idnits 2.17.1 draft-ietf-i2nsf-applicability-02.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- ** The document seems to lack an IANA Considerations section. (See Section 2.2 of https://www.ietf.org/id-info/checklist for how to handle the case when there are no actions for IANA.) Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (March 5, 2018) is 2237 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- -- Obsolete informational reference (is this intentional?): RFC 4566 (Obsoleted by RFC 8866) Summary: 1 error (**), 0 flaws (~~), 1 warning (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group J. Jeong 3 Internet-Draft S. Hyun 4 Intended status: Informational Sungkyunkwan University 5 Expires: September 6, 2018 T. Ahn 6 Korea Telecom 7 S. Hares 8 Huawei 9 D. Lopez 10 Telefonica I+D 11 March 5, 2018 13 Applicability of Interfaces to Network Security Functions to Network- 14 Based Security Services 15 draft-ietf-i2nsf-applicability-02 17 Abstract 19 This document describes the applicability of Interface to Network 20 Security Functions (I2NSF) to network-based security services in 21 Network Functions Virtualization (NFV) environments, such as 22 firewall, deep packet inspection, or attack mitigation engines. 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 https://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 September 6, 2018. 41 Copyright Notice 43 Copyright (c) 2018 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 (https://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 . . . . . . . . . . . . . . . . . . . . . . . . 2 59 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 3 60 3. I2NSF Framework . . . . . . . . . . . . . . . . . . . . . . . 4 61 3.1. Time-dependent Web Access Control Service . . . . . . . . 5 62 4. I2NSF Framework with SDN . . . . . . . . . . . . . . . . . . 7 63 4.1. Firewall: Centralized Firewall System . . . . . . . . . . 10 64 4.2. Deep Packet Inspection: Centralized VoIP/VoLTE Security 65 System . . . . . . . . . . . . . . . . . . . . . . . . . 11 66 4.3. Attack Mitigation: Centralized DDoS-attack Mitigation 67 System . . . . . . . . . . . . . . . . . . . . . . . . . 13 68 5. Security Considerations . . . . . . . . . . . . . . . . . . . 15 69 6. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 15 70 7. Contributors . . . . . . . . . . . . . . . . . . . . . . . . 15 71 8. Informative References . . . . . . . . . . . . . . . . . . . 15 72 Appendix A. Changes from draft-ietf-i2nsf-applicability-01 . . . 19 73 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 19 75 1. Introduction 77 Interface to Network Security Functions (I2NSF) defined a framework 78 and interfaces for interacting with Network Security Functions 79 (NSFs). The I2NSF framework allows heterogeneous NSFs developed by 80 different security solution vendors to be used in the NFV environment 81 by utilizing the capabilities of such products and the virtualization 82 of security functions in the NFV platform. In the I2NSF framework, 83 each NSF initially registers the profile of its own capabilities into 84 the system in order for themselves to be available in the system. In 85 addition, the Security Controller registers itself to the I2NSF user 86 so that the user can request security services to the Security 87 Controller. 89 This document describes the applicability of I2NSF framework to 90 network-based security services with a use case of time-dependent web 91 access control. This document also describes integrating I2NSF 92 framework with Software-Defined Networking (SDN) technology for 93 efficient security services and use cases, such as firewall 94 [opsawg-firewalls], Deep Packet Inspection (DPI), and Distributed 95 Denial of Service (DDoS) attack mitigation. We implemented the I2NSF 96 framework based on SDN for these use cases, and the implementation 97 successfully verified the effectiveness of the I2NSF framework. 99 2. Terminology 101 This document uses the terminology described in [RFC7149], 102 [ITU-T.Y.3300], [ONF-OpenFlow], [ONF-SDN-Architecture], 103 [ITU-T.X.1252], [ITU-T.X.800], [RFC8329], [i2nsf-terminology], 104 [consumer-facing-inf-im], [consumer-facing-inf-dm], 105 [i2nsf-nsf-cap-im], [nsf-facing-inf-dm], [registration-inf-im], 106 [registration-inf-dm], and [nsf-triggered-steering]. In addition, 107 the following terms are defined below: 109 o Software-Defined Networking (SDN): A set of techniques that 110 enables to directly program, orchestrate, control, and manage 111 network resources, which facilitates the design, delivery and 112 operation of network services in a dynamic and scalable manner 113 [ITU-T.Y.3300]. 