idnits 2.17.1 draft-ietf-6man-rfc4941bis-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 : ---------------------------------------------------------------------------- ** 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.) == The 'Obsoletes: ' line in the draft header should list only the _numbers_ of the RFCs which will be obsoleted by this document (if approved); it should not include the word 'RFC' in the list. -- The abstract seems to indicate that this document obsoletes RFC4941, but the header doesn't have an 'Obsoletes:' line to match this. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (March 27, 2020) is 1481 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references to lower-maturity documents in RFCs) ** Obsolete normative reference: RFC 4941 (Obsoleted by RFC 8981) == Outdated reference: A later version (-15) exists of draft-ietf-mboned-ieee802-mcast-problems-11 Summary: 2 errors (**), 0 flaws (~~), 3 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 IPv6 Maintenance (6man) Working Group F. Gont 3 Internet-Draft SI6 Networks / UTN-FRH 4 Obsoletes: rfc4941 (if approved) S. Krishnan 5 Intended status: Standards Track Ericsson Research 6 Expires: September 28, 2020 T. Narten 7 IBM Corporation 8 R. Draves 9 Microsoft Research 10 March 27, 2020 12 Temporary Address Extensions for Stateless Address Autoconfiguration in 13 IPv6 14 draft-ietf-6man-rfc4941bis-08 16 Abstract 18 This document describes an extension that causes nodes to generate 19 global scope addresses with randomized interface identifiers that 20 change over time. Changing global scope addresses over time limits 21 the window of time during which eavesdroppers and other information 22 collectors may trivially perform address-based network activity 23 correlation when the same address is employed for multiple 24 transactions by the same node. Additionally, it reduces the window 25 of exposure of a node via an addresses that becomes revealed as a 26 result of active communication. This document obsoletes RFC4941. 28 Status of This Memo 30 This Internet-Draft is submitted in full conformance with the 31 provisions of BCP 78 and BCP 79. 33 Internet-Drafts are working documents of the Internet Engineering 34 Task Force (IETF). Note that other groups may also distribute 35 working documents as Internet-Drafts. The list of current Internet- 36 Drafts is at https://datatracker.ietf.org/drafts/current/. 38 Internet-Drafts are draft documents valid for a maximum of six months 39 and may be updated, replaced, or obsoleted by other documents at any 40 time. It is inappropriate to use Internet-Drafts as reference 41 material or to cite them other than as "work in progress." 43 This Internet-Draft will expire on September 28, 2020. 45 Copyright Notice 47 Copyright (c) 2020 IETF Trust and the persons identified as the 48 document authors. All rights reserved. 50 This document is subject to BCP 78 and the IETF Trust's Legal 51 Provisions Relating to IETF Documents 52 (https://trustee.ietf.org/license-info) in effect on the date of 53 publication of this document. Please review these documents 54 carefully, as they describe your rights and restrictions with respect 55 to this document. Code Components extracted from this document must 56 include Simplified BSD License text as described in Section 4.e of 57 the Trust Legal Provisions and are provided without warranty as 58 described in the Simplified BSD License. 60 Table of Contents 62 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 63 1.1. Terminology . . . . . . . . . . . . . . . . . . . . . . . 3 64 1.2. Problem Statement . . . . . . . . . . . . . . . . . . . . 4 65 2. Background . . . . . . . . . . . . . . . . . . . . . . . . . 4 66 2.1. Extended Use of the Same Identifier . . . . . . . . . . . 4 67 2.2. Possible Approaches . . . . . . . . . . . . . . . . . . . 6 68 3. Protocol Description . . . . . . . . . . . . . . . . . . . . 6 69 3.1. Design Guidelines . . . . . . . . . . . . . . . . . . . . 6 70 3.2. Assumptions . . . . . . . . . . . . . . . . . . . . . . . 7 71 3.3. Generation of Randomized Interface Identifiers . . . . . 8 72 3.3.1. Simple Randomized Interface Identifiers . . . . . . . 8 73 3.3.2. Hash-based Generation of Randomized Interface 74 Identifiers . . . . . . . . . . . . . . . . . . . . . 9 75 3.4. Generating Temporary Addresses . . . . . . . . . . . . . 10 76 3.5. Expiration of Temporary Addresses . . . . . . . . . . . . 12 77 3.6. Regeneration of Temporary Addresses . . . . . . . . . . . 12 78 3.7. Implementation Considerations . . . . . . . . . . . . . . 13 79 3.8. Defined Constants . . . . . . . . . . . . . . . . . . . . 14 80 4. Implications of Changing Interface Identifiers . . . . . . . 15 81 5. Significant Changes from RFC4941 . . . . . . . . . . . . . . 15 82 6. Future Work . . . . . . . . . . . . . . . . . . . . . . . . . 16 83 7. Security Considerations . . . . . . . . . . . . . . . . . . . 17 84 8. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 17 85 9. References . . . . . . . . . . . . . . . . . . . . . . . . . 18 86 9.1. Normative References . . . . . . . . . . . . . . . . . . 18 87 9.2. Informative References . . . . . . . . . . . . . . . . . 19 88 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 21 90 1. Introduction 92 Stateless address autoconfiguration (SLAAC) [RFC4862] defines how an 93 IPv6 node generates addresses without the need for a Dynamic Host 94 Configuration Protocol for IPv6 (DHCPv6) server. The security and 95 privacy implications of such addresses have been discussed in detail 96 in [RFC7721],[RFC7217], and RFC7707. This document specifies an 97 extension for SLAAC to generate temporary addresses, that can help 98 mitigate some of the aforementioned issues. This is a revision of 99 RFC4941, and formally obsoletes RFC4941. Section 5 describes the 100 changes from [RFC4941]. 