interface | Developer's | 189 | System | < --------- > | Mgmt System | 190 +----------------+-----+ +-------------+ 191 | 192 | I2NSF NSF-Facing Interface 193 | 194 +---------------+----+------------+---------------+ 195 | | | | 196 +---+---+ +---+---+ +---+---+ +---+---+ 197 | NSF-1 | ... | NSF-m | | NSF-1 | ... | NSF-m | ... 198 +-------+ +-------+ +-------+ +-------+ 200 Developer Mgmt System A Developer Mgmt System B 202 Figure 1: I2NSF Reference Model 204 When defining I2NSF interfaces, this framework adheres to the 205 following principles: 207 o Agnostic of network topology and NSF location in the network 209 o Agnostic of provider of the NSF (i.e., independent of the way that 210 the provider makes an NSF available, as well as how the provider 211 allows the NSF to be managed) 213 o Agnostic of any vendor-specific operational, administrative, and 214 management implementation, hosting environment, and form-factor 215 (physical or virtual) 217 o Agnostic to NSF control plane implementation (e.g., signaling 218 capabilities) 220 o Agnostic to NSF data plane implementation (e.g., encapsulation 221 capabilities) 223 In general, all I2NSF interfaces should require at least mutual 224 authentication and authorization for their use. Other security and 225 privacy considerations are specified in Section 11. 227 3.1. I2NSF Consumer-Facing Interface 229 The I2NSF Consumer-Facing Interface is used to enable different users 230 of a given I2NSF system to define, manage, and monitor security 231 policies for specific flows within an administrative domain. The 232 location and implementation of I2NSF policies are irrelevant to the 233 consumer of I2NSF policies. 235 Some examples of I2NSF Consumers include: 237 o A videoconference network manager that needs to dynamically inform 238 the underlay network to allow, rate-limit, or deny flows (some of 239 which are encrypted) based on specific fields in the packets for a 240 certain time span. 242 o Enterprise network administrators and management systems that need 243 to request their provider network to enforce specific I2NSF 244 policies for particular flows. 246 o An IoT management system sending requests to the underlay network 247 to block flows that match a set of specific conditions. 249 3.2. I2NSF NSF-Facing Interface 251 The I2NSF NSF-Facing Interface (NSF-Facing Interface for short) is 252 used to specify and monitor flow-based security policies enforced by 253 one or more NSFs. Note that the I2NSF Management System does not 254 need to use all features of a given NSF, nor does it need to use all 255 available NSFs. Hence, this abstraction enables NSF features to be 256 treated as building blocks by an NSF system; thus, developers are 257 free to use the security functions defined by NSFs independent of 258 vendor and technology. 260 Flow-based NSFs [RFC8192] inspect packets in the order that they 261 are received. Note that all Interface Groups require the NSF to be 262 registered using the Registration Interface. The Interface to 263 flow-based NSFs can be categorized as follows: 265 1. NSF Operational and Administrative Interface: an Interface Group 266 used by the I2NSF Management System to program the operational 267 state of the NSF; this also includes administrative control 268 functions. I2NSF Policy Rules represent one way to change this 269 Interface Group in a consistent manner. Since applications and 270 I2NSF Components need to dynamically control the behavior of 271 traffic that they send and receive, much of the I2NSF effort is 272 focused on this Interface Group. 274 2. Monitoring Interface: an Interface Group used by the the I2NSF 275 Management System to obtain monitoring information from one or 276 more selected NSFs. Each interface in this Interface Group could 277 be a query- or a report-based interface. The difference is that 278 a query-based interface is used by the the I2NSF Management 279 System to obtain information, whereas a report-based interface 280 is used by the NSF to provide information. The functionality of 281 this Interface Group may also be defined by other protocols, 282 such as SYSLOG and DOTS (Distributed Denial-of-Service Open 283 Threat Signaling). The I2NSF Management System may take one or 284 more actions based on the receipt of information; this should be 285 specified by an I2NSF Policy Rule. This Interface Group does 286 NOT change the operational state of the NSF. 288 This document uses the flow-based paradigm to develop the NSF-Facing 289 Interface. A common trait of flow-based NSFs is in the processing 290 of packets based on the content (e.g., header/payload) and/or 291 context (e.g., session state, authentication state) of the 292 received packets. This feature is one of the requirements for 293 defining the behavior of I2NSF. 295 3.3. I2NSF Registration Interface 297 NSFs provided by different vendors may have different capabilities. 