idnits 2.17.1 draft-ietf-tcpm-tcp-soft-errors-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 635. -- Found old boilerplate from RFC 3979, Section 5, paragraph 1 on line 646. -- Found old boilerplate from RFC 3979, Section 5, paragraph 2 on line 653. -- Found old boilerplate from RFC 3979, Section 5, paragraph 3 on line 659. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- ** The document seems to lack an IANA Considerations section. (See Section 2.2 of https://www.ietf.org/id-info/checklist for how to handle the case when there are no actions for IANA.) Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust 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 (June 24, 2007) is 6150 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Unused Reference: 'Guynes' is defined on line 492, but no explicit reference was found in the text ** Obsolete normative reference: RFC 793 (Obsoleted by RFC 9293) ** Obsolete normative reference: RFC 2461 (Obsoleted by RFC 4861) == Outdated reference: A later version (-08) exists of draft-ietf-behave-tcp-07 == Outdated reference: A later version (-12) exists of draft-ietf-tcpm-icmp-attacks-02 -- Obsolete informational reference (is this intentional?): RFC 816 (Obsoleted by RFC 7805) Summary: 4 errors (**), 0 flaws (~~), 4 warnings (==), 8 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 TCP Maintenance and Minor F. Gont 3 Extensions (tcpm) UTN/FRH 4 Internet-Draft June 24, 2007 5 Intended status: Informational 6 Expires: December 26, 2007 8 TCP's Reaction to Soft Errors 9 draft-ietf-tcpm-tcp-soft-errors-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 20 other groups may also distribute working documents as Internet- 21 Drafts. 23 Internet-Drafts are draft documents valid for a maximum of six months 24 and may be updated, replaced, or obsoleted by other documents at any 25 time. It is inappropriate to use Internet-Drafts as reference 26 material or to cite them other than as "work in progress." 28 The list of current Internet-Drafts can be accessed at 29 http://www.ietf.org/ietf/1id-abstracts.txt. 31 The list of Internet-Draft Shadow Directories can be accessed at 32 http://www.ietf.org/shadow.html. 34 This Internet-Draft will expire on December 26, 2007. 36 Copyright Notice 38 Copyright (C) The IETF Trust (2007). 40 Abstract 42 This document describes a non-standard, but widely implemented, 43 modification to TCP's handling of ICMP soft error messages, that 44 rejects pending connection-requests when those error messages are 45 received. This behavior reduces the likelihood of long delays 46 between connection establishment attempts that may arise in a number 47 of scenarios, including one in which dual stack nodes that have IPv6 48 enabled by default are deployed in IPv4 or mixed IPv4 and IPv6 49 environments. 51 Table of Contents 53 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 54 2. Error Handling in TCP . . . . . . . . . . . . . . . . . . . . 3 55 2.1. Reaction to ICMP error messages that indicate hard 56 errors . . . . . . . . . . . . . . . . . . . . . . . . . . 4 57 2.2. Reaction to ICMP error messages that indicate soft 58 errors . . . . . . . . . . . . . . . . . . . . . . . . . . 5 59 3. Problems that may arise from TCP's reaction to soft errors . . 5 60 3.1. General Discussion . . . . . . . . . . . . . . . . . . . . 5 61 3.2. Problems that may arise with Dual Stack IPv6 on by 62 Default . . . . . . . . . . . . . . . . . . . . . . . . . 6 63 4. Deployed workarounds for long delays between 64 connection-establishment attempts . . . . . . . . . . . . . . 7 65 4.1. Context-sensitive ICMP/TCP interaction . . . . . . . . . . 7 66 4.2. Context-sensitive ICMP/TCP interaction with repeated 67 confirmation . . . . . . . . . . . . . . . . . . . . . . . 7 68 5. Possible drawbacks of changing ICMP semantics . . . . . . . . 8 69 5.1. Non-deterministic transient network failures . . . . . . . 9 70 5.2. Deterministic transient network failures . . . . . . . . . 9 71 5.3. Non-compliant Network Address Translators (NATs) . . . . . 9 72 6. Security Considerations . . . . . . . . . . . . . . . . . . . 9 73 7. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 10 74 8. Contributors . . . . . . . . . . . . . . . . . . . . . . . . . 10 75 9. References . . . . . . . . . . . . . . . . . . . . . . . . . . 10 76 9.1. Normative References . . . . . . . . . . . . . . . . . . . 10 77 9.2. Informative References . . . . . . . . . . . . . . . . . . 11 78 Appendix A. Change log (to be removed before publication of 79 the document as an RFC) . . . . . . . . . . . . . . . 