idnits 2.17.1 draft-ietf-rtgwg-spf-uloop-pb-statement-08.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year == The document doesn't use any RFC 2119 keywords, yet seems to have RFC 2119 boilerplate text. -- The document date (November 30, 2018) is 1945 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- No issues found here. Summary: 0 errors (**), 0 flaws (~~), 2 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Routing Area Working Group S. Litkowski 3 Internet-Draft Orange Business Service 4 Intended status: Informational B. Decraene 5 Expires: June 3, 2019 Orange 6 M. Horneffer 7 Deutsche Telekom 8 November 30, 2018 10 Link State protocols SPF trigger and delay algorithm impact on IGP 11 micro-loops 12 draft-ietf-rtgwg-spf-uloop-pb-statement-08 14 Abstract 16 A micro-loop is a packet forwarding loop that may occur transiently 17 among two or more routers in a hop-by-hop packet forwarding paradigm. 19 In this document, we are trying to analyze the impact of using 20 different Link State IGP (Interior Gateway Protocol) implementations 21 in a single network, with respect to micro-loops. The analysis is 22 focused on the SPF (Shortest Path First) delay algorithm. 24 Requirements Language 26 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 27 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 28 "OPTIONAL" in this document are to be interpreted as described in BCP 29 14 [RFC2119] [RFC8174] when, and only when, they appear in all 30 capitals, as shown here. 32 Status of This Memo 34 This Internet-Draft is submitted in full conformance with the 35 provisions of BCP 78 and BCP 79. 37 Internet-Drafts are working documents of the Internet Engineering 38 Task Force (IETF). Note that other groups may also distribute 39 working documents as Internet-Drafts. The list of current Internet- 40 Drafts is at https://datatracker.ietf.org/drafts/current/. 42 Internet-Drafts are draft documents valid for a maximum of six months 43 and may be updated, replaced, or obsoleted by other documents at any 44 time. It is inappropriate to use Internet-Drafts as reference 45 material or to cite them other than as "work in progress." 47 This Internet-Draft will expire on June 3, 2019. 49 Copyright Notice 51 Copyright (c) 2018 IETF Trust and the persons identified as the 52 document authors. All rights reserved. 54 This document is subject to BCP 78 and the IETF Trust's Legal 55 Provisions Relating to IETF Documents 56 (https://trustee.ietf.org/license-info) in effect on the date of 57 publication of this document. Please review these documents 58 carefully, as they describe your rights and restrictions with respect 59 to this document. Code Components extracted from this document must 60 include Simplified BSD License text as described in Section 4.e of 61 the Trust Legal Provisions and are provided without warranty as 62 described in the Simplified BSD License. 64 Table of Contents 66 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 67 2. Problem statement . . . . . . . . . . . . . . . . . . . . . . 4 68 3. SPF trigger strategies . . . . . . . . . . . . . . . . . . . 5 69 4. SPF delay strategies . . . . . . . . . . . . . . . . . . . . 6 70 4.1. Two steps SPF delay . . . . . . . . . . . . . . . . . . . 6 71 4.2. Exponential backoff . . . . . . . . . . . . . . . . . . . 7 72 5. Mixing strategies . . . . . . . . . . . . . . . . . . . . . . 8 73 6. Benefits of standardized SPF delay behavior . . . . . . . . . 11 74 7. Security Considerations . . . . . . . . . . . . . . . . . . . 13 75 8. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 13 76 9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 13 77 10. References . . . . . . . . . . . . . . . . . . . . . . . . . 13 78 10.1. Normative References . . . . . . . . . . . . . . . . . . 13 79 10.2. Informative References . . . . . . . . . . . . . . . . . 14 80 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 14 82 1. Introduction 84 Link State IGP protocols are based on a topology database on which 85 the SPF algorithm is run to find a consistent set of non-looping 86 routing paths. 