298 In order to automate the process of utilizing multiple types of 299 security functions provided by different vendors, it is necessary to 300 have a dedicated interface for vendors to define the capabilities of 301 (i.e., register) their NSFs. This Interface is called the 302 I2NSF Registration Interface. 304 An NSF's capabilities can either be pre-configured or retrieved 305 dynamically through the I2NSF Registration Interface. If a new 306 function that is exposed to the consumer is added to an NSF, then 307 the capabilities of that new function should be registered in the 308 I2NSF Registry via the I2NSF Registration Interface, so that 309 interested management and control entities may be made aware of them. 311 4. Threats Associated with Externally Provided NSFs 313 While associated with a much higher flexibility, and in many cases a 314 necessary approach given the deployment conditions, the usage of 315 externally provided NSFs implies several additional concerns in 316 security. The most relevant threats associated with a security 317 platform of this nature are: 319 o An unknown/unauthorized user can try to impersonate another user 320 that can legitimately access external NSF services. This attack 321 may lead to accessing the I2NSF Policy Rules and applications of 322 the attacked user, and/or to generate network traffic outside the 323 security functions with a falsified identity. 325 o An authorized user may misuse assigned privileges to alter the 326 network traffic processing of other users in the NSF underlay or 327 platform. 329 o A user may try to install malformed elements (e.g., I2NSF Policy 330 Rules, or configuration files), trying to directly take the 331 control of a NSF or the whole provider platform. For example, 332 a user may exploit a vulnerability on one of the functions, or 333 may try to intercept or modify the traffic of other users in the 334 same provider platform. 336 o A malicious provider can modify the software (e.g., the operating 337 system or the specific NSF implementation) to alter the behavior 338 of one or more NSFs. This event has a high impact on all users 339 accessing NSFs, since the provider has the highest level of 340 privileges controlling the operation of the software. 342 o A user that has physical access to the provider platform can 343 modify the behavior of the hardware/software components, or the 344 components themselves. For example, the user can access a serial 345 console (most devices offer this interface for maintenance 346 reasons) to access the NSF software with the same level of 347 privilege of the provider. 349 The use of authentication, authorization, accounting, and audit 350 mechanisms is recommended for all users and applications to access 351 the I2NSF environment. This can be further enhanced by requiring 352 attestation to be used to detect changes to the I2NSF environment 353 by authorized parties. The characteristics of these procedures will 354 define the level of assurance of the I2NSF environment. 356 5. Avoiding NSF Ossification 358 A basic tenet in the introduction of I2NSF standards is that the 359 standards should not make it easier for attackers to compromise the 360 network. Therefore, in constructing standards for I2NSF Interfaces 361 as well as I2NSF Policy Rules, it is equally important to allow 362 support for specific functions, as this enables the introduction of 363 NSFs that evolve to meet new threats. Proposed standards for I2NSF 364 Interfaces to communicate with NSFs, as well as I2NSF Policy Rules 365 to control NSF functionality, should not: 367 o Narrowly define NSF categories, or their roles, when implemented 368 within a network. Security is a constantly evolving discipline. 369 The I2NSF framework relies on an object-oriented information 370 model, which provides an extensible definition of NSF information 371 elements and categories; it is recommended that implementations 372 follow this model. 374 o Attempt to impose functional requirements or constraints, either 375 directly or indirectly, upon NSF developers. Implementations 376 should be free to realize and apply NSFs in a way that best 377 suits the needs of the applications and environment using them. 379 o Be a limited lowest common denominator approach, where interfaces 380 can only support a limited set of standardized functions, without 381 allowing for developer-specific functions. NSFs, interfaces, and 382 the data communicated should be extensible, so that they can 383 evolve to protect against new threats. 385 o Be seen as endorsing a best common practice for the implementation 386 of NSFs; rather, this document describes the conceptual structure 387 and reference model of I2NSF. The purpose of this reference model 388 is to define a common set of concepts in order to facilitate the 389 flexible implementation of an I2NSF system. 