idnits 2.17.1 draft-ietf-intarea-ipv4-id-update-06.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 : ---------------------------------------------------------------------------- == There are 2 instances of lines with non-RFC6890-compliant IPv4 addresses in the document. If these are example addresses, they should be changed. -- The draft header indicates that this document updates RFC2003, but the abstract doesn't seem to directly say this. It does mention RFC2003 though, so this could be OK. -- The draft header indicates that this document updates RFC1122, but the abstract doesn't seem to directly say this. It does mention RFC1122 though, so this could be OK. -- The draft header indicates that this document updates RFC791, but the abstract doesn't seem to directly say this. It does mention RFC791 though, so this could be OK. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year (Using the creation date from RFC791, updated by this document, for RFC5378 checks: 1981-09-01) -- The document seems to contain a disclaimer for pre-RFC5378 work, and may have content which was first submitted before 10 November 2008. The disclaimer is necessary when there are original authors that you have been unable to contact, or if some do not wish to grant the BCP78 rights to the IETF Trust. If you are able to get all authors (current and original) to grant those rights, you can and should remove the disclaimer; otherwise, the disclaimer is needed and you can ignore this comment. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (October 9, 2012) is 4211 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) -- Looks like a reference, but probably isn't: '10' on line 652 -- Obsolete informational reference (is this intentional?): RFC 2460 (Obsoleted by RFC 8200) -- Obsolete informational reference (is this intentional?): RFC 2671 (Obsoleted by RFC 6891) -- Obsolete informational reference (is this intentional?): RFC 4960 (Obsoleted by RFC 9260) -- Obsolete informational reference (is this intentional?): RFC 6145 (Obsoleted by RFC 7915) Summary: 0 errors (**), 0 flaws (~~), 2 warnings (==), 10 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 1 Internet Area WG J. Touch 2 Internet Draft USC/ISI 3 Updates: 791,1122,2003 October 9, 2012 4 Intended status: Proposed Standard 5 Expires: April 2013 7 Updated Specification of the IPv4 ID Field 8 draft-ietf-intarea-ipv4-id-update-06.txt 10 Status of this Memo 12 This Internet-Draft is submitted to IETF in full conformance with the 13 provisions of BCP 78 and BCP 79. 15 This document may contain material from IETF Documents or IETF 16 Contributions published or made publicly available before November 17 10, 2008. The person(s) controlling the copyright in some of this 18 material may not have granted the IETF Trust the right to allow 19 modifications of such material outside the IETF Standards Process. 20 Without obtaining an adequate license from the person(s) controlling 21 the copyright in such materials, this document may not be modified 22 outside the IETF Standards Process, and derivative works of it may 23 not be created outside the IETF Standards Process, except to format 24 it for publication as an RFC or to translate it into languages other 25 than English. 27 Internet-Drafts are working documents of the Internet Engineering 28 Task Force (IETF), its areas, and its working groups. Note that 29 other groups may also distribute working documents as Internet- 30 Drafts. 32 Internet-Drafts are draft documents valid for a maximum of six months 33 and may be updated, replaced, or obsoleted by other documents at any 34 time. It is inappropriate to use Internet-Drafts as reference 35 material or to cite them other than as "work in progress." 37 The list of current Internet-Drafts can be accessed at 38 http://www.ietf.org/ietf/1id-abstracts.txt 40 The list of Internet-Draft Shadow Directories can be accessed at 41 http://www.ietf.org/shadow.html 43 This Internet-Draft will expire on April 9, 2013. 45 Copyright Notice 47 Copyright (c) 2012 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 (http://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 Abstract 62 The IPv4 Identification (ID) field enables fragmentation and 63 reassembly, and as currently specified is required to be unique 64 within the maximum lifetime for all datagrams with a given 65 source/destination/protocol tuple. If enforced, this uniqueness 66 requirement would limit all connections to 6.4 Mbps. Because 67 individual connections commonly exceed this speed, it is clear that 68 existing systems violate the current specification. This document 69 updates the specification of the IPv4 ID field in RFC791, RFC1122, 70 and RFC2003 to more closely reflect current practice and to more 71 closely match IPv6 so that the field's value is defined only when a 72 datagram is actually fragmented. It also discusses the