idnits 2.17.1 draft-ietf-rtgwg-ipfrr-framework-06.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** It looks like you're using RFC 3978 boilerplate. You should update this to the boilerplate described in the IETF Trust License Policy document (see https://trustee.ietf.org/license-info), which is required now. -- Found old boilerplate from RFC 3978, Section 5.1 on line 16. -- Found old boilerplate from RFC 3978, Section 5.5, updated by RFC 4748 on line 667. -- Found old boilerplate from RFC 3979, Section 5, paragraph 1 on line 563. -- Found old boilerplate from RFC 3979, Section 5, paragraph 2 on line 570. -- Found old boilerplate from RFC 3979, Section 5, paragraph 3 on line 576. ** This document has an original RFC 3978 Section 5.4 Copyright Line, instead of the newer IETF Trust Copyright according to RFC 4748. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- == No 'Intended status' indicated for this document; assuming Proposed Standard Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. 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 (Oct 2006) is 6395 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) -- Missing reference section? 'MPLSFRR' on line 178 looks like a reference -- Missing reference section? 'BFD' on line 279 looks like a reference -- Missing reference section? 'BASE' on line 306 looks like a reference -- Missing reference section? 'U-TURNS' on line 318 looks like a reference -- Missing reference section? 'FIFR' on line 318 looks like a reference -- Missing reference section? 'SIMULA' on line 319 looks like a reference -- Missing reference section? 'TUNNELS' on line 323 looks like a reference -- Missing reference section? 'ALT-SP' on line 323 looks like a reference -- Missing reference section? 'NOT-VIA' on line 328 looks like a reference -- Missing reference section? 'MICROLOOP' on line 482 looks like a reference Summary: 2 errors (**), 0 flaws (~~), 2 warnings (==), 17 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 1 Network Working Group M. Shand 2 Internet Draft S. Bryant 3 Expiration Date: May 2007 Cisco Systems 5 Oct 2006 7 IP Fast Reroute Framework 9 draft-ietf-rtgwg-ipfrr-framework-06.txt 11 Status of this Memo 13 By submitting this Internet-Draft, each author represents that any 14 applicable patent or other IPR claims of which he or she is aware 15 have been or will be disclosed, and any of which he or she becomes 16 aware will be disclosed, in accordance with Section 6 of BCP 79. 18 Internet-Drafts are working documents of the Internet Engineering 19 Task Force (IETF), its areas, and its working groups. Note that other 20 groups may also distribute working documents as Internet-Drafts. 22 Internet-Drafts are draft documents valid for a maximum of six months 23 and may be updated, replaced, or obsoleted by other documents at any 24 time. It is inappropriate to use Internet-Drafts as reference 25 material or to cite them other than as "work in progress". 27 The list of current Internet-Drafts can be accessed at 28 http://www.ietf.org/1id-abstracts.txt 30 The list of Internet-Draft Shadow Directories can be accessed at 31 http://www.ietf.org/shadow.html. 33 Abstract 35 This document provides a framework for the development of IP 36 fast-reroute mechanisms which provide protection against link or 37 router failure by invoking locally determined repair paths. Unlike 38 MPLS Fast-reroute, the mechanisms are applicable to a network 39 employing conventional IP routing and forwarding. An essential part 40 of such mechanisms is the prevention of packet loss caused by the 41 loops which normally occur during the re-convergence of the network 42 following a failure. 44 Terminology 46 This section defines words and acronyms used in this draft and other 47 drafts discussing IP Fast-reroute. 49 D Used to denote the destination router under 50 discussion. 52 Distance_opt(A,B) The distance of the shortest path from A 53 to B. 55 Downstream Path This is a subset of the loop-free alternates 56 where the neighbor N meet the following 57 condition:- 59 Distance_opt(N, D) < Distance_opt(S,D) 61 E Used to denote the router which is the 62 primary next-hop neighbor to get from S to 63 the destination D. Where there is an ECMP set 64 for the shortest path from S to D, these are 65 referred to as E_1, E_2, etc. 67 ECMP Equal cost multi-path: Where, for a 68 particular destination D, multiple primary 69 next-hops are used to forward traffic because 70 there exist multiple shortest paths from S 71 via different output layer-3 interfaces. 73 FIB Forwarding Information Base. The database 74 used by the packet forwarder to determine 75 what actions to perform on a packet. 