102 The default address selection for IPv6 has been specified in 103 [RFC6724]. The determination as to whether to use stable versus 104 temporary addresses can in some cases only be made by an application. 105 For example, some applications may always want to use temporary 106 addresses, while others may want to use them only in some 107 circumstances or not at all. An Application Programming Interface 108 (API) such as that specified in [RFC5014] can enable individual 109 applications to indicate a preference for the use of temporary 110 addresses. 112 Section 2 provides background information. Section 3 describes a 113 procedure for generating temporary addresses. Section 4 discusses 114 implications of changing interface identifiers (IIDs). Section 5 115 describes the changes from [RFC4941]. 117 1.1. Terminology 119 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 120 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 121 "OPTIONAL" in this document are to be interpreted as described in BCP 122 14 [RFC2119] [RFC8174] when, and only when, they appear in all 123 capitals, as shown here. 125 The terms "public address", "stable address", "temporary address", 126 "constant IID", "stable IID", and "temporary IID" are to be 127 interpreted as specified in [RFC7721]. 129 The term "global scope addresses" is used in this document to 130 collectively refer to "Global unicast addresses" as defined in 131 [RFC4291] and "Unique local addresses" as defined in [RFC4193], and 132 not to "globally reachable" as defined in [RFC8190]. 134 1.2. Problem Statement 136 Addresses generated using stateless address autoconfiguration 137 [RFC4862] contain an embedded interface identifier, which may remain 138 stable over time. Anytime a fixed identifier is used in multiple 139 contexts, it becomes possible to correlate seemingly unrelated 140 activity using this identifier. 142 The correlation can be performed by 144 o An attacker who is in the path between the node in question and 145 the peer(s) to which it is communicating, and who can view the 146 IPv6 addresses present in the datagrams. 148 o An attacker who can access the communication logs of the peers 149 with which the node has communicated. 151 Since the identifier is embedded within the IPv6 address, it cannot 152 be hidden. This document proposes a solution to this issue by 153 generating interface identifiers that vary over time. 155 Note that an attacker, who is on path, may be able to perform 156 significant correlation on unencrypted packets based on 158 o The payload contents of the packets on the wire 160 o The characteristics of the packets such as packet size and timing 162 Use of temporary addresses will not prevent such payload-based 163 correlation, which can only be addressed by widespread deployment of 164 encryption as advocated in [RFC7624]. Nor will it prevent an on-link 165 observer (e.g. the node's default router) to track all the node's 166 addresses. 168 2. Background 170 This section discusses the problem in more detail, and provides 171 context for evaluating the significance of the concerns in specific 172 environments and makes comparisons with existing practices. 174 2.1. Extended Use of the Same Identifier 176 The use of a non-changing interface identifier to form addresses is a 177 specific instance of the more general case where a constant 178 identifier is reused over an extended period of time and in multiple 179 independent activities. Any time the same identifier is used in 180 multiple contexts, it becomes possible for that identifier to be used 181 to correlate seemingly unrelated activity. For example, a network 182 sniffer placed strategically on a link across which all traffic to/ 183 from a particular host crosses could keep track of which destinations 184 a node communicated with and at what times. Such information can in 185 some cases be used to infer things, such as what hours an employee 186 was active, when someone is at home, etc. Although it might appear 187 that changing an address regularly in such environments would be 188 desirable to lessen privacy concerns, it should be noted that the 189 network prefix portion of an address also serves as a constant 190 identifier. All nodes at, say, a home, would have the same network 191 prefix, which identifies the topological location of those nodes. 192 This has implications for privacy, though not at the same granularity 193 as the concern that this document addresses. Specifically, all nodes 194 within a home could be grouped together for the purposes of 195 collecting information. If the network contains a very small number 196 of nodes, say, just one, changing just the interface identifier will 197 not enhance privacy, since the prefix serves as a constant 198 identifier. 200 One of the requirements for correlating seemingly unrelated 201 activities is the use (and reuse) of an identifier that is 202 recognizable over time within different contexts. IP addresses 203 provide one obvious example, but there are more. 205 For example, web browsers and servers typically exchange "cookies" 206 with each other [RFC6265]. Cookies allow web servers to correlate a 207 current activity with a previous activity. One common usage is to 208 send back targeted advertising to a user by using the cookie supplied 209 by the browser to identify what earlier queries had been made (e.g., 210 for what type of information). Based on the earlier queries, 211 advertisements can be targeted to match the (assumed) interests of 212 the end-user. 214 The use of a constant identifier within an address is of special 215 concern because addresses are a fundamental requirement of 216 communication and cannot easily be hidden from eavesdroppers and 217 other parties. Even when higher layers encrypt their payloads, 218 addresses in packet headers appear in the clear. Consequently, if a 219 mobile host (e.g., laptop) accessed the network from several 220 different locations, an eavesdropper might be able to track the 221 movement of that mobile host from place to place, even if the upper 222 layer payloads were encrypted. 