391 To prevent constraints on NSF developers' creativity and innovation, 392 this document recommends the Flow-based NSF interfaces to be designed 393 from the paradigm of processing packets in the network. Flow-based 394 NSFs ultimately are packet-processing engines that inspect packets 395 traversing networks, either directly or in the context of sessions in 396 which the packet is associated. The goal is to create a workable 397 interface to NSFs that aids in their integration within legacy, SDN, 398 and/or NFV environments, while avoiding potential constraints which 399 could limit their functional capabilities. 401 6. The Network Connecting I2NSF Components 402 6.1. Network Connecting I2NSF Users and the I2NSF Controller 404 As a general principle, in the I2NSF environment, users directly 405 interact with the I2NSF Controller. Given the role of the I2NSF 406 Controller, a mutual authentication of users and the I2NSF 407 Controller is required. I2NSF does not mandate a specific 408 authentication scheme; it is up to the users to choose available 409 authentication schemes based on their needs. 411 Upon successful authentication, a trusted connection between the 412 user and the I2NSF Controller (or an endpoint designated by it) will 413 be established. This means that a direct, physical point-to-point 414 connection, with physical access restricted according to access 415 control, must be used. All traffic to and from the NSF environment 416 will flow through this connection. The connection is intended not 417 only to be secure, but trusted in the sense that it should be bound 418 to the mutual authentication between the user and the I2NSF 419 Controller, as described in [I-D.pastor-i2nsf-remote-attestation]. 420 The only possible exception is when the required level of assurance 421 is lower, (see Section 4.1 of [I-D.pastor-i2nsf-remote-attestation]), 422 in which case the user must be made aware of this circumstance. 424 6.2. Network Connecting the I2NSF Controller and NSFs 426 Most likely the NSFs are not directly attached to the I2NSF 427 Controller; for example, NSFs can be distributed across the network. 428 The network that connects the I2NSF Controller with the NSFs can be 429 the same network that carries the data traffic, or can be a dedicated 430 network for management purposes only. In either case, packet loss 431 could happen due to failure, congestion, or other reasons. 433 Therefore, the transport mechanism used to carry management data and 434 information must be secure. It does not have to be a reliable 435 transport; rather, a transport-independent reliable messaging 436 mechanism is required, where communication can be performed reliably 437 (e.g., by establishing end-to-end communication sessions and by 438 introducing explicit acknowledgement of messages into the 439 communication flow). Latency requirements for control message 440 delivery must also be evaluated. Note that monitoring does not 441 require reliable transport. 443 The network connection between the I2NSF Controller and NSFs can 444 rely either on: 446 o Open environments, where one or more NSFs can be hosted in one or 447 more external administrative domains that are reached via secure 448 external network connections. This requires more restrictive 449 security control to be placed over the I2NSF interface. The 450 information over the I2NSF interfaces shall be exchanged by 451 using the trusted connection described in section 6.1. 453 o Closed environments, where there is only one administrative 454 domain. Such environments provide a more **isolated** 455 environment, but still communicate over the same set of I2NSF 456 interfaces present in open environments (see above). Hence, the 457 security control and access requirements for closed environments 458 are the same as those for open environments. 460 The network connection between the I2NSF Controller and NSFs will 461 use the trusted connection mechanisms described in section 6.1. 462 Following these mechanisms, the connections need to rely on the use 463 of properly verified peer identities (e.g., through an AAA 464 framework). The implementations of identity management functions, as 465 well as the AAA framework, are out of scope for I2NSF. 467 6.3. Interface to vNSFs 469 There are some unique characteristics in interfacing to virtual NSFs: 471 o There could be multiple instantiations of one single NSF that has 472 been distributed across a network. When different instantiations 473 are visible to the I2NSF Controller, different policies may be 474 applied to different instantiations of an individual NSF (e.g., 475 to reflect the different roles that each vNSF is designated for). 