88 Specifications like IS-IS ([RFC1195]) propose some optimizations of 89 the route computation (See Appendix C.1 of [RFC1195]) but not all the 90 implementations follow those non-mandatory optimizations. 92 We will call "SPF triggers", the events that would lead to a new SPF 93 computation based on the topology. 95 Link State IGP protocols, like OSPF ([RFC2328]) and IS-IS 96 ([RFC1195]), are using multiple timers to control the router behavior 97 in case of churn: SPF delay, PRC (Partial Route Computation) delay, 98 LSP (Link State Packet) generation delay, LSP flooding delay, LSP 99 retransmission interval... 101 Some of those timers (values and behavior) are standardized in 102 protocol specifications, while some are not. The SPF computation 103 related timers have generally remained unspecified. 105 For non standardized timers, implementations are free to implement it 106 in any way. For some standardized timers, we can also see that 107 rather than using static configurable values for such timer, 108 implementations may offer dynamically adjusted timers to help 109 controlling the churn. 111 We will call "SPF delay", the timer that exists in most 112 implementations that specifies the required delay before running SPF 113 computation after a SPF trigger is received. 115 A micro-loop is a packet forwarding loop that may occur transiently 116 among two or more routers in a hop-by-hop packet forwarding paradigm. 117 We can observe that these micro-loops are formed when two routers do 118 not update their Forwarding Information Base (FIB) for a certain 119 prefix at the same time. The micro-loop phenomenon is described in 120 [I-D.ietf-rtgwg-microloop-analysis]. 122 Two micro-loop mitigation techniques have been defined by IETF. 123 [RFC6976] has not been widely implemented, presumably due to the 124 complexity of the technique. [RFC8333] has been implemented. 125 However, it does not prevent all micro-loops that can occur for a 126 given topology and failure scenario. 128 In multi-vendor networks, using different implementations of a link 129 state protocol may favor micro-loops creation during the convergence 130 process due to discrepancies of timers. Service Providers are 131 already aware to use similar timers (values and behavior) for all the 132 network as a best practice, but sometimes it is not possible due to 133 limitations of implementations. 135 This document will present why it sounds important for service 136 providers to have consistent implementations of Link State protocols 137 across vendors. We are particularly analyzing the impact of using 138 different Link State IGP implementations in a single network in 139 regards of micro-loops. The analysis is focused on the SPF delay 140 algorithm. 142 [RFC8405] defines a solution that satisfies this problem statement 143 and this document captures the reasoning of the provided solution. 145 2. Problem statement 147 S ---- E 148 | | 149 10 | | 10 150 | | 151 D ---- A 152 | 2 153 Px 155 Figure 1 - Network topology suffering from micro-loops 157 In Figure 1, S uses primarily the SD link to reach the prefixes 158 behind D (Px). When the SD link fails, the IGP convergence occurs. 159 If S converges before E, S will forward the traffic to Px through E, 160 but as E has not converged yet, E will loop back traffic to S, 161 leading to a micro-loop. 163 The micro-loop appears due to the asynchronous convergence of nodes 164 in a network when an event occurs. 166 Multiple factors (or a combination of these factors) may increase the 167 probability for a micro-loop to appear: 169 o the delay of failure notification: the more E is advised of the 170 failure later than S, the more a micro-loop may have a chance to 171 appear. 