idnits 2.17.1 draft-ietf-6man-impatient-nud-07.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 (Using the creation date from RFC4861, updated by this document, for RFC5378 checks: 2004-07-16) -- 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 21, 2013) is 3833 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) No issues found here. Summary: 0 errors (**), 0 flaws (~~), 1 warning (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 6MAN WG E. Nordmark 3 Internet-Draft Arista Networks 4 Updates: 4861 (if approved) I. Gashinsky 5 Intended status: Standards Track Yahoo! 6 Expires: April 24, 2014 October 21, 2013 8 Neighbor Unreachability Detection is too impatient 9 draft-ietf-6man-impatient-nud-07.txt 11 Abstract 13 IPv6 Neighbor Discovery includes Neighbor Unreachability Detection. 14 That function is very useful when a host has an alternative neighbor, 15 for instance when there are multiple default routers, since it allows 16 the host to switch to the alternative neighbor in short time. This 17 time is 3 seconds after the node starts probing by default. However, 18 if there are no alternative neighbors, this is far too impatient. 19 This document specifies relaxed rules for Neighbor Discovery 20 retransmissions that allow an implementation to choose different 21 timeout behavior based on whether or not there are alternative 22 neighbors. This document updates RFC 4861. 24 Status of this Memo 26 This Internet-Draft is submitted in full conformance with the 27 provisions of BCP 78 and BCP 79. 29 Internet-Drafts are working documents of the Internet Engineering 30 Task Force (IETF). Note that other groups may also distribute 31 working documents as Internet-Drafts. The list of current Internet- 32 Drafts is at http://datatracker.ietf.org/drafts/current/. 34 Internet-Drafts are draft documents valid for a maximum of six months 35 and may be updated, replaced, or obsoleted by other documents at any 36 time. It is inappropriate to use Internet-Drafts as reference 37 material or to cite them other than as "work in progress." 39 This Internet-Draft will expire on April 24, 2014. 41 Copyright Notice 43 Copyright (c) 2013 IETF Trust and the persons identified as the 44 document authors. All rights reserved. 46 This document is subject to BCP 78 and the IETF Trust's Legal 47 Provisions Relating to IETF Documents 48 (http://trustee.ietf.org/license-info) in effect on the date of 49 publication of this document. Please review these documents 50 carefully, as they describe your rights and restrictions with respect 51 to this document. Code Components extracted from this document must 52 include Simplified BSD License text as described in Section 4.e of 53 the Trust Legal Provisions and are provided without warranty as 54 described in the Simplified BSD License. 56 Table of Contents 58 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 59 2. Definition Of Terms . . . . . . . . . . . . . . . . . . . . . . 4 60 3. Protocol Updates . . . . . . . . . . . . . . . . . . . . . . . 4 61 4. Example Algorithm . . . . . . . . . . . . . . . . . . . . . . . 6 62 5. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 8 63 6. Security Considerations . . . . . . . . . . . . . . . . . . . . 8 64 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . . 8 65 8. References . . . . . . . . . . . . . . . . . . . . . . . . . . 8 66 8.1. Normative References . . . . . . . . . . . . . . . . . . . 8 67 8.2. Informative References . . . . . . . . . . . . . . . . . . 8 68 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 9 70 1. Introduction 72 IPv6 Neighbor Discovery [RFC4861] includes Neighbor Unreachability 73 Detection (NUD), which detects when a neighbor is no longer 74 reachable. The timeouts specified for NUD are very short (by default 75 three transmissions spaced one second apart). These short timeouts 76 can be appropriate when there are alternative neighbors to which the 77 packets can be sent. For example, if a host has multiple default 78 routers in its Default Router List or if the host has a Neighbor 79 Cache Entry (NCE) created by a Redirect message. In those cases, 80 when NUD fails, the host will try the alternative neighbor by redoing 81 next-hop selection. That implies picking the next router in the 82 Default Router List or discarding the redirect, respectively. 