476 Therefore, it is recommended that Roles, in addition to the use 477 of robust identities, be used to distinguish between different 478 instantiations of the same vNSF. Note that this also applies to 479 physical NSFs. 481 o When multiple instantiations of one single NSF appear as one 482 single entity to the I2NSF Controller, the I2NSF Controller may 483 need to either get assistance from other entities in the I2NSF 484 Management System, and/or delegate the provisioning of the 485 multiple instantiations of the (single) NSF to other entities in 486 the I2NSF Management System. This is shown in Figure 2 below. 488 o Policies enforced by one vNSF instance may need to be retrieved 489 and moved to another vNSF of the same type when user flows are 490 moved from one vNSF to another. 492 o Multiple vNSFs may share the same physical platform. 494 o There may be scenarios where multiple vNSFs collectively perform 495 the security policies needed. 497 +------------------------+ 498 | I2NSF Controller | 499 +------------------------+ 500 ^ ^ 501 | | 502 +-----------+ +------------+ 503 | | 504 v v 505 + - - - - - - - - - - - - - - - + + - - - - - - - - - - - - - - - + 506 | NSF-A +--------------+ | | NSF-B +--------------+ | 507 | | NSF Manager | | | | NSF Manager | | 508 | +--------------+ | | +--------------+ | 509 | + - - - - - - - - - - - - - + | | + - - - - - - - - - - - - - + | 510 | |+---------+ +---------+| | | |+---------+ +---------+| | 511 | || NSF-A#1 | ... | NSF-A#n || | | || NSF-B#1 | ... | NSF-B#m || | 512 | |+---------+ +---------+| | | |+---------+ +---------+| | 513 | | NSF-A cluster | | | | NSF-B cluster | | 514 | + - - - - - - - - - - - - - + | | + - - - - - - - - - - - - - + | 515 + - - - - - - - - - - - - - - - + + - - - - - - - - - - - - - - - + 517 Figure 2: Cluster of NSF Instantiations Management 519 6.4. Consistency 521 There are three basic models of consistency: 523 o centralized, which uses a single manager to impose behavior 524 o decentralized, in which managers make decisions without being 525 aware of each other (i.e., managers do not exchange information) 526 o distributed, in which managers make explicit use of information 527 exchange to arrive at a decision 529 This document does NOT make a recommendation on which of the above 530 three models to use. I2NSF Policy Rules, coupled with an appropriate 531 management strategy, is applicable to the design and integration of 532 any of the above three consistency models. 534 7. I2NSF Flow Security Policy Structure 536 Even though security functions come in a variety of form factors and 537 have different features, provisioning to flow-based NSFs can be 538 standardized by using policy rules. 540 In this version of I2NSF, policy rules are limited to imperative 541 paradigms. I2NSF is using an Event - Condition - Action (ECA) policy, 542 where: 544 o An Event clause is used to trigger the evaluation of the 545 Condition clause of the Policy Rule. 547 o A Condition clause is used to determine whether or not the set of 548 Actions in the I2NSF Policy Rule can be executed or not. 550 o An Action clause defines the type of operations that may be 551 performed on this packet or flow. 553 Each of the above three clauses are defined to be Boolean clauses. 554 This means that each is a logical statement that evaluates to either 555 TRUE or FALSE. 557 The above concepts are described in detail in 558 [I-D.draft-xibassnez-i2nsf-capability]. 560 7.1. Customer-Facing Flow Security Policy Structure 562 This layer is for the user's network management system to express and 563 monitor the needed flow security policies for their specific flows. 565 Some customers may not have the requisite security skills to express 566 security requirements or policies that are precise enough to 567 implement in an NSF. These customers may instead express 568 expectations (e.g., goals, or intent) of the functionality desired 569 by their security policies. Customers may also express guidelines, 570 such as which types of destinations are (or are not) allowed for 571 certain users. As a result, there could be different levels of 572 content and abstractions used in Service Layer policies. Here 573 are some examples of more abstract security Policies that can be 574 developed based on the I2NSF defined customer-facing interface: 576 Enable Internet access for authenticated users 578 Any operation on a HighValueAsset must use the corporate network 580 The use of FTP from any user except the CxOGroup must be audited 582 Streaming media applications are prohibited on the corporate 583 network during business hours 585 Scan email for malware detection protect traffic to corporate 586 network with integrity and confidentiality 588 Remove tracking data from Facebook [website = *.facebook.com] 590 One flow policy over the Customer-Facing Interface may need multiple 591 NSFs at various locations to achieve the desired enforcement. Some 592 flow security policies from users may not be granted because of 593 resource constraints. [I-D.xie-i2nsf-demo-outline-design] describes 594 an implementation of translating a set of user policies to the flow 595 policies to individual