12 80 A.1. Changes from draft-ietf-tcpm-tcp-soft-errors-05 . . . . . 12 81 A.2. Changes from draft-ietf-tcpm-tcp-soft-errors-04 . . . . . 12 82 A.3. Changes from draft-ietf-tcpm-tcp-soft-errors-03 . . . . . 12 83 A.4. Changes from draft-ietf-tcpm-tcp-soft-errors-02 . . . . . 13 84 A.5. Changes from draft-ietf-tcpm-tcp-soft-errors-01 . . . . . 13 85 A.6. Changes from draft-ietf-tcpm-tcp-soft-errors-00 . . . . . 13 86 A.7. Changes from draft-gont-tcpm-tcp-soft-errors-02 . . . . . 13 87 A.8. Changes from draft-gont-tcpm-tcp-soft-errors-01 . . . . . 13 88 A.9. Changes from draft-gont-tcpm-tcp-soft-errors-00 . . . . . 13 89 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . . 14 90 Intellectual Property and Copyright Statements . . . . . . . . . . 15 92 1. Introduction 94 The handling of network failures can be separated into two different 95 actions: fault isolation and fault recovery. Fault isolation 96 consists of the actions that hosts and routers take to determine that 97 there is a network failure. Fault recovery, on the other hand, 98 consists of the actions that hosts and routers perform in an attempt 99 to survive a network failure [RFC0816]. 101 In the Internet architecture, the Internet Control Message Protocol 102 (ICMP) [RFC0792] is one fault isolation technique to report network 103 error conditions to the hosts sending datagrams over the network. 105 When a host is notified of a network error, its network stack will 106 attempt to continue communications, if possible, in the presence of 107 the network failure. The fault recovery strategy may depend on the 108 type of network failure taking place, and the time the error 109 condition is detected. 111 This document analyzes the fault recovery strategy of TCP [RFC0793], 112 and the problems that may arise due to TCP's reaction to ICMP soft 113 errors. It analyzes the problems that may arise when a host tries to 114 establish a TCP connection with a multihomed host for which some of 115 its addresses are unreachable. Additionally, it analyzes the 116 problems that may arise in the specific scenario where dual stack 117 nodes that have IPv6 enabled by default are deployed in IPv4 or mixed 118 IPv4 and IPv6 environments. 120 Finally, we document a modification to TCP's reaction to ICMP 121 messages indicating soft errors during connection startup, that has 122 been implemented in a variety of TCP/IP stacks to help overcome the 123 problems outlined below. We stress that this modification runs 124 contrary to the standard behavior and this document unambiguously 125 does not change the standard reaction. 127 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 128 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 129 document are to be interpreted as described in RFC 2119 [RFC2119]. 131 2. Error Handling in TCP 133 Network errors can be divided into soft and hard errors. Soft errors 134 are considered to be transient network failures, which are likely to 135 be solved in the near term. Hard errors, on the other hand, are 136 considered to reflect network error conditions that are unlikely to 137 be solved in the near future. 139 The Host Requirements RFC [RFC1122] states, in Section 4.2.3.9., that 140 the ICMP messages that indicate soft errors are ICMP "Destination 141 Unreachable" codes 0 (network unreachable), 1 (host unreachable), and 142 5 (source route failed), ICMP "Time Exceeded" codes 0 (time to live 143 exceeded in transit) and 1 (fragment reassembly time exceeded), and 144 ICMP "Parameter Problem". Even though ICMPv6 did not exist when 145 [RFC1122] was written, one could extrapolate the concept of soft 146 errors to ICMPv6 "Destination Unreachable" codes 0 (no route to 147 destination) and 3 (address unreachable), ICMPv6 "Time Exceeded" 148 codes 0 (Hop limit exceeded in transit) and 1 (Fragment reassembly 149 time exceeded), and ICMPv6 "Parameter Problem" codes 0 (Erroneous 150 header field encountered), 1 (Unrecognized Next Header type 151 encountered) and 2 (Unrecognized IPv6 option encountered) [RFC4443]. 