idnits 2.17.1 draft-ietf-forces-framework-10.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** Looks like you're using RFC 2026 boilerplate. This must be updated to follow RFC 3978/3979, as updated by RFC 4748. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- ** The document seems to lack a 1id_guidelines paragraph about Internet-Drafts being working documents. ** The document seems to lack a 1id_guidelines paragraph about 6 months document validity -- however, there's a paragraph with a matching beginning. Boilerplate error? == No 'Intended status' indicated for this document; assuming Proposed Standard == The page length should not exceed 58 lines per page, but there was 1 longer page, the longest (page 2) being 107 lines == It seems as if not all pages are separated by form feeds - found 0 form feeds but 39 pages 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.) ** There is 1 instance of too long lines in the document, the longest one being 20 characters in excess of 72. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the RFC 3978 Section 5.4 Copyright Line does not match the current year -- The document seems to lack a disclaimer for pre-RFC5378 work, but may have content which was first submitted before 10 November 2008. If you have contacted all the original authors and they are all willing to grant the BCP78 rights to the IETF Trust, then this is fine, and you can ignore this comment. If not, you may need to add the pre-RFC5378 disclaimer. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (October 2003) is 7499 days in the past. Is this intentional? -- Found something which looks like a code comment -- if you have code sections in the document, please surround them with '' and '' lines. Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Unused Reference: '2' is defined on line 1664, but no explicit reference was found in the text == Unused Reference: '12' is defined on line 1698, but no explicit reference was found in the text == Unused Reference: '16' is defined on line 1710, but no explicit reference was found in the text -- Possible downref: Non-RFC (?) normative reference: ref. '3' == Outdated reference: A later version (-13) exists of draft-ietf-idr-restart-06 -- Obsolete informational reference (is this intentional?): RFC 2246 (ref. '13') (Obsoleted by RFC 4346) -- Obsolete informational reference (is this intentional?): RFC 2401 (ref. '14') (Obsoleted by RFC 4301) -- Obsolete informational reference (is this intentional?): RFC 2409 (ref. '15') (Obsoleted by RFC 4306) Summary: 5 errors (**), 0 flaws (~~), 8 warnings (==), 7 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 1 Internet Draft L. Yang 2 Expiration: April 2004 Intel Corp. 3 File: draft-ietf-forces-framework-10.txt R. Dantu 4 Working Group: ForCES Univ. of North Texas 5 T. Anderson 6 Intel Corp. 7 R. Gopal 8 Nokia 10 October 2003 12 Forwarding and Control Element Separation (ForCES) Framework 14 draft-ietf-forces-framework-10.txt 16 Status of this Memo 18 This document is an Internet-Draft and is in full conformance with 19 all provisions of Section 10 of RFC2026. Internet-Drafts are 20 working documents of the Internet Engineering Task Force (IETF), 21 its areas, and its working groups. Note that other groups may also 22 distribute working documents as Internet-Drafts. 24 Internet-Drafts are draft documents valid for a maximum of six 25 months and may be updated, replaced, or obsoleted by other 26 documents at any time. It is inappropriate to use Internet-Drafts 27 as reference material or to cite them other than as ``work in 28 progress.'' 30 The list of current Internet-Drafts can be accessed at 31 http://www.ietf.org/ietf/1id-abstracts.txt. 33 The list of Internet-Draft Shadow Directories can be accessed at 34 http://www.ietf.org/shadow.html. 36 Copyright Notice 38 Copyright (C) The Internet Society (2003). All Rights Reserved. 40 Abstract 42 This document defines the architectural framework for the ForCES 43 (Forwarding and Control Element Separation) network elements, and 44 identifies the associated entities and the interactions among them. 46 Table of Contents 47 1. Definitions...................................................3 48 1.1. Conventions used in this document........................3 49 1.2. Terminologies............................................3 50 2. Introduction to Forwarding and Control Element Separation 51 (ForCES).........................................................5 52 3. Architecture..................................................9 53 3.1. Control Elements and Fr Reference Point.................10 54 3.2. Forwarding Elements and Fi reference point..............11 55 3.3. CE Managers.............................................14 56 3.4. FE Managers.............................................15 57 4. Operational Phases...........................................15 58 4.1. Pre-association Phase...................................15 59 4.1.1. Fl Reference Point.................................15 60 4.1.2. Ff Reference Point.................................16 61 4.1.3. Fc Reference Point.................................17 62 4.2. Post-association Phase and Fp reference point...........17 63 4.2.1. Proximity and Interconnect between CEs and FEs.....17 64 4.2.2. Association Establishment..........................18 65 4.2.3. Steady-state Communication.........................19 66 4.2.4. Data Packets across Fp reference point.............20 67 4.2.5. Proxy FE...........................................21 68 4.3. Association Re-establishment............................21 69 4.3.1. CE graceful restart................................22 70 4.3.2. FE restart.........................................23 71 5. Applicability to RFC1812.....................................24 72 5.1. General Router Requirements.............................25 73 5.2. Link Layer..............................................26 74 5.3. Internet Layer Protocols................................26 75 5.4. Internet Layer Forwarding...............................27 76 5.5. Transport Layer.........................................28 77 5.6. Application Layer -- Routing Protocols..................28 78 5.7. Application Layer -- Network Management Protocol........29 79 6. Summary......................................................29 80 7. Acknowledgements.............................................29 81 8. Security Considerations......................................29 82 8.1. Analysis of Potential Threats Introduced by ForCES......30 83 8.1.1. "Join" or "Remove" Message Flooding on CEs.........30 84 8.1.2. Impersonation Attack...............................30 85 8.1.3. Replay Attack......................................31 86 8.1.4. Attack during Fail Over............................31 87 8.1.5. Data Integrity.....................................31 88 8.1.6. Data Confidentiality...............................32 89 8.1.7. Sharing security parameters........................32 90 8.1.8. Denial of Service Attack via External Interface....32 91 8.2. Security Recommendations for ForCES.....................32 92 8.2.1. Security Configuration.............................33 93 8.2.2. Using TLS with ForCES..............................34 94 8.2.3. Using IPsec with ForCES............................35 95 9. Normative References.........................................36 96 10. Informative References......................................37 97 11. Authors' Addresses..........................................37 98 12. Intellectual Property Right.................................38 99 13. Full Copyright Statement....................................39 101 1. Definitions 103 1.1. Conventions used in this document 105 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 106 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in 107 this document are to be interpreted as described in RFC 2119. 109 1.2. Terminologies 111 A set of terminology associated with the ForCES requirements is 112 defined in [3] and we only include the definitions that are most 113 relevant to this document here. 115 Addressable Entity (AE) - An entity that is directly addressable 116 given some interconnect technology. For example, on IP networks, 117 it is a device to which we can communicate using an IP address; on 118 a switch fabric, it is a device to which we can communicate using a 119 switch fabric port number. 121 Physical Forwarding Element (PFE) - An AE that includes hardware 122 used to provide per-packet processing and handling. This hardware 123 may consist of (but is not limited to) network processors, ASICs 124 (Application-Specific Integrated Circuits), or general purpose 125 processors, installed on line cards, daughter boards, mezzanine 126 cards, or in stand-alone boxes. 128 PFE Partition - A logical partition of a PFE consisting of some 129 subset of each of the resources (e.g., ports, memory, forwarding 130 table entries) available on the PFE. This concept is analogous to 131 that of the resources assigned to a virtual switching element as 132 described in [8]. 134 Physical Control Element (PCE) - An AE that includes hardware used 135 to provide control functionality. This hardware typically includes 136 a general purpose processor. 138 PCE Partition - A logical partition of a PCE consisting of some 139 subset of each of the resources available on the PCE. 141 Forwarding Element (FE) - A logical entity that implements the 142 ForCES protocol. FEs use the underlying hardware to provide per- 143 packet processing and handling as directed by a CE via the ForCES 144 protocol. FEs may happen to be a single blade (or PFE), a 145 partition of a PFE or multiple PFEs. 147 Control Element (CE) - A logical entity that implements the ForCES 148 protocol and uses it to instruct one or more FEs how to process 149 packets. CEs handle functionality such as the execution of control 150 and signaling protocols. CEs may consist of PCE partitions or 151 whole PCEs. 153 ForCES Network Element (NE) - An entity composed of one or more CEs 154 and one or more FEs. To entities outside an NE, the NE represents 155 a single point of management. Similarly, an NE usually hides its 156 internal organization from external entities. 158 Pre-association Phase - The period of time during which an FE 159 Manager (see below) and a CE Manager (see below) are determining 160 which FE and CE should be part of the same network element. 162 Post-association Phase - The period of time during which an FE does 163 know which CE is to control it and vice versa, including the time 164 during which the CE and FE are establishing communication with one 165 another. 167 ForCES Protocol - While there may be multiple protocols used within 168 the overall ForCES architecture, the term "ForCES protocol" refers 169 only to the ForCES post-association phase protocol (see below). 171 ForCES Post-Association Phase Protocol - The protocol used for 172 post-association phase communication between CEs and FEs. This 173 protocol does not apply to CE-to-CE communication, FE-to-FE 174 communication, nor to communication between FE and CE managers. 