115 o Firewall: A service function at the junction of two network 116 segments that inspects every packet that attempts to cross the 117 boundary. It also rejects any packet that does not satisfy 118 certain criteria for, for example, disallowed port numbers or IP 119 addresses. 121 o Centralized Firewall System: A centralized firewall that can 122 establish and distribute policy rules into network resources for 123 efficient firewall management. These rules can be managed 124 dynamically by a centralized server for firewall. SDN can work as 125 a network-based firewall system through a standard interface 126 between an SDN switch and a firewall function as a vitual network 127 function (VNF). 129 o Centralized VoIP Security System: A centralized security system 130 that handles the security functions required for VoIP and VoLTE 131 services. SDN can work as a network-based security system through 132 a standard interface between an SDN switch and a VoIP/VoLTE 133 security function as a VNF. 135 o Centralized DDoS-attack Mitigation System: A centralized mitigator 136 that can establish and distribute access control policy rules into 137 network resources for efficient DDoS-attack mitigation. These 138 rules can be managed dynamically by a centralized server for DDoS- 139 attack mitigation. The SDN controller and switches can 140 cooperatively work as a network-based firewall system through a 141 standard interface between an SDN switch and a firewall function 142 as a VNF running in the SDN controller. 144 3. I2NSF Framework 146 This section describes an I2NSF framework and its use case. Figure 1 147 shows an I2NSF framework [RFC8329] to support network-based security 148 services. As shown in Figure 1, I2NSF User can use security 149 functions by delivering high-level security policies, which specify 150 security requirements the I2NSF user wants to enforce, to the 151 Security Controller via the Consumer-Facing Interface 152 [consumer-facing-inf-im][consumer-facing-inf-dm]. 154 The Security Controller receives and analyzes the high-level security 155 policies from an I2NSF User, and identifies what types of security 156 capabilities are required to meet these high-level security policies. 157 The Security Controller then identifies NSFs that have the required 158 security capabilities, and generates low-level security policies for 159 each of the NSFs so that the high-level security policies are 160 eventually enforced by those NSFs. Finally, the Security Controller 161 sends the generated low-level security policies to the NSFs 162 [i2nsf-nsf-cap-im][nsf-facing-inf-dm]. 164 The Security Controller requests NSFs to perform low-level security 165 services via the NSF-Facing Interface. The NSFs are enabled as 166 Virtual Network Functions (VNFs) on top of virtual machines through 167 Network Functions Virtualization (NFV) [ETSI-NFV]. In addition, the 168 Security Controller uses the I2NSF Registration Interface 169 [registration-inf-im][registration-inf-dm] to communicate with 170 Developer's Management System (called Developer's Mgmt System) for 171 registering (or deregistering) the developer's NSFs into (or from) 172 the NFV system using the I2NSF framework. 174 The Consumer-Facing Interface between an I2NSF User and the Security 175 Controller can be implemented using, for example, RESTCONF [RFC8040]. 176 Data models specified by YANG [RFC6020] describe high-level security 177 policies to be specified by an I2NSF User. The data model defined in 178 [consumer-facing-inf-dm] can be used for the I2NSF Consumer-Facing 179 Interface. 181 +------------+ 182 | I2NSF User | 183 +------------+ 184 ^ 185 | Consumer-Facing Interface 186 v 187 +-------------------+ Registration +-----------------------+ 188 |Security Controller|<-------------------->|Developer's Mgmt System| 189 +-------------------+ Interface +-----------------------+ 190 ^ 191 | NSF-Facing Interface 192 v 193 +----------------+ +---------------+ +-----------------------+ 194 | NSF-1 |-| NSF-2 |...| NSF-n | 195 | (Firewall) | | (Web Filter) | |(DDoS-Attack Mitigator)| 196 +----------------+ +---------------+ +-----------------------+ 198 Figure 1: I2NSF Framework 200 The NSF-Facing Interface between Security Controller and NSFs can be 201 implemented using NETCONF [RFC6241]. YANG data models describe low- 202 level security policies for the sake of NSFs, which are translated 203 from the high-level security policies by the Security Controller. 