153 +----------------------------------+--------------------------------+ 154 | ICMP | ICMPv6 | 155 +----------------------------------+--------------------------------+ 156 | Destination Unreachable (codes | Destination Unreachable (codes | 157 | 0, 1, and 5) | 0 and 3) | 158 +----------------------------------+--------------------------------+ 159 | Time Exceeded (codes 0 and 1) | Time exceeded (codes 0 and 1) | 160 +----------------------------------+--------------------------------+ 161 | Parameter Problem | Parameter Problem (codes 0, 1, | 162 | | and 2) | 163 +----------------------------------+--------------------------------+ 165 Table 1: Extrapolating the concept of soft errors to ICMPv6 167 When there is a network failure that is not signaled to the sending 168 host, such as a gateway corrupting packets, TCP's fault recovery 169 action is to repeatedly retransmit the segment until either it gets 170 acknowledged, or the connection times out. 172 In the case that a host does receive an ICMP error message referring 173 to an ongoing TCP connection, the IP layer will pass this message up 174 to the corresponding TCP instance to raise awareness of the network 175 failure [RFC1122]. TCP's reaction to ICMP messages will depend on 176 the type of error being signaled. 178 2.1. Reaction to ICMP error messages that indicate hard errors 180 When receiving an ICMP error message that indicates a hard error 181 condition, TCP will simply abort the corresponding connection, 182 regardless of the connection state. 184 The Host Requirements RFC [RFC1122] states, in Section 4.2.3.9, that 185 TCP SHOULD abort connections when receiving ICMP error messages that 186 indicate hard errors. This policy is based on the premise that, as 187 hard errors indicate network error conditions that will not change in 188 the near term, it will not be possible for TCP to usefully recover 189 from this type of network failure. 191 2.2. Reaction to ICMP error messages that indicate soft errors 193 If an ICMP error message is received that indicates a soft error, TCP 194 will repeatedly retransmit the segment until it either gets 195 acknowledged or the connection times out. In addition, the TCP 196 sender may record the information for possible later use [Stevens] 197 (pp. 317-319). 199 The Host Requirements RFC [RFC1122] states, in Section 4.2.3.9, that 200 TCP MUST NOT abort connections when receiving ICMP error messages 201 that indicate soft errors. This policy is based on the premise that, 202 as soft errors are transient network failures that will hopefully be 203 solved in the near term, one of the retransmissions will succeed. 205 When the connection timer expires, and an ICMP soft error message has 206 been received before the timeout, TCP can use this information to 207 provide the user with a more specific error message [Stevens] (pp. 208 317-319). 210 This reaction to soft errors exploits the valuable feature of the 211 Internet that for many network failures, the network can be 212 dynamically reconstructed without any disruption of the endpoints. 214 3. Problems that may arise from TCP's reaction to soft errors 216 3.1. General Discussion 218 Even though TCP's fault recovery strategy in the presence of soft 219 errors allows for TCP connections to survive transient network 220 failures, there are scenarios in which this policy may cause 221 undesirable effects. 223 For example, consider a scenario in which an application on a local 224 host is trying to communicate with a destination whose name resolves 225 to several IP addresses. The application on the local host will try 226 to establish a connection with the destination host, usually cycling 227 through the list of IP addresses, until one succeeds [RFC1123]. 228 Suppose that some (but not all) of the addresses in the returned list 229 are permanently unreachable. If such a permanently unreachable 230 address is the first in the list, the application will likely try to 231 use the permanently unreachable address first and block waiting for a 232 timeout before trying an alternate address. 234 As discussed in Section 2, this unreachability condition may or may 235 not be signaled to the sending host. If the local TCP is not 236 signaled concerning the error condition, there is very little that 237 can be done other than repeatedly retransmit the SYN segment, and 238 wait for the existing timeout mechanism in TCP, or an application 239 timeout, to be triggered. However, even if unreachability is 240 signaled by some intermediate router to the local TCP by means of an 241 ICMP soft error message, the local TCP will still repeatedly 242 retransmit the SYN segment until the connection timer expires (in the 243 hopes that the error is transient). The Host Requirements RFC 244 [RFC1122] states that this timer MUST be large enough to provide 245 retransmission of the SYN segment for at least 3 minutes. This would 246 mean that the application on the local host would spend several 247 minutes for each unreachable address it uses for trying to establish 248 the TCP connection. These long delays between connection 249 establishment attempts would be inappropriate for many interactive 250 applications such as the web ([Shneiderman] and [Thadani] offer some 251 insight into interactive systems). This highlights that there is no 252 one definition of a "transient error" and that the level of 253 persistence in the face of failure represents a tradeoff. 