224 Changing global scope addresses over time limits the time window over 225 which eavesdroppers and other information collectors may trivially 226 correlate network activity when the same address is employed for 227 multiple transactions by the same node. Additionally, it reduces the 228 window of exposure of a node via an address that gets revealed as a 229 result of active communication. 231 The security and privacy implications of IPv6 addresses are discussed 232 in detail in [RFC7721], [RFC7707], and [RFC7217]. 234 2.2. Possible Approaches 236 One approach, compatible with the stateless address autoconfiguration 237 architecture, would be to change the interface identifier portion of 238 an address over time. Changing the interface identifier can make it 239 more difficult to look at the IP addresses in independent 240 transactions and identify which ones actually correspond to the same 241 node, both in the case where the routing prefix portion of an address 242 changes and when it does not. 244 Many machines function as both clients and servers. In such cases, 245 the machine would need a DNS name for its use as a server. Whether 246 the address stays fixed or changes has little privacy implication 247 since the DNS name remains constant and serves as a constant 248 identifier. When acting as a client (e.g., initiating 249 communication), however, such a machine may want to vary the 250 addresses it uses. In such environments, one may need multiple 251 addresses: a stable address registered in the DNS, that is used to 252 accept incoming connection requests from other machines, and a 253 temporary address used to shield the identity of the client when it 254 initiates communication. 256 On the other hand, a machine that functions only as a client may want 257 to employ only temporary addresses for public communication. 259 To make it difficult to make educated guesses as to whether two 260 different interface identifiers belong to the same node, the 261 algorithm for generating alternate identifiers must include input 262 that has an unpredictable component from the perspective of the 263 outside entities that are collecting information. 265 3. Protocol Description 267 The following subsections define the procedures for the generation of 268 IPv6 temporary addresses. 270 3.1. Design Guidelines 272 Temporary addresses observe the following properties: 274 1. Temporary addresses are typically employed for initiating 275 outgoing sessions. 277 2. Temporary addresses are used for a short period of time 278 (typically hours to days) and are subsequently deprecated. 280 Deprecated addresses can continue to be used for established 281 connections, but are not used to initiate new connections. 283 3. New temporary addresses are generated periodically to replace 284 temporary addresses that expire. 286 4. Temporary addresses must have a limited lifetime (limited "valid 287 lifetime" and "preferred lifetime" from [RFC4862]), that should 288 be statistically different for different addresses. The lifetime 289 of an address should be further reduced when privacy-meaningful 290 events (such as a node attaching to a different network, or the 291 regeneration of a new randomized MAC address) takes place. 293 5. By default, one address is generated for each prefix advertised 294 by stateless address autoconfiguration. The resulting Interface 295 Identifiers must be statistically different when addresses are 296 configured for different prefixes. That is, when temporary 297 addresses are generated for different autoconfiguration prefixes 298 for the same network interface, the resulting Interface 299 Identifiers must be statistically different. This means that, 300 given two addresses that employ different prefixes, it must be 301 difficult for an outside entity to tell whether the addresses 302 correspond to the same network interface or even whether they 303 have been generated by the same host. 305 6. It must be difficult for an outside entity to predict the 306 Interface Identifiers that will be employed for temporary 307 addresses, even with knowledge of the algorithm/method employed 308 to generate them and/or knowledge of the Interface Identifiers 309 previously employed for other temporary addresses. These 310 Interface Identifiers must be semantically opaque [RFC7136] and 311 must not follow any specific patterns. 313 3.2. Assumptions 315 The following algorithm assumes that for a given temporary address, 316 an implementation can determine the prefix from which it was 317 generated. When a temporary address is deprecated, a new temporary 318 address is generated. The specific valid and preferred lifetimes for 319 the new address are dependent on the corresponding lifetime values 320 set for the prefix from which it was generated. 322 Finally, this document assumes that when a node initiates outgoing 323 communication, temporary addresses can be given preference over 324 stable addresses (if available), when the device is configured to do 325 so. [RFC6724] mandates implementations to provide a mechanism, which 326 allows an application to configure its preference for temporary 327 addresses over stable addresses. It also allows for an 328 implementation to prefer temporary addresses by default, so that the 329 connections initiated by the node can use temporary addresses without 330 requiring application-specific enablement. This document also 331 assumes that an API will exist that allows individual applications to 332 indicate whether they prefer to use temporary or stable addresses and 333 override the system defaults (see e.g. [RFC5014]). 