175 The ForCES protocol is a master-slave protocol in which FEs are 176 slaves and CEs are masters. This protocol includes both the 177 management of the communication channel (e.g., connection 178 establishment, heartbeats) and the control messages themselves. 179 This protocol could be a single protocol or could consist of 180 multiple protocols working together, and may be unicast based or 181 multicast based. A separate protocol document will specify this 182 information. 184 FE Manager - A logical entity that operates in the pre-association 185 phase and is responsible for determining to which CE(s) an FE 186 should communicate. This process is called CE discovery and may 187 involve the FE manager learning the capabilities of available CEs. 188 An FE manager may use anything from a static configuration to a 189 pre-association phase protocol (see below) to determine which CE(s) 190 to use; however, this is currently out of scope. Being a logical 191 entity, an FE manager might be physically combined with any of the 192 other logical entities mentioned in this section. 194 CE Manager - A logical entity that operates in the pre-association 195 phase and is responsible for determining to which FE(s) a CE should 196 communicate. This process is called FE discovery and may involve 197 the CE manager learning the capabilities of available FEs. A CE 198 manager may use anything from a static configuration to a pre- 199 association phase protocol (see below) to determine which FE to 200 use, however this is currently out of scope. Being a logical 201 entity, a CE manager might be physically combined with any of the 202 other logical entities mentioned in this section. 204 Pre-association Phase Protocol - A protocol between FE managers and 205 CE managers that is used to determine which CEs or FEs to use. A 206 pre-association phase protocol may include a CE and/or FE 207 capability discovery mechanism. Note that this capability 208 discovery process is wholly separate from (and does not replace) 209 that used within the ForCES protocol. However, the two capability 210 discovery mechanisms may utilize the same FE model. 212 FE Model - A model that describes the logical processing functions 213 of an FE. 215 ForCES Protocol Element - An FE or CE. 217 Intra-FE topology - - Representation of how a single FE is realized 218 by combining possibly multiple logical functional blocks along 219 multiple data path. This is defined by the FE model. 221 FE Topology -- Representation of how the multiple FEs in a single 222 NE are interconnected. Sometimes it is called inter-FE topology, 223 to be distinguished from intra-FE topology used by the FE model. 225 Inter-FE topology - see FE Topology. 227 2. Introduction to Forwarding and Control Element Separation (ForCES) 229 An IP network element (NE) appears to external entities as a 230 monolithic piece of network equipment, e.g., a router, NAT, 231 firewall, or load balancer. Internally, however, an IP network 232 element (NE) (such as a router) is composed of numerous logically 233 separated entities that cooperate to provide a given functionality 234 (such as routing). Two types of network element components exist: 235 control element (CE) in control plane and forwarding element (FE) 236 in forwarding plane (or data plane). Forwarding elements typically 237 are ASIC, network-processor, or general-purpose processor-based 238 devices that handle data path operations for each packet. Control 239 elements are typically based on general-purpose processors that 240 provide control functionality like routing and signaling protocols. 242 ForCES aims to define a framework and associated protocol(s) to 243 standardize information exchange between the control and forwarding 244 plane. Having standard mechanisms allows CEs and FEs to become 245 physically separated standard components. This physical separation 246 accrues several benefits to the ForCES architecture. Separate 247 components would allow component vendors to specialize in one 248 component without having to become experts in all components. 249 Standard protocol also allows the CEs and FEs from different 250 component vendors to interoperate with each other and hence it 251 becomes possible for system vendors to integrate together the CEs 252 and FEs from different component suppliers. This interoperability 253 translates into more design choices and flexibility for the system 254 vendors. Overall, ForCES will enable rapid innovation in both the 255 control and forwarding planes while maintaining interoperability. 256 Scalability is also easily provided by this architecture in that 257 additional forwarding or control capacity can be added to existing 258 network elements without the need for forklift upgrades. 260 ------------------------- ------------------------- 261 | Control Blade A | | Control Blade B | 262 | (CE) | | (CE) | 263 ------------------------- ------------------------- 264 ^ | ^ | 265 | | | | 266 | V | V 267 --------------------------------------------------------- 268 | Switch Fabric Backplane | 269 --------------------------------------------------------- 270 ^ | ^ | ^ | 271 | | | | . . . | | 272 | V | V | V 273 ------------ ------------ ------------ 274 |Router | |Router | |Router | 275 |Blade #1 | |Blade #2 | |Blade #N | 276 | (FE) | | (FE) | | (FE) | 277 ------------ ------------ ------------ 278 ^ | ^ | ^ | 279 | | | | . . . | | 280 | V | V | V 282 Figure 1. A router configuration example with separate blades. 284 One example of such physical separation is at the blade level. 285 Figure 1 shows such an example configuration of a router, with two 286 control blades and multiple forwarding blades, all interconnected 287 into a switch fabric backplane. In such a chassis configuration, 288 the control blades are the CEs while the router blades are FEs, and 289 the switch fabric backplane provides the physical interconnect for 290 all the blades. Control blade A may be the primary CE while 291 control blade B may be the backup CE providing redundancy. It is 292 also possible to have a redundant switch fabric for high 293 availability support. Routers today with this kind of 294 configuration use proprietary interfaces for messaging between CEs 295 and FEs. The goal of ForCES is to replace such proprietary 296 interfaces with a standard protocol. With a standard protocol like 297 ForCES implemented on all blades, it becomes possible for control 298 blades from vendor X and forwarding blades from vendor Y to work 299 seamlessly together in one chassis. 301 ------- ------- 302 | CE1 | | CE2 | 303 ------- ------- 304 ^ ^ 305 | | 306 V V 307 ============================================ Ethernet 308 ^ ^ . . . ^ 309 | | | 310 V V V 311 ------- ------- -------- 312 | FE#1| | FE#2| | FE#n | 313 ------- ------- -------- 314 ^ | ^ | ^ | 315 | | | | | | 316 | V | V | V 318 Figure 2. A router configuration example with separate boxes. 320 Another level of physical separation between the CEs and FEs can be 321 at the box level. In such configuration, all the CEs and FEs are 322 physically separated boxes, interconnected with some kind of high 323 speed LAN connection (like Gigabit Ethernet). These separated CEs 324 and FEs are only one hop away from each other within a local area 325 network. The CEs and FEs communicate to each other by running 326 ForCES, and the collection of these CEs and FEs together become one 327 routing unit to the external world. Figure 2 shows such an example. 329 In both examples shown here, the same physical interconnect is used 330 for both CE-to-FE and FE-to-FE communication. However, that does 331 not have to be the case. One reason to use different interconnects 332 is that CE-to-FE interconnect does not have to be as fast as the 333 FE-to-FE interconnect, so the more expensive fast connections can 334 be saved for FE-to-FE. The separate interconnects may also provide 335 reliability and redundancy benefits for the NE. 337 Some examples of control functions that can be implemented in the 338 CE include routing protocols like RIP, OSPF and BGP, control and 339 signaling protocols like RSVP (Resource Reservation Protocol), LDP 340 (Label Distribution Protocol) for MPLS, etc. Examples of 341 forwarding functions in the FE include LPM (longest prefix match) 342 forwarder, classifiers, traffic shaper, meter, NAT (Network Address 343 Translators), etc. Figure 3 provides example functions in both CE 344 and FE. Any given NE may contain one or many of these CE and FE 345 functions in it. The diagram also shows that ForCES protocol is 346 used to transport both the control messages for ForCES itself and 347 the data packets that are originated/destined from/to the control 348 functions in CE (e.g., routing packets). Section 4.2.4 provides 349 more detail on this. 351 ------------------------------------------------- 352 | | | | | | | 353 |OSPF |RIP |BGP |RSVP |LDP |. . . | 354 | | | | | | | 355 ------------------------------------------------- 356 | ForCES Interface | 357 ------------------------------------------------- 358 ^ ^ 359 ForCES | |data 360 control | |packets 361 messages| |(e.g., routing packets) 362 v v 363 ------------------------------------------------- 364 | ForCES Interface | 365 ------------------------------------------------- 366 | | | | | | | 367 |LPM Fwd|Meter |Shaper |NAT |Classi-|. . . | 368 | | | | |fier | | 369 ------------------------------------------------- 370 | FE resources | 371 ------------------------------------------------- 372 Figure 3. Examples of CE and FE functions 374 A set of requirements for control and forwarding separation is 375 identified in [3]. This document describes a ForCES architecture 376 that satisfies the architectural requirements of that document and 377 defines a framework for ForCES network elements and the associated 378 entities to facilitate protocol definition. Whenever necessary, 379 this document uses many examples to illustrate the issues and/or 380 possible solutions in ForCES. These examples are intended to be 381 just examples, and should not be taken as the only or definite ways 382 of doing certain things. It is expected that separate document 383 will be produced by the ForCES working group to specify the ForCES 384 protocol(s). 386 3. Architecture 388 This section defines the ForCES architectural framework and the 389 associated logical components. This ForCES framework defines 390 components of ForCES NEs including several ancillary components. 391 These components may be connected in different kinds of topologies 392 for flexible packet processing. 