255 It is worth noting that while most applications try the addresses 256 returned by the name-to-address function in serial, this is certainly 257 not the only possible approach. For example, applications could try 258 multiple addresses in parallel until one succeeds, possibly avoiding 259 the problem of long delays between connection establishment attempts 260 described in this document. 262 3.2. Problems that may arise with Dual Stack IPv6 on by Default 264 A particular scenario in which the above sketched type of problem may 265 occur regularly is that where dual stack nodes that have IPv6 enabled 266 by default are deployed in IPv4 or mixed IPv4 and IPv6 environments, 267 and the IPv6 connectivity is non-existent 268 [I-D.ietf-v6ops-v6onbydefault]. 270 As discussed in [I-D.ietf-v6ops-v6onbydefault], there are two 271 possible variants of this scenario, which differ in whether the lack 272 of connectivity is signaled to the sending node, or not. 274 In those scenarios in which packets sent to a destination are 275 silently dropped and no ICMPv6 [RFC4443] errors are generated, there 276 is little that can be done other than waiting for the existing 277 connection timeout mechanism in TCP, or an application timeout, to be 278 triggered. 280 In scenarios where a node has no default routers and Neighbor 281 Unreachability Detection (NUD) [RFC2461] fails for destinations 282 assumed to be on-link, or where firewalls or other systems that 283 enforce scope boundaries send ICMPv6 errors, the sending node will be 284 signaled of the unreachability problem. However, as discussed in 285 Section 2.2, standard TCP implementations will not abort connections 286 when receiving ICMP error messages that indicate soft errors. 288 4. Deployed workarounds for long delays between connection- 289 establishment attempts 291 The following subsections describe a number of workarounds for the 292 problem of long delays between connection-establishment attempts that 293 have been implemented in a variety of TCP/IP stacks. We note that 294 treating soft errors as hard errors during connection establishment, 295 while widespread, is not part of standard TCP behavior and this 296 document does not change that state of affairs. The TCPM WG 297 consensus was to document this widespread implementation of 298 nonstandard TCP behavior, but to not change the TCP standard. 300 4.1. Context-sensitive ICMP/TCP interaction 302 As discussed in Section 1, it may make sense for the fault recovery 303 action to depend not only on the type of error being reported, but 304 also on the state of the connection against which the error is 305 reported. For example, one could infer that when an error arrives in 306 response to opening a new connection, it is probably caused by 307 opening the connection improperly, rather than by a transient network 308 failure [RFC0816]. 310 A number of TCP implementations have modified their reaction to soft 311 errors, to treat the errors as hard errors in the SYN-SENT or SYN- 312 RECEIVED states. However, this change violates section 4.2.3.9 of 313 [RFC1122], which states that these Unreachable messages indicate soft 314 error conditions and TCP MUST NOT abort the corresponding connection. 316 This workaround has been implemented, for example, in the Linux 317 kernel since version 2.0.0 (released in 1996) [Linux]. Section 4.2 318 discusses a more conservative approach than that sketched above that 319 is implemented in FreeBSD. 321 4.2. Context-sensitive ICMP/TCP interaction with repeated confirmation 323 A more conservative approach than simply treating soft errors as hard 324 errors as described above would be to abort a connection in the SYN- 325 SENT or SYN-RECEIVED states only after an ICMP Destination 326 Unreachable has been received a specified number of times, and the 327 SYN segment has been retransmitted more than some specified number of 328 times. 330 Two new parameters would have to be introduced to TCP, to be used 331 only during the connection-establishment phase: MAXSYNREXMIT and 332 MAXSOFTERROR. MAXSYNREXMIT would specify the number of times the SYN 333 segment would have to be retransmitted before a connection is 334 aborted. MAXSOFTERROR would specify the number of ICMP messages 335 indicating soft errors that would have to be received before a 336 connection is aborted. 