335 3.3. Generation of Randomized Interface Identifiers 337 The following subsections specify example algorithms for generating 338 temporary interface identifiers that follow the guidelines in 339 Section 3.1 of this document. The algorithm specified in 340 Section 3.3.1 benefits from a Pseudo-Random Number Generator (PRNG) 341 available on the system. The algorithm specified in Section 3.3.2 342 allows for code reuse by nodes that implement [RFC7217]. 344 3.3.1. Simple Randomized Interface Identifiers 346 One approach is to select a pseudorandom number of the appropriate 347 length. A node employing this algorithm should generate IIDs as 348 follows: 350 1. Obtain a random number (see [RFC4086] for randomness requirements 351 for security). 353 2. The Interface Identifier is obtained by taking as many bits from 354 the random number obtained in the previous step as necessary. 355 Note: there are no special bits in an Interface Identifier 356 [RFC7136]. 358 We note that [RFC4291] requires that the Interface IDs of all 359 unicast addresses (except those that start with the binary 360 value 000) be 64 bits long. However, the method discussed in 361 this document could be employed for generating Interface IDs 362 of any arbitrary length, albeit at the expense of reduced 363 entropy (when employing Interface IDs smaller than 64 bits). 364 The privacy implications of the IID length are discussed in 365 [RFC7421]. 367 3. The resulting Interface Identifier SHOULD be compared against the 368 reserved IPv6 Interface Identifiers [RFC5453] [IANA-RESERVED-IID] 369 and against those Interface Identifiers already employed in an 370 address of the same network interface and the same network 371 prefix. In the event that an unacceptable identifier has been 372 generated, a new interface identifier should be generated, by 373 repeating the algorithm from the first step. 375 3.3.2. Hash-based Generation of Randomized Interface Identifiers 377 The algorithm in [RFC7217] can be augmented for the generation of 378 temporary addresses. The benefit of this would be that a node could 379 employ a single algorithm for generating stable and temporary 380 addresses, by employing appropriate parameters. 382 Nodes would employ the following algorithm for generating the 383 temporary IID: 385 1. Compute a random identifier with the expression: 387 RID = F(Prefix, Net_Iface, Network_ID, Time, DAD_Counter, 388 secret_key) 390 Where: 392 RID: 393 Random Identifier 395 F(): 396 A pseudorandom function (PRF) that MUST NOT be computable from 397 the outside (without knowledge of the secret key). F() MUST 398 also be difficult to reverse, such that it resists attempts to 399 obtain the secret_key, even when given samples of the output 400 of F() and knowledge or control of the other input parameters. 401 F() SHOULD produce an output of at least 64 bits. F() could 402 be implemented as a cryptographic hash of the concatenation of 403 each of the function parameters. SHA-256 [FIPS-SHS] is one 404 possible option for F(). Note: MD5 [RFC1321] is considered 405 unacceptable for F() [RFC6151]. 407 Prefix: 408 The prefix to be used for SLAAC, as learned from an ICMPv6 409 Router Advertisement message. 411 Net_Iface: 412 The MAC address corresponding to the underlying network 413 interface card, in the case the link uses IEEE802 link-layer 414 identifiers. Employing the MAC address for this parameter 415 (over the other suggested options in RFC7217) means that the 416 re-generation of a randomized MAC address will result in a 417 different temporary address. 419 Network_ID: 420 Some network-specific data that identifies the subnet to which 421 this interface is attached -- for example, the IEEE 802.11 422 Service Set Identifier (SSID) corresponding to the network to 423 which this interface is associated. Additionally, Simple DNA 424 [RFC6059] describes ideas that could be leveraged to generate 425 a Network_ID parameter. This parameter is SHOULD be employed 426 if some form of "Network_ID" is available. 428 Time: 429 An implementation-dependent representation of time. One 430 possible example is the representation in UNIX-like systems 431 [OPEN-GROUP], that measure time in terms of the number of 432 seconds elapsed since the Epoch (00:00:00 Coordinated 433 Universal Time (UTC), 1 January 1970). The addition of the 434 "Time" argument results in (statistically) different interface 435 identifiers over time. 437 DAD_Counter: 438 A counter that is employed to resolve Duplicate Address 439 Detection (DAD) conflicts. 441 secret_key: 442 A secret key that is not known by the attacker. The secret 443 key SHOULD be of at least 128 bits. It MUST be initialized to 444 a pseudo-random number (see [RFC4086] for randomness 445 requirements for security) when the operating system is 446 "bootstrapped". 448 2. The Interface Identifier is finally obtained by taking as many 449 bits from the RID value (computed in the previous step) as 450 necessary, starting from the least significant bit. The 451 resulting Interface Identifier SHOULD be compared against the 452 reserved IPv6 Interface Identifiers [RFC5453] [IANA-RESERVED-IID] 453 and against those Interface Identifiers already employed in an 454 address of the same network interface and the same network 455 prefix. In the event that an unacceptable identifier has been 456 generated, the value DAD_Counter should be incremented by 1, and 457 the algorithm should be restarted from the first step. 459 3.4. Generating Temporary Addresses 461 [RFC4862] describes the steps for generating a link-local address 462 when an interface becomes enabled as well as the steps for generating 463 addresses for other scopes. This document extends [RFC4862] as 464 follows. When processing a Router Advertisement with a Prefix 465 Information option carrying a prefix for the purposes of address 466 autoconfiguration (i.e., the A bit is set), the node MUST perform the 467 following steps: 469 1. Process the Prefix Information Option as defined in [RFC4862], 470 adjusting the lifetimes of existing temporary addresses. If a 471 received option may extend the lifetimes of temporary addresses, 472 with the overall constraint that no temporary addresses should 473 ever remain "valid" or "preferred" for a time longer than 474 (TEMP_VALID_LIFETIME) or (TEMP_PREFERRED_LIFETIME - 475 DESYNC_FACTOR) respectively. The configuration variables 476 TEMP_VALID_LIFETIME and TEMP_PREFERRED_LIFETIME correspond to 477 approximate target lifetimes for temporary addresses. 479 2. One way an implementation can satisfy the above constraints is to 480 associate with each temporary address a creation time (called 481 CREATION_TIME) that indicates the time at which the address was 482 created. When updating the preferred lifetime of an existing 483 temporary address, it would be set to expire at whichever time is 484 earlier: the time indicated by the received lifetime or 485 (CREATION_TIME + TEMP_PREFERRED_LIFETIME - DESYNC_FACTOR). A 486 similar approach can be used with the valid lifetime. 488 3. If the node has not configured any temporary address for the 489 corresponding prefix, the node SHOULD create a new temporary 490 address for such prefix. 492 Note: 493 For example, a host might implement prefix-specific policies 494 such as not configuring temporary addresses for the Unique 495 Local IPv6 Unicast Addresses (ULA) [RFC4193] prefix. 497 4. When creating a temporary address, the lifetime values MUST be 498 derived from the corresponding prefix as follows: 500 * Its Valid Lifetime is the lower of the Valid Lifetime of the 501 prefix and TEMP_VALID_LIFETIME 503 * Its Preferred Lifetime is the lower of the Preferred Lifetime 504 of the prefix and TEMP_PREFERRED_LIFETIME - DESYNC_FACTOR. 506 5. A temporary address is created only if this calculated Preferred 507 Lifetime is greater than REGEN_ADVANCE time units. In 508 particular, an implementation MUST NOT create a temporary address 509 with a zero Preferred Lifetime. 511 6. New temporary addresses MUST be created by appending a randomized 512 interface identifier (generates as described in Section 3.3 of 513 this document) to the prefix that was received. 515 7. The node MUST perform duplicate address detection (DAD) on the 516 generated temporary address. If DAD indicates the address is 517 already in use, the node MUST generate a new randomized interface 518 identifier, and repeat the previous steps as appropriate up to 519 TEMP_IDGEN_RETRIES times. If after TEMP_IDGEN_RETRIES 520 consecutive attempts no non-unique address was generated, the 521 node MUST log a system error and MUST NOT attempt to generate 522 temporary addresses for that interface. This allows hosts to 523 recover from occasional DAD failures, or otherwise log the 524 recurrent address collisions. 526 3.5. Expiration of Temporary Addresses 528 When a temporary address becomes deprecated, a new one MUST be 529 generated. This is done by repeating the actions described in 530 Section 3.4, starting at step 4). Note that, except for the 531 transient period when a temporary address is being regenerated, in 532 normal operation at most one temporary address per prefix should be 533 in a non-deprecated state at any given time on a given interface. 534 Note that if a temporary address becomes deprecated as result of 535 processing a Prefix Information Option with a zero Preferred 536 Lifetime, then a new temporary address MUST NOT be generated. To 537 ensure that a preferred temporary address is always available, a new 538 temporary address SHOULD be regenerated slightly before its 539 predecessor is deprecated. This is to allow sufficient time to avoid 540 race conditions in the case where generating a new temporary address 541 is not instantaneous, such as when duplicate address detection must 542 be run. The node SHOULD start the address regeneration process 543 REGEN_ADVANCE time units before a temporary address would actually be 544 deprecated. 546 As an optional optimization, an implementation MAY remove a 547 deprecated temporary address that is not in use by applications or 548 upper layers as detailed in Section 6. 550 3.6. Regeneration of Temporary Addresses 552 The frequency at which temporary addresses change depends on how a 553 device is being used (e.g., how frequently it initiates new 554 communication) and the concerns of the end user. The most egregious 555 privacy concerns appear to involve addresses used for long periods of 556 time (weeks to months to years). The more frequently an address 557 changes, the less feasible collecting or coordinating information 558 keyed on interface identifiers becomes. Moreover, the cost of 559 collecting information and attempting to correlate it based on 560 interface identifiers will only be justified if enough addresses 561 contain non-changing identifiers to make it worthwhile. Thus, having 562 large numbers of clients change their address on a daily or weekly 563 basis is likely to be sufficient to alleviate most privacy concerns. 565 There are also client costs associated with having a large number of 566 addresses associated with a node (e.g., in doing address lookups, the 567 need to join many multicast groups, etc.). Thus, changing addresses 568 frequently (e.g., every few minutes) may have performance 569 implications. 571 Nodes following this specification SHOULD generate new temporary 572 addresses on a periodic basis. This can be achieved by generating a 573 new temporary address at least once every (TEMP_PREFERRED_LIFETIME - 574 REGEN_ADVANCE - DESYNC_FACTOR) time units. As described above, 575 generating a new temporary address REGEN_ADVANCE time units before a 576 temporary address becomes deprecated produces addresses with a 577 preferred lifetime no larger than TEMP_PREFERRED_LIFETIME. The value 578 DESYNC_FACTOR is a random value (different for each client) that 579 ensures that clients don't synchronize with each other and generate 580 new addresses at exactly the same time. When the preferred lifetime 581 expires, a new temporary address MUST be generated using the new 582 randomized interface identifier. 584 Because the precise frequency at which it is appropriate to generate 585 new addresses varies from one environment to another, implementations 586 SHOULD provide end users with the ability to change the frequency at 587 which addresses are regenerated. The default value is given in 588 TEMP_PREFERRED_LIFETIME and is one day. In addition, the exact time 589 at which to invalidate a temporary address depends on how 590 applications are used by end users. Thus, the suggested default 591 value of two days (TEMP_VALID_LIFETIME) may not be appropriate in all 592 environments. Implementations SHOULD provide end users with the 593 ability to override both of these default values. 595 Finally, when an interface connects to a new (different) link, a new 596 set of temporary addresses MUST be generated immediately for use on 597 the new link. If a device moves from one link to another, generating 598 a new set of temporary addresses ensures that the device uses 599 different randomized interface identifiers for the temporary 600 addresses associated with the two links, making it more difficult to 601 correlate addresses from the two different links as being from the 602 same node. The node MAY follow any process available to it, to 603 determine that the link change has occurred. One such process is 604 described by "Simple Procedures for Detecting Network Attachment in 605 IPv6" [RFC6059]. Detecting link changes would prevent link down/up 606 events from causing temporary addresses to be (unnecessarily) 607 regenerated. 609 3.7. Implementation Considerations 611 Devices implementing this specification MUST provide a way for the 612 end user to explicitly enable or disable the use of temporary 613 addresses. In addition, a site might wish to disable the use of 614 temporary addresses in order to simplify network debugging and 615 operations. Consequently, implementations SHOULD provide a way for 616 trusted system administrators to enable or disable the use of 617 temporary addresses. 619 Additionally, sites might wish to selectively enable or disable the 620 use of temporary addresses for some prefixes. For example, a site 621 might wish to disable temporary address generation for "Unique local" 622 [RFC4193] prefixes while still generating temporary addresses for all 623 other global prefixes. Another site might wish to enable temporary 624 address generation only for the prefixes 2001:db8:1::/48 and 625 2001:db8:2::/48 while disabling it for all other prefixes. To 626 support this behavior, implementations SHOULD provide a way to enable 627 and disable generation of temporary addresses for specific prefix 628 subranges. This per-prefix setting SHOULD override the global 629 settings on the node with respect to the specified prefix subranges. 630 Note that the per-prefix setting can be applied at any granularity, 631 and not necessarily on a per subnet basis. 633 Use of the extensions defined in this document may complicate 634 debugging and other operational troubleshooting activities. 635 Consequently, it may be site policy that temporary addresses should 636 not be used. Consequently, implementations MUST provide a method for 637 the end user or trusted administrator to override the use of 638 temporary addresses. 640 3.8. Defined Constants 642 Constants defined in this document include: 644 TEMP_VALID_LIFETIME -- Default value: 2 days. Users should be able 645 to override the default value. 647 TEMP_PREFERRED_LIFETIME -- Default value: 1 day. Users should be 648 able to override the default value. 650 REGEN_ADVANCE -- 5 seconds 652 MAX_DESYNC_FACTOR -- 10 minutes. Upper bound on DESYNC_FACTOR. 654 DESYNC_FACTOR -- A random value within the range 0 - 655 MAX_DESYNC_FACTOR. It is computed once at system start (rather than 656 each time it is used) and must never be greater than 657 (TEMP_PREFERRED_LIFETIME - REGEN_ADVANCE). 659 TEMP_IDGEN_RETRIES -- Default value: 3 661 4. Implications of Changing Interface Identifiers 663 The desires of protecting individual privacy versus the desire to 664 effectively maintain and debug a network can conflict with each 665 other. Having clients use addresses that change over time will make 666 it more difficult to track down and isolate operational problems. 667 For example, when looking at packet traces, it could become more 668 difficult to determine whether one is seeing behavior caused by a 669 single errant machine, or by a number of them. 671 Network deployments are currently recommended to provide multiple 672 IPv6 addresses from each prefix to general-purpose hosts [RFC7934]. 673 However, in some scenarios, use of a large number of IPv6 addresses 674 may have negative implications on network devices that need to 675 maintain entries for each IPv6 address in some data structures (e.g., 676 [RFC7039]). Additionally, concurrent active use of multiple IPv6 677 addresses will increase neighbour discovery traffic if Neighbour 678 Caches in network devices are not large enough to store all addresses 679 on the link. This can impact performance and energy efficiency on 680 networks on which multicast is expensive (e.g. 681 [I-D.ietf-mboned-ieee802-mcast-problems]). 