NSFs. 597 I2NSF will first focus on user policies that can be modeled as 598 closely as possible to the flow security policies used by individual 599 NSFs. An I2NSF user flow policy should be similar in structure to 600 the structure of an I2NSF Policy Rule, but with more of a user- 601 oriented expression for the packet content, context, and other parts 602 of an ECA policy rule. This enables the user to construct an I2NSF 603 Policy Rule without having to know the exact syntax of the desired 604 content (e.g., actual tags or addresses) to match in the packets. For 605 example, when used in the context of policy rules over the Client 606 Facing Interface: 608 An Event can be "the client has passed the AAA process" 610 A Condition can be matching user identifier, or from specific 611 ingress or egress points 613 An action can be establishing a IPsec tunnel 615 7.2. NSF-Facing Flow Security Policy Structure 617 The NSF-Facing Interface is to pass explicit rules to individual NSFs 618 to treat packets, as well as methods to monitor the execution status 619 of those functions. 621 Here are some examples of events over the NSF facing interface: 623 time == 08:00 625 notification that a NSF state changes from standby to active 627 user logon or logoff 629 Here are some examples of conditions over the NSF facing interface 631 o Packet content values that look for one or more packet headers, 632 data from the packet payload, bits in the packet, or data that 633 are derived from the packet. 635 o Context values that are based on measured and/or inferred 636 knowledge, which can be used to define the state and environment 637 in which a managed entity exists or has existed. In addition to 638 state data, this includes data from sessions, direction of the 639 traffic, time, and geo-location information. State refers to the 640 behavior of a managed entity at a particular point in time. 641 Hence, it may refer to situations in which multiple pieces of 642 information that are not available at the same time must be 643 analyzed. For example, tracking established TCP connections 644 (connections that have gone through the initial three-way 645 handshake). 647 Actions to individual flow-based NSFs include: 649 o Actions performed on ingress packets, such as pass, drop, 650 rate limiting, and mirroring. 652 o Actions performed on egress packets, such as invoke signaling, 653 tunnel encapsulation, packet forwarding and/or transformation. 655 o Applying a specific functional profile or signature - e.g., an IPS 656 Profile, a signature file, an anti-virus file, or a URL filtering 657 file. Many flow-based NSFs utilize profile and/or signature files 658 to achieve more effective threat detection and prevention. It is 659 not uncommon for a NSF to apply different profiles and/or 660 signatures for different flows. Some profiles/signatures do not 661 require any knowledge of past or future activities, while others 662 are stateful, and may need to maintain state for a specific length 663 of time. 665 The functional profile or signature file is one of the key properties 666 that determine the effectiveness of the NSF, and is mostly NSF- 667 specific today. The rulesets and software interfaces of I2NSF aim to 668 specify the format to pass profile and signature files while 669 supporting specific functionalities of each. 671 Policy consistency among multiple security function instances is very 672 critical because security policies are no longer maintained by one 673 central security device, but instead are enforced by multiple 674 security functions instantiated at various locations. 676 7.3. Differences from ACL Data Models 678 Policy rules are very different from ACLs. An ACL is NOT a policy. 679 Rather, policies are used to manage the construction and lifecycle 680 of an ACL. 682 [I-D.ietf-netmod-acl-model] has defined rules for the Access 683 Control List supported by most routers/switches that forward packets 684 based on packets' L2, L3, or sometimes L4 headers. The actions for 685 Access Control Lists include Pass, Drop, or Redirect. 687 The functional profiles (or signatures) for NSFs are not present in 688 [I-D.ietf-netmod-acl-model] because the functional profiles are 689 unique to specific NSFs. For example, most IPS/IDS implementations 690 have their proprietary functions/profiles. One of the goals of I2NSF 691 is to define a common envelop format for exchanging or sharing 692 profiles among different organizations to achieve more effective 693 protection against threats. 695 The "packet content matching" of the I2NSF policies should not only 696 include the matching criteria specified by 697 [I-D.ietf-netmod-acl-model], but also the L4-L7 fields depending 698 on the NSFs selected. 700 Some Flow-based NSFs need matching criteria that include the context 701 associated with the packets. This may also include metadata. 