338 Two additional state variables would need to be introduced to store 339 additional state information during the connection-establishment 340 phase: "nsynrexmit" and "nsofterror". Both would be initialized to 341 zero when a connection attempt is initiated, with "nsynrexmit" being 342 incremented by one every time the SYN segment is retransmitted and 343 "nsofterror" being incremented by one every time an ICMP message that 344 indicates a soft error is received. 346 A connection in the SYN-SENT or SYN-RECEIVED states would be aborted 347 if "nsynrexmit" was greater than MAXSYNREXMIT and "nsofterror" was 348 simultaneously greater than MAXSOFTERROR. 350 This approach would give the network more time to solve the 351 connectivity problem than simply aborting a connection attempt upon 352 reception of the first soft error. However, it should be noted that 353 depending on the values chosen for the MAXSYNREXMIT and MAXSOFTERROR 354 parameters, this approach could still lead to long delays between 355 connection establishment attempts, thus not solving the problem. For 356 example, BSD systems abort connections in the SYN-SENT or the SYN- 357 RECEIVED state when a second ICMP error is received, and the SYN 358 segment has been retransmitted more than three times. They also set 359 up a "connection-establishment timer" that imposes an upper limit on 360 the time the connection establishment attempt has to succeed, which 361 expires after 75 seconds [Stevens2] (pp. 828-829). Even when this 362 policy may be better than the three-minutes timeout policy specified 363 in [RFC1122], it may still be inappropriate for handling the 364 potential problems described in this document. This more 365 conservative approach has been implemented in BSD systems for more 366 than ten years [Stevens2]. 368 We also note that the approach given in this section is a generalized 369 version of the approach sketched in the previous section. In 370 particular, with MAXSOFTERROR set to 1 and MAXSYNREXMIT set to zero 371 the schemes are identical. 373 5. Possible drawbacks of changing ICMP semantics 375 The following subsections discuss some of the possible drawbacks 376 arising from the use of the non-standard modifications to TCP's 377 reaction to soft errors described in Section 4.1 and Section 4.2. 379 5.1. Non-deterministic transient network failures 381 In scenarios where a transient network failure affects all of the 382 addresses returned by the name-to-address translation function, all 383 destinations could be unreachable for some short period of time. For 384 example, a mobile system consisting of a cell and a repeater may pass 385 through a tunnel, leading to a loss of connectivity at the repeater, 386 with the repeater sending ICMP soft errors back to the cell. In such 387 scenarios, the application could quickly cycle through all the IP 388 addresses in the list and return an error, when it could have let TCP 389 retry a destination a few seconds later, when the transient problem 390 could have disappeared. In this case, the modifications described 391 here make TCP less robust than a standards-compliant implementation. 393 Additionally, in many cases a domain name maps to a single IP 394 address. In such a case, it might be better to try that address 395 persistently according to normal TCP rules, instead of just aborting 396 the pending connection upon receipt of an ICMP soft error. 398 5.2. Deterministic transient network failures 400 There are some scenarios in which transient network failures could be 401 deterministic. For example, consider a scenario in which upstream 402 network connectivity is triggered by network use. That is, network 403 connectivity is instantiated only on an "as needed" basis. In this 404 scenario, the connection triggering the upstream connectivity could 405 deterministically receive ICMP Destination Unreachables while the 406 upstream connectivity is being activated, and thus would be aborted. 407 Again, in this case, the modifications described here make TCP less 408 robust than a standards-compliant implementation. 410 5.3. Non-compliant Network Address Translators (NATs) 412 Some NATs respond to an unsolicited inbound SYN segment with an ICMP 413 soft error message. If the system sending the unsolicited SYN 414 segment implements the workaround described in this document, it will 415 abort the connection upon receipt of the ICMP error message, thus 416 probably preventing TCP's simultaneous open through the NAT from 417 succeeding. However, it must be stressed that those NATs described 418 in this section are not BEHAVE-compliant, and therefore should 419 implement REQ-4 of [I-D.ietf-behave-tcp] instead. 