683 The use of temporary addresses may cause unexpected difficulties with 684 some applications. For example, some servers refuse to accept 685 communications from clients for which they cannot map the IP address 686 into a DNS name. That is, they perform a DNS PTR query to determine 687 the DNS name, and may then also perform an AAAA query on the returned 688 name to verify that the returned DNS name maps back into the address 689 being used. Consequently, clients not properly registered in the DNS 690 may be unable to access some services. However, a node's DNS name 691 (if non-changing) would serve as a constant identifier. The wide 692 deployment of the extension described in this document could 693 challenge the practice of inverse-DNS-based "validation", which has 694 little validity, though it is widely implemented. In order to meet 695 server challenges, nodes could register temporary addresses in the 696 DNS using random names (for example, a string version of the random 697 address itself), albeit at the expense of increased complexity. 699 In addition, some applications may not behave robustly if temporary 700 addresses are used and an address expires before the application has 701 terminated, or if it opens multiple sessions, but expects them to all 702 use the same addresses. 704 5. Significant Changes from RFC4941 706 This section summarizes the changes in this document relative to RFC 707 4941 that an implementer of RFC 4941 should be aware of. 709 Broadly speaking, this document introduces the following changes: 711 o Addresses a number of flaws in the algorithm for generating 712 temporary addresses: The aforementioned flaws include the use of 713 MD5 for computing the temporary IIDs, and reusing the same IID for 714 multiple prefixes (see [RAID2015] and [RFC7721] for further 715 details). 717 o Allows hosts to employ only temporary addresses: 718 [RFC4941] assumed that temporary addresses were configured in 719 addition to stable addresses. This document does not imply or 720 require the configuration of stable addresses, and thus 721 implementations can now configure both stable and temporary 722 addresses, or temporary addresses only. 724 o Removes the recommendation that temporary addresses be disabled by 725 default: 726 This is in line with BCP188 ([RFC7258]), and also with BCP204 727 ([RFC7934]). 729 o Reduces the default Valid Lifetime for temporary addresses: 730 The default Valid Lifetime for temporary addresses has been 731 reduced from 1 week to 2 days, decreasing the typical number of 732 concurrent temporary addresses from 7 to 2. This reduces the 733 possible stress on network elements (see Section 4 for further 734 details). 736 o Addresses all errata submitted for [RFC4941]. 738 6. Future Work 740 An implementation might want to keep track of which addresses are 741 being used by upper layers so as to be able to remove a deprecated 742 temporary address from internal data structures once no upper layer 743 protocols are using it (but not before). This is in contrast to 744 current approaches where addresses are removed from an interface when 745 they become invalid [RFC4862], independent of whether or not upper 746 layer protocols are still using them. For TCP connections, such 747 information is available in control blocks. For UDP-based 748 applications, it may be the case that only the applications have 749 knowledge about what addresses are actually in use. Consequently, an 750 implementation generally will need to use heuristics in deciding when 751 an address is no longer in use. 753 7. Security Considerations 755 If a very small number of nodes (say, only one) use a given prefix 756 for extended periods of time, just changing the interface identifier 757 part of the address may not be sufficient to mitigate address-based 758 network activity correlation, since the prefix acts as a constant 759 identifier. The procedures described in this document are most 760 effective when the prefix is reasonably non static or is used by a 761 fairly large number of nodes. Additionally, if a temporary address 762 is used in a session where the user authenticates, any notion of 763 "privacy" for that address is compromised. 765 While this document discusses ways of obscuring a user's IP address, 766 the method described is believed to be ineffective against 767 sophisticated forms of traffic analysis. To increase effectiveness, 768 one may need to consider the use of more advanced techniques, such as 769 Onion Routing [ONION]. 771 Ingress filtering has been and is being deployed as a means of 772 preventing the use of spoofed source addresses in Distributed Denial 773 of Service (DDoS) attacks. In a network with a large number of 774 nodes, new temporary addresses are created at a fairly high rate. 775 This might make it difficult for ingress filtering mechanisms to 776 distinguish between legitimately changing temporary addresses and 777 spoofed source addresses, which are "in-prefix" (using a 778 topologically correct prefix and non-existent interface ID). This 779 can be addressed by using access control mechanisms on a per-address 780 basis on the network egress point. 782 8. Acknowledgments 784 The authors would like to thank (in alphabetical order) Fred Baker, 785 Brian Carpenter, Tim Chown, Lorenzo Colitti, David Farmer, Tom 786 Herbert, Bob Hinden, Christian Huitema, Erik Kline, Gyan Mishra, Dave 787 Plonka, Michael Richardson, Mark Smith, Pascal Thubert, Ole Troan, 788 Johanna Ullrich, and Timothy Winters, for providing valuable comments 789 on earlier versions of this document. 791 This document incorporates errata submitted for [RFC4941] by Jiri 792 Bohac and Alfred Hoenes. 794 This document is based on [RFC4941] (a revision of RFC3041). Suresh 795 Krishnan was the sole author of RFC4941. He would like to 796 acknowledge the contributions of the IPv6 working group and, in 797 particular, Jari Arkko, Pekka Nikander, Pekka Savola, Francis Dupont, 798 Brian Haberman, Tatuya Jinmei, and Margaret Wasserman for their 799 detailed comments. 