703 The I2NSF "actions" should extend the actions specified by 704 [I-D.ietf-netmod-acl-model] to include applying statistics 705 functions, threat profiles, or signature files that clients provide. 707 8. Capability Negotiation 709 It is very possible that the underlay network (or provider network) 710 does not have the capability or resource to enforce the flow security 711 policies requested by the overlay network (or enterprise network). 712 Therefore, it is required that the I2NSF system support dynamic 713 discovery capabilities, as well as a query mechanism, so that the 714 I2NSF system can expose appropriate security services using 715 I2NSF capabilities. This may also be used to support negotiation 716 between a user and the I2NSF system. Such dynamic negotiation 717 facilitates the delivery of the required security service(s). The 718 outcome of the negotiation would feed the I2NSF Management System, 719 which would then dynamically allocate appropriate NSFs (along with 720 any resources needed by the allocated NSFs) and configure the set of 721 security services that meet the requirements of the user. 723 When an NSF cannot perform the desired provisioning (e.g., due to 724 resource constraints), it must inform the I2NSF Management System. 726 The protocol needed for this security function/capability negotiation 727 may be somewhat correlated to the dynamic service parameter 728 negotiation procedure described in [RFC7297]. The Connectivity 729 Provisioning Profile (CPP) template, even though currently covering 730 only Connectivity requirements, includes security clauses such as 731 isolation requirements and non-via nodes. Hence, it could be extended 732 as a basis for the negotiation procedure. Likewise, the companion 733 Connectivity Provisioning Negotiation Protocol (CPNP) could be a 734 candidate for the negotiation procedure. 736 "Security-as-a-Service" would be a typical example of the kind of 737 (CPP-based) negotiation procedures that could take place between a 738 corporate customer and a service provider. However, more security 739 specific parameters have to be considered. 741 [I.D.-draft-xibassnez-i2nsf-capability] describes the concepts of 742 capabilities in detail. 744 9. Registration Considerations 746 9.1. Flow-Based NSF Capability Characterization 748 There are many types of flow-based NSFs. Firewall, IPS, and IDS are 749 the commonly deployed flow-based NSFs. However, the differences 750 among them are definitely blurring, due to more powerful technology, 751 integration of platforms, and new threats. Basic types of 752 flow-based NSFs include: 754 o Firewall - A device or a function that analyzes packet headers and 755 enforces policy based on protocol type, source address, 756 destination address, source port, destination port, and/or other 757 attributes of the packet header. Packets that do not match policy 758 are rejected. Note that additional functions, such as logging and 759 notification of a system administrator, could optionally be 760 enforced as well. 761 o IDS (Intrusion Detection System) - A device or function that 762 analyzes packets, both header and payload, looking for known 763 events. When a known event is detected, a log message is 764 generated detailing the event. Note that additional functions, 765 such as notification of a system administrator, could optionally 766 be enforced as well. 768 o IPS (Intrusion Prevention System) - A device or function that 769 analyzes packets, both header and payload, looking for known 770 events. When a known event is detected, the packet is rejected. 771 Note that additional functions, such as logging and notification 772 of a system administrator, could optionally be enforced as well. 774 Flow-based NSFs differ in the depth of packet header or payload they 775 can inspect, the various session/context states they can maintain, 776 and the specific profiles and the actions they can apply. An example 777 of a session is "allowing outbound connection requests and only 778 allowing return traffic from the external network". 780 9.2. Registration Categories 782 Developers can register their NSFs using Packet Content Match 783 categories. The IDR (Inter-Domain Routing) Flow Specification 784 [RFC5575] has specified 12 different packet header matching types. 786 IPFIX data [IPFIX-D] defines IP flow information and mechanisms to 787 transmit such information. This includes flow attributes as well as 788 information about the metering and exporting processes is also 789 included. Such contain may be stored in a IPFIX registry [IPFIX-R]. 790 As such, IPFIX information should be considered for defining 791 categories of registration information. 