421 6. Security Considerations 423 This document describes a non-standard modification to TCP's reaction 424 to soft errors that has been implemented in a variety of TCP 425 implementations. This modification makes TCP abort a connection in 426 the SYN-SENT or the SYN-RECEIVED states when it receives an ICMP 427 "Destination Unreachable" message that indicates a soft error. 428 Therefore, the modification could be exploited to reset valid 429 connections during the connection-establishment phase. 431 The non-standard workaround described in this document makes TCP more 432 vulnerable to attack, even if only slightly. However, we note that 433 an attacker wishing to reset ongoing TCP connections could send any 434 of the ICMP hard error messages in any connection state. 436 A discussion of the use of ICMP to perform a variety of attacks 437 against TCP, and a number of counter-measures that minimize the 438 impact of these attacks can be found in [I-D.ietf-tcpm-icmp-attacks]. 440 A discussion of the security issues arising from the use of ICMPv6 441 can be found in [RFC4443]. 443 7. Acknowledgements 445 The author wishes to thank Mark Allman, Ron Bonica, Ted Faber, Gorry 446 Fairhurst, Sally Floyd, Tomohiro Fujisaki, Guillermo Gont, Saikat 447 Guha, Alfred Hoenes, Michael Kerrisk, Eddie Kohler, Mika Liljeberg, 448 Arifumi Matsumoto, Carlos Pignataro, Pasi Sarolahti, Pekka Savola, 449 Pyda Srisuresh, and Joe Touch, for contributing many valuable 450 comments on earlier versions of this document. 452 The author wishes to express deep and heartfelt gratitude to Jorge 453 Oscar Gont and Nelida Garcia, for their precious motivation and 454 guidance. 456 8. Contributors 458 Mika Liljeberg was the first to describe how their implementation 459 treated soft errors. Based on that, the solution discussed in 460 Section 4.1 was documented in [I-D.ietf-v6ops-v6onbydefault] by 461 Sebastien Roy, Alain Durand and James Paugh. 463 9. References 465 9.1. Normative References 467 [RFC0792] Postel, J., "Internet Control Message Protocol", STD 5, 468 RFC 792, September 1981. 470 [RFC0793] Postel, J., "Transmission Control Protocol", STD 7, 471 RFC 793, September 1981. 473 [RFC1122] Braden, R., "Requirements for Internet Hosts - 474 Communication Layers", STD 3, RFC 1122, October 1989. 476 [RFC1123] Braden, R., "Requirements for Internet Hosts - Application 477 and Support", STD 3, RFC 1123, October 1989. 479 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 480 Requirement Levels", BCP 14, RFC 2119, March 1997. 482 [RFC2461] Narten, T., Nordmark, E., and W. Simpson, "Neighbor 483 Discovery for IP Version 6 (IPv6)", RFC 2461, 484 December 1998. 486 [RFC4443] Conta, A., Deering, S., and M. Gupta, "Internet Control 487 Message Protocol (ICMPv6) for the Internet Protocol 488 Version 6 (IPv6) Specification", RFC 4443, March 2006. 490 9.2. Informative References 492 [Guynes] Guynes, J., "Impact of System Response Time on State 493 Anxiety", Communications of the ACM , 1988. 495 [I-D.ietf-behave-tcp] 496 Guha, S., "NAT Behavioral Requirements for TCP", 497 draft-ietf-behave-tcp-07 (work in progress), April 2007. 499 [I-D.ietf-tcpm-icmp-attacks] 500 Gont, F., "ICMP attacks against TCP", 501 draft-ietf-tcpm-icmp-attacks-02 (work in progress), 502 May 2007. 504 [I-D.ietf-v6ops-v6onbydefault] 505 Roy, S., Durand, A., and J. Paugh, "Issues with Dual Stack 506 IPv6 on by Default", draft-ietf-v6ops-v6onbydefault-03 507 (work in progress), July 2004. 509 [Linux] The Linux Project, "http://www.kernel.org". 511 [RFC0816] Clark, D., "Fault isolation and recovery", RFC 816, 512 July 1982. 514 [Shneiderman] 515 Shneiderman, B., "Response Time and Display Rate in Human 516 Performance with Computers", ACM Computing Surveys , 1984. 518 [Stevens] Stevens, W., "TCP/IP Illustrated, Volume 1: The 519 Protocols", Addison-Wesley , 1994. 521 [Stevens2] 522 Wright, G. and W. Stevens, "TCP/IP Illustrated, Volume 2: 523 The Implementation", Addison-Wesley , 1994. 525 [Thadani] Thadani, A., "Interactive User Productivity", IBM Systems 526 Journal No. 1, 1981. 528 Appendix A. Change log (to be removed before publication of the 529 document as an RFC) 531 A.1. Changes from draft-ietf-tcpm-tcp-soft-errors-05 533 o Miscellaneous edits, clarifications, and reorganization of both 534 workarounds into a single top-level section, as suggested by Pasi 535 Sarolahti. 537 o Added note on non-compliant NATs, as suggested by Ted Faber and 538 Saikat Guha 540 o Miscellaneous edits suggested by Gorry Fairhurst 542 o Added a table to clarify how to extrapolate the concept of ICMPv4 543 "soft errors" to ICMPv6 (as suggested by Arifumi Matsumoto and 544 Gorry Fairhurst). 546 o Miscellaneous edits, clarification on alternative approach by 547 sending connection requests in parallel, example of mobile system 548 (for non-deterministic errors), and note on the possible impact of 549 the workarounds on TCP's robusteness (as suggested by Joe Touch) 551 A.2. Changes from draft-ietf-tcpm-tcp-soft-errors-04 553 o Addresses feedback sent by Carlos Pignataro (adds missing error 554 codes in Section 2, and fixes a number of typos/writeos). 556 A.3. Changes from draft-ietf-tcpm-tcp-soft-errors-03 558 o Addresses feedback sent by Ted Faber and Gorry Fairhurst 559 (miscellaneous editorial changes). 561 A.4. Changes from draft-ietf-tcpm-tcp-soft-errors-02 563 o Moved appendix on FreeBSD's approach to the body of the draft. 565 o Removed rest of the appendix, as suggested by Ron Bonica and Mark 566 Allman. 568 o Reworded some parts of the document to make the text more neutral. 570 o Miscellaneous editorial changes. 572 A.5. Changes from draft-ietf-tcpm-tcp-soft-errors-01 574 o Addressed feedback posted by Sally Floyd (remove sentence in 575 Section 2.1 regarding processing of RST segments) 577 A.6. Changes from draft-ietf-tcpm-tcp-soft-errors-00 579 o Miscellaneous editorial changes 581 A.7. Changes from draft-gont-tcpm-tcp-soft-errors-02 583 o Draft resubmitted as draft-ietf. 585 o Miscellaneous editorial changes 587 A.8. Changes from draft-gont-tcpm-tcp-soft-errors-01 589 o Changed wording to describe the mechanism, rather than proposing 590 it 592 o Miscellaneous editorial changes 594 A.9. Changes from draft-gont-tcpm-tcp-soft-errors-00 596 o Added reference to the Linux implementation in Section 4.1 598 o Added Section 5 600 o Added section on Higher-Level API 602 o Added Section 4.2 604 o Moved section "Asynchronous Application Notification" to Appendix 606 o Added section on parallel connection requests 607 o Miscellaneous editorial changes 609 Author's Address 611 Fernando Gont 612 Universidad Tecnologica Nacional / Facultad Regional Haedo 613 Evaristo Carriego 2644 614 Haedo, Provincia de Buenos Aires 1706 615 Argentina 617 Phone: +54 11 4650 8472 618 Email: fernando@gont.com.ar 619 URI: http://www.gont.com.ar 621 Full Copyright Statement 623 Copyright (C) The IETF Trust (2007). 625 This document is subject to the rights, licenses and restrictions 626 contained in BCP 78, and except as set forth therein, the authors 627 retain all their rights. 629 This document and the information contained herein are provided on an 630 "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS 631 OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND 632 THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS 633 OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF 634 THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED 635 WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. 637 Intellectual Property 639 The IETF takes no position regarding the validity or scope of any 640 Intellectual Property Rights or other rights that might be claimed to 641 pertain to the implementation or use of the technology described in 642 this document or the extent to which any license under such rights 643 might or might not be available; nor does it represent that it has 644 made any independent effort to identify any such rights. Information 645 on the procedures with respect to rights in RFC documents can be 646 found in BCP 78 and BCP 79. 648 Copies of IPR disclosures made to the IETF Secretariat and any 649 assurances of licenses to be made available, or the result of an 650 attempt made to obtain a general license or permission for the use of 651 such proprietary rights by implementers or users of this 652 specification can be obtained from the IETF on-line IPR repository at 653 http://www.ietf.org/ipr. 655 The IETF invites any interested party to bring to its attention any 656 copyrights, patents or patent applications, or other proprietary 657 rights that may cover technology that may be required to implement 658 this standard. Please address the information to the IETF at 659 ietf-ipr@ietf.org. 661 Acknowledgment 663 Funding for the RFC Editor function is provided by the IETF 664 Administrative Support Activity (IASA).