394 --------------------------------------- 395 | ForCES Network Element | 396 -------------- Fc | -------------- -------------- | 397 | CE Manager |---------+-| CE 1 |------| CE 2 | | 398 -------------- | | | Fr | | | 399 | | -------------- -------------- | 400 | Fl | | | Fp / | 401 | | Fp| |----------| / | 402 | | | |/ | 403 | | | | | 404 | | | Fp /|----| | 405 | | | /--------/ | | 406 -------------- Ff | -------------- -------------- | 407 | FE Manager |---------+-| FE 1 | Fi | FE 2 | | 408 -------------- | | |------| | | 409 | -------------- -------------- | 410 | | | | | | | | | | 411 ----+--+--+--+----------+--+--+--+----- 412 | | | | | | | | 413 | | | | | | | | 414 Fi/f Fi/f 416 Figure 4. ForCES Architectural Diagram 418 The diagram in Figure 4 shows the logical components of the ForCES 419 architecture and their relationships. There are two kinds of 420 components inside a ForCES network element: control element (CE) 421 and forwarding element (FE). The framework allows multiple 422 instances of CE and FE inside one NE. Each FE contains one or more 423 physical media interfaces for receiving and transmitting packets 424 from/to the external world. The aggregation of these FE interfaces 425 becomes the NE's external interfaces. In addition to the external 426 interfaces, there must also exist some kind of interconnect within 427 the NE so that the CE and FE can communicate with each other, and 428 one FE can forward packets to another FE. The diagram also shows 429 two entities outside of the ForCES NE: CE Manager and FE Manager. 430 These two entities provide configuration to the corresponding CE or 431 FE in the pre-association phase (see Section 4.1). There is no 432 defined role for FE Manager and CE Manager in post-association 433 phase, thus these logical components are not considered part of the 434 ForCES NE. 436 For convenience, the logical interactions between these components 437 are labeled by reference points Fp, Fc, Ff, Fr, Fl, and Fi, as 438 shown in Figure 4. The FE external interfaces are labeled as Fi/f. 439 More detail is provided in Section 3 and 4 for each of these 440 reference points. All these reference points are important in 441 understanding the ForCES architecture, however, the ForCES protocol 442 is only defined over one reference point -- Fp. 444 The interface between two ForCES NEs is identical to the interface 445 between two conventional routers and these two NEs exchange the 446 protocol packets through the external interfaces at Fi/f. ForCES 447 NEs connect to existing routers transparently. 449 3.1. Control Elements and Fr Reference Point 451 It is not necessary to define any protocols across the Fr reference 452 point to enable control and forwarding separation for simple 453 configurations like single CE and multiple FEs. However, this 454 architecture permits multiple CEs to be present in a network 455 element. In cases where an implementation uses multiple CEs, the 456 invariant that the CEs and FEs together appear as a single NE must 457 be maintained. 459 Multiple CEs may be used for redundancy, load sharing, distributed 460 control, or other purposes. Redundancy is the case where one or 461 more CEs are prepared to take over should an active CE fail. Load 462 sharing is the case where two or more CEs are concurrently active 463 and any request that can be serviced by one of the CEs can also be 464 serviced by any of the other CEs. For both redundancy and load 465 sharing, the CEs involved are equivalently capable. The only 466 difference between these two cases is in terms of how many active 467 CEs there are. Distributed control is the case where two or more 468 CEs are concurrently active but certain requests can only be 469 serviced by certain CEs. 471 When multiple CEs are employed in a ForCES NE, their internal 472 organization is considered an implementation issue that is beyond 473 the scope of ForCES. CEs are wholly responsible for coordinating 474 amongst themselves via the Fr reference point to provide 475 consistency and synchronization. However, ForCES does not define 476 the implementation or protocols used between CEs, nor does it 477 define how to distribute functionality among CEs. Nevertheless, 478 ForCES will support mechanisms for CE redundancy or fail over, and 479 it is expected that vendors will provide redundancy or fail over 480 solutions within this framework. 482 3.2. Forwarding Elements and Fi reference point 484 An FE is a logical entity that implements the ForCES protocol and 485 uses the underlying hardware to provide per-packet processing and 486 handling as directed by a CE. It is possible to partition one 487 physical FE into multiple logical FEs. It is also possible for one 488 FE to use multiple physical FEs. The mapping between physical 489 FE(s) and the logical FE(s) is beyond the scope of ForCES. For 490 example, a logical partition of a physical FE can be created by 491 assigning some portion of each of the resources (e.g., ports, 492 memory, forwarding table entries) available on the physical FE to 493 each of the logical FEs. Such concept of FE virtualization is 494 analogous to a virtual switching element as described in [8]. If 495 FE virtualization occurs only in the pre-association phase, it has 496 no impact on ForCES. However, if FE virtualization results in 497 dynamic resource change on FE during post-association phase, the FE 498 model needs to be able to report such capability and the ForCES 499 protocol needs to be able to inform the CE of such change via 500 asynchronous messages (see [3], Section 5, requirement #6). 502 FEs perform all packet processing functions as directed by CEs. 503 FEs have no initiative of their own. Instead, FEs are slaves and 504 only do as they are told. FEs may communicate with one or more CEs 505 concurrently across reference point Fp. FEs have no notion of CE 506 redundancy, load sharing, or distributed control. Instead, FEs 507 accept commands from any CE authorized to control them, and it is 508 up to the CEs to coordinate among themselves to achieve redundancy, 509 load sharing or distributed control. The idea is to keep FEs as 510 simple and dumb as possible so that FEs can focus their resource on 511 the packet processing functions. 513 For example, in Figure 5, FE1 and FE2 can be configured to accept 514 commands from both the primary CE (CE1) and the backup CE (CE2). 515 Upon detection of CE1 failure, perhaps across the Fr or Fp 516 reference point, CE2 is configured to take over activities of CE1. 517 This is beyond the scope of ForCES and is not discussed further. 519 Distributed control can be achieved in a similar fashion, without 520 much intelligence on the part of FEs. For example, FEs can be 521 configured to detect RSVP and BGP protocol packets, and forward 522 RSVP packets to one CE and BGP packets to another CE. Hence, FEs 523 may need to do packet filtering for forwarding packets to specific 524 CEs. 526 ------- Fr ------- 527 | CE1 | ------| CE2 | 528 ------- ------- 529 | \ / | 530 | \ / | 531 | \ / | 532 | \/Fp | 533 | /\ | 534 | / \ | 535 | / \ | 536 ------- Fi ------- 537 | FE1 |<----->| FE2 | 538 ------- ------- 540 Figure 5. CE redundancy example. 542 This architecture permits multiple FEs to be present in an NE. [3] 543 dictates that the ForCES protocol must be able to scale to at least 544 hundreds of FEs (see [3] Section 5, requirement #11). Each of 545 these FEs may potentially have a different set of packet processing 546 functions, with different media interfaces. FEs are responsible 547 for basic maintenance of layer-2 connectivity with other FEs and 548 with external entities. Many layer-2 media include sophisticated 549 control protocols. The FORCES protocol (over the Fp reference 550 point) will be able to carry messages for such protocols so that, 551 in keeping with the dumb FE model, the CE can provide appropriate 552 intelligence and control over these media. 554 When multiple FEs are present, ForCES requires that packets must be 555 able to arrive at the NE by one FE and leave the NE via a different 556 FE (See [3], Section 5, Requirement #3). Packets that enter the NE 557 via one FE and leave the NE via a different FE are transferred 558 between FEs across the Fi reference point. Fi reference point 559 could be used by FEs to discovery their (inter-FE) topology, 560 perhaps during pre-association phase. The Fi reference point is a 561 separate protocol from the Fp reference point and is not currently 562 defined by the ForCES architecture. 564 FEs could be connected in different kinds of topologies and packet 565 processing may spread across several FEs in the topology. Hence, 566 logical packet flow may be different from physical FE topology. 567 Figure 6 provides some topology examples. When it is necessary to 568 forward packets between FEs, the CE needs to understand the FE 569 topology. The FE topology may be queried from the FEs by the CEs 570 via ForCES protocol, but the FEs are not required to provide that 571 information to the CEs. So, the FE topology information may also be 572 gathered by other means outside of the ForCES protocol (like inter- 573 FE topology discovery protocol). 575 ----------------- 576 | CE | 577 ----------------- 578 ^ ^ ^ 579 / | \ 580 / v \ 581 / ------- \ 582 / +->| FE3 |<-+ \ 583 / | | | | \ 584 v | ------- | v 585 ------- | | ------- 586 | FE1 |<-+ +->| FE2 | 587 | |<--------------->| | 588 ------- ------- 589 ^ | ^ | 590 | | | | 591 | v | v 593 (a) Full mesh among FE1, FE2 and FE3. 595 ----------- 596 | CE | 597 ----------- 598 ^ ^ ^ ^ 599 / | | \ 600 /------ | | ------\ 601 v v v v 602 ------- ------- ------- ------- 603 | FE1 |<->| FE2 |<->| FE3 |<->| FE4 | 604 ------- ------- ------- ------- 605 ^ | ^ | ^ | ^ | 606 | | | | | | | | 607 | v | v | v | v 609 (b) Multiple FEs in a daisy chain 611 ^ | 612 | v 613 ----------- 614 | FE1 |<-----------------------| 615 ----------- | 616 ^ ^ | 617 / \ | 618 | ^ / \ ^ | V 619 v | v v | v ---------- 620 --------- --------- | | 621 | FE2 | | FE3 |<------------>| CE | 622 --------- --------- | | 623 ^ ^ ^ ---------- 624 | \ / ^ ^ 625 | \ / | | 626 | v v | | 627 | ----------- | | 628 | | FE4 |<----------------------| | 629 | ----------- | 630 | | ^ | 631 | v | | 632 | | 633 |----------------------------------------| 635 (c) Multiple FEs connected by a ring 637 Figure 6. Some examples of FE topology. 639 3.3.CE Managers 641 CE managers are responsible for determining which FEs a CE should 642 control. It is legitimate for CE managers to be hard-coded with 643 the knowledge of with which FEs its CEs should communicate with. A 644 CE manager may also be physically embedded into a CE and be 645 implemented as a simple keypad or other direct configuration 646 mechanism on the CE. Finally, CE managers may be physically and 647 logically separate entities that configure the CE with FE 648 information via such mechanisms as COPS-PR [6] or SNMP [4]. 650 3.4. FE Managers 652 FE managers are responsible for determining with which CE any 653 particular FE should initially communicate. Like CE managers, no 654 restrictions are placed on how an FE manager decides with which CE 655 its FEs should communicate, nor are restrictions placed on how FE 656 managers are implemented. Each FE should have one and only one FE 657 manager, while different FEs may have the same or different FE 658 manager(s). Each manager can choose to exist and operate 659 independently of other manager. 