793 More packet content matching types have been proposed in the IDR WG. 794 I2NSF should re-use the packet matching types being specified as much 795 as possible. More matching types might be added for Flow-based NSFS. 797 Tables 1-4 below list the applicable packet content categories that 798 can be potentially used as packet matching types by Flow-based NSFs: 800 +-----------------------------------------------------------+ 801 | Packet Content Matching Capability Index | 802 +---------------+-------------------------------------------+ 803 | Layer 2 | Layer 2 header fields: | 804 | Header | Source | 805 | | Destination | 806 | | s-VID | 807 | | c-VID | 808 | | Ethertype | 809 |---------------+-------------------------------------------+ 810 | Layer 3 | Layer header fields: | 811 | | protocol | 812 | IPv4 Header | dest port | 813 | | src port | 814 | | src address | 815 | | dest address | 816 | | dscp | 817 | | length | 818 | | flags | 819 | | ttl | 820 | IPv6 Header | | 821 | | protocol/nh | 822 | | src port | 823 | | dest port | 824 | | src address | 825 | | dest address | 826 | | length | 827 | | traffic class | 828 | | hop limit | 829 | | flow label | 830 | | dscp | 831 |---------------+-------------------------------------------+ 832 | Layer 4 | Layer header fields: | 833 | TCP | Port | 834 | SCTP | syn | 835 | DCCP | ack | 836 | | fin | 837 | | rst | 838 | | ? psh | 839 | | ? urg | 840 | | ? window | 841 | | sockstress | 842 | | Note: bitmap could be used to | 843 | | represent all the fields | 844 | UDP | | 845 | | flood abuse | 846 | | fragment abuse | 847 | | Port | 848 |---------------+-------------------------------------------+ 849 | HTTP layer | | 850 | | | hash collision | 851 | | | http - get flood | 852 | | | http - post flood | 853 | | | http - random/invalid url | 854 | | | http - slowloris | 855 | | | http - slow read | 856 | | | http - r-u-dead-yet (rudy) | 857 | | | http - malformed request | 858 | | | http - xss | 859 | | | https - ssl session exhaustion | 860 +---------------+----------+--------------------------------+ 861 | IETF PCP | Configurable | 862 | | Ports | 863 +---------------+-------------------------------------------+ 864 | IETF TRAM | profile | 865 +---------------+-------------------------------------------+ 867 Table 1: Packet Content Matching Capability Index 869 Notes: DCCP: Datagram Congestion Control Protocol 870 PCP: Port Control Protocol 871 TRAM: TURN Revised and Modernized, where TURN stands for 872 Traversal Using Relays around NAT 874 +-----------------------------------------------------------+ 875 | Context Matching Capability Index | 876 +---------------+-------------------------------------------+ 877 | Session | Session state, | 878 | | bidirectional state | 879 +---------------+-------------------------------------------+ 880 | Time | time span | 881 | | time occurrence | 882 +---------------+-------------------------------------------+ 883 | Events | Event URL, variables | 884 +---------------+-------------------------------------------+ 885 | Location | Text string, GPS coords, URL | 886 +---------------+-------------------------------------------+ 887 | Connection | Internet (unsecured), Internet | 888 | Type | (secured by VPN, etc.), Intranet, ... | 889 +---------------+-------------------------------------------+ 890 | Direction | Inbound, Outbound | 892 +---------------+-------------------------------------------+ 893 | State | Authentication State | 894 | | Authorization State | 895 | | Accounting State | 896 | | Session State | 897 +---------------+-------------------------------------------+ 899 Table 2: Context Matching Capability Index 901 Note: These fields are used to provide context information for I2NSF 902 Policy Rules to make decisions on how to handle traffic. For 903 example, GPS coordinates define the location of the traffic that 904 is entering and exiting an I2NSF system; this enables the 905 developer to apply different rules for ingress and egress 906 traffic handling. 908 +-----------------------------------------------------------+ 909 | Action Capability Index | 910 +---------------+-------------------------------------------+ 911 | Ingress port | SFC header termination, | 912 | | VxLAN header termination | 913 +---------------+-------------------------------------------+ 914 | | Pass | 915 | Actions | Deny | 916 | | Mirror | 917 | | Simple Statistics: Count (X min; Day;..)| 918 | | Client specified Functions: URL | 919 +---------------+-------------------------------------------+ 920 | Egress | Encap SFC, VxLAN, or other header | 921 +---------------+-------------------------------------------+ 923 Table 3: Action Capability Index 925 +-----------------------------------------------------------+ 926 | Functional Profile Index | 927 +---------------+-------------------------------------------+ 928 | Profile types | Name, type, or Flexible | 929 | Signature | Profile/signature URL Command for | 930 | | I2NSF Controller to enable/disable | 931 +---------------+-------------------------------------------+ 933 Table 4: Function Profile Index 935 10. Manageability Considerations 937 Management of NSFs includes: 939 o Lifecycle management and resource management of NSFs 941 o Configuration of devices, such as address configuration, device 942 internal attributes configuration, etc. 944 o Signaling 946 o Policy rules provisioning 948 Currently, I2NSF only focuses on the policy rule provisioning part. 950 11. Security Considerations 952 The configuration, control, and monitoring of NSFs provide access to 953 and information about security functions that are critical for 954 delivering network security and for protecting end-to-end traffic. 