204 The data model defined in [nsf-facing-inf-dm] can be used for the 205 I2NSF NSF-Facing Interface. 207 The Registration Interface between the Security Controller and the 208 Developer's Mgmt System can be implemented by RESTCONF [RFC8040]. 209 The data model defined in [registration-inf-dm] can be used for the 210 I2NSF Registration Interface. 212 Also, the I2NSF framework can enforce multiple chained NSFs for the 213 low-level security policies by means of service function chaining 214 (SFC) techniques for the I2NSF architecture described in 215 [nsf-triggered-steering]. 217 The following describes a security service scenario using the I2NSF 218 framework. 220 3.1. Time-dependent Web Access Control Service 222 This service scenario assumes that an enterprise network 223 administrator wants to control the staff members' access to Facebook 224 during business hours. The following is an example high-level 225 security policy rule that the administrator requests: Block the staff 226 members' access to Facebook from 9 am to 6 pm. The administrator 227 sends this high-level security policy to the security controller, 228 then the security controller identifies required secuity 229 capabilities, e.g., IP address and port number inspection 230 capabilities and URL inspection capability. In this scenario, it is 231 assumed that the IP address and port number inspection capabilities 232 are required to check whether a received packet is an HTTP packet 233 from a staff member. The URL inspection capability is required to 234 check whether the target URL of a received packet is facebook.com or 235 not. 237 The Security Controller maintains the security capabilities of each 238 NSF running in the I2NSF system, which have been reported by the 239 Developer's Management System via the Registation interface. Based 240 on this information, the Security Controller identifies NSFs that can 241 perform the IP address and port number inspection and URL inspection. 242 In this scenario, it is assumed that an NSF of firewall has the IP 243 address and port number inspection capabilities and an NSF of web 244 filter has URL inspection capability. 246 The Security Controller generates low-level security rules for the 247 NSFs to perform IP address and port number inspection, URL 248 inspection, and time checking. Specifically, the Security Controller 249 may interoperate with an access control server in the enterprise 250 network in order to retrieve the information (e.g., IP address in 251 use, company ID, and role) of each employee that is currently using 252 the network. Based on the retrieved information, the Security 253 Controller generates low-level security rules to check whether the 254 source IP address of a received packet matches any one being used by 255 a staff member. In addition, the low-level security rules should be 256 able to determine that a received packet is of HTTP protocol. The 257 low-level security rules for web filter checks that the target URL 258 field of a received packet is equal to facebook.com. Finally, the 259 Security Controller sends the low-level security rules of the IP 260 address and port number inspection to the NSF of firewall and the 261 low-level rules for URL inspection to the NSF of web filter. 263 The following describes how the time-dependent web access control 264 service is enforced by the NSFs of firewall and web filter. 266 1. A staff member tries to access Fackbook.com during business 267 hours, e.g., 10 am. 269 2. The packet is forwarded from the staff member's device to the 270 firewall, and the firewall checks the source IP address and port 271 number. Now the firewall identifies the received packet is an 272 HTTP packet from the staff member. 274 3. The firewall triggers the web filter to further inspect the 275 packet, and the packet is forwarded from the firewall to the web 276 filter. Service Function Chaining (SFC) technology can be 277 utilized to support such packet forwarding in the I2NSF framework 278 [nsf-triggered-steering]. 280 4. The web filter checks the target URL field of the received 281 packet, and realizes the packet is toward Facebook.com. The web 282 filter then checks that the current time is in business hours. 283 If so, the web filter drops the packet, and consequently the 284 staff member's access to Facebook during business hours is 285 blocked. 