661 4. Operational Phases 663 Both FEs and CEs require some configuration in place before they 664 can start information exchange and function as a coherent network 665 element. Two operational phases are identified in this framework: 666 pre-association and post-association. 668 4.1.Pre-association Phase 670 Pre-association phase is the period of time during which an FE 671 Manager and a CE Manager are determining which FE and CE should be 672 part of the same network element. The protocols used during this 673 phase may include all or some of the message exchange over Fl, Ff 674 and Fc reference points. However, all these may be optional and 675 none of this is within the scope of ForCES protocol. 677 4.1.1. Fl Reference Point 679 CE managers and FE managers may communicate across the Fl reference 680 point in the pre-association phase in order to determine which CEs 681 and FEs should communicate with each other. Communication across 682 the Fl reference point is optional in this architecture. No 683 requirements are placed on this reference point. 685 CE managers and FE managers may be operated by different entities. 686 The operator of the CE manager may not want to divulge, except to 687 specified FE managers, any characteristics of the CEs it manages. 688 Similarly, the operator of the FE manager may not want to divulge 689 FE characteristics, except to authorized entities. As such, CE 690 managers and FE managers may need to authenticate one another. 691 Subsequent communication between CE managers and FE managers may 692 require other security functions such as privacy, non-repudiation, 693 freshness, and integrity. 695 FE Manager FE CE Manager CE 696 | | | | 697 | | | | 698 |(security exchange) | | 699 1|<------------------------------>| | 700 | | | | 701 |(a list of CEs and their attributes) | 702 2|<-------------------------------| | 703 | | | | 704 |(a list of FEs and their attributes) | 705 3|------------------------------->| | 706 | | | | 707 | | | | 708 |<----------------Fl------------>| | 710 Figure 7. An example of message exchange over Fl reference point 712 Once the necessary security functions have been performed, the CE 713 and FE managers communicate to determine which CEs and FEs should 714 communicate with each other. At the very minimum, the CE and FE 715 managers need to learn of the existence of available FEs and CEs 716 respectively. This discovery process may entail one or both 717 managers learning the capabilities of the discovered ForCES 718 protocol elements. Figure 7 shows an example of possible message 719 exchange between CE manager and FE manager over Fl reference point. 721 4.1.2. Ff Reference Point 723 The Ff reference point is used to inform forwarding elements of the 724 association decisions made by the FE manager in pre-association 725 phase. Only authorized entities may instruct an FE with respect to 726 which CE should control it. Therefore, privacy, integrity, 727 freshness, and authentication are necessary between the FE manager 728 and FEs when the FE manager is remote to the FE. Once the 729 appropriate security has been established, the FE manager instructs 730 the FEs across this reference point to join a new NE or to 731 disconnect from an existing NE. The FE Manager could also assign 732 unique FE identifiers to the FEs using this reference point. The 733 FE identifiers are useful in post association phase to express FE 734 topology. Figure 8 shows example of message exchange over Ff 735 reference point. 737 FE Manager FE CE Manager CE 738 | | | | 739 | | | | 740 |(security exchange) |(security exchange) 741 1|<------------>|authentication 1|<----------->|authentication 742 | | | | 743 |(FE ID, attributes) |(CE ID, attributes) 744 2|<-------------|request 2|<----------->|request 745 | | | | 746 3|------------->|response 3|------------>|response 747 |(corresponding CE ID) |(corresponding FE ID) 748 | | | | 749 | | | | 750 |<-----Ff----->| |<-----Fc---->| 752 Figure 8. Examples of message exchange 753 over Ff and Fc reference points. 755 Note that the FE manager function may be co-located with the FE 756 (such as by manual keypad entry of the CE IP address), in which 757 case this reference point is reduced to a built-in function. 759 4.1.3. Fc Reference Point 761 The Fc reference point is used to inform control elements of the 762 association decisions made by CE managers in pre-association phase. 763 When the CE manager is remote, only authorized entities may 764 instruct a CE to control certain FEs. Privacy, integrity, 765 freshness and authentication are also required across this 766 reference point in such a configuration. Once appropriate security 767 has been established, the CE manager instructs CEs as to which FEs 768 they should control and how they should control them. Figure 8 769 shows example of message exchange over Fc reference point. 771 As with the FE manager and FEs, configurations are possible where 772 the CE manager and CE are co-located and no protocol is used for 773 this function. 775 4.2. Post-association Phase and Fp reference point 777 Post-association phase is the period of time during which an FE and 778 CE have been configured with information necessary to contact each 779 other and includes both association establishment and steady-state 780 communication. The communication between CE and FE is performed 781 across the Fp ("p" meaning protocol) reference point. ForCES 782 protocol is exclusively used for all communication across the Fp 783 reference point. 785 4.2.1. Proximity and Interconnect between CEs and FEs 786 The ForCES Working Group has made a conscious decision that the 787 first version of ForCES will be focused on "very close" CE/FE 788 localities in IP networks. Very Close localities consist of 789 control and forwarding elements that either are components in the 790 same physical box, or are separated at most by one local network 791 hop ([7]). CEs and FEs can be connected by a variety of 792 interconnect technologies, including Ethernet connections, 793 backplanes, ATM (cell) fabrics, etc. ForCES should be able to 794 support each of these interconnects (see [3] Section 5, requirement 795 #1). When the CEs and FEs are separated beyond a single L3 routing 796 hop, the ForCES protocol will make use of an existing RFC2914 797 compliant L4 protocol with adequate reliability, security and 798 congestion control (e.g. TCP, SCTP) for transport purposes. 800 4.2.2. Association Establishment 802 FE CE 803 | | 804 |(Security exchange.) | 805 1|<--------------------->| 806 | | 807 |(Let me join the NE please.) 808 2|---------------------->| 809 | | 810 |(What kind of FE are you? -- capability query) 811 3|<----------------------| 812 | | 813 |(Here is my FE functions/state: use model to 814 describe) 815 4|---------------------->| 816 | | 817 |(How are you connected with other FEs?) 818 5|<----------------------| 819 | | 820 |(Here is the FE topology info) 821 6|---------------------->| 822 | | 823 |(Initial config for FE -- optional) 824 7|<----------------------| 825 | | 826 |(I am ready to go. Shall I?) 827 8|---------------------->| 828 | | 829 |(Go ahead!) | 830 9|<----------------------| 831 | | 833 Figure 9. Example of message exchange between CE and FE 834 over Fp to establish NE association 836 As an example, figure 9 shows some of the message exchange that may 837 happen before the association between the CE and FE is fully 838 established. Either the CE or FE can initiate the connection. 839 Security handshake is necessary to authenticate the two 840 communication endpoints to each other before any further message 841 exchange can happen. The exact details of the security handshake 842 depend on the security solution chosen by ForCES protocol. Section 843 8 provides more details on the security considerations for ForCES. 844 After the successful security handshake, the FE needs to inform the 845 CE of its own capability and its topology in relation to other FEs. 846 The capability of the FE is represented by the FE model, described 847 in a separate document. The model would allow an FE to describe 848 what kind of packet processing functions it contains, in what order 849 the processing happens, what kinds of configurable parameters it 850 allows, what statistics it collects and what events it might throw, 851 etc. Once such information is available to the CE, the CE may 852 choose to send some initial or default configuration to the FE so 853 that the FE can start receiving and processing packets correctly. 854 Such initialization may not be necessary if the FE already obtains 855 the information from its own bootstrap process. Once FE starts 856 accepting packets for processing, we say the association of this FE 857 with its CE is now established. From then on, the CE and FE enter 858 steady-state communication. 860 4.2.3. Steady-state Communication 862 Once an association is established between the CE and FE, the 863 ForCES protocol is used by the CE and FE over Fp reference point to 864 exchange information to facilitate packet processing. 866 FE CE 867 | | 868 |(Add these new routes.)| 869 1|<----------------------| 870 | | 871 |(Successful.) | 872 2|---------------------->| 873 | | 874 | | 875 |(Query some stats.) | 876 1|<----------------------| 877 | | 878 |(Reply with stats collected.) 879 2|---------------------->| 880 | | 881 | | 882 |(My port is down, with port #.) 883 1|---------------------->| 884 | | 885 |(Here is a new forwarding table) 886 2|<----------------------| 887 | | 888 Figure 10. Examples of message exchange between CE and FE 889 over Fp during steady-state communication 891 Based on the information acquired through CEs' control processing, 892 CEs will frequently need to manipulate the packet-forwarding 893 behaviors of their FE(s) by sending instructions to FEs. For 894 example, Figure 10 shows message exchange examples in which the CE 895 sends new routes to the FE so that the FE can add them to its 896 forwarding table. The CE may query the FE for statistics collected 897 by the FE and the FE may notify the CE of important events such as 898 port failure. 