84 The timeouts specified in [RFC4861] were chosen to be short in order 85 to optimize for the scenarios where alternative neighbors are 86 available. 88 However, when there is no alternative neighbor there are several 89 benefits in making NUD try probing for a longer time. One of those 90 benefits is to make NUD more robust against transient failures, such 91 as spanning tree reconvergence and other layer 2 issues that can take 92 many seconds to resolve. Marking the NCE as unreachable in that case 93 causes additional multicast on the network. Assuming there are IP 94 packets to send, the lack of an NCE will result in multicast Neighbor 95 Solicitations being sent (to the solicited-node multicast address) 96 every second instead of the unicast Neighbor Solicitations that NUD 97 sends. 99 As a result IPv6 Neighbor Discovery is operationally more brittle 100 than IPv4 ARP. For IPv4 there is no mandatory time limit on the 101 retransmission behavior for ARP [RFC0826] which allows implementors 102 to pick more robust schemes. 104 The following constant values in [RFC4861] seem to have been made 105 part of IPv6 conformance testing: MAX_MULTICAST_SOLICIT, 106 MAX_UNICAST_SOLICIT, and RETRANS_TIMER. While such strict 107 conformance testing seems consistent with [RFC4861], it means that 108 the standard needs to be updated to allow IPv6 Neighbor Discovery to 109 be as robust as ARP. 111 This document updates RFC 4861 to relax the retransmission rules. 113 Additional motivations for making IPv6 Neighbor Discovery more robust 114 in the face of degenerate conditions are covered in [RFC6583]. 116 2. Definition Of Terms 118 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 119 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 120 document are to be interpreted as described in [RFC2119]. 122 3. Protocol Updates 124 Discarding the NCE after after three packets spaced one second apart 125 is only needed when an alternative neighbor is available, such as an 126 additional default router or a redirect. 128 If an implementation transmits more than MAX_UNICAST_SOLICIT/ 129 MAX_MULTICAST_SOLICIT packets then it SHOULD use exponential backoff 130 of the retransmit timer. This is to avoid any significant load due 131 to a steady background level of retransmissions from implementations 132 that retransmit a large number of NSes before discarding the NCE. 134 Even if there is no alternative neighbor, the protocol needs to be 135 able to handle the case when the link-layer address of the neighbor/ 136 target has changed by switching to multicast Neighbor Solicitations 137 at some point in time. 139 In order to capture all the cases above this document introduces a 140 new UNREACHABLE state in the conceptual model described in [RFC4861]. 141 A NCE in the UNREACHABLE state retains the link-layer address, and 142 IPv6 packets continue to be sent to that link-layer address. But in 143 the UNREACHABLE state the NUD Neighbor Solicitations are multicast 144 (to the solicited-node multicast address), using a timeout that 145 follows an exponential backoff. 147 In the places where RFC4861 says to to discard/delete the NCE after N 148 probes (Section 7.3, 7.3.3 and Appendix C) this document instead 149 specifies a transition to the UNREACHABLE state. 151 If the Neighbor Cache Entry was created by a redirect, a node MAY 152 delete the NCE instead of changing its state to UNREACHABLE. In any 153 case, the node SHOULD NOT use an NCE created by a Redirect to send 154 packets if that NCE is in UNREACHABLE state. Packets should be sent 155 following the next-hop selection algorithm in [RFC4861], Section 5.2, 156 which disregards NCEs that are not reachable. 158 The default router selection in [RFC4861], Section 6.3.6 says to 159 prefer default routers that are "known to be reachable". For the 160 purposes of that section, if the NCE for the router is in UNREACHABLE 161 state, it is not known to be reachable. Thus the particular text in 162 section 6.3.6 which says "in any state other than INCOMPLETE" needs 163 to be extended to say "in any state other than INCOMPLETE or 164 UNREACHABLE". 