955 Therefore, it is important that the messages that are exchanged 956 within this architecture utilize a trustworthy, robust, and fully 957 secure communication channel. The mechanisms adopted within the 958 solution space must include proper secure communication channels 959 that are carefully specified for carrying the controlling and 960 monitoring information between the NSFs and their management entity 961 or entities. The threats associated with remotely managed NSFs are 962 discussed in Section 4, and solutions must address those concerns. 964 This framework is intended for enterprise users, with or without 965 cloud service offerings. Privacy of users must be provided by 966 using existing standard mechanisms, such as encryption; 967 anonymization of data should also be done if possible (depending 968 on the transport used). Such mechanisms require confidentiality 969 and integrity. 971 12. IANA Considerations 973 This document requires no IANA actions. RFC Editor: Please remove 974 this section before publication. 976 13. Acknowledgements 978 This document includes significant contributions from Christian 979 Jacquenet (Orange), Seetharama Rao Durbha (Cablelabs), Mohamed 980 Boucadair (Orange), Ramki Krishnan (Dell), Anil Lohiya (Juniper 981 Networks), Joe Parrott (BT), Frank Xialing (Huawei), and 982 XiaoJun Zhuang (China Mobile). 984 Some of the results leading to this work have received funding from 985 the European Union Seventh Framework Programme (FP7/2007-2013) under 986 grant agreement no. 611458. 988 14. References 990 14.1. Normative References 992 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 993 Requirement Levels", BCP 14, RFC 2119, March 1997, 994 . 996 [RFC5575] Marques, P., Sheth, N., Raszuk, R., Greene, B., Mauch, J., 997 and D. McPherson, "Dissemination of Flow Specification 998 Rules", RFC 5575, August 2009, 999 1001 [RFC7297] Boucadair, M., Jacquenet, C., and N. Wang, "IP 1002 Connectivity Provisioning Profile (CPP)", RFC 7297, 1003 July 2014, 1004 1006 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in 1007 RFC 2119 Key Words", RFC 8174, May 2017, 1008 1010 [IPFIX-D] https://datatracker.ietf.org/wg/ipfix/documents/ 1012 [IPFIX-R] https://www.iana.org/assignments/ipfix/ipfix.xhtml 1014 14.2. Informative References 1016 [RFC8192] Hares, S., Dunbar, L., Lopez, D., Zarny, M., and C. 1017 Jacquenet, "I2NSF Problem Statement and Use cases", 1018 RFC 8192, July 2017 1019 https://datatracker.ietf.org/doc/rfc8192/. 1021 [I-D.ietf-netmod-acl-model] 1022 Bogdanovic, D., Sreenivasa, K., Huang, L., and D. Blair, 1023 "Network Access Control List (ACL) YANG Data Model", 1024 draft-ietf-netmod-acl-model-14 (work in progress), 1025 October, 2017. 1027 [I-D.ietf-i2nsf-terminology] 1028 Hares, S., Strassner, J., Lopez, D., Xia, L., and H. 1029 Birkholz, "Interface to Network Security Functions (I2NSF) 1030 Terminology", draft-ietf-i2nsf-terminology-04 (work in 1031 progress), July 2017. 1033 [I-D.draft-xibassnez-i2nsf-capability] 1034 Xia, L., Strassner, J., Basile, C., and Lopez, D., 1035 "Information Model of NSFs Capabilities", 1036 draft-xibassnez-i2nsf-capability-02.txt (work in 1037 progress), July, 2017. 1039 [I-D.pastor-i2nsf-remote-attestation] 1040 Pastor, A., Lopez, D., and A. Shaw, "Remote Attestation 1041 Procedures for Network Security Functions (NSFs) through 1042 the I2NSF Security Controller", 1043 draft-pastor-i2nsf-nsf-remote-attestation-02 (work in 1044 progress), September 2017. 1046 [I-D.xie-i2nsf-demo-outline-design] 1047 Xie, Y., Xia, L., and J. Wu, "Interface to Network 1048 Security Functions Demo Outline Design", 1049 draft-xie-i2nsf-demo-outline-design-00 (work in progress), 1050 April 2015. 1052 [gs_NFV] "ETSI NFV Group Specification; Network Functions 1053 Virtualization (NFV) Use Cases. ETSI GS NFV 001v1.1.1", 1054 2013. 1056 Authors' Addresses 1058 Diego R. Lopez 1059 Telefonica I+D 1060 Editor Jose Manuel Lara, 9 1061 Seville, 41013 1062 Spain 1063 Email: diego.r.lopez@telefonica.com 1065 Edward Lopez 1066 Curveball Networks 1067 Chantilly, Virgina 1068 USA 1069 Email: elopez@fortinet.com 1071 Linda Dunbar 1072 Huawei Technologies 1073 USA 1074 Email: Linda.Dunbar@huawei.com 1076 John Strassner 1077 Huawei Technologies 1078 Santa Clara, CA 1079 USA 1080 Email: John.sc.Strassner@huawei.com 1082 Rakesh Kumar 1083 Juniper Networks 1084 India 1085 Email: rkkumar@juniper.net