287 4. I2NSF Framework with SDN 289 This section describes an I2NSF framework with SDN for I2NSF 290 applicability and use cases, such as firewall, deep packet 291 inspection, and DDoS-attack mitigation functions. SDN enables some 292 packet filtering rules to be enforced in the network switches by 293 controlling their packet forwarding rules. By taking advantage of 294 this capability of SDN, it is possible to optimize the process of 295 security service enforcement in the I2NSF system. 297 Figure 2 shows an I2NSF framework [RFC8329] with SDN networks to 298 support network-based security services. In this system, the 299 enforcement of security policy rules is divided into the SDN switches 300 and NSFs. Especially, SDN switches enforce simple packet filtering 301 rules that can be translated into their packet forwarding rules, 302 whereas NSFs enforce NSF-related security rules requiring the 303 security capabilities of the NSFs. For this purpose, the Security 304 Controller instructs the Switch Controller via NSF-Facing Interface 305 so that SDN switches can perform the required security services with 306 flow tables under the supervision of the Switch Controller (i.e., SDN 307 Controller). 309 As an example, let us consider two different types of security rules: 310 Rule A is a simple packet fltering rule that checks only the IP 311 address and port number of a given packet, whereas rule B is a time- 312 consuming packet inspection rule for analyzing whether an attached 313 file being transmitted over a flow of packets contains malware. Rule 314 A can be translated into packet forwarding rules of SDN switches and 315 thus be enforced by the switches. In contrast, rule B cannot be 316 enforced by switches, but it can be enforced by NSFs with anti- 317 malware capability. Specifically, a flow of packets is forwarded to 318 and reassembled by an NSF to reconstruct the attached file stored in 319 the flow of packets. The NSF then analyzes the file to check the 320 existence of malware. If the file contains malware, the NSF drops 321 the packets. 323 In an I2NSF framework with SDN, the Security Controller can analyze 324 given security policy rules and automatically determine which of the 325 given security policy rules should be enforced by SDN switches and 326 which should be enforced by NSFs. If some of the given rules 327 requires security capabilities that can be provided by SDN switches, 328 then the Security Controller instructs the Switch Controller via NSF- 329 Facing Interface so that SDN switches can enforce those security 330 policy rules with flow tables under the supervision of the Switch 331 Controller (i.e., SDN Controller). Or if some rules require security 332 capabilities that can be provided by not SDN switches but NSFs, then 333 the Security Controller instructs relevant NSFs to enforce those 334 rules. 336 +------------+ 337 | I2NSF User | 338 +------------+ 339 ^ 340 | Consumer-Facing Interface 341 v 342 +-------------------+ Registration +-----------------------+ 343 |Security Controller|<-------------------->|Developer's Mgmt System| 344 +-------------------+ Interface +-----------------------+ 345 ^ ^ 346 | | NSF-Facing Interface 347 | v 348 | +----------------+ +---------------+ +-----------------------+ 349 | | NSF-1 |-| NSF-2 |...| NSF-n | 350 | | (Firewall) | | (DPI) | |(DDoS-Attack Mitigator)| 351 | +----------------+ +---------------+ +-----------------------+ 352 | ^ 353 | | 354 | v 355 | +--------+ 356 | | SFF | 357 | +--------+ 358 | ^ 359 | | 360 | V SDN Network 361 +--|----------------------------------------------------------------+ 362 | V NSF-Facing Interface | 363 | +-----------------+ | 364 | |Switch Controller| | 365 | +-----------------+ | 366 | ^ | 367 | | SDN Southbound Interface | 368 | v | 369 | +--------+ +--------+ +--------+ +--------+ | 370 | |Switch 1|-|Switch 2|-|Switch 3|......|Switch m| | 371 | +--------+ +--------+ +--------+ +--------+ | 372 +-------------------------------------------------------------------+ 374 Figure 2: An I2NSF Framework with SDN Network 376 The following subsections introduce three use cases for cloud-based 377 security services: (i) firewall system, (ii) deep packet inspection 378 system, and (iii) attack mitigation system. [RFC8192] 380 4.1. Firewall: Centralized Firewall System 382 A centralized network firewall can manage each network resource and 383 firewall rules can be managed flexibly by a centralized server for 384 firewall (called Firewall). The centralized network firewall 385 controls each switch for the network resource management and the 386 firewall rules can be added or deleted dynamically. 388 The procedure of firewall operations in this system is as follows: 390 1. A switch forwards an unknown flow's packet to one of the Switch 391 Controllers. 