77 IPFRR IP fast-reroute 79 Link(A->B) A link connecting router A to router B. 81 Loop-Free This is a neighbor N, that is not a primary 82 Alternate next-hop neighbor E, whose shortest path to 83 the destination D does not go back through 84 the router S. The neighbor N must meet the 85 following condition:- 87 Distance_opt(N, D) < Distance_opt(N, S) + 88 Distance_opt(S, D) 90 Loop-Free A neighbor N_i, which is not the particular 91 Neighbor primary neighbor E_k under discussion, and 92 whose shortest path to D does not traverse S. 93 For example, if there are two primary 94 neighbors E_1 and E_2, E_1 is a loop-free 95 neighbor with regard to E_2 and vice versa. 97 Loop-Free This is a path via a Loop-Free Neighbor N_i 98 Link-Protecting which does not go through the particular link 99 Alternate of S which is being protected to reach the 100 destination D. 102 Loop-Free This is a path via a Loop-Free Neighbor N_i 103 Node-Protecting which does not go through the particular 104 Alternate primary neighbor of S which is being 105 protected to reach the destination D. 107 Micro-loop A temporary forwarding loop which exists 108 during a routing transition as a result of 109 temporary inconsistencies between FIBs. 111 N_i The ith neighbor of S. 113 Primary Neighbor A neighbor N_i of S which is one of the next 114 hops for destination D in S's FIB prior to 115 any failure. 117 R_i_j The jth neighbor of N_i. 119 Routing The process whereby routers converge on a new 120 transition topology. In conventional networks this 121 process frequently causes some disruption to 122 packet delivery. 124 RPF Reverse Path Forwarding. I.e. checking that a 125 packet is received over the interface which 126 would be used to send packets addressed to 127 the source address of the packet. 129 S Used to denote a router that is the source of 130 a repair that is computed in anticipation of 131 the failure of a neighboring router denoted 132 as E, or of the link between S and E. It is 133 the viewpoint from which IP Fast-Reroute is 134 described. 136 S_i The set of neighbors of E, in addition to S, 137 which will independently take the role of S 138 for the traffic they carry. 140 SPF Shortest Path First, e.g. Dijkstra's 141 algorithm. 143 SPT Shortest path tree 145 Upstream This is a forwarding loop which involves a 146 Forwarding Loop set of routers, none of which are directly 147 connected to the link which has caused the 148 topology change that triggered a new SPF in 149 any of the routers. 151 1. Introduction 153 When a link or node failure occurs in a routed network, there is 154 inevitably a period of disruption to the delivery of traffic until 155 the network re-converges on the new topology. Packets for 156 destinations which were previously reached by traversing the failed 157 component may be dropped or may suffer looping. Traditionally such 158 disruptions have lasted for periods of at least several seconds, and 159 most applications have been constructed to tolerate such a quality of 160 service. 162 Recent advances in routers have reduced this interval to under a 163 second for carefully configured networks using link state IGPs. 164 However, new Internet services are emerging which may be sensitive to 165 periods of traffic loss which are orders of magnitude shorter than 166 this. 168 Addressing these issues is difficult because the distributed nature 169 of the network imposes an intrinsic limit on the minimum convergence 170 time which can be achieved. 172 However, there is an alternative approach, which is to compute backup 173 routes that allow the failure to be repaired locally by the router(s) 174 detecting the failure without the immediate need to inform other 175 routers of the failure. In this case, the disruption time can be 176 limited to the small time taken to detect the adjacent failure and 177 invoke the backup routes. This is analogous to the technique employed 178 by MPLS Fast-Reroute [MPLSFRR], but the mechanisms employed for the 179 backup routes in pure IP networks are necessarily very different. 181 This document provides a framework for the development of this 182 approach. 184 2. Problem Analysis 186 The duration of the packet delivery disruption caused by a 187 conventional routing transition is determined by a number of factors: 189 1. The time taken to detect the failure. This may be of the order 190 of a few mS when it can be detected at the physical layer, up to 191 several tens of seconds when a routing protocol hello is 192 employed. During this period packets will be unavoidably lost. 194 2. The time taken for the local router to react to the failure. 195 This will typically involve generating and flooding new routing 196 updates, perhaps after some hold-down delay, and re-computing 197 the router's FIB. 199 3. The time taken to pass the information about the failure to 200 other routers in the network. In the absence of routing protocol 201 packet loss, this is typically between 10mS and 100mS per hop. 203 4. The time taken to re-compute the forwarding tables. This is 204 typically a few mS for a link state protocol using Dijkstra's 205 algorithm. 