801 Rich Draves and Thomas Narten were the authors of RFC 3041. They 802 would like to acknowledge the contributions of the IPv6 working group 803 and, in particular, Ran Atkinson, Matt Crawford, Steve Deering, 804 Allison Mankin, and Peter Bieringer. 806 9. References 808 9.1. Normative References 810 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 811 Requirement Levels", BCP 14, RFC 2119, 812 DOI 10.17487/RFC2119, March 1997, 813 . 815 [RFC4086] Eastlake 3rd, D., Schiller, J., and S. Crocker, 816 "Randomness Requirements for Security", BCP 106, RFC 4086, 817 DOI 10.17487/RFC4086, June 2005, 818 . 820 [RFC4193] Hinden, R. and B. Haberman, "Unique Local IPv6 Unicast 821 Addresses", RFC 4193, DOI 10.17487/RFC4193, October 2005, 822 . 824 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 825 Architecture", RFC 4291, DOI 10.17487/RFC4291, February 826 2006, . 828 [RFC4862] Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless 829 Address Autoconfiguration", RFC 4862, 830 DOI 10.17487/RFC4862, September 2007, 831 . 833 [RFC4941] Narten, T., Draves, R., and S. Krishnan, "Privacy 834 Extensions for Stateless Address Autoconfiguration in 835 IPv6", RFC 4941, DOI 10.17487/RFC4941, September 2007, 836 . 838 [RFC5453] Krishnan, S., "Reserved IPv6 Interface Identifiers", 839 RFC 5453, DOI 10.17487/RFC5453, February 2009, 840 . 842 [RFC6724] Thaler, D., Ed., Draves, R., Matsumoto, A., and T. Chown, 843 "Default Address Selection for Internet Protocol Version 6 844 (IPv6)", RFC 6724, DOI 10.17487/RFC6724, September 2012, 845 . 847 [RFC7136] Carpenter, B. and S. Jiang, "Significance of IPv6 848 Interface Identifiers", RFC 7136, DOI 10.17487/RFC7136, 849 February 2014, . 851 [RFC7217] Gont, F., "A Method for Generating Semantically Opaque 852 Interface Identifiers with IPv6 Stateless Address 853 Autoconfiguration (SLAAC)", RFC 7217, 854 DOI 10.17487/RFC7217, April 2014, 855 . 857 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 858 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 859 May 2017, . 861 [RFC8190] Bonica, R., Cotton, M., Haberman, B., and L. Vegoda, 862 "Updates to the Special-Purpose IP Address Registries", 863 BCP 153, RFC 8190, DOI 10.17487/RFC8190, June 2017, 864 . 866 9.2. Informative References 868 [FIPS-SHS] 869 NIST, "Secure Hash Standard (SHS)", FIPS 870 Publication 180-4, August 2015, 871 . 874 [I-D.ietf-mboned-ieee802-mcast-problems] 875 Perkins, C., McBride, M., Stanley, D., Kumari, W., and J. 876 Zuniga, "Multicast Considerations over IEEE 802 Wireless 877 Media", draft-ietf-mboned-ieee802-mcast-problems-11 (work 878 in progress), December 2019. 880 [IANA-RESERVED-IID] 881 IANA, "Reserved IPv6 Interface Identifiers", 882 . 884 [ONION] Reed, MGR., Syverson, PFS., and DMG. Goldschlag, "Proxies 885 for Anonymous Routing", Proceedings of the 12th Annual 886 Computer Security Applications Conference, San Diego, CA, 887 December 1996. 889 [OPEN-GROUP] 890 The Open Group, "The Open Group Base Specifications Issue 891 7 / IEEE Std 1003.1-2008, 2016 Edition", 892 Section 4.16 Seconds Since the Epoch, 2016, 893 . 896 [RAID2015] 897 Ullrich, J. and E. Weippl, "Privacy is Not an Option: 898 Attacking the IPv6 Privacy Extension", International 899 Symposium on Recent Advances in Intrusion Detection 900 (RAID), 2015, . 903 [RFC1321] Rivest, R., "The MD5 Message-Digest Algorithm", RFC 1321, 904 DOI 10.17487/RFC1321, April 1992, 905 . 907 [RFC5014] Nordmark, E., Chakrabarti, S., and J. Laganier, "IPv6 908 Socket API for Source Address Selection", RFC 5014, 909 DOI 10.17487/RFC5014, September 2007, 910 . 912 [RFC6059] Krishnan, S. and G. Daley, "Simple Procedures for 913 Detecting Network Attachment in IPv6", RFC 6059, 914 DOI 10.17487/RFC6059, November 2010, 915 . 917 [RFC6151] Turner, S. and L. Chen, "Updated Security Considerations 918 for the MD5 Message-Digest and the HMAC-MD5 Algorithms", 919 RFC 6151, DOI 10.17487/RFC6151, March 2011, 920 . 922 [RFC6265] Barth, A., "HTTP State Management Mechanism", RFC 6265, 923 DOI 10.17487/RFC6265, April 2011, 924 . 926 [RFC7039] Wu, J., Bi, J., Bagnulo, M., Baker, F., and C. Vogt, Ed., 927 "Source Address Validation Improvement (SAVI) Framework", 928 RFC 7039, DOI 10.17487/RFC7039, October 2013, 929 . 931 [RFC7258] Farrell, S. and H. Tschofenig, "Pervasive Monitoring Is an 932 Attack", BCP 188, RFC 7258, DOI 10.17487/RFC7258, May 933 2014, . 935 [RFC7421] Carpenter, B., Ed., Chown, T., Gont, F., Jiang, S., 936 Petrescu, A., and A. Yourtchenko, "Analysis of the 64-bit 937 Boundary in IPv6 Addressing", RFC 7421, 938 DOI 10.17487/RFC7421, January 2015, 939 . 941 [RFC7624] Barnes, R., Schneier, B., Jennings, C., Hardie, T., 942 Trammell, B., Huitema, C., and D. Borkmann, 943 "Confidentiality in the Face of Pervasive Surveillance: A 944 Threat Model and Problem Statement", RFC 7624, 945 DOI 10.17487/RFC7624, August 2015, 946 . 948 [RFC7707] Gont, F. and T. Chown, "Network Reconnaissance in IPv6 949 Networks", RFC 7707, DOI 10.17487/RFC7707, March 2016, 950 . 952 [RFC7721] Cooper, A., Gont, F., and D. Thaler, "Security and Privacy 953 Considerations for IPv6 Address Generation Mechanisms", 954 RFC 7721, DOI 10.17487/RFC7721, March 2016, 955 . 957 [RFC7934] Colitti, L., Cerf, V., Cheshire, S., and D. Schinazi, 958 "Host Address Availability Recommendations", BCP 204, 959 RFC 7934, DOI 10.17487/RFC7934, July 2016, 960 . 962 Authors' Addresses 964 Fernando Gont 965 SI6 Networks / UTN-FRH 966 Evaristo Carriego 2644 967 Haedo, Provincia de Buenos Aires 1706 968 Argentina 970 Phone: +54 11 4650 8472 971 Email: fgont@si6networks.com 972 URI: https://www.si6networks.com 974 Suresh Krishnan 975 Ericsson Research 976 8400 Decarie Blvd. 977 Town of Mount Royal, QC 978 Canada 980 Email: suresh.krishnan@ericsson.com 981 Thomas Narten 982 IBM Corporation 983 P.O. Box 12195 984 Research Triangle Park, NC 985 USA 987 Email: narten@us.ibm.com 989 Richard Draves 990 Microsoft Research 991 One Microsoft Way 992 Redmond, WA 993 USA 995 Email: richdr@microsoft.com