393 2. The Switch Controller forwards the unknown flow's packet to an 394 appropriate security service application, such as the Firewall. 396 3. The Firewall analyzes, typically, the headers and contents of the 397 packet. 399 4. If the Firewall regards the packet as a malicious one with a 400 suspicious pattern, it reports the malicious packet to the Switch 401 Controller. 403 5. The Switch Controller installs new rules (e.g., drop packets with 404 the suspicious pattern) into underlying switches. 406 6. The suspected packets are dropped by these switches. 408 Existing SDN protocols can be used through standard interfaces 409 between the firewall application and switches 410 [RFC7149][ITU-T.Y.3300][ONF-OpenFlow] [ONF-SDN-Architecture]. 412 Legacy firewalls have some challenges such as the expensive cost, 413 performance, management of access control, establishment of policy, 414 and packet-based access mechanism. The proposed framework can 415 resolve the challenges through the above centralized firewall system 416 based on SDN as follows: 418 o Cost: The cost of adding firewalls to network resources such as 419 routers, gateways, and switches is substantial due to the reason 420 that we need to add firewall on each network resource. To solve 421 this, each network resource can be managed centrally such that a 422 single firewall is manipulated by a centralized server. 424 o Performance: The performance of firewalls is often slower than the 425 link speed of network interfaces. Every network resource for 426 firewall needs to check firewall rules according to network 427 conditions. Firewalls can be adaptively deployed among network 428 switches, depending on network conditions in the framework. 430 o The management of access control: Since there may be hundreds of 431 network resources in a network, the dynamic management of access 432 control for security services like firewall is a challenge. In 433 the framework, firewall rules can be dynamically added for new 434 malware. 436 o The establishment of policy: Policy should be established for each 437 network resource. However, it is difficult to describe what flows 438 are permitted or denied for firewall within a specific 439 organization network under management. Thus, a centralized view 440 is helpful to determine security policies for such a network. 442 o Packet-based access mechanism: Packet-based access mechanism is 443 not enough for firewall in practice since the basic unit of access 444 control is usually users or applications. Therefore, application 445 level rules can be defined and added to the firewall system 446 through the centralized server. 448 4.2. Deep Packet Inspection: Centralized VoIP/VoLTE Security System 450 A centralized VoIP/VoLTE security system can monitor each VoIP/VoLTE 451 flow and manage VoIP/VoLTE security rules controlled by a centralized 452 server for VoIP/VoLTE security service called VoIP Intrusion 453 Prevention System (IPS). The VoIP/VoLTE security system controls 454 each switch for the VoIP/VoLTE call flow management by manipulating 455 the rules that can be added, deleted or modified dynamically. 457 A centralized VoIP/VoLTE security system can cooperate with a network 458 firewall to realize VoIP/VoLTE security service. Specifically, a 459 network firewall performs basic security checks of an unknown flow's 460 packet observed by a switch. If the network firewall detects that 461 the packet is an unknown VoIP call flow's packet that exhibits some 462 suspicious patterns, then it triggers the VoIP/VoLTE security system 463 for more specialized security analysis of the suspicious VoIP call 464 packet. 466 The procedure of VoIP/VoLTE security operations in this system is as 467 follows: 469 1. A switch forwards an unknown flow's packet to the Switch 470 Controller, and the Switch Controller further forwards the 471 unknown flow's packet to the Firewall for basic security 472 inspection. 474 2. The Firewall analyzes the header fields of the packet, and 475 figures out that this is an unknown VoIP call flow's signal 476 packet (e.g., SIP packet) of a suspicious pattern. 478 3. The Firewall triggers an appropriate security service function, 479 such as VoIP IPS, for detailed security analysis of the 480 suspicious signal packet. That is, the firewall sends the packet 481 to the Service Function Forwarder (SFF) in the I2NSF framework 482 [nsf-triggered-steering], as shown in Figure 2. The SFF forwards 483 the suspicious signal packet to the VoIP IPS. 485 4. The VoIP IPS analyzes the headers and contents of the signal 486 packet, such as calling number and session description headers 487 [RFC4566]. 