207 5. The time taken to load the revised forwarding tables into the 208 forwarding hardware. This time is very implementation dependant 209 and also depends on the number of prefixes affected by the 210 failure, but may be several hundred mS. 212 The disruption will last until the routers adjacent to the failure 213 have completed steps 1 and 2, and then all the routers in the network 214 whose paths are affected by the failure have completed the remaining 215 steps. 217 The initial packet loss is caused by the router(s) adjacent to the 218 failure continuing to attempt to transmit packets across the failure 219 until it is detected. This loss is unavoidable, but the detection 220 time can be reduced to a few tens of mS as described in section 3.1. 222 Subsequent packet loss is caused by the "micro-loops" which form 223 because of temporary inconsistencies between routers' forwarding 224 tables. These occur as a result of the different times at which 225 routers update their forwarding tables to reflect the failure. These 226 variable delays are caused by steps 3, 4 and 5 above and in many 227 routers it is step 5 which is both the largest factor and which has 228 the greatest variance between routers. The large variance arises from 229 implementation differences and from the differing impact that a 230 failure has on each individual router. For example, the number of 231 prefixes affected by the failure may vary dramatically from one 232 router to another. 234 In order to achieve packet disruption times which are commensurate 235 with the failure detection times it is necessary to perform two 236 distinct tasks: 238 1. Provide a mechanism for the router(s) adjacent to the failure to 239 rapidly invoke a repair path, which is unaffected by any 240 subsequent re-convergence. 242 2. Provide a mechanism to prevent the effects of micro-loops during 243 subsequent re-convergence. 245 Performing the first task without the second will result in the 246 repair path being starved of traffic and hence being redundant. 247 Performing the second without the first will result in traffic being 248 discarded by the router(s) adjacent to the failure. Both tasks are 249 necessary for an effective solution to the problem. 251 However, repair paths can be used in isolation where the failure is 252 short-lived. The repair paths can be kept in place until the failure 253 is repaired and there is no need to advertise the failure to other 254 routers. 256 Similarly, micro-loop avoidance can be used in isolation to prevent 257 loops arising from pre-planned management action, because the link or 258 node being shut down can remain in service for a short time after its 259 removal has been announced into the network, and hence it can 260 function as its own "repair path". 262 Note that micro-loops can also occur when a link or node is restored 263 to service and thus a micro-loop avoidance mechanism is required for 264 both link up and link down cases. 266 3. Mechanisms for IP Fast-reroute 268 The set of mechanisms required for an effective solution to the 269 problem can be broken down into the following sub-problems. 271 3.1. Mechanisms for fast failure detection 273 It is critical that the failure detection time is minimized. A number 274 of approaches are possible, such as: 276 1. Physical detection; for example, loss of light. 278 2. Routing protocol independent protocol detection; for example, 279 The Bidirectional Failure Detection protocol [BFD]. 281 3. Routing protocol detection; for example, use of "fast hellos". 283 3.2. Mechanisms for repair paths 285 Once a failure has been detected by one of the above mechanisms, 286 traffic which previously traversed the failure is transmitted over 287 one or more repair paths. The design of the repair paths should be 288 such that they can be pre-calculated in anticipation of each local 289 failure and made available for invocation with minimal delay. There 290 are three basic categories of repair paths: 292 1. Equal cost multi-paths (ECMP). Where such paths exist, and one 293 or more of the alternate paths do not traverse the failure, they 294 may trivially be used as repair paths. 296 2. Loop free alternate paths. Such a path exists when a direct 297 neighbor of the router adjacent to the failure has a path to the 298 destination which can be guaranteed not to traverse the failure. 300 3. Multi-hop repair paths. When there is no feasible loop free 301 alternate path it may still be possible to locate a router, 302 which is more than one hop away from the router adjacent to the 303 failure, from which traffic will be forwarded to the destination 304 without traversing the failure. 306 ECMP and loop free alternate paths (as described in [BASE]) offer the 307 simplest repair paths and would normally be used when they are 308 available. It is anticipated that around 80% of failures (see section 309 3.2.2) can be repaired using these basic methods alone. 