166 Apart from the use of multicast NS instead of unicast NS, and the 167 exponential backoff of the timer, the UNREACHABLE state works the 168 same as the current PROBE state. 170 A node MAY garbage collect a Neighbor Cache Entry at any time as 171 specified in RFC 4861. This freedom to garbage collect does not 172 change with the introduction of the UNREACHABLE state in the 173 conceptual model. An implementation MAY prefer garbage collecting 174 UNREACHABLE NCEs over other NCEs. 176 There is a non-obvious extension to the state machine description in 177 Appendix C in RFC 4861 in the case for "NA, Solicited=1, Override=0. 178 Different link-layer address than cached". There we need to add 179 "UNREACHABLE" to the current list of "STALE, PROBE, Or DELAY". That 180 is, the NCE would be unchanged. Note that there is no corresponding 181 change necessary to the text in [RFC4861], Section 7.2.5, since it is 182 phrased using "Otherwise" instead of explicitly listing the three 183 states. 185 The other state transitions described in Appendix C handle the 186 introduction of the UNREACHABLE state without any change, since they 187 are described using "not INCOMPLETE". 189 There is also the more obvious change already described above. RFC 190 4861 has this: 192 State Event Action New state 194 PROBE Retransmit timeout, Discard entry - 195 N or more 196 retransmissions. 198 That needs to be replaced by: 200 State Event Action New state 202 PROBE Retransmit timeout, Increase timeout UNREACHABLE 203 N retransmissions. Send multicast NS 205 UNREACHABLE Retransmit timeout Increase timeout UNREACHABLE 206 Send multicast NS 208 The exponential backoff SHOULD be clamped at some reasonable maximum 209 retransmit timeout, such as 60 seconds (see MAX_RETRANS_TIMER below). 210 If there is no IPv6 packet sent using the UNREACHABLE NCE, then it is 211 RECOMMENDED to stop the retransmits of the multicast NS until either 212 the NCE is garbage collected or there are IPv6 packets sent using the 213 NCE. The multicast NS and associated exponential backoff can be 214 applied on the condition of the continued use of the NCE to send IPv6 215 packets to the recorded link-layer address. 217 A node can unicast the first few Neighbor Solicitation messages even 218 while in UNREACHABLE state, but it MUST switch to multicast Neighbor 219 Solicitations within 60 seconds of the initial retransmission to be 220 able to handle a link-layer address change for the target. The 221 example below shows such behavior. 223 4. Example Algorithm 225 This section is NOT normative, but specifies a simple implementation 226 which conforms with this document. The implementation is described 227 using operator configurable values that allows it to be configured in 228 a way to be compatible with the retransmission behavior in [RFC4861]. 229 The operator can configure the values for MAX_UNICAST_SOLICIT, 230 MAX_MULTICAST_SOLICIT, RETRANS_TIMER, and the new BACKOFF_MULTIPLE, 231 MAX_RETRANS_TIMER and MARK_UNREACHABLE. This allows the 232 implementation to be as simple as: 234 next_retrans = ($BACKOFF_MULTIPLE ^ $solicit_retrans_num) * 235 $RetransTimer * $JitterFactor where solicit_retrans_num is zero for 236 the first transmission, and JitterFactor is a random value between 237 MIN_RANDOM_FACTOR and MAX_RANDOM_FACTOR [RFC4861] to avoid any 238 synchronization of transmissions from different hosts. 240 After MARK_UNREACHABLE transmissions the implementation would mark 241 the NCE UNREACHABLE and as result explore alternate next hops. After 242 MAX_UNICAST_SOLICIT the implementation would switch to multicast NUD 243 probes. 245 The behavior of this example algorithm is to have 5 attempts, with 246 timing spacing of 0 (initial request), 1 second later, 3 seconds 247 after the first retransmission, then 9, then 27, and switch to 248 UNREACHABLE after the first three transmissions. Thus relative to 249 the time of the first transmissions the retransmissions would occur 250 at 1 second, 4 seconds, 13 seconds, and finally 40 seconds. At 4 251 seconds from the first transmission the NCE would be marked 252 UNREACHABLE. That behavior corresponds to: 254 MAX_UNICAST_SOLICIT=5 256 RETRANS_TIMER=1 (default) 257 MAX_RETRANS_TIMER=60 259 BACKOFF_MULTIPLE=3 261 MARK_UNREACHABLE=3 263 After 3 retransmissions the implementation would mark the NCE 264 UNREACHABLE. That results in trying an alternative neighbor, such as 265 another default router or ignoring a redirect as specified in 266 [RFC4861]. With the above values that would occur after 4 seconds 267 after the first transmission compared to the 2 seconds using the 268 fixed scheme in [RFC4861]. That additional delay is small compared 269 to the default 30,000 milliseconds ReachableTime. 