489 5. If, for example, the VoIP IPS regards the packet as a spoofed 490 packet by hackers or a scanning packet searching for VoIP/VoLTE 491 devices, it drops the packet. In addition, the VoIP IPS requests 492 the Switch Controller to block that packet and the subsequent 493 packets that have the same call-id. 495 6. The Switch Controller installs new rules (e.g., drop packets) 496 into underlying switches. 498 7. The illegal packets are dropped by these switches. 500 Existing SDN protocols can be used through standard interfaces 501 between the VoIP IPS application and switches [RFC7149][ITU-T.Y.3300] 502 [ONF-OpenFlow][ONF-SDN-Architecture]. 504 Legacy hardware based VoIP IPS has some challenges, such as 505 provisioning time, the granularity of security, expensive cost, and 506 the establishment of policy. The I2NSF framework can resolve the 507 challenges through the above centralized VoIP/VoLTE security system 508 based on SDN as follows: 510 o Provisioning: The provisioning time of setting up a legacy VoIP 511 IPS to network is substantial because it takes from some hours to 512 some days. By managing the network resources centrally, VoIP IPS 513 can provide more agility in provisioning both virtual and physical 514 network resources from a central location. 516 o The granularity of security: The security rules of a legacy VoIP 517 IPS are compounded considering the granularity of security. The 518 proposed framework can provide more granular security by 519 centralizing security control into a switch controller. The VoIP 520 IPS can effectively manage security rules throughout the network. 522 o Cost: The cost of adding VoIP IPS to network resources, such as 523 routers, gateways, and switches is substantial due to the reason 524 that we need to add VoIP IPS on each network resource. To solve 525 this, each network resource can be managed centrally such that a 526 single VoIP IPS is manipulated by a centralized server. 528 o The establishment of policy: Policy should be established for each 529 network resource. However, it is difficult to describe what flows 530 are permitted or denied for VoIP IPS within a specific 531 organization network under management. Thus, a centralized view 532 is helpful to determine security policies for such a network. 534 4.3. Attack Mitigation: Centralized DDoS-attack Mitigation System 536 A centralized DDoS-attack mitigation can manage each network resource 537 and manipulate rules to each switch through a centralized server for 538 DDoS-attack mitigation (called DDoS-attack Mitigator). The 539 centralized DDoS-attack mitigation system defends servers against 540 DDoS attacks outside private network, that is, from public network. 542 Servers are categorized into stateless servers (e.g., DNS servers) 543 and stateful servers (e.g., web servers). For DDoS-attack 544 mitigation, traffic flows in switches are dynamically configured by 545 traffic flow forwarding path management according to the category of 546 servers [AVANT-GUARD]. Such a managenent should consider the load 547 balance among the switches for the defense against DDoS attacks. 549 The procedure of DDoS-attack mitigation operations in this system is 550 as follows: 552 1. A Switch periodically reports an inter-arrival pattern of a 553 flow's packets to one of the Switch Controllers. 555 2. The Switch Controller forwards the flow's inter-arrival pattern 556 to an appropriate security service application, such as DDoS- 557 attack Mitigator. 559 3. The DDoS-attack Mitigator analyzes the reported pattern for the 560 flow. 562 4. If the DDoS-attack Mitigator regards the pattern as a DDoS 563 attack, it computes a packet dropping probability corresponding 564 to suspiciousness level and reports this DDoS-attack flow to 565 Switch Controller. 567 5. The Switch Controller installs new rules into switches (e.g., 568 forward packets with the suspicious inter-arrival pattern with a 569 dropping probability). 571 6. The suspicious flow's packets are randomly dropped by switches 572 with the dropping probability. 574 For the above centralized DDoS-attack mitigation system, the existing 575 SDN protocols can be used through standard interfaces between the 576 DDoS-attack mitigator application and switches [RFC7149] 577 [ITU-T.Y.3300][ONF-OpenFlow][ONF-SDN-Architecture]. 