900 4.2.4. Data Packets across Fp reference point 902 --------------------- ---------------------- 903 | | | | 904 | +--------+ | | +--------+ | 905 | |CE(BGP) | | | |CE(BGP) | | 906 | +--------+ | | +--------+ | 907 | | | | ^ | 908 | |Fp | | |Fp | 909 | v | | | | 910 | +--------+ | | +--------+ | 911 | | FE | | | | FE | | 912 | +--------+ | | +--------+ | 913 | | | | ^ | 914 | Router | | | Router | | 915 | A | | | B | | 916 ---------+----------- -----------+---------- 917 v ^ 918 | | 919 | | 920 ------------------->--------------- 922 Figure 11. Example to show data packet flow between two NEs. 924 Control plane protocol packets (such as RIP, OSPF messages) 925 addressed to any of NE's interfaces are typically redirected by the 926 receiving FE to its CE, and CE may originate packets and have its 927 FE deliver them to other NEs. Therefore, ForCES protocol over Fp 928 not only transports the ForCES protocol messages between CEs and 929 FEs, but also encapsulates the data packets from control plane 930 protocols. Moreover, one FE may be controlled by multiple CEs for 931 distributed control. In this configuration, the control protocols 932 supported by the FORCES NEs may spread across multiple CEs. For 933 example, one CE may support routing protocols like OSPF and BGP, 934 while a signaling and admission control protocol like RSVP is 935 supported in another CE. FEs are configured to recognize and 936 filter these protocol packets and forward them to the corresponding 937 CE. 939 Figure 11 shows one example of how the BGP packets originated by 940 router A are passed to router B. In this example, the ForCES 941 protocol is used to transport the packets from the CE to the FE 942 inside router A, and then from the FE to the CE inside router B. 943 In light of the fact that the ForCES protocol is responsible for 944 transporting both the control messages and the data packets between 945 the CE and FE over Fp reference point, it is possible to use either 946 a single protocol or multiple protocols to achieve this. 948 4.2.5. Proxy FE 950 In the case where a physical FE cannot implement (e.g., due to the 951 lack of a general purpose CPU) the ForCES protocol directly, a 952 proxy FE can be used in the middle of Fp reference point. This 953 allows the CE communicate to the physical FE via the proxy by using 954 ForCES, while the proxy manipulates the physical FE using some 955 intermediary form of communication (e.g., a non-ForCES protocol or 956 DMA). In such an implementation, the combination of the proxy and 957 the physical FE becomes one logical FE entity. 959 4.3. Association Re-establishment 961 FEs and CEs may join and leave NEs dynamically (see [3] Section 5, 962 requirements #12). When an FE or CE leaves the NE, the association 963 with the NE is broken. If the leaving party rejoins an NE later, 964 to re-establish the association, it may need to re-enter the pre- 965 association phase. Loss of association can also happen 966 unexpectedly due to loss of connection between the CE and the FE. 967 Therefore, the framework allows the bi-directional transition 968 between these two phases, but the ForCES protocol is only 969 applicable for the post-association phase. However, the protocol 970 should provide mechanisms to support association re-establishment. 972 This includes the ability for CEs and FEs to determine when there 973 is a loss of association between them, ability to restore 974 association and efficient state (re)synchronization mechanisms (see 975 [3] Section 5, requirement #7). Note that security association and 976 state must be also re-established to guarantee the same level of 977 security exists before and after the association re-establishment. 979 When an FE leaves or joins an existing NE that is already in post- 980 association phase, the CE needs to be aware of the impact on FE 981 topology and deals with the change accordingly. 983 4.3.1. CE graceful restart 985 The failure and restart of the CE in a router can potentially cause 986 much stress and disruption on the control plane throughout a 987 network. Because when a CE has to restart for any reason, the 988 router loses routing adjacencies or sessions with its routing 989 neighbors. Neighbors who detect the lost adjacency normally re- 990 compute new routes and then send routing updates to their own 991 neighbors to communicate the lost adjacency. Their neighbors do 992 the same thing to propagate throughout the network. In the 993 meantime, the restarting router cannot receive traffic from other 994 routers because the neighbors have stopped using the router's 995 previously advertised routes. When the restarting router restores 996 adjacencies, neighbors must once again re-compute new routes and 997 send out additional routing updates. The restarting router is 998 unable to forward packets until it has re-established routing 999 adjacencies with neighbors, received route updates through these 1000 adjacencies, and computed new routes. Until convergence takes 1001 place throughout the network, packets may be lost in transient 1002 black holes or forwarding loops. 1004 A high availability mechanism known as the "graceful restart" has 1005 been used by the IP routing protocols (OSPF [10], BGP [11], BGP 1006 [11]) and MPLS label distribution protocol (LDP [9]) to help 1007 minimize the negative effects on routing throughout an entire 1008 network caused by a restarting router. Route flap on neighboring 1009 routers is avoided, and a restarting router can continue to forward 1010 packets that would otherwise be dropped. 1012 While the details differ from protocol to protocol, the general 1013 idea behind the graceful restart mechanism remains the same. With 1014 the graceful restart, a restarting router can inform its neighbors 1015 when it restarts. The neighbors may detect the lost adjacency but 1016 do not recompute new routes or send routing updates to their 1017 neighbors. The neighbors also hold on to the routes received from 1018 the restarting router before restart and assume they are still 1019 valid for a limited time. By doing so, the restarting router's FEs 1020 can also continue to receive and forward traffic from other 1021 neighbors for a limited time by using the routes they already have. 1022 The restarting router then re-establishes routing adjacencies, 1023 downloads updated routes from all its neighbors, recomputes new 1024 routes and uses them to replace the older routes it was using. It 1025 then sends these updated routes to its neighbors and signals the 1026 completion of the graceful restart process. 1028 Non-stop forwarding is a requirement for graceful restart. It is 1029 necessary so a router can continue to forward packets while it is 1030 downloading routing information and recomputing new routes. This 1031 ensures that packets will not be dropped. As one can see, one of 1032 the benefits afforded by the separation of CE and FE is exactly the 1033 ability of non-stop forwarding in the face of the CE failure and 1034 restart. The support of dynamic changes to CE/FE association in 1035 ForCES also makes it compatible with high availability mechanisms 1036 such as graceful restart. 1038 ForCES should be able to support CE graceful restart easily. When 1039 the association is established the first time, the CE must inform 1040 the FEs what to do in the case of CE failure. If graceful restart 1041 is not supported, the FEs may be told to stop packet processing all 1042 together if its CE fails. If graceful restart is supported, the 1043 FEs should be told to cache and hold on to its FE state including 1044 the forwarding tables across the restarts. A timer must be 1045 included so that the timeout causes such cached state to expire 1046 eventually. Those timers should be settable by the CE. 1048 4.3.2. FE restart 1050 In the same example in Figure 5, assuming CE1 is the working CE for 1051 the moment, what would happen if one of the FEs, say FE1, leaves 1052 the NE temporarily? FE1 may voluntarily decide to leave the 1053 association. Alternatively, FE1 may stop functioning simply due to 1054 unexpected failure. In the former case, CE1 receives a "leave- 1055 association request" from FE1. In the latter, CE1 detects the 1056 failure of FE1 by some other means. In both cases, CE1 must inform 1057 the routing protocols of such an event, most likely prompting a 1058 reachability and SPF (Shortest Path First) recalculation and 1059 associated downloading of new FIBs from CE1 to the other remaining 1060 FEs (only FE2 in this example). Such recalculation and FIB update 1061 will also be propagated from CE1 to the NE's neighbors that are 1062 affected by the connectivity of FE1. 1064 When FE1 decides to rejoin again, or when it restarts again from 1065 the failure, FE1 needs to re-discover its master (CE). This can be 1066 achieved by several means. It may re-enter the pre-association 1067 phase and get that information from its FE manager. It may 1068 retrieve the previous CE information from its cache, if it can 1069 validate the information freshness. Once it discovers its CE, it 1070 starts message exchange with the CE to re-establish the association 1071 just as outlined in Figure 9, with the possible exception that it 1072 might be able to bypass the transport of the complete initial 1073 configuration. Suppose that FE1 still has its routing table and 1074 other state information from the last association. Instead of 1075 sending all the information again from scratch, it may be able to 1076 use more efficient mechanism to re-sync up the state with its CE if 1077 such mechanism is supported by the ForCES protocol. For example, 1078 CRC-32 of the state might give a quick indication of whether or not 1079 the state is in-sync with its CE. By comparing its state with the 1080 CE first, it sends an information update only if it is needed. 1081 ForCES protocol may choose to implement similar optimization 1082 mechanisms, but it may also choose not to, as this is not a 1083 requirement. 1085 5. Applicability to RFC1812 1087 [3] Section 5, requirement #9 dictates "Any proposed ForCES 1088 architecture must explain how that architecture supports all of the 1089 router functions as defined in RFC1812." RFC1812 discusses many 1090 important requirements for IPv4 routers from the link layer to the 1091 application layer. This section addresses the relevant 1092 requirements in RFC1812 for implementing IPv4 routers based on 1093 ForCES architecture and explains how ForCES satisfies these 1094 requirements by providing guidelines on how to separate the 1095 functionalities required into forwarding plane and control plane. 1097 In general, the forwarding plane carries out the bulk of the per- 1098 packet processing that is required at line speed, while the control 1099 plane carries most of the computationally complex operations that 1100 are typical of the control and signaling protocols. However, it is 1101 impossible to draw a rigid line to divide the processing into CEs 1102 and FEs cleanly. Nor should the ForCES architecture limit the 1103 innovative approaches in control and forwarding plane separation. 