173 o the SPF delay: most implementations support a delay for the SPF 174 computation to try to catch as many events as possible. If S uses 175 an SPF delay timer of x msec and E uses an SPF delay timer of y 176 msec and x < y, E would start converging after S leading to a 177 potential micro-loop. 179 o the SPF computation time: mostly a matter of CPU power and 180 optimizations like incremental SPF. If S computes its SPF faster 181 than E, there is a chance for a micro-loop to appear. CPUs are 182 today fast enough to consider SPF computation time as negligible 183 (on the order of milliseconds in a large network). 185 o the SPF computation order: an SPF trigger can be common to 186 multiple IGP areas or levels (e.g., IS-IS Level1/Level2) or for 187 multiple address families with multi-topologies. There is no 188 specified order for SPF computation today and it is implementation 189 dependent. In such scenarios, if the order of SPF computation 190 done in S and E for each area/level/topology/SPF-algorithm is 191 different, there is a possibility for a micro-loop to appear. 193 o the RIB and FIB prefix insertion speed or ordering. This is 194 highly dependent on the implementation. 196 Even if all of these factors may increase the probability for a 197 micro-loop to appear, the SPF delay, especially in case of churn, 198 plays a significant role. As the number of IGP events increase, the 199 delta between SPF delay values used by routers becomes significant 200 and the major part (especially when one router increases its timer 201 exponentially while another one increases it in a more smoother way). 202 Another important factor is the time to update the FIB. As of today, 203 total FIB update time is the major factor for IGP convergence. 204 However, for micro-loops, what's matter is not the total time, but 205 the difference to install the same prefix between nodes. That part 206 may be the main part for the first iteration but is not for 207 subsequent IGP events. In addition, this part is very implementation 208 specific and difficult/impossible to standardize, while the SPF delay 209 algorithm may be standardized. 211 As a consequence, this document will focus on the analysis of the SPF 212 delay behavior and associated triggers. 214 3. SPF trigger strategies 216 Depending on the change advertised in LSPDU (Link State Protocol Data 217 Unit) or LSA (Link State Advertisement), the topology may be affected 218 or not. An implementation may avoid running the SPF computation (and 219 may only run IP reachability computation instead) if the advertised 220 change does not affect the topology. 222 Different strategies exists to trigger the SPF computation: 224 1. An implementation may always run a full SPF for any type of 225 change. 227 2. An implementation may run a full SPF only when required. For 228 example, if a link fails, a local node will run an SPF for its 229 local LSP update. If the LSP from the neighbor (describing the 230 same failure) is received after SPF has started, the local node 231 can decide that a new full SPF is not required as the topology 232 has not changed. 234 3. If the topology does not change, an implementation may only 235 recompute the IP reachability. 237 As noted in Section 1, SPF optimizations are not mandatory in 238 specifications. This has led to the implementation of different 239 strategies. 241 4. SPF delay strategies 243 Implementations of link state routing protocols use different 244 strategies to delay the SPF computation. The two most common SPF 245 delay behaviors are the following: 247 1. Two phase SPF delay. 249 2. Exponential backoff delay. 251 These behaviors will be explained in the next sections. 253 4.1. Two steps SPF delay 255 The SPF delay is managed by four parameters: 257 o Rapid delay: amount of time to wait before running SPF, after the 258 initial SPF trigger event. 260 o Rapid runs: the number of consecutive SPF runs that can use the 261 rapid delay. When the number is exceeded, the delay moves to the 262 slow delay value. 264 o Slow delay: amount of time to wait before running SPF. 266 o Wait time: amount of time to wait without receiving SPF trigger 267 events before going back to the rapid delay. 