579 The centralized DDoS-attack mitigation system has challenges similar 580 to the centralized firewall system. The proposed framework can 581 resolve the challenges through the above centralized DDoS-attack 582 mitigation system based on SDN as follows: 584 o Cost: The cost of adding DDoS-attack mitigators to network 585 resources such as routers, gateways, and switches is substantial 586 due to the reason that we need to add DDoS-attack mitigator on 587 each network resource. To solve this, each network resource can 588 be managed centrally such that a single DDoS-attack mitigator is 589 manipulated by a centralized server. 591 o Performance: The performance of DDoS-attack mitigators is often 592 slower than the link speed of network interfaces. The checking of 593 DDoS attacks may reduce the performance of the network interfaces. 594 DDoS-attack mitigators can be adaptively deployed among network 595 switches, depending on network conditions in the framework. 597 o The management of network resources: Since there may be hundreds 598 of network resources in an administered network, the dynamic 599 management of network resources for performance (e.g., load 600 balancing) is a challenge for DDoS-attack mitigation. In the 601 framework, as dynamic network resource management, traffic flow 602 forwarding path management can handle the load balancing of 603 network switches [AVANT-GUARD]. With this management, the current 604 and near-future workload can be spread among the network switches 605 for DDoS-attack mitigation. In addition, DDoS-attack mitigation 606 rules can be dynamically added for new DDoS attacks. 608 o The establishment of policy: Policy should be established for each 609 network resource. However, it is difficult to describe what flows 610 are permitted or denied for new DDoS-attacks (e.g., DNS reflection 611 attack) within a specific organization network under management. 612 Thus, a centralized view is helpful to determine security policies 613 for such a network. 615 So far this document has described the procedure and impact of the 616 three use cases for network-based security services using the I2NSF 617 framework with SDN networks. To support these use cases in the 618 proposed data-driven security service framework, YANG data models 619 described in [consumer-facing-inf-dm], [nsf-facing-inf-dm], and 620 [registration-inf-dm] can be used as Consumer-Facing Interface, NSF- 621 Facing Interface, and Registration Interface, respectively, along 622 with RESTCONF [RFC8040] and NETCONF [RFC6241]. 624 5. Security Considerations 626 The I2NSF framework with SDN networks in this document is derived 627 from the I2NSF framework [RFC8329], so the security considerations of 628 the I2NSF framework should be included in this document. Therefore, 629 proper secure communication channels should be used the delivery of 630 control or management messages among the components in the proposed 631 framework. 633 This document shares all the security issues of SDN that are 634 specified in the "Security Considerations" section of [ITU-T.Y.3300]. 636 6. Acknowledgments 638 This work was supported by Institute for Information & communications 639 Technology Promotion (IITP) grant funded by the Korea government 640 (MSIP) (No.R-20160222-002755, Cloud based Security Intelligence 641 Technology Development for the Customized Security Service 642 Provisioning). 644 7. Contributors 646 I2NSF is a group effort. I2NSF has had a number of contributing 647 authors. The following are considered co-authors: 649 o Hyoungshick Kim (Sungkyunkwan University) 651 o Jung-Soo Park (ETRI) 653 o Se-Hui Lee (Korea Telecom) 655 o Mohamed Boucadair (Orange) 657 8. Informative References 659 [AVANT-GUARD] 660 Shin, S., Yegneswaran, V., Porras, P., and G. Gu, "AVANT- 661 GUARD: Scalable and Vigilant Switch Flow Management in 662 Software-Defined Networks", ACM CCS, November 2013. 664 [consumer-facing-inf-dm] 665 Jeong, J., Kim, E., Ahn, T., Kumar, R., and S. Hares, 666 "I2NSF Consumer-Facing Interface YANG Data Model", draft- 667 ietf-i2nsf-consumer-facing-interface-dm-00 (work in 668 progress), March 2018. 670 [consumer-facing-inf-im] 671 Kumar, R., Lohiya, A., Qi, D., Bitar, N., Palislamovic, 672 S., Xia, L., and J. Jeong, "Information Model for 673 Consumer-Facing Interface to Security Controller", draft- 674 kumar-i2nsf-client-facing-interface-im-04 (work in 675 progress), October 2017. 677 [ETSI-NFV] 678 ETSI GS NFV 002 V1.1.1, "Network Functions Virtualisation 679 (NFV); Architectural Framework", October 2013. 