1104 As more and more processing power is available in the FEs, some of 1105 the control functions that traditionally are performed by CEs may 1106 now be moved to FEs for better performance and scalability. Such 1107 offloaded functions may include part of ICMP or TCP processing, or 1108 part of routing protocols. Once off-loaded onto the forwarding 1109 plane, such CE functions, even though logically belonging to the 1110 control plane, now become part of the FE functions. Just like the 1111 other logical functions performed by FEs, such off-loaded functions 1112 must be expressed as part of the FE model so that the CEs can 1113 decide how to best take advantage of these off-loaded functions 1114 when present on the FEs. 1116 5.1. General Router Requirements 1118 Routers have at least two or more logical interfaces. When CEs and 1119 FEs are separated by ForCES within a single NE, some additional 1120 interfaces are needed for intra-NE communications. Figure 12 shows 1121 an example to illustrate that. This NE contains one CE and two 1122 FEs. Each FE has four interfaces; two of them are used for 1123 receiving and transmitting packets to the external world, while the 1124 other two are for intra-NE connections. CE has two logical 1125 interfaces #9 and #10, connected to interfaces #3 and #6 from FE1 1126 and FE2, respectively. Interface #4 and #5 are connected for FE1- 1127 FE2 communication. Therefore, this router NE provides four 1128 external interfaces (#1, 2, 7 and 8). 1130 --------------------------------- 1131 | router NE | 1132 | ----------- ----------- | 1133 | | FE1 | | FE2 | | 1134 | ----------- ----------- | 1135 | 1| 2| 3| 4| 5| 6| 7| 8| | 1136 | | | | | | | | | | 1137 | | | | +----+ | | | | 1138 | | | | | | | | 1139 | | | 9| 10| | | | 1140 | | | -------------- | | | 1141 | | | | CE | | | | 1142 | | | -------------- | | | 1143 | | | | | | 1144 -----+--+----------------+--+---- 1145 | | | | 1146 | | | | 1148 Figure 12. A router NE example with four interfaces. 1150 IPv4 routers must implement IP to support its packet forwarding 1151 function, which is driven by its FIB (Forwarding Information Base). 1152 This Internet layer forwarding (see RFC1812 [1] Section 5) 1153 functionality naturally belongs to FEs in the ForCES architecture. 1155 A router may implement transport layer protocols (like TCP and UDP) 1156 that are required to support application layer protocols (see 1157 RFC1812 [1] Section 6). One important class of application 1158 protocols is routing protocols (see RFC1812 [1] Section 7). In 1159 ForCES architecture, routing protocols are naturally implemented by 1160 CEs. Routing protocols require routers communicate with each 1161 other. This communication between CEs in different routers is 1162 supported in ForCES by FEs' ability to redirect data packets 1163 addressed to routers (i.e., NEs) and CEs' ability to originate 1164 packets and have them delivered by their FEs. This communication 1165 occurs across Fp reference point inside each router and between 1166 neighboring routers' external interfaces, as illustrated in Figure 1167 11. 1169 5.2.Link Layer 1171 Since FEs own all the external interfaces for the router, FEs need 1172 to conform to the link layer requirements in RFC1812. Arguably, 1173 ARP support may be implemented in either CEs or FEs. As we will 1174 see later, a number of behaviors that RFC1812 mandates fall into 1175 this category -- they may be performed by the FE and may be 1176 performed by the CE. A general guideline is needed to ensure 1177 interoperability between separated control and forwarding planes. 1178 The guideline we offer here is that CEs MUST be capable of these 1179 kind of operations while FEs MAY choose to implement them. FE 1180 model should indicate its capabilities in this regard so that CEs 1181 can decide where these functions are implemented. 1183 Interface parameters, including MTU, IP address, etc., must be 1184 configurable by CEs via ForCES. CEs must be able to determine 1185 whether a physical interface in an FE is available to send packets 1186 or not. FEs must also inform CEs the status change of the 1187 interfaces (like link up/down) via ForCES. 1189 5.3.Internet Layer Protocols 1191 Both FEs and CEs must implement IP protocol and all mandatory 1192 extensions as RFC1812 specified. CEs should implement IP options 1193 like source route and record route while FEs may choose to 1194 implement those as well. The timestamp option should be 1195 implemented by FEs to insert the timestamp most accurately. The FE 1196 must interpret the IP options that it understands and preserve the 1197 rest unchanged for use by CEs. Both FEs and CEs might choose to 1198 silently discard packets without sending ICMP errors, but such 1199 events should be logged and counted. FEs may report statistics for 1200 such events to CEs via ForCES. 1202 When multiple FEs are involved to process packets, the appearance 1203 of single NE must be strictly maintained. For example, Time-To- 1204 Live (TTL) must be decremented only once within a single NE. For 1205 example, it can be always decremented by the last FE with egress 1206 function. 1208 FEs must receive and process normally any packets with a broadcast 1209 destination address or a multicast destination address that the 1210 router has asked to receive. When IP multicast is supported in 1211 routers, IGMP is implemented in CEs. CEs are also required of ICMP 1212 support, while it is optional for FEs to support ICMP. Such an 1213 option can be communicated to CEs as part of the FE model. 1214 Therefore, FEs can always rely upon CEs to send out ICMP error 1215 messages, but FEs also have the option to generate ICMP error 1216 messages themselves. 1218 5.4.Internet Layer Forwarding 1220 IP forwarding is implemented by FEs. When the routing table is 1221 updated at CEs, ForCES is used to send the new route entries from 1222 CEs to FEs. Each FE has its own forwarding table and uses this 1223 table to direct packets to the next hop interface. 1225 Upon receiving IP packets, the FE verifies the IP header and 1226 processes most of the IP options. Some options cannot be processed 1227 until the routing decision has been made. The routing decision is 1228 made after examining the destination IP address. If the 1229 destination address belongs to the router itself, the packets are 1230 filtered and either processed locally or forwarded to CE, depending 1231 upon the instructions set-up by CE. Otherwise, the FE determines 1232 the next hop IP address by looking up in its forwarding table. The 1233 FE also determines the network interface it uses to send the 1234 packets. Sometimes an FE may need to forward the packets to 1235 another FE before packets can be forwarded out to the next hop. 1236 Right before packets are forwarded out to the next hop, the FE 1237 decrements TTL by 1 and processes any IP options that cannot be 1238 processed before. The FE performs any IP fragmentation if 1239 necessary, determines link layer address (e.g., by ARP), and 1240 encapsulates the IP datagram (or each of the fragments thereof) in 1241 an appropriate link layer frame and queues it for output on the 1242 interface selected. 1244 Other options mentioned in RFC1812 for IP forwarding may also be 1245 implemented at FEs, for example, packet filtering. 1247 FEs typically forward packets destined locally to CEs. FEs may 1248 also forward exceptional packets (packets that FEs do not know how 1249 to handle) to CEs. CEs are required to handle packets forwarded by 1250 FEs for whatever different reasons. It might be necessary for 1251 ForCES to attach some meta-data with the packets to indicate the 1252 reasons of forwarding from FEs to CEs. Upon receiving packets with 1253 meta-data from FEs, CEs can decide to either process the packets 1254 themselves, or pass the packets to the upper layer protocols 1255 including routing and management protocols. If CEs are to process 1256 the packets by themselves, CEs may choose to discard the packets, 1257 or modify and re-send the packets. CEs may also originate new 1258 packets and deliver them to FEs for further forwarding. 1260 Any state change during router operation must also be handled 1261 correctly according to RFC1812. For example, when an FE ceases 1262 forwarding, the entire NE may continue forwarding packets, but it 1263 needs to stop advertising routes that are affected by the failed 1264 FE. 1266 5.5. Transport Layer 1268 Transport layer is typically implemented at CEs to support higher 1269 layer application protocols like routing protocols. In practice, 1270 this means that most CEs implement both the Transmission Control 1271 Protocol (TCP) and the User Datagram Protocol (UDP). 1273 Both CEs and FEs need to implement ForCES protocol. If some layer- 1274 4 transport is used to support ForCES, then both CEs and FEs need 1275 to implement the L4 transport and ForCES protocols. 1277 5.6. Application Layer -- Routing Protocols 1279 Interior and exterior routing protocols are implemented on CEs. 1280 The routing packets originated by CEs are forwarded to FEs for 1281 delivery. The results of such protocols (like forwarding table 1282 updates) are communicated to FEs via ForCES. 1284 For performance or scalability reasons, portions of the control 1285 plane functions that need faster response may be moved from the CEs 1286 and off-loaded onto the FEs. For example in OSPF, the Hello 1287 protocol packets are generated and processed periodically. When 1288 done at CEs, the inbound Hello packets have to traverse from the 1289 external interfaces at the FEs to the CEs via the internal CE-FE 1290 channel. Similarly, the outbound Hello packets have to go from the 1291 CEs to the FEs and to the external interfaces. Frequent Hello 1292 updates place heavy processing overhead on the CEs and can 1293 overwhelm the CE-FE channel as well. Since typically there are far 1294 more FEs than CEs in a router, the off-loaded Hello packets are 1295 processed in a much more distributed and scalable fashion. By 1296 expressing such off-loaded functions in the FE model, we can ensure 1297 interoperability. However, the exact description of the off-loaded 1298 functionality corresponding to the off-loaded functions expressed 1299 in the FE model are not part of the model itself and will need to 1300 be worked out as a separate specification. 1302 5.7. Application Layer -- Network Management Protocol 1304 RFC1812 also dictates "Routers MUST be manageable by SNMP." In 1305 general, for post-association phase, most external management tasks 1306 (including SNMP) should be done through interaction with the CE in 1307 order to support the appearance of a single functional device. 1308 Therefore, it is recommended that SNMP agent be implemented by CEs 1309 and the SNMP messages received by FEs be redirected to their CEs. 1310 AgentX framework defined in RFC2741 ([5]) may be applied here such 1311 that CEs act in the role of master agent to process SNMP protocol 1312 messages while FEs act in the role of subagent to provide access to 1313 the MIB objects residing on FEs. AgentX protocol messages between 1314 the master agent (CE) and the subagent (FE) are encapsulated and 1315 transported via ForCES, just like data packets from any other 1316 application layer protocols. 