impact of 73 these changes on how datagrams are used. 75 Table of Contents 77 1. Introduction...................................................3 78 2. Conventions used in this document..............................3 79 3. The IPv4 ID Field..............................................4 80 3.1. Uses of the IPv4 ID Field.................................4 81 3.2. Background on IPv4 ID Reassembly Issues...................5 82 4. Updates to the IPv4 ID Specification...........................6 83 4.1. IPv4 ID Used Only for Fragmentation.......................7 84 4.2. Encourage Safe IPv4 ID Use................................8 85 4.3. IPv4 ID Requirements That Persist.........................8 86 5. Impact of Proposed Changes.....................................9 87 5.1. Impact on Legacy Internet Devices.........................9 88 5.2. Impact on Datagram Generation............................10 89 5.3. Impact on Middleboxes....................................11 90 5.3.1. Rewriting Middleboxes...............................11 91 5.3.2. Filtering Middleboxes...............................12 92 5.4. Impact on Header Compression.............................13 93 6. Updates to Existing Standards.................................13 94 6.1. Updates to RFC 791.......................................13 95 6.2. Updates to RFC 1122......................................14 96 6.3. Updates to RFC 2003......................................15 97 7. Security Considerations.......................................15 98 8. IANA Considerations...........................................15 99 9. References....................................................16 100 9.1. Normative References.....................................16 101 9.2. Informative References...................................16 102 10. Acknowledgments..............................................18 104 1. Introduction 106 In IPv4, the Identification (ID) field is a 16-bit value that is 107 unique for every datagram for a given source address, destination 108 address, and protocol, such that it does not repeat within the 109 maximum datagram lifetime (MDL) [RFC791][RFC1122]. As currently 110 specified, all datagrams between a source and destination of a given 111 protocol must have unique IPv4 ID values over a period of this MDL, 112 which is typically interpreted as two minutes, and is related to the 113 recommended reassembly timeout [RFC1122]. This uniqueness is 114 currently specified as for all datagrams, regardless of fragmentation 115 settings. 117 Uniqueness of the IPv4 ID is commonly violated by high speed devices; 118 if strictly enforced, it would limit the speed of a single protocol 119 between two IP endpoints to 6.4 Mbps for typical MTUs of 1500 bytes 120 [RFC4963]. It is common for a single connection to operate far in 121 excess of these rates, which strongly indicates that the uniqueness 122 of the IPv4 ID as specified is already moot. Further, some sources 123 have been generating non-varying IPv4 IDs for many years (e.g., 124 cellphones), which resulted in support for such in ROHC [RFC5225]. 126 This document updates the specification of the IPv4 ID field to more 127 closely reflect current practice, and to include considerations taken 128 into account during the specification of the similar field in IPv6. 130 2. Conventions used in this document 132 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 133 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 134 document are to be interpreted as described in RFC-2119 [RFC2119]. 136 In this document, the characters ">>" proceeding an indented line(s) 137 indicates a requirement using the key words listed above. This 138 convention aids reviewers in quickly identifying or finding this 139 document's explicit requirements. 141 3. The IPv4 ID Field 143 IP supports datagram fragmentation, where large datagrams are split 144 into smaller components to traverse links with limited maximum 145 transmission units (MTUs). Fragments are indicated in different ways 146 in IPv4 and IPv6: 148 o In IPv4, fragments are indicated using four fields of the basic 149 header: Identification (ID), Fragment Offset, a "Don't Fragment" 150 flag (DF), and a "More Fragments" flag (MF) [RFC791] 152 o In IPv6, fragments are indicated in an extension header that 153 includes an ID, Fragment Offset, and M (more fragments) flag 154 similar to their counterparts in IPv4 [RFC2460] 156 IPv4 and IPv6 fragmentation differs in a few important ways. IPv6 157 fragmentation occurs only at the source, so a DF bit is not needed to 158 prevent downstream devices from initiating fragmentation (i.e., IPv6 159 always acts as if DF=1). The IPv6 fragment header is present only 160 when a datagram has been fragmented, or when the source has received 161 a "packet too big" ICMPv6 error message indicating that the path 162 cannot support the required minimum 1280-byte IPv6 MTU and is thus 163 subject to translation [RFC2460][RFC4443]. The latter case is 164 relevant only for IPv6 datagrams sent to IPv4 destinations to support 165 subsequent fragmentation after translation to IPv4. 167 With the exception of these two cases, the ID field is not present 168 for non-fragmented datagrams, and thus is meaningful only for 169 datagrams that are already fragmented or datagrams intended to be 170 fragmented as part of IPv4 translation. Finally, the IPv6 ID field is 171 32 bits, and required unique per source/destination address pair for 172 IPv6, whereas for IPv4 it is only 16 bits and required unique per 173 source/destination/protocol triple. 175 This document focuses on the IPv4 ID field issues, because in IPv6 176 the field is larger and present only in fragments. 178 3.1. Uses of the IPv4 ID Field 180 The IPv4 ID field was originally intended for fragmentation and 181 reassembly [RFC791]. Within a given source address, destination 182 address, and protocol, fragments of an original datagram are matched 183 based on their IPv4 ID. This requires that IDs are unique within the 184 address/protocol triple when fragmentation is possible (e.g., DF=0) 185 or when it has already occurred (e.g., frag_offset>0 or MF=1). 187 Other uses have been envisioned for the IPv4 ID field. The field has 188 been proposed as a way to detect and remove duplicate datagrams, 189 e.g., at congested routers (noted in Sec. 3.2.1.5 of [RFC1122]) or in 190 network accelerators. It has similarly been proposed for use at end 191 hosts to reduce the impact of duplication on higher-layer protocols 192 (e.g., additional processing in TCP, or the need for application- 193 layer duplicate suppression in UDP). This is also discussed further 194 in Section 5.1. 196 The IPv4 ID field is used in some diagnostic tools to correlate 197 datagrams measured at various locations along a network path. This is 198 already insufficient in IPv6 because unfragmented datagrams lack an 199 ID, so these tools are already being updated to avoid such reliance 200 on the ID field. This is also discussed further in Section 5.1. 202 The ID clearly needs to be unique (within MDL, within the 203 src/dst/protocol tuple) to support fragmentation and reassembly, but 204 not all datagrams are fragmented or allow fragmentation. This 205 document deprecates non-fragmentation uses, allowing the ID to be 206 repeated (within MDL, within the src/dst/protocol tuple) in those 207 cases. 209 3.2. Background on IPv4 ID Reassembly Issues 211 The following is a summary of issues with IPv4 fragment reassembly in 212 high speed environments raised previously [RFC4963]. Readers are 213 encouraged to consult RFC 4963 for a more detailed discussion of 214 these issues. 216 With the maximum IPv4 datagram size of 64KB, a 16-bit ID field that 217 does not repeat within 120 seconds means that the aggregate of all 218 TCP connections of a given protocol between two IP endpoints is 219 limited to roughly 286 Mbps; at a more typical MTU of 1500 bytes, 220 this speed drops to 6.4 Mbps [RFC791][RFC1122][RFC4963]. This limit 221 currently applies for all IPv4 datagrams within a single protocol 222 (i.e., the IPv4 protocol field) between two IP addresses, regardless 223 of whether fragmentation is enabled or inhibited, and whether a 224 datagram is fragmented or not. 226 IPv6, even at typical MTUs, is capable of 18.7 Tbps with 227 fragmentation between two IP endpoints as an aggregate across all 228 protocols, due to the larger 32-bit ID field (and the fact that the 229 IPv6 next-header field, the equivalent of the IPv4 protocol field, is 230 not considered in differentiating fragments). When fragmentation is 231 not used the field is absent, and in that case IPv6 speeds are not 232 limited by the ID field uniqueness. 234 Note also that 120 seconds is only an estimate on the MDL. It is 235 related to the reassembly timeout as a lower bound and the TCP 236 Maximum Segment Lifetime as an upper bound (both as noted in 237 [RFC1122]). Network delays are incurred in other ways, e.g., 238 satellite links, which can add seconds of delay even though the TTL 239 is not decremented by a corresponding amount. There is thus no 240 enforcement mechanism to ensure that datagrams older than 120 seconds 241 are discarded. 243 Wireless Internet devices are frequently connected at speeds over 54 244 Mbps, and wired links of 1 Gbps have been the default for several 245 years. Although many end-to-end transport paths are congestion 246 limited, these devices easily achieve 100+ Mbps application-layer 247 throughput over LANs (e.g., disk-to-disk file transfer rates), and 248 numerous throughput demonstrations with COTS systems over wide-area 249 paths exhibit these speeds for over a decade. This strongly suggests 250 that IPv4 ID uniqueness has been moot for a long time. 252 4. Updates to the IPv4 ID Specification 254 This document updates the specification of the IPv4 ID field in three 255 distinct ways, as discussed in subsequent subsections: 257 o Use the IPv4 ID field only for fragmentation 259 o Avoiding a performance impact when the IPv4 ID field is used 261 o Encourage safe operation when the IPv4 ID field is used 263 There are two kinds of datagrams used in the following discussion, 264 named as follows: 266 o Atomic datagrams are datagrams not yet fragmented and for which 267 further fragmentation has been inhibited. 269 o Non-atomic datagrams are datagrams that either already have been 270 fragmented or for which fragmentation remains possible. 272 This same definition can be expressed in pseudo code as using common 273 logical operators (equals is ==, logical 'and' is &&, logical 'or' is 274 ||, greater than is >, and parenthesis function typically) as: 276 o Atomic datagrams: (DF==1)&&(MF==0)&&(frag_offset==0) 277 o Non-atomic datagrams: (DF==0)||(MF==1)||(frag_offset>0) 279 The test for non-atomic datagrams is the logical negative of the test 280 for atomic datagrams, thus all possibilities are considered. 282 4.1. IPv4 ID Used Only for Fragmentation 284 Although RFC1122 suggests the IPv4 ID field has other uses, including 285 datagram de-duplication, such uses are already not interoperable with 286 known implementations of sources that do not vary their ID. This 287 document thus defines this field's value only for fragmentation and 288 reassembly: 290 >> IPv4 ID field MUST NOT be used for purposes other than 291 fragmentation and reassembly. 293 Datagram de-duplication is accomplished using hash-based duplicate 294 detection for cases where the ID field is absent (IPv6 unfragmented 295 datagrams), which can also be applied to IPv4 atomic datagrams 296 without utilizing the ID field [RFC6621]. 298 In atomic datagrams, the IPv4 ID field has no meaning, and thus can 299 be set to an arbitrary value, i.e., the requirement for non-repeating 300 IDs within the address/protocol triple is no longer required for 301 atomic datagrams: 303 >> Originating sources MAY set the IPv4 ID field of atomic datagrams 304 to any value. 306 Second, all network nodes, whether at intermediate routers, 307 destination hosts, or other devices (e.g., NATs and other address 308 sharing mechanisms, firewalls, tunnel egresses), cannot rely on the 309 field: 311 >> All devices that examine IPv4 headers MUST ignore the IPv4 ID 312 field of atomic datagrams. 314 The IPv4 ID field is thus meaningful only for non-atomic datagrams - 315 datagrams that have either already been fragmented, or those for 316 which fragmentation remains permitted. Atomic datagrams are detected 317 by their DF, MF, and fragmentation offset fields as explained in 318 Section 4, because such a test is completely backward compatible; 319 this document thus does not reserve any IPv4 ID values, including 0, 320 as distinguished. 322 Deprecating the use of the IPv4 ID field for non-reassembly uses 323 should have little - if any - impact. IPv4 IDs are already frequently 324 repeated, e.g., over even moderately fast connections and from some 325 sources that do not vary the ID at all, and no adverse impact has 326 been observed. Duplicate suppression was suggested [RFC1122] and has 327 been implemented in some protocol accelerators, but no impacts of 328 IPv4 ID reuse have been noted to date. Routers are not required to 329 issue ICMPs on any particular timescale, and so IPv4 ID repetition 330 should not have been used for validation and has not been observed, 331 and again repetition already occurs and would have been noticed 332 [RFC1812]. ICMP relaying at tunnel ingresses is specified to use soft 333 state rather than a datagram cache, and should have been noted if the 334 latter for similar reasons [RFC2003]. These and other legacy issues 335 are discussed further in Section 5.1. 337 4.2. Encourage Safe IPv4 ID Use 339 This document makes further changes to the specification of the IPv4 340 ID field and its use to encourage its safe use as corollary 341 requirements changes as follows. 343 RFC 1122 discusses that if TCP retransmits a segment it may be 344 possible to reuse the IPv4 ID (see Section 6.2). This can make it 345 difficult for a source to avoid IPv4 ID repetition for received 346 fragments. RFC 1122 concludes that this behavior "is not useful"; 347 this document formalizes that conclusion as follows: 349 >> The IPv4 ID of non-atomic datagrams MUST NOT be reused when 350 sending a copy of an earlier non-atomic datagram. 352 RFC 1122 also suggests that fragments can overlap [RFC1122]. Such 353 overlap can occur if successive retransmissions are fragmented in 354 different ways but with the same reassembly IPv4 ID. This overlap is 355 noted as the result of reusing IPv4 IDs when retransmitting 356 datagrams, which this document deprecates. However, it is also the 357 result of in-network datagram duplication, which can still occur. As 358 a result this document does not change the need to support 359 overlapping fragments. 361 4.3. IPv4 ID Requirements That Persist 363 This document does not relax the IPv4 ID field uniqueness 364 requirements of [RFC791] for non-atomic datagrams, i.e.: 366 >> Sources emitting non-atomic datagrams MUST NOT repeat IPv4 ID 367 values within one MDL for a given source address/destination 368 address/protocol triple. 370 Such sources include originating hosts, tunnel ingresses, and NATs 371 (including other address sharing mechanisms) (see Section 5.3). 373 This document does not relax the requirement that all network devices 374 honor the DF bit, i.e.: 376 >> IPv4 datagrams whose DF=1 MUST NOT be fragmented. 378 >> IPv4 datagram transit devices MUST NOT clear the DF bit. 380 In specific, DF=1 prevents fragmenting atomic datagrams. DF=1 also 381 prevents further fragmenting received fragments. In-network 382 fragmentation is permitted only when DF=0; this document does not 383 change that requirement. 385 5. Impact of Proposed Changes 387 This section discusses the impact of the proposed changes on legacy 388 devices, datagram generation in updated devices, middleboxes, and 389 header compression. 391 5.1. Impact on Legacy Internet Devices 393 Legacy uses of the IPv4 ID field consist of fragment generation, 394 fragment reassembly, duplicate datagram detection, and "other" uses. 396 Current devices already generate ID values that are reused within the 397 source address, destination address, protocol, and ID tuple in less 398 than the current estimated Internet MDL of two minutes. They assume 399 that the MDL over their end-to-end path is much lower. 401 Existing devices have been known to generate non-varying IDs for 402 atomic datagrams for nearly a decade, notably some cell phones. Such 403 constant ID values are the reason for their support as an 404 optimization of ROHC [RFC5225]. This is discussed further in Section 405 5.4. Generation of IPv4 datagrams with constant (zero) IDs is also 406 described as part of the IP/ICMP translation standard [RFC6145]. 408 Many current devices support fragmentation that ignores the IPv4 409 Don't Fragment (DF) bit. Such devices already transit traffic from 410 sources that reuse the ID. If fragments of different datagrams 411 reusing the same ID (within the source/destination/protocol tuple) 412 arrive at the destination interleaved, fragmentation would fail and 413 traffic would be dropped. Either such interleaving is uncommon, or 414 traffic from such devices is not widely traversing these DF-ignoring 415 devices, because significant occurrence of reassembly errors has not 416 been reported. DF-ignoring devices do not comply with existing 417 standards, and it is not feasible to update the standards to allow 418 them as compliant. 420 The ID field has been envisioned for use in duplicate detection, as 421 discussed in Section 4.1 [RFC1122]. Although this document now allows 422 IPv4 ID reuse for atomic datagrams, such reuse is already common (as 423 noted above). Protocol accelerators are known to implement IPv4 424 duplicate detection, but such devices are also known to violate other 425 Internet standards to achieve higher end-to-end performance. These 426 devices would already exhibit erroneous drops for this current 427 traffic, and this has not been reported. 429 There are other potential uses of the ID field, such as for 430 diagnostic purposes. Such uses already need to accommodate atomic 431 datagrams with reused ID fields. There are no reports of such uses 432 having problems with current datagrams that reuse IDs. These and any 433 other uses of the ID field are encouraged to apply IPv6-compatible 434 methods for IPv4 as well. 436 Thus, as a result of previous requirements, this document recommends 437 that IPv4 duplicate detection and diagnostic mechanisms apply IPv6- 438 compatible methods, i.e., that do not rely on the ID field (e.g., as 439 suggested in [RFC6621]). This is a consequence of using the ID field 440 only for reassembly, as well as the known hazard of existing devices 441 already reusing the ID field. 443 5.2. Impact on Datagram Generation 445 The following is a summary of the recommendations that are the result 446 of the previous changes to the IPv4 ID field specification. 448 Because atomic datagrams can use arbitrary IPv4 ID values, the ID 449 field no longer imposes a performance impact in those cases. However, 450 the performance impact remains for non-atomic datagrams. As a result: 452 >> Sources of non-atomic IPv4 datagrams MUST rate-limit their output 453 to comply with the ID uniqueness requirements. 455 Such sources include, in particular, DNS over UDP [RFC2671]. 457 Because there is no strict definition of the MDL, reassembly hazards 458 exist regardless of the IPv4 ID reuse interval or the reassembly 459 timeout. As a result: 461 >> Higher layer protocols SHOULD verify the integrity of IPv4 462 datagrams, e.g., using a checksum or hash that can detect reassembly 463 errors (the UDP checksum is weak in this regard, but better than 464 nothing). 466 Additional integrity checks can be employed using tunnels, as 467 supported by SEAL, IPsec, or SCTP [RFC4301][RFC4960][RFC5320]. Such 468 checks can avoid the reassembly hazards that can occur when using UDP 469 and TCP checksums [RFC4963], or when using partial checksums as in 470 UDP-Lite [RFC3828]. Because such integrity checks can avoid the 471 impact of reassembly errors: 473 >> Sources of non-atomic IPv4 datagrams using strong integrity checks 474 MAY reuse the ID within MDL values smaller than is typical. 476 Note, however, that such frequent reuse can still result in corrupted 477 reassembly and poor throughput, although it would not propagate 478 reassembly errors to higher layer protocols. 480 5.3. Impact on Middleboxes 482 Middleboxes include rewriting devices that include network address 483 translators (NATs), address/port translators (NAPTs), and other 484 address sharing mechanisms (ASMs). They also include devices that 485 inspect and filter datagrams that are not routers, such as 486 accelerators and firewalls. 488 The changes proposed in this document may not be implemented by 489 middleboxes, however these changes are more likely to make current 490 middlebox behavior compliant than to affect the service provided by 491 those devices. 493 5.3.1. Rewriting Middleboxes 495 NATs and NAPTs rewrite IP fields, and tunnel ingresses (using IPv4 496 encapsulation) copy and modify some IPv4 fields, so all are 497 considered sources, as do any devices that rewrite any portion of the 498 source address, destination address, protocol, and ID tuple for any 499 datagrams [RFC3022]. This is also true for other ASMs, including 4rd, 500 IVI, and others in the "A+P" (address plus port) family [Bo11] [De11] 501 [RFC6219]. It is equally true for any other datagram rewriting 502 mechanism. As a result, they are subject to all the requirements of 503 any source, as has been noted. 505 NATs/ASMs/rewriters present a particularly challenging situation for 506 fragmentation. Because they overwrite portions of the reassembly 507 tuple in both directions, they can destroy tuple uniqueness and 508 result in a reassembly hazard. Whenever IPv4 source address, 509 destination address, or protocol fields are modified, a 510 NAT/ASM/rewriter needs to ensure that the ID field is generated 511 appropriately, rather than simply copied from the incoming datagram. 512 In specific: 514 >> Address sharing or rewriting devices MUST ensure that the IPv4 ID 515 field of datagrams whose address or protocol are translated comply 516 with these requirements as if the datagram were sourced by that 517 device. 519 This compliance means that the IPv4 ID field of non-atomic datagrams 520 translated at a NAT/ASM/rewriter needs to obey the uniqueness 521 requirements of any IPv4 datagram source. Unfortunately, fragments 522 already violate that requirement, as they repeat an IPv4 ID within 523 the MDL for a given source address, destination address, and protocol 524 triple. 526 Such problems with transmitting fragments through NATs/ASMs/rewriters 527 are already known; translation is based on the transport port number, 528 which is present in only the first fragment anyway [RFC3022]. This 529 document underscores the point that not only is reassembly (and 530 possibly subsequent fragmentation) required for translation, it can 531 be used to avoid issues with IPv4 ID uniqueness. 533 Note that NATs/ASMs already need to exercise special care when 534 emitting datagrams on their public side, because merging datagrams 535 from many sources onto a single outgoing source address can result in 536 IPv4 ID collisions. This situation precedes this document, and is not 537 affected by it. It is exacerbated in large-scale, so-called "carrier 538 grade" NATs [Pe11]. 540 Tunnel ingresses act as sources for the outermost header, but tunnels 541 act as routers for the inner headers (i.e., the datagram as arriving 542 at the tunnel ingress). Ingresses can always fragment as originating 543 sources of the outer header, because they control the uniqueness of 544 that IPv4 ID field and the value of DF on the outer header 545 independent of those values on the inner (arriving datagram) header. 547 5.3.2. Filtering Middleboxes 549 Middleboxes also include devices that filter datagrams, including 550 network accelerators and firewalls. Some such devices reportedly 551 feature datagram de-duplication that relies on IP ID uniqueness to 552 identify duplicates, which has been discussed in Section 5.1. 554 5.4. Impact on Header Compression 556 Header compression algorithms already accommodate various ways in 557 which the IPv4 ID changes between sequential datagrams [RFC1144] 558 [RFC2508] [RFC3545] [RFC5225]. Such algorithms currently assume that 559 the IPv4 ID is preserved end-to-end. Some algorithms already allow 560 assuming the ID does not change (e.g., ROHC [RFC5225]), where others 561 include non-changing IDs via zero deltas (e.g., ECRTP [RFC3545]). 563 When compression assumes a changing ID as a default, having a non- 564 changing ID can make compression less efficient. Such non-changing 565 IDs have been described in various RFCs (e.g., footnote 21 of 566 [RFC1144] and cRTP [RFC2508]). When compression can assume a non- 567 changing IPv4 ID - as with ROHC and ECRTP - efficiency can be 568 increased. 570 6. Updates to Existing Standards 572 The following sections address the specific changes to existing 573 protocols indicated by this document. 575 6.1. Updates to RFC 791 577 RFC 791 states that: 579 The originating protocol module of an internet datagram sets the 580 identification field to a value that must be unique for that 581 source-destination pair and protocol for the time the datagram 582 will be active in the internet system. 584 And later that: 586 Thus, the sender must choose the Identifier to be unique for this 587 source, destination pair and protocol for the time the datagram 588 (or any fragment of it) could be alive in the internet. 590 It seems then that a sending protocol module needs to keep a table 591 of Identifiers, one entry for each destination it has communicated 592 with in the last maximum datagram lifetime for the internet. 594 However, since the Identifier field allows 65,536 different 595 values, some host may be able to simply use unique identifiers 596 independent of destination. 598 It is appropriate for some higher level protocols to choose the 599 identifier. For example, TCP protocol modules may retransmit an 600 identical TCP segment, and the probability for correct reception 601 would be enhanced if the retransmission carried the same 602 identifier as the original transmission since fragments of either 603 datagram could be used to construct a correct TCP segment. 605 This document changes RFC 791 as follows: 607 o IPv4 ID uniqueness applies to only non-atomic datagrams. 609 o Retransmitted non-atomic IPv4 datagrams are no longer permitted to 610 reuse the ID value. 612 6.2. Updates to RFC 1122 614 RFC 1122 states that: 616 3.2.1.5 Identification: RFC-791 Section 3.2 618 When sending an identical copy of an earlier datagram, a 619 host MAY optionally retain the same Identification field in 620 the copy. 622 DISCUSSION: 624 Some Internet protocol experts have maintained that when a 625 host sends an identical copy of an earlier datagram, the new 626 copy should contain the same Identification value as the 627 original. There are two suggested advantages: (1) if the 628 datagrams are fragmented and some of the fragments are lost, 629 the receiver may be able to reconstruct a complete datagram 630 from fragments of the original and the copies; (2) a 631 congested gateway might use the IP Identification field (and 632 Fragment Offset) to discard duplicate datagrams from the 633 queue. 635 This document changes RFC 1122 as follows: 637 o The IPv4 ID field is no longer permitted to be used for duplicate 638 detection. This applies to both atomic and non-atomic datagrams. 640 o Retransmitted non-atomic IPv4 datagrams are no longer permitted to 641 reuse the ID value. 643 6.3. Updates to RFC 2003 645 This document updates how IPv4-in-IPv4 tunnels create IPv4 ID values 646 for the IPv4 outer header [RFC2003], but only in the same way as for 647 any other IPv4 datagram source. In specific, RFC 2003 states the 648 following, where ref. [10] is RFC 791: 650 Identification, Flags, Fragment Offset 652 These three fields are set as specified in [10]... 654 This document changes RFC 2003 as follows: 656 o The IPv4 ID field is set as permitted by RFCXXXX. 658 7. Security Considerations 660 When the IPv4 ID is ignored on receipt (e.g., for atomic datagrams), 661 its value becomes unconstrained; that field then can more easily be 662 used as a covert channel. For some atomic datagrams it is now 663 possible, and may be desirable, to rewrite the IPv4 ID field to avoid 664 its use as such a channel. Rewriting would be prohibited for 665 datagrams protected by IPsec Authentication Header (AH), although we 666 do not recommend use of AH to achieve this result [RFC4302]. 668 The IPv4 ID also now adds much less to the entropy of the header of a 669 datagram. Such entropy might be used as input to cryptographic 670 algorithms or pseudorandom generators, although IDs have never been 671 assured sufficient entropy for such purposes. The IPv4 ID had 672 previously been unique (for a given source/address pair, and protocol 673 field) within one MDL, although this requirement was not enforced and 674 clearly is typically ignored. The IPv4 ID of atomic datagrams is not 675 required unique, and so contributes no entropy to the header. 677 The deprecation of the IPv4 ID field's uniqueness for atomic 678 datagrams can defeat the ability to count devices behind a 679 NAT/ASM/rewriter [Be02]. This is not intended as a security feature, 680 however. 682 8. IANA Considerations 684 There are no IANA considerations in this document. 686 The RFC Editor should remove this section prior to publication 688 9. References 690 9.1. Normative References 692 [RFC791] Postel, J., "Internet Protocol", RFC 791 / STD 5, September 693 1981. 695 [RFC1122] Braden, R., Ed., "Requirements for Internet Hosts - 696 Communication Layers", RFC 1122 / STD 3, October 1989. 698 [RFC1812] Baker, F. (Ed.), "Requirements for IP Version 4 Routers", 699 RFC 1812 / STD 4, Jun. 1995. 701 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 702 Requirement Levels", RFC 2119 / BCP 14, March 1997. 704 [RFC2003] Perkins, C., "IP Encapsulation within IP", RFC 2003, 705 October 1996. 707 9.2. Informative References 709 [Be02] Bellovin, S., "A Technique for Counting NATted Hosts", 710 Internet Measurement Conference, Proceedings of the 2nd ACM 711 SIGCOMM Workshop on Internet Measurement, Nov. 2002. 713 [Bo11] Boucadair, M., J. Touch, P. Levis, R. Penno, "Analysis of 714 Solution Candidates to Reveal a Host Identifier in Shared 715 Address Deployments", (work in progress), draft-boucadair- 716 intarea-nat-reveal-analysis, Sept. 2011. 718 [De11] Despres, R. (Ed.), S. Matsushima, T. Murakami, O. Troan, 719 "IPv4 Residual Deployment across IPv6-Service networks 720 (4rd)", (work in progress), draft-despres-intarea-4rd, Mar. 721 2011. 723 [Pe11] Perreault, S., (Ed.), I. Yamagata, S. Miyakawa, A. 724 Nakagawa, H. Ashida, "Common requirements of IP address 725 sharing schemes", (work in progress), draft-ietf-behave- 726 lsn-requirements, Mar. 2011. 728 [RFC1144] Jacobson, V., "Compressing TCP/IP Headers", RFC 1144, Feb. 729 1990. 731 [RFC2460] Deering, S., R. Hinden, "Internet Protocol, Version 6 732 (IPv6) Specification", RFC 2460, Dec. 1998. 734 [RFC2508] Casner, S., V. Jacobson. "Compressing IP/UDP/RTP Headers 735 for Low-Speed Serial Links", RFC 2508, Feb. 1999. 737 [RFC2671] Vixie,P., "Extension Mechanisms for DNS (EDNS0)", RFC 2671, 738 Aug. 1999. 740 [RFC3022] Srisuresh, P. and K. Egevang, "Traditional IP Network 741 Address Translator (Traditional NAT)", RFC 3022, Jan. 2001. 743 [RFC3545] Koren, T., S. Casner, J. Geevarghese, B. Thompson, P. 744 Ruddy, "Enhanced Compressed RTP (CRTP) for Links with High 745 Delay, Packet Loss and Reordering", RFC 3545, Jul. 2003. 747 [RFC3828] Larzon, L-A., M. Degermark, S. Pink, L-E. Jonsson, Ed., G. 748 Fairhurst, Ed., "The Lightweight User Datagram Protocol 749 (UDP-Lite)", RFC 3828, Jul. 2004. 751 [RFC4301] Kent, S., K. Seo, "Security Architecture for the Internet 752 Protocol", RFC 4301, Dec. 2005. 754 [RFC4302] Kent, S., "IP Authentication Header", RFC 4302, Dec. 2005. 756 [RFC4443] Conta, A., S. Deering, M. Gupta (Ed.), "Internet Control 757 Message Protocol (ICMPv6) for the Internet Protocol Version 758 6 (IPv6) Specification", RFC 4443, March. 2006. 760 [RFC4960] Stewart, R. (Ed.), "Stream Control Transmission Protocol", 761 RFC 4960, Sep. 2007. 763 [RFC4963] Heffner, J., M. Mathis, B. Chandler, "IPv4 Reassembly 764 Errors at High Data Rates," RFC 4963, Jul. 2007. 766 [RFC5225] Pelletier, G., K. Sandlund, "RObust Header Compression 767 Version 2 (ROHCv2): Profiles for RTP, UDP, IP, ESP and UDP- 768 Lite", RFC 5225, Apr. 2008. 770 [RFC5320] Templin, F., Ed., "The Subnetwork Encapsulation and 771 Adaptation Layer (SEAL)", RFC 5320, Feb. 2010. 773 [RFC6145] Li, X., C. Bao, F. Baker, "IP/ICMP Translation Algorithm," 774 RFC 6145, Apr. 2011. 776 [RFC6219] Li, X., C. Bao, M. Chen, H. Zhang, J. Wu, "The China 777 Education and Research Network (CERNET) IVI Translation 778 Design and Deployment for the IPv4/IPv6 Coexistence and 779 Transition", RFC 6219, May 2011. 781 [RFC6621] Macker, J. (Ed.), "Simplified Multicast Forwarding," RFC 782 6621, May 2012. 784 10. Acknowledgments 786 This document was inspired by of numerous discussions among the 787 authors, Jari Arkko, Lars Eggert, Dino Farinacci, and Fred Templin, 788 as well as members participating in the Internet Area Working Group. 789 Detailed feedback was provided by Gorry Fairhurst, Brian Haberman, 790 Ted Hardie, Mike Heard, Erik Nordmark, Carlos Pignataro, and Dan 791 Wing. This document originated as an Independent Stream draft co- 792 authored by Matt Mathis, PSC, and his contributions are greatly 793 appreciated. 795 This document was prepared using 2-Word-v2.0.template.dot. 797 Author's Address 799 Joe Touch 800 USC/ISI 801 4676 Admiralty Way 802 Marina del Rey, CA 90292-6695 803 U.S.A. 805 Phone: +1 (310) 448-9151 806 Email: touch@isi.edu