681 [i2nsf-nsf-cap-im] 682 Xia, L., Strassner, J., Basile, C., and D. Lopez, 683 "Information Model of NSFs Capabilities", draft-ietf- 684 i2nsf-capability-00 (work in progress), September 2017. 686 [i2nsf-terminology] 687 Hares, S., Strassner, J., Lopez, D., Xia, L., and H. 688 Birkholz, "Interface to Network Security Functions (I2NSF) 689 Terminology", draft-ietf-i2nsf-terminology-05 (work in 690 progress), January 2018. 692 [ITU-T.X.1252] 693 Recommendation ITU-T X.1252, "Baseline Identity Management 694 Terms and Definitions", April 2010. 696 [ITU-T.X.800] 697 Recommendation ITU-T X.800, "Security Architecture for 698 Open Systems Interconnection for CCITT Applications", 699 March 1991. 701 [ITU-T.Y.3300] 702 Recommendation ITU-T Y.3300, "Framework of Software- 703 Defined Networking", June 2014. 705 [nsf-facing-inf-dm] 706 Kim, J., Jeong, J., Park, J., Hares, S., and Q. Lin, 707 "I2NSF Network Security Function-Facing Interface YANG 708 Data Model", draft-ietf-i2nsf-nsf-facing-interface-data- 709 model-00 (work in progress), March 2018. 711 [nsf-triggered-steering] 712 Hyun, S., Jeong, J., Park, J., and S. Hares, "Service 713 Function Chaining-Enabled I2NSF Architecture", draft-hyun- 714 i2nsf-nsf-triggered-steering-05 (work in progress), March 715 2018. 717 [ONF-OpenFlow] 718 ONF, "OpenFlow Switch Specification (Version 1.4.0)", 719 October 2013. 721 [ONF-SDN-Architecture] 722 ONF, "SDN Architecture", June 2014. 724 [opsawg-firewalls] 725 Baker, F. and P. Hoffman, "On Firewalls in Internet 726 Security", draft-ietf-opsawg-firewalls-01 (work in 727 progress), October 2012. 729 [registration-inf-dm] 730 Hyun, S., Jeong, J., Roh, T., Wi, S., and J. Park, "I2NSF 731 Registration Interface YANG Data Model", draft-hyun-i2nsf- 732 registration-dm-03 (work in progress), March 2018. 734 [registration-inf-im] 735 Hyun, S., Jeong, J., Roh, T., Wi, S., and J. Park, "I2NSF 736 Registration Interface Information Model", draft-hyun- 737 i2nsf-registration-interface-im-04 (work in progress), 738 March 2018. 740 [RFC4566] Handley, M., Jacobson, V., and C. Perkins, "SDP: Session 741 Description Protocol", RFC 4566, July 2006. 743 [RFC6020] Bjorklund, M., "YANG - A Data Modeling Language for the 744 Network Configuration Protocol (NETCONF)", RFC 6020, 745 October 2010. 747 [RFC6241] Enns, R., Bjorklund, M., Schoenwaelder, J., and A. 748 Bierman, "Network Configuration Protocol (NETCONF)", 749 RFC 6241, June 2011. 751 [RFC7149] Boucadair, M. and C. Jacquenet, "Software-Defined 752 Networking: A Perspective from within a Service Provider 753 Environment", RFC 7149, March 2014. 755 [RFC8040] Bierman, A., Bjorklund, M., and K. Watsen, "RESTCONF 756 Protocol", RFC 8040, January 2017. 758 [RFC8192] Hares, S., Lopez, D., Zarny, M., Jacquenet, C., Kumar, R., 759 and J. Jeong, "Interface to Network Security Functions 760 (I2NSF): Problem Statement and Use Cases", RFC 8192, July 761 2017. 763 [RFC8329] Lopez, D., Lopez, E., Dunbar, L., Strassner, J., and R. 764 Kumar, "Framework for Interface to Network Security 765 Functions", RFC 8329, February 2018. 767 Appendix A. Changes from draft-ietf-i2nsf-applicability-01 769 The following changes have been made from draft-ietf-i2nsf- 770 applicability-01: 772 o In Section 4, it is clarified what types of security policy rules 773 can be enforced by SDN switches or NSFs in the environment of 774 I2NSF framework with SDN. 776 o In Section 4, it is explained what should be done by the Security 777 Controller in order to divide the enforcement of security policy 778 rules into the SDN switches and NSFs in the I2NSF framework with 779 SDN. 781 Authors' Addresses 783 Jaehoon Paul Jeong 784 Department of Software 785 Sungkyunkwan University 786 2066 Seobu-Ro, Jangan-Gu 787 Suwon, Gyeonggi-Do 16419 788 Republic of Korea 790 Phone: +82 31 299 4957 791 Fax: +82 31 290 7996 792 EMail: pauljeong@skku.edu 793 URI: http://iotlab.skku.edu/people-jaehoon-jeong.php 795 Sangwon Hyun 796 Department of Software 797 Sungkyunkwan University 798 2066 Seobu-Ro, Jangan-Gu 799 Suwon, Gyeonggi-Do 16419 800 Republic of Korea 802 Phone: +82 31 290 7222 803 Fax: +82 31 299 6673 804 EMail: swhyun77@skku.edu 805 URI: http://imtl.skku.ac.kr/ 806 Tae-Jin Ahn 807 Korea Telecom 808 70 Yuseong-Ro, Yuseong-Gu 809 Daejeon 305-811 810 Republic of Korea 812 Phone: +82 42 870 8409 813 EMail: taejin.ahn@kt.com 815 Susan Hares 816 Huawei 817 7453 Hickory Hill 818 Saline, MI 48176 819 USA 821 Phone: +1-734-604-0332 822 EMail: shares@ndzh.com 824 Diego R. Lopez 825 Telefonica I+D 826 Jose Manuel Lara, 9 827 Seville 41013 828 Spain 830 Phone: +34 682 051 091 831 EMail: diego.r.lopez@telefonica.com