271 After 5 transmissions, i.e., 40 seconds after the initial 272 transmission, the example behavior is to switch to multicast NUD 273 probes. In the language of the state machine in [RFC4861] that 274 corresponds to the action "Discard entry". Thus any attempts to send 275 future packets would result in sending multicast NS packets. An 276 implementation MAY retain the backoff value as it switches to 277 multicast NUD probes. The potential downside of deferring switching 278 to multicast is that it would take longer for NUD to handle a change 279 in a link-layer address i.e., the case when a host or a router 280 changes their link-layer address while keeping the same IPv6 address. 281 However, [RFC4861] says that a node MAY send unsolicited NS to handle 282 that case, which is rather infrequent in operational networks. In 283 any case, the implementation needs to follow the "SHOULD" in section 284 Section 3 to switch to multicast solutions within 60 seconds after 285 the initial transmission. 287 If BACKOFF_MULTIPLE=1, MARK_UNREACHABLE=3 and MAX_UNICAST_SOLICIT=3, 288 you would get the same behavior as in [RFC4861]. 290 An implementation following this algorithm would, if the request was 291 not answered at first due for example to a transitory condition, 292 retry immediately, and then back off for progressively longer 293 periods. This would allow for a reasonably fast resolution time when 294 the transitory condition clears. 296 Note that RetransTimer and ReachableTime are by default set from the 297 protocol constants RETRANS_TIMER and REACHABLE_TIME, but are 298 overridden by values advertised in Router Advertisements as specified 299 in [RFC4861]. That remains the case even with the protocol updates 300 specified in this document. The key values that the operator would 301 configure are BACKOFF_MULTIPLE, MAX_RETRANS_TIMER, 302 MAX_UNICAST_SOLICIT and MAX_MULTICAST_SOLICIT. 304 It is be useful to have a maximum value for 305 ($BACKOFF_MULTIPLE^$solicit_attempt_num)*$RetransTimer so that the 306 retransmissions are not too far apart. The above value of 60 seconds 307 for this MAX_RETRANS_TIMER is consistent with DHCPv6. 309 5. Acknowledgements 311 The comments from Thomas Narten, Philip Homburg, Joel Jaeggli, Hemant 312 Singh, Tina Tsou, Suresh Krishnan, and Murray Kucherawy have helped 313 improve this draft. 315 6. Security Considerations 317 Relaxing the retransmission behavior for NUD is believed to have no 318 impact on security. In particular, it doesn't impact the application 319 Secure Neighbor Discovery [RFC3971]. 321 7. IANA Considerations 323 This are no IANA considerations for this document. 325 8. References 327 8.1. Normative References 329 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 330 Requirement Levels", BCP 14, RFC 2119, March 1997. 332 [RFC3971] Arkko, J., Kempf, J., Zill, B., and P. Nikander, "SEcure 333 Neighbor Discovery (SEND)", RFC 3971, March 2005. 335 [RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman, 336 "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861, 337 September 2007. 339 8.2. Informative References 341 [RFC0826] Plummer, D., "Ethernet Address Resolution Protocol: Or 342 converting network protocol addresses to 48.bit Ethernet 343 address for transmission on Ethernet hardware", STD 37, 344 RFC 826, November 1982. 346 [RFC6583] Gashinsky, I., Jaeggli, J., and W. Kumari, "Operational 347 Neighbor Discovery Problems", RFC 6583, March 2012. 349 Authors' Addresses 351 Erik Nordmark 352 Arista Networks 353 Santa Clara, CA 354 USA 356 Email: nordmark@acm.org 358 Igor Gashinsky 359 Yahoo! 360 45 W 18th St 361 New York, NY 362 USA 364 Email: igor@yahoo-inc.com