1318 6. Summary 1320 This document defines an architectural framework for ForCES. It 1321 identifies the relevant components for a ForCES network element, 1322 including (one or more) FEs, (one or more) CEs, one optional FE 1323 manager, and one optional CE manager. It also identifies the 1324 interaction among these components and discusses all the major 1325 reference points. It is important to point out that, among all the 1326 reference points, only the Fp interface between CEs and FEs is 1327 within the scope of ForCES. ForCES alone may not be enough to 1328 support all desirable NE configurations. However, we believe that 1329 ForCES over Fp interface is the most important element in realizing 1330 physical separation and interoperability of CEs and FEs, and hence 1331 the first interface that ought to be standardized. Simple and 1332 useful configurations can still be implemented with only CE-FE 1333 interface being standardized, e.g., single CE with full-meshed FEs. 1335 7. Acknowledgements 1337 Joel M. Halpern gave us many insightful comments and suggestions 1338 and pointed out several major issues. T. Sridhar suggested that 1339 the AgentX protocol could be used with SNMP to manage the ForCES 1340 network elements. Many of our colleagues and people in the ForCES 1341 mailing list also provided valuable feedback. 1343 8. Security Considerations 1344 In general, the physical separation of two entities usually results 1345 in a potentially insecure link between the two entities and hence 1346 much stricter security measurements are required. For example, we 1347 pointed out in Section 4.1 that authentication becomes necessary 1348 between CE manager and FE manager, between CE and CE manager, 1349 between FE and FE manager in some configurations. The physical 1350 separation of CE and FE also imposes serious security requirement 1351 for ForCES protocol over Fp interface. This section first attempts 1352 to describe the security threats that may be introduced by the 1353 physical separation of the FEs and the CEs, and then it provides 1354 recommendation and guidelines for secure operation and management 1355 of ForCES protocol over Fp interface based on existing standard 1356 security solutions. 1358 8.1. Analysis of Potential Threats Introduced by ForCES 1360 This section provides the threat analysis for ForCES, with a focus 1361 on Fp interface. Each threat is described in details with the 1362 effects on the ForCES protocol entities or/and the NE as a whole, 1363 and the required functionalities that need to be in place to defend 1364 the threat. 1366 8.1.1. "Join" or "Remove" Message Flooding on CEs 1368 Threats: A malicious node could send a stream of false "join NE" 1369 or "remove from NE" requests on behalf of non-existent or 1370 unauthorized FE to legitimate CEs at a very rapid rate and thereby 1371 create unnecessary state in the CEs. 1373 Effects: If by maintaining state for non-existent or unauthorized 1374 FEs, a CE may become unavailable for other processing and hence 1375 suffer from denial of service (DoS) attack similar to the TCP SYN 1376 DoS. If multiple CEs are used, the unnecessary state information 1377 may also be conveyed to multiple CEs via Fr interface (e.g., from 1378 the active CE to the stand-by CE) and hence subject multiple CEs to 1379 DoS attack. 1381 Requirement: A CE that receives a "join" or "remove" request 1382 should not create any state information until it has authenticated 1383 the FE endpoint. 1385 8.1.2. Impersonation Attack 1387 Threats: A malicious node can impersonate a CE or FE and send out 1388 false messages. 1390 Effects: The whole NE could be compromised. 1392 Requirement: The CE or FE must authenticate the message before 1393 accepting and processing it. 1395 8.1.3. Replay Attack 1397 Threat: A malicious node could replay the entire message previously 1398 sent by an FE or CE entity to get around authentication. 1400 Effect: The NE could be compromised. 1402 Requirement: Replay protection mechanism needs to be part of the 1403 security solution to defend against this attack. 1405 8.1.4. Attack during Fail Over 1407 Threat: A malicious node may exploit the CE fail-over mechanism to 1408 take over the control of NE. For example, suppose two CEs, say CE-A 1409 and CE-B, are controlling several FEs. CE-A is active and CE-B is 1410 stand-by. When CE-A fails, CE-B is taking over the active CE 1411 position. The FEs already had a trusted relationship with CE-A, 1412 but the FEs may not have the same trusted relationship established 1413 with CE-B prior to the fail-over. A malicious node can take over 1414 as CE-B if such trusted relationship is not established during the 1415 fail-over. 1417 Effect: The NE may be compromised after such insecure fail-over. 1419 Requirement: The level of trust relationship between the stand-by 1420 CE and the FEs must be as strong as the one between the active CE 1421 and the FEs. The security association between the FEs and the 1422 stand-by CE may be established prior to fail-over. If not already 1423 in place, such security association must be re-established before 1424 the stand-by CE takes over. 1426 8.1.5. Data Integrity 1428 Threats: A malicious node may inject false messages to legitimate 1429 CE or FE. 1431 Effect: An FE or CE receives the fabricated packet and performs 1432 incorrect or catastrophic operation. 1434 Requirement: Protocol messages require integrity protection. 1436 8.1.6. Data Confidentiality 1438 Threat: When FE and CE are physically separated, a malicious node 1439 may eavesdrop the messages in transit. Some of the messages are 1440 critical to the functioning of the whole network, while others may 1441 contain confidential business data. Leaking of such information 1442 may result in compromise even beyond the immediate CE or FE. 1444 Effect: Sensitive information might be exposed between CE and FE. 1446 Requirement: Data confidentiality between FE and CE must be 1447 available for sensitive information. 1449 8.1.7. Sharing security parameters 1451 Threat: Consider a scenario where several FEs communicating to the 1452 same CE share the same authentication keys for the Fp interface. 1453 If any FE or the CE is compromised, all other entities are 1454 compromised. 1456 Effect: The whole NE is compromised. 1458 Recommendation: To avoid this side effect, it's better to configure 1459 different security parameters for each FE-CE communication over Fp 1460 interface. 1462 8.1.8. Denial of Service Attack via External Interface 1464 Threat: When an FE receives a packet that is destined for its CE, 1465 the FE forwards the packet over the Fp interface. Malicious node 1466 can generate huge message storm like routing protocol packets etc. 1467 through the external Fi/f interface so that the FE has to process 1468 and forward all packets to CE through Fp interface. 1470 Effect: CE encounters resource exhaustion and bandwidth starvation 1471 on Fp interface due to an overwhelming number of packets from FEs. 1473 Requirement: Some sort of rate limiting mechanism MUST to be in 1474 place at both FE and CE. Examples of such mechanisms include 1475 explicit rate limiters or congestion control algorithms. Rate 1476 Limiter SHOULD be configured at FE for each message type that are 1477 being received through Fi/F interface. 1479 8.2. Security Recommendations for ForCES 1481 The requirements document [3] suggested that ForCES protocol should 1482 support reliability over Fp interface, but no particular transport 1483 protocol is yet specified for ForCES. This framework document does 1484 not intend to specify the particular transport either, and so we 1485 only provide recommendations and guidelines based on the existing 1486 standard security protocols that can work with the common transport 1487 candidates suitable for ForCES. 1489 We review two existing security protocol solutions, namely IPsec 1490 (IP Security) [14] or TLS (Transport Layer Security) [13]. TLS 1491 works with reliable transports such as TCP or SCTP for unicast, 1492 while IPsec can be used with any transport (UDP, TCP, SCTP) and 1493 supports both unicast and multicast. Both TLS and IPsec can be 1494 used potentially to satisfy all of the security requirements for 1495 ForCES protocol. Other approaches may be used as well but are not 1496 documented here, including using L2 security mechanisms for a given 1497 L2 interconnect technology, as long as the requirements can be 1498 satisfied. 1500 When ForCES is deployed between CEs and FEs inside a box, 1501 authentication, confidentiality and integrity may be provided by 1502 the physical security of the box and so the security mechanisms may 1503 be turned off, depending on the networking topology and its 1504 administration policy. However, it is important to realize that 1505 even if the NE is in a single-box, the DoS attacks as described in 1506 Section 8.1.8 can still be launched through Fi/f interfaces. 1507 Therefore, it is important to have the corresponding counter- 1508 measurement in place even for single-box deployment. 1510 8.2.1. Security Configuration 1512 The NE administrator has the freedom to determine the exact 1513 security configuration that is needed for the specific deployment. 1514 For example, ForCES may be deployed between CEs and FEs connected 1515 to each other inside a box over a backplane. In such scenario, 1516 physical security of the box ensures that most of the attacks such 1517 as man-in-the-middle, snooping, and impersonation are not possible, 1518 and hence ForCES architecture may rely on the physical security of 1519 the box to defend against these attacks and protocol mechanisms may 1520 be turned off. However, it is also shown that denial of service 1521 attack via external interface as described in Section 8.1.8 is 1522 still a potential threat even for such "all-in-one-box" deployment 1523 scenario and hence the rate limiting mechanism is still necessary. 1524 This is just one example to show that it is important to assess the 1525 security needs of the ForCES-enabled network elements under 1526 different deployment scenarios. It should be possible for the 1527 administrator to configure the level of security needed for the 1528 ForCES protocol. 1530 8.2.2. Using TLS with ForCES 1532 TLS [13] can be used if a reliable unicast transport such as TCP or 1533 SCTP is used for ForCES over the Fp interface. The TLS handshake 1534 protocol is used during association establishment or re- 1535 establishment phase to negotiate a TLS session between the CE and 1536 FE. Once the session is in place, the TLS record protocol is used 1537 to secure ForCES communication messages between the CE and FE. 1539 A basic outline of how TLS can be used with ForCES is described 1540 below. Steps 1) till 7) complete the security handshake as 1541 illustrated in Figure 9 while step 8) is for all the further 1542 communication between the CE and FE, including the rest of messages 1543 after the security handshake shown in Figure 9 and the steady-state 1544 communication shown in Figure 10. 1546 1) During Pre-association phase all FEs are configured with 1547 the CEs (including both the active CE and the standby CE). 