269 Example: Rapid delay = 50msec, Rapid runs = 3, Slow delay = 1sec, 270 Wait time = 2sec 272 SPF delay time 273 ^ 274 | 275 | 276 SD- | x xx x 277 | 278 | 279 | 280 RD- | x x x x 281 | 282 +---------------------------------> Events 283 | | | | || | | 284 < wait time > 286 Figure 2 - Two phase delay algorithm 288 4.2. Exponential backoff 290 The algorithm has two modes: the fast mode and the backoff mode. In 291 the fast mode, the SPF delay is usually delayed by a very small 292 amount of time (fast reaction). When an SPF computation has run in 293 the fast mode, the algorithm automatically moves to the backoff mode 294 (a single SPF run is authorized in the fast mode). In the backoff 295 mode, the SPF delay is increasing exponentially at each run. When 296 the network becomes stable, the algorithm moves back to the fast 297 mode. The SPF delay is managed by four parameters: 299 o First delay: amount of time to wait before running SPF. This 300 delay is used only when SPF is in fast mode. 302 o Incremental delay: amount of time to wait before running SPF. 303 This delay is used only when SPF is in backoff mode and increments 304 exponentially at each SPF run. 306 o Maximum delay: maximum amount of time to wait before running SPF. 308 o Wait time: amount of time to wait without events before going back 309 to the fast mode. 311 Example: First delay = 50msec, Incremental delay = 50msec, Maximum 312 delay = 1sec, Wait time = 2sec 314 SPF delay time 315 ^ 316 MD- | xx x 317 | 318 | 319 | 320 | 321 | 322 | x 323 | 324 | 325 | 326 | x 327 | 328 FD- | x x x 329 ID | 330 +---------------------------------> Events 331 | | | | || | | 332 < wait time > 333 FM->BM -------------------->FM 335 Figure 3 - Exponential delay algorithm 337 5. Mixing strategies 339 In Figure 1, we consider a flow of packet from S to D. We consider 340 that S is using optimized SPF triggering (Full SPF is triggered only 341 when necessary), and two steps SPF delay (rapid=150ms,rapid-runs=3, 342 slow=1s). As implementation of S is optimized, Partial Reachability 343 Computation (PRC) is available. We consider the same timers as SPF 344 for delaying PRC. We consider that E is using a SPF trigger strategy 345 that always compute a Full SPF for any change, and uses the 346 exponential backoff strategy for SPF delay (start=150ms, inc=150ms, 347 max=1s) 349 We also consider the following sequence of events: 351 o t0=0 ms: a prefix is declared down in the network. We consider 352 this event to happen at time=0. 354 o 200ms: the prefix is declared as up. 356 o 400ms: a prefix is declared down in the network. 358 o 1000ms: S-D link fails. 360 +--------+--------------------+------------------+------------------+ 361 | Time | Network Event | Router S events | Router E events | 362 +--------+--------------------+------------------+------------------+ 363 | t0=0 | Prefix DOWN | | | 364 | 10ms | | Schedule PRC (in | Schedule SPF (in | 365 | | | 150ms) | 150ms) | 366 | | | | | 367 | | | | | 368 | 160ms | | PRC starts | SPF starts | 369 | 161ms | | PRC ends | | 370 | 162ms | | RIB/FIB starts | | 371 | 163ms | | | SPF ends | 372 | 164ms | | | RIB/FIB starts | 373 | 175ms | | RIB/FIB ends | | 374 | 178ms | | | RIB/FIB ends | 375 | | | | | 376 | 200ms | Prefix UP | | | 377 | 212ms | | Schedule PRC (in | | 378 | | | 150ms) | | 379 | 214ms | | | Schedule SPF (in | 380 | | | | 150ms) | 381 | | | | | 382 | | | | | 383 | 370ms | | PRC starts | | 384 | 372ms | | PRC ends | | 385 | 373ms | | | SPF starts | 386 | 373ms | | RIB/FIB starts | | 387 | 375ms | | | SPF ends | 388 | 376ms | | | RIB/FIB starts | 389 | 383ms | | RIB/FIB ends | | 390 | 385ms | | | RIB/FIB ends | 391 | | | | | 392 | 400ms | Prefix DOWN | | | 393 | 410ms | | Schedule PRC (in | Schedule SPF (in | 394 | | | 300ms) | 300ms) | 395 | | | | | 396 | | | | | 397 | | | | | 398 | | | | | 399 | 710ms | | PRC starts | SPF starts | 400 | 711ms | | PRC ends | | 401 | 712ms | | RIB/FIB starts | | 402 | 713ms | | | SPF ends | 403 | 714ms | | | RIB/FIB starts | 404 | 716ms | | RIB/FIB ends | RIB/FIB ends | 405 | | | | | 406 | 1000ms | S-D link DOWN | | | 407 | 1010ms | | Schedule SPF (in | Schedule SPF (in | 408 | | | 150ms) | 600ms) | 409 | | | | | 410 | | | | | 411 | 1160ms | | SPF starts | | 412 | 1161ms | | SPF ends | | 413 | 1162ms | Micro-loop may | RIB/FIB starts | | 414 | | start from here | | | 415 | 1175ms | | RIB/FIB ends | | 416 | | | | | 417 | | | | | 418 | | | | | 419 | | | | | 420 | 1612ms | | | SPF starts | 421 | 1615ms | | | SPF ends | 422 | 1616ms | | | RIB/FIB starts | 423 | 1626ms | Micro-loop ends | | RIB/FIB ends | 424 +--------+--------------------+------------------+------------------+ 426 Table 1 - Route computation when S and E use the different behaviors 427 and multiple events appear 429 In the Table 1, we can see that due to discrepancies in the SPF 430 management, after multiple events of a different type, the values of 431 the SPF delay are completely misaligned between node S and node E, 432 leading to the creation of micro-loops. 434 The same issue can also appear with only a single type of event as 435 shown below: 437 +--------+--------------------+------------------+------------------+ 438 | Time | Network Event | Router S events | Router E events | 439 +--------+--------------------+------------------+------------------+ 440 | t0=0 | Link DOWN | | | 441 | 10ms | | Schedule SPF (in | Schedule SPF (in | 442 | | | 150ms) | 150ms) | 443 | | | | | 444 | | | | | 445 | 160ms | | SPF starts | SPF starts | 446 | 161ms | | SPF ends | | 447 | 162ms | | RIB/FIB starts | | 448 | 163ms | | | SPF ends | 449 | 164ms | | | RIB/FIB starts | 450 | 175ms | | RIB/FIB ends | | 451 | 178ms | | | RIB/FIB ends | 452 | | | | | 453 | 200ms | Link DOWN | | | 454 | 212ms | | Schedule SPF (in | | 455 | | | 150ms) | | 456 | 214ms | | | Schedule SPF (in | 457 | | | | 150ms) | 458 | | | | | 459 | | | | | 460 | 370ms | | SPF starts | | 461 | 372ms | | SPF ends | | 462 | 373ms | | | SPF starts | 463 | 373ms | | RIB/FIB starts | | 464 | 375ms | | | SPF ends | 465 | 376ms | | | RIB/FIB starts | 466 | 383ms | | RIB/FIB ends | | 467 | 385ms | | | RIB/FIB ends | 468 | | | | | 469 | 400ms | Link DOWN | | | 470 | 410ms | | Schedule SPF (in | Schedule SPF (in | 471 | | | 150ms) | 300ms) | 472 | | | | | 473 | | | | | 474 | 560ms | | SPF starts | | 475 | 561ms | | SPF ends | | 476 | 562ms | Micro-loop may | RIB/FIB starts | | 477 | | start from here | | | 478 | 568ms | | RIB/FIB ends | | 479 | | | | | 480 | | | | | 481 | 710ms | | | SPF starts | 482 | 713ms | | | SPF ends | 483 | 714ms | | | RIB/FIB starts | 484 | 716ms | Micro-loop ends | | RIB/FIB ends | 485 | | | | | 486 | 1000ms | Link DOWN | | | 487 | 1010ms | | Schedule SPF (in | Schedule SPF (in | 488 | | | 1s) | 600ms) | 489 | | | | | 490 | | | | | 491 | | | | | 492 | | | | | 493 | 1612ms | | | SPF starts | 494 | 1615ms | | | SPF ends | 495 | 1616ms | Micro-loop may | | RIB/FIB starts | 496 | | start from here | | | 497 | 1626ms | | | RIB/FIB ends | 498 | | | | | 499 | | | | | 500 | | | | | 501 | | | | | 502 | 2012ms | | SPF starts | | 503 | 2014ms | | SPF ends | | 504 | 2015ms | | RIB/FIB starts | | 505 | 2025ms | Micro-loop ends | RIB/FIB ends | | 506 | | | | | 507 | | | | | 508 +--------+--------------------+------------------+------------------+ 510 Table 2 - Route computation upon multiple link down events when S and 511 E use the different behaviors 513 6. Benefits of standardized SPF delay behavior 515 Using the same event sequence as in Table 1, we may expect fewer and/ 516 or shorter micro-loops using a standardized SPF delay. 518 +--------+--------------------+------------------+------------------+ 519 | Time | Network Event | Router S events | Router E events | 520 +--------+--------------------+------------------+------------------+ 521 | t0=0 | Prefix DOWN | | | 522 | 10ms | | Schedule PRC (in | Schedule PRC (in | 523 | | | 150ms) | 150ms) | 524 | | | | | 525 | | | | | 526 | 160ms | | PRC starts | PRC starts | 527 | 161ms | | PRC ends | | 528 | 162ms | | RIB/FIB starts | PRC ends | 529 | 163ms | | | RIB/FIB starts | 530 | 175ms | | RIB/FIB ends | | 531 | 176ms | | | RIB/FIB ends | 532 | | | | | 533 | 200ms | Prefix UP | | | 534 | 212ms | | Schedule PRC (in | | 535 | | | 150ms) | | 536 | 213ms | | | Schedule PRC (in | 537 | | | | 150ms) | 538 | | | | | 539 | | | | | 540 | 370ms | | PRC starts | PRC starts | 541 | 372ms | | PRC ends | | 542 | 373ms | | RIB/FIB starts | PRC ends | 543 | 374ms | | | RIB/FIB starts | 544 | 383ms | | RIB/FIB ends | | 545 | 384ms | | | RIB/FIB ends | 546 | | | | | 547 | 400ms | Prefix DOWN | | | 548 | 410ms | | Schedule PRC (in | Schedule PRC (in | 549 | | | 300ms) | 300ms) | 550 | | | | | 551 | | | | | 552 | | | | | 553 | | | | | 554 | 710ms | | PRC starts | PRC starts | 555 | 711ms | | PRC ends | PRC ends | 556 | 712ms | | RIB/FIB starts | | 557 | 713ms | | | RIB/FIB starts | 558 | 716ms | | RIB/FIB ends | RIB/FIB ends | 559 | | | | | 560 | 1000ms | S-D link DOWN | | | 561 | 1010ms | | Schedule SPF (in | Schedule SPF (in | 562 | | | 150ms) | 150ms) | 563 | | | | | 564 | | | | | 565 | 1160ms | | SPF starts | | 566 | 1161ms | | SPF ends | SPF starts | 567 | 1162ms | Micro-loop may | RIB/FIB starts | SPF ends | 568 | | start from here | | | 569 | 1163ms | | | RIB/FIB starts | 570 | 1175ms | | RIB/FIB ends | | 571 | 1177ms | Micro-loop ends | | RIB/FIB ends | 572 +--------+--------------------+------------------+------------------+ 574 Table 3 - Route computation when S and E use the same standardized 575 behavior 577 As displayed above, there could be some other parameters like router 578 computation power, flooding timers that may also influence micro- 579 loops. In all the examples in this document comparing the SPF timer 580 behavior of router S and router E, we have made router E a bit slower 581 than router S. This can lead to micro-loops even when both S and E 582 use a common standardized SPF behavior. However, we expect that by 583 aligning implementations of the SPF delay, service providers may 584 reduce the number and the duration of micro-loops. 586 7. Security Considerations 588 This document does not introduce any security consideration. 590 8. Acknowledgements 592 Authors would like to thank Mike Shand and Chris Bowers for their 593 useful comments. 595 9. IANA Considerations 597 This document has no action for IANA. 599 10. References 601 10.1. Normative References 603 [RFC1195] Callon, R., "Use of OSI IS-IS for routing in TCP/IP and 604 dual environments", RFC 1195, DOI 10.17487/RFC1195, 605 December 1990, . 607 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 608 Requirement Levels", BCP 14, RFC 2119, 609 DOI 10.17487/RFC2119, March 1997, 610 . 612 [RFC2328] Moy, J., "OSPF Version 2", STD 54, RFC 2328, 613 DOI 10.17487/RFC2328, April 1998, 614 . 616 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 617 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 618 May 2017, . 620 [RFC8405] Decraene, B., Litkowski, S., Gredler, H., Lindem, A., 621 Francois, P., and C. Bowers, "Shortest Path First (SPF) 622 Back-Off Delay Algorithm for Link-State IGPs", RFC 8405, 623 DOI 10.17487/RFC8405, June 2018, 624 . 626 10.2. Informative References 628 [I-D.ietf-rtgwg-microloop-analysis] 629 Zinin, A., "Analysis and Minimization of Microloops in 630 Link-state Routing Protocols", draft-ietf-rtgwg-microloop- 631 analysis-01 (work in progress), October 2005. 633 [RFC6976] Shand, M., Bryant, S., Previdi, S., Filsfils, C., 634 Francois, P., and O. Bonaventure, "Framework for Loop-Free 635 Convergence Using the Ordered Forwarding Information Base 636 (oFIB) Approach", RFC 6976, DOI 10.17487/RFC6976, July 637 2013, . 639 [RFC8333] Litkowski, S., Decraene, B., Filsfils, C., and P. 640 Francois, "Micro-loop Prevention by Introducing a Local 641 Convergence Delay", RFC 8333, DOI 10.17487/RFC8333, March 642 2018, . 644 Authors' Addresses 646 Stephane Litkowski 647 Orange Business Service 649 Email: stephane.litkowski@orange.com 651 Bruno Decraene 652 Orange 654 Email: bruno.decraene@orange.com 656 Martin Horneffer 657 Deutsche Telekom 659 Email: martin.horneffer@telekom.de