311 Multi-hop repair paths are more complex, both in the computations 312 required to determine their existence, and in the mechanisms required 313 to invoke them. They can be further classified as: 315 1. Mechanisms where one or more alternate FIBs are pre-computed in 316 all routers and the repaired packet is instructed to be 317 forwarded using a "repair FIB" by some method of per packet 318 signaling such as detecting a "U-turn" [U-TURNS, FIFR] or by 319 marking the packet [SIMULA]. 321 2. Mechanisms functionally equivalent to a loose source route which 322 is invoked using the normal FIB. These include tunnels 323 [TUNNELS], alternative shortest paths [ALT-SP] and label based 324 mechanisms. 326 3. Mechanisms employing special addresses or labels which are 327 installed in the FIBs of all routers with routes pre-computed to 328 avoid certain components of the network. For example [NOT-VIA]. 330 In many cases a repair path which reaches two hops away from the 331 router detecting the failure will suffice, and it is anticipated that 332 around 98% of failures (see section 3.2.2) can be repaired by this 333 method. However, to provide complete repair coverage some use of 334 longer multi-hop repair paths is generally necessary. 336 3.2.1. Scope of repair paths 338 A particular repair path may be valid for all destinations which 339 require repair or may only be valid for a subset of destinations. If 340 a repair path is valid for a node immediately downstream of the 341 failure, then it will be valid for all destinations previously 342 reachable by traversing the failure. However, in cases where such a 343 repair path is difficult to achieve because it requires a high order 344 multi-hop repair path, it may still be possible to identify lower 345 order repair paths (possibly even loop free alternate paths) which 346 allow the majority of destinations to be repaired. When IPFRR is 347 unable to provide complete repair, it is desirable that the extent of 348 the repair coverage can be determined and reported via network 349 management. 351 There is a tradeoff to be achieved between minimizing the number of 352 repair paths to be computed, and minimizing the overheads incurred in 353 using higher order multi-hop repair paths for destinations for which 354 they are not strictly necessary. However, the computational cost of 355 determining repair paths on an individual destination basis can be 356 very high. 358 It will frequently be the case that the majority of destinations may 359 be repaired using only the "basic" repair mechanism, leaving a 360 smaller subset of the destinations to be repaired using one of the 361 more complex multi-hop methods. Such a hybrid approach may go some 362 way to resolving the conflict between completeness and complexity. 364 The use of repair paths may result in excessive traffic passing over 365 a link, resulting in congestion discard. This reduces the 366 effectiveness of IPFRR. Mechanisms to influence the distribution of 367 repaired traffic to minimize this effect are therefore desirable. 369 3.2.2. Analysis of repair coverage 371 In some cases the repair strategy will permit the repair of all 372 single link or node failures in the network for all possible 373 destinations. This can be defined as 100% coverage. However, where 374 the coverage is less than 100% it is important for the purposes of 375 comparisons between different proposed repair strategies to define 376 what is meant by such a percentage. There are four possibilities: 378 1. The percentage of links (or nodes) which can be fully protected 379 for all destinations. This is appropriate where the requirement 380 is to protect all traffic, but some percentage of the possible 381 failures may be identified as being un-protectable. 383 2. The percentage of destinations which can be fully protected for 384 all link (or node) failures. This is appropriate where the 385 requirement is to protect against all possible failures, but 386 some percentage of destinations may be identified as being 387 un-protectable. 389 3. For all destinations (d) and for all failures (f), the 390 percentage of the total potential failure cases (d*f) which are 391 protected. This is appropriate where the requirement is an 392 overall "best effort" protection. 394 4. The percentage of packets normally passing though the network 395 that will continue to reach their destination. This requires a 396 traffic matrix for the network as part of the analysis. 398 The coverage obtained is dependent on the repair strategy and highly 399 dependent on the detailed topology and metrics. Any figures quoted in 400 this document are for illustrative purposes only. 402 3.2.3. Link or node repair 404 A repair path may be computed to protect against failure of an 405 adjacent link, or failure of an adjacent node. In general, link 406 protection is simpler to achieve. A repair which protects against 407 node failure will also protect against link failure for all 408 destinations except those for which the adjacent node is a single 409 point of failure. 411 In some cases it may be necessary to distinguish between a link or 412 node failure in order that the optimal repair strategy is invoked. 413 Methods for link/node failure determination may be based on 414 techniques such as BFD. This determination may be made prior to 415 invoking any repairs, but this will increase the period of packet 416 loss following a failure unless the determination can be performed as 417 part of the failure detection mechanism itself. Alternatively, a 418 subsequent determination can be used to optimise an already invoked 419 default strategy. 421 3.2.4. Maintenance of Repair paths 423 In order to meet the response time goals, it is expected (though not 424 required) that repair paths, and their associated FIB entries, will 425 be pre-computed and installed ready for invocation when a failure is 426 detected. Following invocation the repair paths remain in effect 427 until they are no longer required. This will normally be when the 428 routing protocol has re-converged on the new topology taking into 429 account the failure, and traffic will no longer be using the repair 430 paths. 432 The repair paths have the property that they are unaffected by any 433 topology changes resulting from the failure which caused their 434 instantiation. Therefore there is no need to re-compute them during 435 the convergence period. They may be affected by an unrelated 436 simultaneous topology change, but such events are out of scope of 437 this work (see section 3.2.5). 439 Once the routing protocol has re-converged it is necessary for all 440 repair paths to take account of the new topology. Various 441 optimizations may permit the efficient identification of repair paths 442 which are unaffected by the change, and hence do not require full 443 re-computation. Since the new repair paths will not be required until 444 the next failure occurs, the re-computation may be performed as a 445 background task and be subject to a hold-down, but excessive delay in 446 completing this operation will increase the risk of a new failure 447 occurring before the repair paths are in place. 449 3.2.5. Multiple failures and Shared Risk Link Groups 451 Complete protection against multiple unrelated failures is out of 452 scope of this work. However, it is important that the occurrence of a 453 second failure while one failure is undergoing repair should not 454 result in a level of service which is significantly worse than that 455 which would have been achieved in the absence of any repair strategy. 457 Shared Risk Link Groups are an example of multiple related failures, 458 and the more complex aspects of their protection is a matter for 459 further study. 461 One specific example of an SRLG which is clearly within the scope of 462 this work is a node failure. This causes the simultaneous failure of 463 multiple links, but their closely defined topological relationship 464 makes the problem more tractable. 466 3.3. Local Area Networks 468 Protection against partial or complete failure of LANs is more 469 complex than the point to point case. In general there is a tradeoff 470 between the simplicity of the repair and the ability to provide 471 complete and optimal repair coverage. 473 3.4. Mechanisms for micro-loop prevention 475 Control of micro-loops is important not only because they can cause 476 packet loss in traffic which is affected by the failure, but because 477 by saturating a link with looping packets they can also cause 478 congestion loss of traffic flowing over that link which would 479 otherwise be unaffected by the failure. 481 A number of solutions to the problem of micro-loop formation have 482 been proposed and are summarized in [MICROLOOP]. The following 483 factors are significant in their classification: 485 1. Partial or complete protection against micro-loops. 487 2. Delay imposed upon convergence. 489 3. Tolerance of multiple failures (from node failures, and in 490 general). 492 4. Computational complexity (pre-computed or real time). 494 5. Applicability to scheduled events. 496 6. Applicability to link/node reinstatement. 498 4. Management Considerations 500 While many of the management requirements will be specific to 501 particular IPFRR solutions, the following general aspects need to be 502 addressed: 504 1. Configuration 506 a. Enabling/disabling IPFRR support. 508 b. Enabling/disabling protection on a per link/node basis. 510 c. Expressing preferences regarding the links/nodes used for 511 repair paths. 513 d. Configuration of failure detection mechanisms. 515 e. Configuration of loop avoidance strategies. 517 2. Monitoring 519 a. Notification of links/nodes/destinations which cannot be 520 protected. 522 b. Notification of pre-computed repair paths, and anticipated 523 traffic patterns. 525 c. Counts of failure detections, protection invocations and 526 packets forwarded over repair paths. 528 5. Scope and applicability 530 The initial scope of this work is in the context of link state IGPs. 531 Link state protocols provide ubiquitous topology information, which 532 facilitates the computation of repairs paths. 534 Provision of similar facilities in non-link state IGPs and BGP is a 535 matter for further study, but the correct operation of the repair 536 mechanisms for traffic with a destination outside the IGP domain is 537 an important consideration for solutions based on this framework 539 6. IANA considerations 541 There are no IANA considerations that arise from this framework 542 document. 544 7. Security Considerations 546 This framework document does not itself introduce any security 547 issues, but attention must be paid to the security implications of 548 any proposed solutions to the problem. 550 8. IPR Disclosure Acknowledgement 552 Certain IPR may be applicable to the mechanisms outlined in this 553 document. Please check the detailed specifications for possible IPR 554 notices. 556 The IETF takes no position regarding the validity or scope of any 557 Intellectual Property Rights or other rights that might be claimed to 558 pertain to the implementation or use of the technology described in 559 this document or the extent to which any license under such rights 560 might or might not be available; nor does it represent that it has 561 made any independent effort to identify any such rights. Information 562 on the procedures with respect to rights in RFC documents can be 563 found in BCP 78 and BCP 79. 565 Copies of IPR disclosures made to the IETF Secretariat and any 566 assurances of licenses to be made available, or the result of an 567 attempt made to obtain a general license or permission for the use of 568 such proprietary rights by implementers or users of this 569 specification can be obtained from the IETF on-line IPR repository at 570 http://www.ietf.org/ipr. 572 The IETF invites any interested party to bring to its attention any 573 copyrights, patents or patent applications, or other proprietary 574 rights that may cover technology that may be required to implement 575 this standard. Please address the information to the IETF at 576 ietf-ipr@ietf.org. 578 9. Acknowledgements 580 The authors would like to acknowledge contributions made by Alia 581 Atlas and Alex Zinin. 583 10. Normative References 585 Internet-drafts are works in progress available from 586 http://www.ietf.org/internet-drafts/ 588 11. Informative References 590 Internet-drafts are works in progress available from 591 http://www.ietf.org/internet-drafts/ 593 ALT-SP Tian, A., Chen, N., "Fast Reroute using 594 Alternative Shortest Paths", draft-tian-frr- 595 alt-shortest-path-01.txt, (work in progress) 597 BASE Atlas, A., Zinin, A., "Basic Specification 598 for IP Fast-Reroute: Loop-free Alternates", 599 draft-ietf-rtgwg-ipfrr-spec-base-05.txt, 600 (work in progress) 602 BFD Katz, D. and Ward, D., "Bidirectional 603 Forwarding Detection", 604 draft-ietf-bfd-base-04.txt, (work in 605 progress). 607 FIFR S. Nelakuditi, S. Lee, Y. Yu, Z.-L. Zhang, 608 and C.-N. Chuah, "Fast local rerouting for 609 handling transient link failures.," Tech. 610 Rep. TR-2004-004, University of South 611 Carolina, 2004. 613 MPLSFRR Pan, P. et al, "Fast Reroute Extensions to 614 RSVP-TE for LSP Tunnels", RFC 4090. 616 MICROLOOP Bryant, S. and Shand, M., "A Framework for 617 Loop-free Convergence", 618 draft-bryant-shand-lf-conv-frmwk-03.txt, 619 (work in progress). 621 NOT-VIA Bryant, S., Previdi, S., Shand, M., "IP Fast 622 Reroute Using Notvia Addresses", 623 draft-bryant-shand-ipfrr-notvia-addresses- 624 03.txt, (work in progress). 626 SIMULA Lysne, O., et al, "Fast IP Network Recovery 627 using Multiple Routing Configurations", 628 http://folk.uio.no/amundk/infocom06.pdf 630 TUNNELS Bryant, S. et al, "IP Fast Reroute using 631 tunnels", draft-bryant-ipfrr-tunnels-02.txt, 632 (work in progress). 634 U-TURNS Atlas, A. et al, "IP/LDP Local Protection", 635 draft-atlas-ip-local-protect-03.txt, (work in 636 progress). 638 12. Authors' Addresses 640 Stewart Bryant 641 Cisco Systems, 642 250, Longwater Avenue, 643 Green Park, 644 Reading, RG2 6GB, 645 United Kingdom. Email: stbryant@cisco.com 647 Mike Shand 648 Cisco Systems, 649 250, Longwater Avenue, 650 Green Park, 651 Reading, RG2 6GB, 652 United Kingdom. Email: mshand@cisco.com 654 Full copyright statement 656 Copyright (C) The Internet Society (2006). This document is subject 657 to the rights, licenses and restrictions contained in BCP 78, and 658 except as set forth therein, the authors retain all their rights. 660 This document and the information contained herein are provided on an 661 "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS 662 OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF 663 TRUST AND THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL 664 WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY 665 WARRANTY THAT THE USE OF THE INFORMATION HEREIN WILL NOT INFRINGE ANY 666 RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A 667 PARTICULAR PURPOSE.