1548 2) The FE establishes a TLS connection with the CE (master) 1549 and negotiates a cipher suite. 1550 3) The FE (slave) gets the CE certificate, validates the 1551 signature, checks the expiration date, checks if the 1552 certificate has been revoked. 1553 4) The CE (master) gets the FE certificate and performs the 1554 same validation as the FE in step 3). 1555 5) If any of the check fails in step 3) or step 4), endpoint 1556 must generate an error message and abort. 1557 6) After successful mutual authentication, a TLS session is 1558 established between CE and FE. 1559 7) The FE sends a "join NE" message to the CE. 1560 8) The FE and CE use TLS session for further communication. 1562 Note that there are different ways for the CE and FE to validate a 1563 received certificate. One way is to configure the FE Manager or CE 1564 Manager or other central component as CA, so that the CE or FE can 1565 query this pre-configured CA to validate that the certificate has 1566 not been revoked. Another way is to have the CE and the FE 1567 configured directly a list of valid certificates in the pre- 1568 association phase. 1570 In the case of fail-over, it is the responsibility of the active CE 1571 and the standby CE to synchronize ForCES states including the TLS 1572 states to minimize the state reestablishment during fail-over. 1573 Care must be taken to ensure that the standby CE is also 1574 authenticated in the same way as the active CE, either before or 1575 during the fail-over. 1577 8.2.3. Using IPsec with ForCES 1579 IPsec [14] can be used with any transport protocol, such as UDP, 1580 SCTP and TCP over Fp interface for ForCES. When using IPsec, we 1581 recommend using ESP in transport mode for ForCES because message 1582 confidentiality is required for ForCES. 1584 IPsec can be used with both manual and automated SA and 1585 cryptographic key management. But Ipsec's replay protection 1586 mechanisms are not available if manual key management is used. 1587 Hence, automatic key management is recommended if replay protection 1588 is deemed important. Otherwise, manual key management might be 1589 sufficient for some deployment scenarios, esp. when the number of 1590 CEs and FEs is relatively small. It is recommended that the keys 1591 be changed periodically even for manual key management. 1593 IPsec can support both unicast and multicast transport. At the time 1594 this document was published, MSEC working group is actively working 1595 on standardizing protocols to provide multicast security [17]. 1596 Multicast-based solutions relying on IPsec should specify how to 1597 meet the security requirements in [3]. 1599 Unlike TLS, IPsec provides security services between the CE and FE 1600 at IP level, and so the security handshake as illustrated in Figure 1601 9 amounts to a "no-op" when manual key management is used. The 1602 following outline the steps taken for ForCES in such a case. 1604 1) During Pre-association phase all FEs are configured with 1605 the CEs (including active CE and standby CE) and SA parameters 1606 manually. 1607 2) The FE sends a "join NE" message to the CE. This message 1608 and all others that follow are afforded security service 1609 according to the manually configured IPsec SA parameters, but 1610 replay protection is not available. 1612 It is up to the administrator to decide whether to share the same 1613 key across multiple FE-CE communication, but it is recommended that 1614 different keys be used. Similarly, it is recommended that 1615 different keys be used for inbound and outbound traffic. 1617 If automatic key management is needed, IKE [15] can be used for 1618 that purpose. Other automatic key distribution techniques such as 1619 Kerberos may be used as well. The key exchange process 1620 constitutes the security handshake as illustrated in Figure 9. The 1621 following shows the steps involved in using IKE with IPsec for 1622 ForCES. Steps 1) to 6) constitute the security handshake in Figure 1623 9. 1625 1) During Pre-association phase all FEs are configured with 1626 the CEs (including active CE and standby CE), IPsec policy 1627 etc. 1628 2) The FE kicks off IKE process and tries to establish an 1629 IPsec SA with the CE (master). The FE (Slave) gets the CE 1630 certificate as part of the IKE negotiation. The FE validates 1631 signature, checks the expiration date, checks if the 1632 certificate has been revoked. 1633 3) The CE (master) gets the FE certificate and performs the 1634 same check as the FE in step 2). 1635 4) If any of the check fails in step 2) or step 3), the 1636 endpoint must generate an error message and abort. 1637 5) After successful mutual authentication, IPsec session is 1638 established between the CE and FE. 1639 6) The FE sends a "join NE" message to CE. No SADB entry is 1640 created in FE yet. 1641 7) The FE and CE use the IPsec session for further 1642 communication. 1644 FE Manager or CE Manager or other central component can be used as 1645 CA for validating CE and FE certificates during the IKE process. 1646 Alternatively, during the pre-association phase, the CE and FE can 1647 be configured directly with the required information such as 1648 certificates or passwords etc depending upon the type of 1649 authentication that administrator wants to configure. 1651 In the case of fail-over, it is the responsibility of active CE and 1652 standby CE to synchronize ForCES states and IPsec states to 1653 minimize the state reestablishment during fail-over. 1654 Alternatively, the FE needs to establish different IPsec SA during 1655 the startup operation itself with each CE. This will minimize the 1656 periodic state transfer across IPsec layer though Fr (CE-CE) 1657 Interface. 1659 9. Normative References 1661 [1] Baker, F., "Requirements for IP Version 4 Routers", RFC 1812, 1662 June 1995. 1664 [2] Floyd, S., "Congestion Control Principles", RFC 2914, September 1665 2000. 1667 [3] Khosravi, H. et al., "Requirements for Separation of IP Control 1668 and Forwarding", work in progress, May 2003, . 1671 10. Informative References 1673 [4] Case, J., et al., "Introduction and Applicability Statements 1674 for Internet Standard Management Framework", RFC 3410, December 1675 2002. 1677 [5] Daniele, M. et al., "Agent Extensibility (AgentX) Protocol 1678 Version 1", RFC 2741, January 2000. 1680 [6] Chan, K. et al., "COPS Usage for Policy Provisioning (COPS- 1681 PR)", RFC 3084, March 2001. 1683 [7] Crouch, A. et al., "ForCES Applicability Statement", work in 1684 progress, June 2002, . 1686 [8] Anderson, T. and J. Buerkle, "Requirements for the Dynamic 1687 Partitioning of Switching Elements", RFC 3532, May 2003. 1689 [9] Leelanivas, M. et al., "Graceful Restart Mechanism for Label 1690 Distribution Protocol", RFC 3478, February 2003. 1692 [10] Moy, J. et al., "Graceful OSPF Restart", work in progress, 1693 March 2003, . 1695 [11] Sangli, S. et al., "Graceful Restart Mechanism for BGP", work 1696 in progress, January 2003, < draft-ietf-idr-restart-06.txt>. 1698 [12] Shand, M. and L. Ginsberg, "Restart Signaling for IS-IS", work 1699 in progress, March 2003, . 1701 [13] Dierks, T. and C. Allen, "The TLS Protocol, version 1.0", RFC 1702 2246, January 1999. 1704 [14] Kent, S. and R. Atkinson, "Security Architecture for the 1705 Internet Protocol", RFC 2401, November 1998. 1707 [15] Harkins, D. and D. Carrel, "The Internet Key Exchange (IKE) ", 1708 RFC 2409, November 1998. 1710 [16] Bellovin, S., "Guidelines for Mandating the Use of Ipsec", 1711 work in progress, October 2002, . 1713 [17] Hardjono, T. and Weis, B. "The Multicast Security 1714 Architecture", work in progress, August 2003, . 1717 11. Authors' Addresses 1718 L. Lily Yang 1719 Intel Corp., MS JF3-206, 1720 2111 NE 25th Avenue 1721 Hillsboro, OR 97124, USA 1722 Phone: +1 503 264 8813 1723 Email: lily.l.yang@intel.com 1725 Ram Dantu 1726 Department of Computer Science, 1727 University of North Texas, 1728 Denton, TX 76203, USA 1729 Phone: +1 940 565 2822 1730 Email: rdantu@unt.edu 1732 Todd A. Anderson 1733 Intel Corp. 1734 2111 NE 25th Avenue 1735 Hillsboro, OR 97124, USA 1736 Phone: +1 503 712 1760 1737 Email: todd.a.anderson@intel.com 1739 Ram Gopal 1740 Nokia Research Center 1741 5, Wayside Road, 1742 Burlington, MA 01803, USA 1743 Phone: +1 781 993 3685 1744 Email: ram.gopal@nokia.com 1746 12. Intellectual Property Right 1747 The IETF takes no position regarding the validity or scope of any 1748 intellectual property or other rights that might be claimed to 1749 pertain to the implementation or use of the technology described in 1750 this document or the extent to which any license under such rights 1751 might or might not be available; neither does it represent that it 1752 has made any effort to identify any such rights. Information on 1753 the IETF's procedures with respect to rights in standards-track and 1754 standards-related documentation can be found in RFC 2026. Copies 1755 of claims of rights made available for publication and any 1756 assurances of licenses to be made available, or the result of an 1757 attempt made to obtain a general license or permission for the use 1758 of such proprietary rights by implementors or users of this 1759 specification can be obtained from the IETF Secretariat. 1761 The IETF invites any interested party to bring to its attention any 1762 copyrights, patents or patent applications, or other proprietary 1763 rights which may cover technology that may be required to practice 1764 this standard. Please address the information to the IETF 1765 Executive Director. 1767 13. Full Copyright Statement 1769 Copyright (C) The Internet Society (2003). All Rights Reserved. 1771 This document and translations of it may be copied and furnished to 1772 others, and derivative works that comment on or otherwise explain 1773 it or assist in its implementation may be prepared, copied, 1774 published and distributed, in whole or in part, without restriction 1775 of any kind, provided that the above copyright notice and this 1776 paragraph are included on all such copies and derivative works. 1777 However, this document itself may not be modified in any way, such 1778 as by removing the copyright notice or references to the Internet 1779 Society or other Internet organizations, except as needed for the 1780 purpose of developing Internet standards in which case the 1781 procedures for copyrights defined in the Internet Standards process 1782 must be followed, or as required to translate it into languages 1783 other than English. 1785 The limited permissions granted above are perpetual and will not be 1786 revoked by the Internet Society or its successors or assigns. 1787 This document and the information contained herein is provided on 1788 an "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET 1789 ENGINEERING TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR 1790 IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF 1791 THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED 1792 WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE."