idnits 2.17.1 draft-ietf-v6ops-mech-v2-03.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** It looks like you're using RFC 3978 boilerplate. You should update this to the boilerplate described in the IETF Trust License Policy document (see https://trustee.ietf.org/license-info), which is required now. -- Found old boilerplate from RFC 3667, Section 5.1 on line 15. -- Found old boilerplate from RFC 3978, Section 5.5 on line 1370. -- Found old boilerplate from RFC 3979, Section 5, paragraph 1 on line 1347. -- Found old boilerplate from RFC 3979, Section 5, paragraph 2 on line 1354. -- Found old boilerplate from RFC 3979, Section 5, paragraph 3 on line 1360. ** The document seems to lack an RFC 3978 Section 5.1 IPR Disclosure Acknowledgement -- however, there's a paragraph with a matching beginning. Boilerplate error? ** This document has an original RFC 3978 Section 5.4 Copyright Line, instead of the newer IETF Trust Copyright according to RFC 4748. ** This document has an original RFC 3978 Section 5.5 Disclaimer, instead of the newer disclaimer which includes the IETF Trust according to RFC 4748. ** The document uses RFC 3667 boilerplate or RFC 3978-like boilerplate instead of verbatim RFC 3978 boilerplate. After 6 May 2005, submission of drafts without verbatim RFC 3978 boilerplate is not accepted. The following non-3978 patterns matched text found in the document. That text should be removed or replaced: By submitting this Internet-Draft, I certify that any applicable patent or other IPR claims of which I am aware have been disclosed, or will be disclosed, and any of which I become aware will be disclosed, in accordance with RFC 3668. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- == No 'Intended status' indicated for this document; assuming Proposed Standard == The page length should not exceed 58 lines per page, but there was 31 longer pages, the longest (page 14) being 75 lines == It seems as if not all pages are separated by form feeds - found 0 form feeds but 32 pages 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.) ** There is 1 instance of too long lines in the document, the longest one being 2 characters in excess of 72. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the RFC 3978 Section 5.4 Copyright Line does not match the current year == Line 1233 has weird spacing: '...arified that ...' == The document seems to lack the recommended RFC 2119 boilerplate, even if it appears to use RFC 2119 keywords. (The document does seem to have the reference to RFC 2119 which the ID-Checklist requires). -- 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.) -- Couldn't find a document date in the document -- date freshness check skipped. Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Unused Reference: 'RFC1122' is defined on line 1020, but no explicit reference was found in the text == Unused Reference: 'RFC1812' is defined on line 1026, but no explicit reference was found in the text == Unused Reference: 'RFC2462' is defined on line 1035, but no explicit reference was found in the text ** Obsolete normative reference: RFC 1981 (Obsoleted by RFC 8201) ** Obsolete normative reference: RFC 2460 (Obsoleted by RFC 8200) ** Obsolete normative reference: RFC 2463 (Obsoleted by RFC 4443) == Outdated reference: A later version (-12) exists of draft-ietf-dnsop-ipv6-dns-issues-07 == Outdated reference: A later version (-03) exists of draft-savola-v6ops-security-overview-02 == Outdated reference: A later version (-03) exists of draft-tschofenig-v6ops-secure-tunnels-00 -- Obsolete informational reference (is this intentional?): RFC 2401 (Obsoleted by RFC 4301) -- Obsolete informational reference (is this intentional?): RFC 2461 (Obsoleted by RFC 4861) -- Obsolete informational reference (is this intentional?): RFC 2462 (Obsoleted by RFC 4862) -- Obsolete informational reference (is this intentional?): RFC 2667 (Obsoleted by RFC 4087) -- Obsolete informational reference (is this intentional?): RFC 3484 (Obsoleted by RFC 6724) -- Obsolete informational reference (is this intentional?): RFC 3513 (Obsoleted by RFC 4291) Summary: 10 errors (**), 0 flaws (~~), 12 warnings (==), 13 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 INTERNET-DRAFT E. Nordmark 3 June 15, 2004 Sun Microsystems, Inc. 4 Obsoletes: 2893 R. E. Gilligan 5 Intransa, Inc. 7 Basic Transition Mechanisms for IPv6 Hosts and Routers 8 10 Status of this Memo 12 By submitting this Internet-Draft, I certify that any applicable 13 patent or other IPR claims of which I am aware have been disclosed, 14 and any of which I become aware will be disclosed, in accordance with 15 RFC 3668. 17 Internet-Drafts are working documents of the Internet Engineering 18 Task Force (IETF), its areas, and its working groups. Note that 19 other groups may also distribute working documents as Internet- 20 Drafts. 22 Internet-Drafts are draft documents valid for a maximum of six months 23 and may be updated, replaced, or obsoleted by other documents at any 24 time. It is inappropriate to use Internet-Drafts as reference 25 material or to cite them other than as "work in progress." 27 The list of current Internet-Drafts can be accessed at 28 http://www.ietf.org/ietf/1id-abstracts.txt 30 The list of Internet-Draft Shadow Directories can be accessed at 31 http://www.ietf.org/shadow.html. 33 This draft expires on December 15, 2004. 35 Abstract 37 This document specifies IPv4 compatibility mechanisms that can be 38 implemented by IPv6 hosts and routers. Two mechanisms are specified, 39 "dual stack" and configured tunneling. Dual stack implies providing 40 complete implementations of both versions of the Internet Protocol 41 (IPv4 and IPv6) and configured tunneling provides a means to carry 42 IPv6 packets over unmodified IPv4 routing infrastructures. 44 This document obsoletes RFC 2893. 46 Contents 48 Status of this Memo.......................................... 1 50 1. Introduction............................................. 3 51 1.1. Terminology......................................... 3 53 2. Dual IP Layer Operation.................................. 5 54 2.1. Address Configuration............................... 5 55 2.2. DNS................................................. 5 57 3. Configured Tunneling Mechanisms.......................... 7 58 3.1. Encapsulation....................................... 8 59 3.2. Tunnel MTU and Fragmentation........................ 9 60 3.2.1. Static Tunnel MTU.............................. 9 61 3.2.2. Dynamic Tunnel MTU............................. 10 62 3.3. Hop Limit........................................... 11 63 3.4. Handling ICMPv4 errors.............................. 12 64 3.5. IPv4 Header Construction............................ 14 65 3.6. Decapsulation....................................... 15 66 3.7. Link-Local Addresses................................ 18 67 3.8. Neighbor Discovery over Tunnels..................... 19 69 4. Threat Related to Source Address Spoofing................ 20 71 5. Security Considerations.................................. 21 73 6. Acknowledgments.......................................... 22 75 7. References............................................... 23 76 7.1. Normative References................................ 23 77 7.2. Informative References.............................. 23 79 8. Authors' Addresses....................................... 25 81 9. Changes from RFC 2893.................................... 25 82 9.1. Changes from draft-ietf-v6ops-mech-v2-00............ 27 83 9.2. Changes from draft-ietf-v6ops-mech-v2-01............ 29 84 9.3. Changes from draft-ietf-v6ops-mech-v2-02............ 30 86 1. Introduction 88 The key to a successful IPv6 transition is compatibility with the 89 large installed base of IPv4 hosts and routers. Maintaining 90 compatibility with IPv4 while deploying IPv6 will streamline the task 91 of transitioning the Internet to IPv6. This specification defines 92 two mechanisms that IPv6 hosts and routers may implement in order to 93 be compatible with IPv4 hosts and routers. 95 The mechanisms in this document are designed to be employed by IPv6 96 hosts and routers that need to interoperate with IPv4 hosts and 97 utilize IPv4 routing infrastructures. We expect that most nodes in 98 the Internet will need such compatibility for a long time to come, 99 and perhaps even indefinitely. 101 The mechanisms specified here are: 103 - Dual IP layer (also known as Dual Stack): A technique for 104 providing complete support for both Internet protocols -- IPv4 105 and IPv6 -- in hosts and routers. 107 - Configured tunneling of IPv6 over IPv4: A technique for 108 establishing point-to-point tunnels by encapsulating IPv6 109 packets within IPv4 headers to carry them over IPv4 routing 110 infrastructures. 112 The mechanisms defined here are intended to be the core of a 113 "transition toolbox" -- a growing collection of techniques which 114 implementations and users may employ to ease the transition. The 115 tools may be used as needed. Implementations and sites decide which 116 techniques are appropriate to their specific needs. 118 This document defines the basic set of transition mechanisms, but 119 these are not the only tools available. Additional transition and 120 compatibility mechanisms are specified in other documents. 122 1.1. Terminology 124 The following terms are used in this document: 126 Types of Nodes 128 IPv4-only node: 130 A host or router that implements only IPv4. An IPv4- 131 only node does not understand IPv6. The installed base 132 of IPv4 hosts and routers existing before the transition 133 begins are IPv4-only nodes. 135 IPv6/IPv4 node: 137 A host or router that implements both IPv4 and IPv6. 139 IPv6-only node: 141 A host or router that implements IPv6, and does not 142 implement IPv4. The operation of IPv6-only nodes is not 143 addressed in this memo. 145 IPv6 node: 147 Any host or router that implements IPv6. IPv6/IPv4 and 148 IPv6-only nodes are both IPv6 nodes. 150 IPv4 node: 152 Any host or router that implements IPv4. IPv6/IPv4 and 153 IPv4-only nodes are both IPv4 nodes. 155 Techniques Used in the Transition 157 IPv6-over-IPv4 tunneling: 159 The technique of encapsulating IPv6 packets within IPv4 160 so that they can be carried across IPv4 routing 161 infrastructures. 163 Configured tunneling: 165 IPv6-over-IPv4 tunneling where the IPv4 tunnel endpoint 166 address is determined by configuration information on 167 the encapsulator. All tunnels are assumed to be 168 bidirectional, behaving as virtual point-to-point links. 170 Other transition mechanisms, including other tunneling mechanisms, 171 are outside the scope of this document. 173 The keywords MUST, MUST NOT, REQUIRED, SHALL, SHALL NOT, SHOULD, 174 SHOULD NOT, RECOMMENDED, MAY, and OPTIONAL, when they appear in this 175 document, are to be interpreted as described in [RFC2119]. 177 2. Dual IP Layer Operation 179 The most straightforward way for IPv6 nodes to remain compatible with 180 IPv4-only nodes is by providing a complete IPv4 implementation. IPv6 181 nodes that provide complete IPv4 and IPv6 implementations are called 182 "IPv6/IPv4 nodes." IPv6/IPv4 nodes have the ability to send and 183 receive both IPv4 and IPv6 packets. They can directly interoperate 184 with IPv4 nodes using IPv4 packets, and also directly interoperate 185 with IPv6 nodes using IPv6 packets. 187 Even though a node may be equipped to support both protocols, one or 188 the other stack may be disabled for operational reasons. Here we use 189 a rather loose notion of "stack". A stack being enabled has IP 190 addresses assigned etc, but whether or not any particular application 191 is available on the stacks is explicitly not defined. Thus IPv6/IPv4 192 nodes may be operated in one of three modes: 194 - With their IPv4 stack enabled and their IPv6 stack disabled. 196 - With their IPv6 stack enabled and their IPv4 stack disabled. 198 - With both stacks enabled. 200 IPv6/IPv4 nodes with their IPv6 stack disabled will operate like 201 IPv4-only nodes. Similarly, IPv6/IPv4 nodes with their IPv4 stacks 202 disabled will operate like IPv6-only nodes. IPv6/IPv4 nodes MAY 203 provide a configuration switch to disable either their IPv4 or IPv6 204 stack. 206 The configured tunneling technique, which is described in section 3, 207 may or may not be used in addition to the dual IP layer operation. 209 2.1. Address Configuration 211 Because the nodes support both protocols, IPv6/IPv4 nodes may be 212 configured with both IPv4 and IPv6 addresses. IPv6/IPv4 nodes use 213 IPv4 mechanisms (e.g., DHCP) to acquire their IPv4 addresses, and 214 IPv6 protocol mechanisms (e.g., stateless address autoconfiguration 215 and/or DHCPv6) to acquire their IPv6 addresses. 217 2.2. DNS 219 The Domain Naming System (DNS) is used in both IPv4 and IPv6 to map 220 between hostnames and IP addresses. A new resource record type named 221 "AAAA" has been defined for IPv6 addresses [RFC3596]. Since 222 IPv6/IPv4 nodes must be able to interoperate directly with both IPv4 223 and IPv6 nodes, they must provide resolver libraries capable of 224 dealing with IPv4 "A" records as well as IPv6 "AAAA" records. Note 225 that the lookup of A versus AAAA records is independent of whether 226 the DNS packets are carried in IPv4 or IPv6 packets, and that there 227 is no assumption that the DNS servers know the IPv4/IPv6 capabilities 228 of the requesting node. 230 The issues and operational guidelines for using IPv6 with DNS are 231 described at more length in other documents [DNSOPV6]. 233 DNS resolver libraries on IPv6/IPv4 nodes MUST be capable of handling 234 both AAAA and A records. However, when a query locates an AAAA 235 record holding an IPv6 address, and an A record holding an IPv4 236 address, the resolver library MAY order the results returned to the 237 application in order to influence the version of IP packets used to 238 communicate with that node -- IPv6 first, or IPv4 first. 240 The applications SHOULD be able to specify whether they want IPv4, 241 IPv6 or both records [RFC3493]. That defines which address families 242 the resolver looks up. If there isn't an application choice, or if 243 the application has requested both, the resolver library MUST NOT 244 filter out any records. 246 Since most applications try the addresses in the order they are 247 returned by the resolver, this can affect the IP version "preference" 248 of applications. 250 A resolver library performing ordering of addresses might also want 251 to take into account external factors such as, whether IPv6 252 interfaces have been configured on the node. 254 The decision to order DNS results is implementation specific. 255 IPv6/IPv4 nodes MAY provide policy configuration to control ordering 256 of addresses returned by the resolver -- i.e., which order to sort -- 257 or leave the decision entirely up to the application. 259 This is a bare minimum DNS processing implementation for dual-stack 260 hosts. The more extensive algorithm is specified in [RFC3484]. 262 3. Configured Tunneling Mechanisms 264 In most deployment scenarios, the IPv6 routing infrastructure will be 265 built up over time. While the IPv6 infrastructure is being deployed, 266 the existing IPv4 routing infrastructure can remain functional, and 267 can be used to carry IPv6 traffic. Tunneling provides a way to 268 utilize an existing IPv4 routing infrastructure to carry IPv6 269 traffic. 271 IPv6/IPv4 hosts and routers can tunnel IPv6 datagrams over regions of 272 IPv4 routing topology by encapsulating them within IPv4 packets. 273 Tunneling can be used in a variety of ways: 275 - Router-to-Router. IPv6/IPv4 routers interconnected by an IPv4 276 infrastructure can tunnel IPv6 packets between themselves. In 277 this case, the tunnel spans one segment of the end-to-end path 278 that the IPv6 packet takes. 280 - Host-to-Router. IPv6/IPv4 hosts can tunnel IPv6 packets to an 281 intermediary IPv6/IPv4 router that is reachable via an IPv4 282 infrastructure. This type of tunnel spans the first segment of 283 the packet's end-to-end path. 285 - Host-to-Host. IPv6/IPv4 hosts that are interconnected by an 286 IPv4 infrastructure can tunnel IPv6 packets between themselves. 287 In this case, the tunnel spans the entire end-to-end path that 288 the packet takes. 290 - Router-to-Host. IPv6/IPv4 routers can tunnel IPv6 packets to 291 their final destination IPv6/IPv4 host. This tunnel spans only 292 the last segment of the end-to-end path. 294 Configured tunneling can be used in all of the above cases, but is 295 most likely to be used router-to-router due to the need to explicitly 296 configure the tunneling endpoints. 298 The underlying mechanisms for tunneling are: 300 - The entry node of the tunnel (the encapsulator) creates an 301 encapsulating IPv4 header and transmits the encapsulated packet. 303 - The exit node of the tunnel (the decapsulator) receives the 304 encapsulated packet, reassembles the packet if needed, removes 305 the IPv4 header, and processes the received IPv6 packet. 307 - The encapsulator may need to maintain soft state information for 308 each tunnel recording such parameters as the MTU of the tunnel 309 in order to process IPv6 packets forwarded into the tunnel. 311 In configured tunneling, the tunnel endpoint address is determined 312 from configuration information in the encapsulator. For each tunnel, 313 the encapsulator must store the tunnel endpoint address. When an 314 IPv6 packet is transmitted over a tunnel, the tunnel endpoint address 315 configured for that tunnel is used as the destination address for the 316 encapsulating IPv4 header. 318 The determination of which packets to tunnel is usually made by 319 routing information on the encapsulator. This is usually done via a 320 routing table, which directs packets based on their destination 321 address using the prefix mask and match technique. 323 3.1. Encapsulation 325 The encapsulation of an IPv6 datagram in IPv4 is shown below: 327 +-------------+ 328 | IPv4 | 329 | Header | 330 +-------------+ +-------------+ 331 | IPv6 | | IPv6 | 332 | Header | | Header | 333 +-------------+ +-------------+ 334 | Transport | | Transport | 335 | Layer | ===> | Layer | 336 | Header | | Header | 337 +-------------+ +-------------+ 338 | | | | 339 ~ Data ~ ~ Data ~ 340 | | | | 341 +-------------+ +-------------+ 343 Encapsulating IPv6 in IPv4 345 In addition to adding an IPv4 header, the encapsulator also has to 346 handle some more complex issues: 348 - Determine when to fragment and when to report an ICMPv6 "packet 349 too big" error back to the source. 351 - How to reflect ICMPv4 errors from routers along the tunnel path 352 back to the source as ICMPv6 errors. 354 Those issues are discussed in the following sections. 356 3.2. Tunnel MTU and Fragmentation 358 Naively the encapsulator could view encapsulation as IPv6 using IPv4 359 as a link layer with a very large MTU (65535-20 bytes to be exact; 20 360 bytes "extra" are needed for the encapsulating IPv4 header). The 361 encapsulator would only need to report ICMPv6 "packet too big" errors 362 back to the source for packets that exceed this MTU. However, such a 363 scheme would be inefficient or non-interoperable for three reasons 364 and therefore MUST NOT be used: 366 1) It would result in more fragmentation than needed. IPv4 layer 367 fragmentation should be avoided due to the performance problems 368 caused by the loss unit being smaller than the retransmission 369 unit [KM97]. 371 2) Any IPv4 fragmentation occurring inside the tunnel, i.e. between 372 the encapsulator and the decapsulator, would have to be 373 reassembled at the tunnel endpoint. For tunnels that terminate 374 at a router, this would require additional memory and other 375 resources to reassemble the IPv4 fragments into a complete IPv6 376 packet before that packet could be forwarded onward. 378 3) The encapsulator has no way of knowing that the decapsulator is 379 able to defragment such IPv4 packets (see Section 3.7 for 380 details), and has no way of knowing that the decapsulator is 381 able to handle such a large IPv6 Maximum Receive Unit (MRU). 383 Hence, the encapsulator MUST NOT treat the tunnel as an interface 384 with an MTU of 64 kilobytes, but instead either use the fixed static 385 MTU or OPTIONAL dynamic MTU determination based on the IPv4 path MTU 386 to the tunnel endpoint. 388 If both the mechanisms are implemented, the decision which to use 389 SHOULD be configurable on a per-tunnel endpoint basis. 391 3.2.1. Static Tunnel MTU 393 A node using static tunnel MTU treats the tunnel interface as having 394 a fixed interface MTU. By default, the MTU MUST be between 1280 and 395 1480 bytes (inclusive), but it SHOULD be 1280 bytes. If the default 396 is not 1280 bytes, the implementation MUST have a configuration knob 397 which can be used to change the MTU value. 399 A node must be able to accept a fragmented IPv6 packet that, after 400 reassembly, is as large as 1500 octets [RFC2460]. This memo also 401 includes requirements (see Section 3.6) for the amount of IPv4 402 reassembly and IPv6 MRU that MUST be supported by all the 403 decapsulators. These ensure correct interoperability with any fixed 404 MTUs between 1280 and 1480 bytes. 406 A larger fixed MTU than supported by these requirements, must not be 407 configured unless it has been administratively ensured that the 408 decapsulator can reassemble or receive packets of that size. 410 The selection of a good tunnel MTU depends on many factors; at least: 412 - Whether the IPv4 protocol-41 packets will be transported over 413 media which may have a lower path MTU (e.g., IPv4 Virtual 414 Private Networks); then picking too high a value might lead to 415 IPv4 fragmentation. 417 - Whether the tunnel is used to transport IPv6 tunneled packets 418 (e.g., a mobile node with an IPv4-in-IPv6 configured tunnel, and 419 an IPv6-in-IPv6 tunnel interface); then picking too low a value 420 might lead to IPv6 fragmentation. 422 If layered encapsulation is believed to be present, it may be prudent 423 to consider supporting dynamic MTU determination instead as it is 424 able to minimize fragmentation and optimize packet sizes. 426 When using the static tunnel MTU the Don't Fragment bit MUST NOT be 427 set in the encapsulating IPv4 header. As a result the encapsulator 428 should not receive any ICMPv4 "packet too big" messages as a result 429 of the packets it has encapsulated. 431 3.2.2. Dynamic Tunnel MTU 433 The dynamic MTU determination is OPTIONAL. However, if it is 434 implemented, it SHOULD have the behavior described in this document. 436 The fragmentation inside the tunnel can be reduced to a minimum by 437 having the encapsulator track the IPv4 Path MTU across the tunnel, 438 using the IPv4 Path MTU Discovery Protocol [RFC1191] and recording 439 the resulting path MTU. The IPv6 layer in the encapsulator can then 440 view a tunnel as a link layer with an MTU equal to the IPv4 path MTU, 441 minus the size of the encapsulating IPv4 header. 443 Note that this does not eliminate IPv4 fragmentation in the case when 444 the IPv4 path MTU would result in an IPv6 MTU less than 1280 bytes. 445 (Any link layer used by IPv6 has to have an MTU of at least 1280 446 bytes [RFC2460].) In this case the IPv6 layer has to "see" a link 447 layer with an MTU of 1280 bytes and the encapsulator has to use IPv4 448 fragmentation in order to forward the 1280 byte IPv6 packets. 450 The encapsulator SHOULD employ the following algorithm to determine 451 when to forward an IPv6 packet that is larger than the tunnel's path 452 MTU using IPv4 fragmentation, and when to return an ICMPv6 "packet 453 too big" message per [RFC1981]: 455 if (IPv4 path MTU - 20) is less than 1280 456 if packet is larger than 1280 bytes 457 Send ICMPv6 "packet too big" with MTU = 1280. 458 Drop packet. 459 else 460 Encapsulate but do not set the Don't Fragment 461 flag in the IPv4 header. The resulting IPv4 462 packet might be fragmented by the IPv4 layer on 463 the encapsulator or by some router along 464 the IPv4 path. 465 endif 466 else 467 if packet is larger than (IPv4 path MTU - 20) 468 Send ICMPv6 "packet too big" with 469 MTU = (IPv4 path MTU - 20). 470 Drop packet. 471 else 472 Encapsulate and set the Don't Fragment flag 473 in the IPv4 header. 474 endif 475 endif 477 Encapsulators that have a large number of tunnels may choose between 478 dynamic versus static tunnel MTU on a per-tunnel endpoint basis. In 479 cases where the number of tunnels that any one node is using is 480 large, it is helpful to observe that this state information can be 481 cached and discarded when not in use. 483 Note that using dynamic tunnel MTU is subject to IPv4 PMTU blackholes 484 should the ICMPv4 "packet too big" messages be dropped by firewalls 485 or not generated by the routers. [RFC1435, RFC2923] 487 3.3. Hop Limit 489 IPv6-over-IPv4 tunnels are modeled as "single-hop" from the IPv6 490 perspective. The tunnel is opaque to users of the network, and is not 491 detectable by network diagnostic tools such as traceroute. 493 The single-hop model is implemented by having the encapsulators and 494 decapsulators process the IPv6 hop limit field as they would if they 495 were forwarding a packet on to any other datalink. That is, they 496 decrement the hop limit by 1 when forwarding an IPv6 packet. (The 497 originating node and final destination do not decrement the hop 498 limit.) 500 The TTL of the encapsulating IPv4 header is selected in an 501 implementation dependent manner. The current suggested value is 502 published in the "Assigned Numbers" RFC [RFC3232][ASSIGNED]. The 503 implementations MAY also consider using the value 255. 504 Implementations MAY provide a mechanism to allow the administrator to 505 configure the IPv4 TTL as the IP Tunnel MIB [RFC2667]. 507 3.4. Handling ICMPv4 errors 509 In response to encapsulated packets it has sent into the tunnel, the 510 encapsulator might receive ICMPv4 error messages from IPv4 routers 511 inside the tunnel. These packets are addressed to the encapsulator 512 because it is the IPv4 source of the encapsulated packet. 514 ICMPv4 error handling is only applicable to dynamic MTU 515 determination, even though the functions could be used with static 516 MTU tunnels as well. 518 The ICMPv4 "packet too big" error messages are handled according to 519 IPv4 Path MTU Discovery [RFC1191] and the resulting path MTU is 520 recorded in the IPv4 layer. The recorded path MTU is used by IPv6 to 521 determine if an ICMPv6 "packet too big" error has to be generated as 522 described in section 3.2.2. 524 The handling of other types of ICMPv4 error messages depends on how 525 much information is included in the "packet in error" field, which 526 holds the encapsulated packet that caused the error. 528 Many older IPv4 routers return only 8 bytes of data beyond the IPv4 529 header of the packet in error, which is not enough to include the 530 address fields of the IPv6 header. More modern IPv4 routers are 531 likely to return enough data beyond the IPv4 header to include the 532 entire IPv6 header and possibly even the data beyond that. 534 If the offending packet includes enough data, the encapsulator MAY 535 extract the encapsulated IPv6 packet and use it to generate an ICMPv6 536 message directed back to the originating IPv6 node, as shown below: 538 +--------------+ 539 | IPv4 Header | 540 | dst = encaps | 541 | node | 542 +--------------+ 543 | ICMPv4 | 544 | Header | 545 - - +--------------+ 546 | IPv4 Header | 547 | src = encaps | 548 IPv4 | node | 549 +--------------+ - - 550 Packet | IPv6 | 551 | Header | Original IPv6 552 in +--------------+ Packet - 553 | Transport | Can be used to 554 Error | Header | generate an 555 +--------------+ ICMPv6 556 | | error message 557 ~ Data ~ back to the source. 558 | | 559 - - +--------------+ - - 561 ICMPv4 Error Message Returned to Encapsulating Node 563 When receiving ICMPv4 errors as above and the errors are not "packet 564 too big" it would be useful to log the error as an error related to 565 the tunnel. Also, if sufficient headers are included in the error, 566 then the originating node MAY send an ICMPv6 error of type 567 "unreachable" with code "address unreachable" to the IPv6 source. 568 (The "address unreachable" code is appropriate since, from the 569 perspective of IPv6, the tunnel is a link and that code is used for 570 link-specific errors [RFC2463]). 572 Note that when IPv4 path MTU is exceeded, and ICMPv4 errors of only 8 573 bytes of payload are generated, or ICMPv4 errors do not cause the 574 generation of ICMPv6 errors in case there is enough payload, there 575 will be at least two packet drops instead of at least one (the case 576 of a single layer of MTU discovery). Consider a case where an IPv6 577 host is connected to an IPv4/IPv6 router, which is connected to a 578 network where an ICMPv4 error about too big packet size is generated. 579 First the router needs to learn the tunnel (IPv4) MTU which causes at 580 least one packet loss, and then the host needs to learn the (IPv6) 581 MTU from the router which causes at least one packet loss. Still, in 582 all cases there can be more than one packet loss if there are 583 multiple large packets in flight at the same time. 585 3.5. IPv4 Header Construction 587 When encapsulating an IPv6 packet in an IPv4 datagram, the IPv4 588 header fields are set as follows: 590 Version: 592 4 594 IP Header Length in 32-bit words: 596 5 (There are no IPv4 options in the encapsulating 597 header.) 599 Type of Service: 601 0 unless otherwise specified. (See [RFC2983] and 602 [RFC3168] section 9.1 for issues relating to the Type- 603 of-Service byte and tunneling.) 605 Total Length: 607 Payload length from IPv6 header plus length of IPv6 and 608 IPv4 headers (i.e., IPv6 payload length plus a constant 609 60 bytes). 611 Identification: 613 Generated uniquely as for any IPv4 packet transmitted by 614 the system. 616 Flags: 618 Set the Don't Fragment (DF) flag as specified in section 619 3.2. Set the More Fragments (MF) bit as necessary if 620 fragmenting. 622 Fragment offset: 624 Set as necessary if fragmenting. 626 Time to Live: 628 Set in an implementation-specific manner, as described 629 in section 3.3. 631 Protocol: 633 41 (Assigned payload type number for IPv6). 635 Header Checksum: 637 Calculate the checksum of the IPv4 header. [RFC791] 639 Source Address: 641 IPv4 address of outgoing interface of the encapsulator 642 or an administratively specified address as described 643 below. 645 Destination Address: 647 IPv4 address of the tunnel endpoint. 649 When encapsulating the packets, the nodes must ensure that they will 650 use the source address that the tunnel peer has configured, so that 651 the source addresses are acceptable to the decapsulator. This may be 652 a problem with multi-addressed, and in particular, multi-interface 653 nodes, especially when the routing is changed from a stable 654 condition, as the source address selection may be adversely affected. 655 Therefore, it SHOULD be possible to administratively specify the 656 source address of a tunnel. 658 3.6. Decapsulation 660 When an IPv6/IPv4 host or a router receives an IPv4 datagram that is 661 addressed to one of its own IPv4 addresses or a joined multicast 662 group address, and the value of the protocol field is 41, the packet 663 is potentially part of a tunnel and needs to be verified to belong to 664 one of the configured tunnel interfaces (by checking 665 source/destination addresses), reassembled (if fragmented at the IPv4 666 level), have the IPv4 header removed and the resulting IPv6 datagram 667 be submitted to the IPv6 layer code on the node. 669 The decapsulator MUST verify that the tunnel source address is 670 correct before further processing packets, to mitigate the problems 671 with address spoofing (see section 4). This check also applies to 672 packets which are delivered to transport protocols on the 673 decapsulator. This is done by verifying that the source address is 674 the IPv4 address of the other end of a tunnel configured on the node. 675 Packets for which the IPv4 source address does not match MUST be 676 discarded and an ICMP message SHOULD NOT be generated; however, if 677 the implementation normally sends an ICMP message when receiving an 678 unknown protocol packet, such an error message MAY be sent (e.g., 679 ICMPv4 Protocol 41 Unreachable). 681 A side effect of this address verification is that the node will 682 silently discard packets with a wrong source address, and packets 683 which were received by the node but not directly addressed to it 684 (e.g., broadcast addresses). 686 Independent of any other forms of IPv4 ingress filtering the 687 administrator of the node may have configured to perform, the 688 implementation MAY perform ingress filtering, i.e., check that the 689 packet is arriving from the interface in the direction of the route 690 towards the tunnel end-point, similar to a Strict Reverse Path 691 Forwarding (RPF) check [RFC3704]. As this may cause problems on 692 tunnels which are routed through multiple links, it is RECOMMENDED 693 that this check, if done, is disabled by default. The packets caught 694 by this check SHOULD be discarded; an ICMP message SHOULD NOT be 695 generated by default. 697 The decapsulator MUST be capable of having, on the tunnel interfaces, 698 an IPv6 MRU of at least the maximum of of 1500 bytes and the largest 699 (IPv6) interface MTU on the decapsulator. 701 The decapsulator MUST be capable of reassembling an IPv4 packet that 702 is (after the reassembly) the maximum of 1500 bytes and the largest 703 (IPv4) interface MTU on the decapsulator. The 1500 byte number is a 704 result of encapsulators that use the static MTU scheme in section 705 3.2.1, while encapsulators that use the dynamic scheme in section 706 3.2.2 can cause up to the largest interface MTU on the decapsulator 707 to be received. (Note that it is strictly the interface MTU on the 708 last IPv4 router *before* the decapsulator that matters, but for most 709 links the MTU is the same between all neighbors.) 711 This reassembly limit allows dynamic tunnel MTU determination by the 712 encapsulator to take advantage of larger IPv4 path MTUs. An 713 implementation MAY have a configuration knob which can be used to set 714 a larger value of the tunnel reassembly buffers than the above 715 number, but it MUST NOT be set below the above number. 717 The decapsulation is shown below: 719 +-------------+ 720 | IPv4 | 721 | Header | 722 +-------------+ +-------------+ 723 | IPv6 | | IPv6 | 724 | Header | | Header | 725 +-------------+ +-------------+ 726 | Transport | | Transport | 727 | Layer | ===> | Layer | 728 | Header | | Header | 729 +-------------+ +-------------+ 730 | | | | 731 ~ Data ~ ~ Data ~ 732 | | | | 733 +-------------+ +-------------+ 735 Decapsulating IPv6 from IPv4 737 When decapsulating the packet, the IPv6 header is not modified. 738 (However, see [RFC2983] and [RFC3168] section 9.1 for issues relating 739 to the Type of Service byte and tunneling.) If the packet is 740 subsequently forwarded, its hop limit is decremented by one. 742 The decapsulator performs IPv4 reassembly before decapsulating the 743 IPv6 packet. 745 The encapsulating IPv4 header is discarded. When reconstructing the 746 IPv6 packet the length MUST be determined from the IPv6 payload 747 length since the IPv4 packet might be padded (thus have a length 748 which is larger than the IPv6 packet plus the IPv4 header being 749 removed). 751 After the decapsulation the node MUST silently discard a packet with 752 an invalid IPv6 source address. The list of invalid source addresses 753 SHOULD include at least: 755 - all multicast addresses (FF00::/8) 757 - the loopback address (::1) 759 - all the IPv4-compatible IPv6 addresses [RFC3513] (::/96), 760 excluding the unspecified address for Duplicate Address 761 Detection (::/128) 763 - all the IPv4-mapped IPv6 addresses (::ffff:0:0/96) 765 In addition, the node should be configured to perform ingress 766 filtering [RFC2827][RFC3704] on the IPv6 source address, similar to 767 on any of its interfaces, e.g.: 769 1) if the tunnel is towards the Internet, the node should be 770 configured to check that the site's IPv6 prefixes are not used 771 as the source addresses, or 773 2) if the tunnel is towards an edge network, the node should be 774 configured to check that the source address belongs to that edge 775 network. 777 The prefix lists in the former typically need to be manually 778 configured; the latter could be verified automatically, e.g., by 779 using a strict unicast RPF check, as long as an interface can be 780 designated to be towards an edge. 782 It is RECOMMENDED that the implementations provide a single knob to 783 make it easier to for the administrators to enable strict ingress 784 filtering towards edge networks. 786 3.7. Link-Local Addresses 788 The configured tunnels are IPv6 interfaces (over the IPv4 "link 789 layer") and thus MUST have link-local addresses. The link-local 790 addresses are used by, e.g., routing protocols operating over the 791 tunnels. 793 The interface identifier [RFC3513] for such an interface may be based 794 on the 32-bit IPv4 address of an underlying interface, or formed 795 using some other means, as long as it's unique from the other tunnel 796 endpoint with a reasonably high probability. 798 Note that it may be desirable to form the link-local address in a 799 fashion that minimizes the probability and the effect of having to 800 renumber the link-local address in the event of a topology or 801 hardware change. 803 If an IPv4 address is used for forming the IPv6 link-local address, 804 the interface identifier is the IPv4 address, prepended by zeros. 805 Note that the "Universal/Local" bit is zero, indicating that the 806 interface identifier is not globally unique. The link-local address 807 is formed by appending the interface identifier to the prefix 808 FE80::/64. 810 When the host has more than one IPv4 address in use on the physical 811 interface concerned, a choice of one of these IPv4 addresses is made 812 by the administrator or the implementation when forming the link- 813 local address. 815 +-------+-------+-------+-------+-------+-------+------+------+ 816 | FE 80 00 00 00 00 00 00 | 817 +-------+-------+-------+-------+-------+-------+------+------+ 818 | 00 00 00 00 | IPv4 Address | 819 +-------+-------+-------+-------+-------+-------+------+------+ 821 3.8. Neighbor Discovery over Tunnels 823 Configured tunnel implementations MUST at least accept and respond to 824 the probe packets used by Neighbor Unreachability Detection (NUD) 825 [RFC2461]. The implementations SHOULD also send NUD probe packets to 826 detect when the configured tunnel fails at which point the 827 implementation can use an alternate path to reach the destination. 828 Note that Neighbor Discovery allows that the sending of NUD probes be 829 omitted for router to router links if the routing protocol tracks 830 bidirectional reachability. 832 For the purposes of Neighbor Discovery the configured tunnels 833 specified in this document are assumed to NOT have a link-layer 834 address, even though the link-layer (IPv4) does have an address. 835 This means that: 837 - the sender of Neighbor Discovery packets SHOULD NOT include 838 Source Link Layer Address options or Target Link Layer Address 839 options on the tunnel link. 841 - the receiver MUST, while otherwise processing the Neighbor 842 Discovery packet, silently ignore the content of any Source Link 843 Layer Address options or Target Link Layer Address options 844 received on the tunnel link. 846 Not using a link layer address options is consistent with how 847 Neighbor Discovery is used on other point-to-point links. 849 4. Threat Related to Source Address Spoofing 851 The specification above contains rules that apply tunnel source 852 address verification in particular and ingress filtering 853 [RFC2827][RFC3704] in general to packets before they are 854 decapsulated. When IP-in-IP tunneling (independent of IP versions) 855 is used it is important that this can not be used to bypass any 856 ingress filtering in use for non-tunneled packets. Thus the rules in 857 this document are derived based on should ingress filtering be used 858 for IPv4 and IPv6, the use of tunneling should not provide an easy 859 way to circumvent the filtering. 861 In this case, without specific ingress filtering checks in the 862 decapsulator, it would be possible for an attacker to inject a packet 863 with: 865 - Outer IPv4 source: real IPv4 address of attacker 867 - Outer IPv4 destination: IPv4 address of decapsulator 869 - Inner IPv6 source: Alice which is either the decapsulator or a 870 node close to it. 872 - Inner IPv6 destination: Bob 874 Even if all IPv4 routers between the attacker and the decapsulator 875 implement IPv4 ingress filtering, and all IPv6 routers between the 876 decapsulator and Bob implement IPv6 ingress filtering, the above 877 spoofed packets will not be filtered out. As a result Bob will 878 receive a packet that looks like it was sent from Alice even though 879 the sender was some unrelated node. 881 The solution to this is to have the decapsulator only accept 882 encapsulated packets from the explicitly configured source address 883 (i.e., the other end of the tunnel) as specified in section 3.6. 884 While this does not provide complete protection in the case ingress 885 filtering has not been deployed, it does provide a significant 886 increase in security. The issue and the remainder threats are 887 discussed at more length in Security Considerations. 889 5. Security Considerations 891 Generic security considerations of using IPv6 are discussed in a 892 separate document [V6SEC]. 894 An implementation of tunneling needs to be aware that while a tunnel 895 is a link (as defined in [RFC2460]), the threat model for a tunnel 896 might be rather different than for other links, since the tunnel 897 potentially includes all of the Internet. 899 Several mechanisms (e.g., Neighbor Discovery) depend on Hop Count 900 being 255 and/or the addresses being link-local for ensuring that a 901 packet originated on-link, in a semi-trusted environment. Tunnels 902 are more vulnerable to a breach of this assumption than physical 903 links, as an attacker anywhere in the Internet can send an IPv6-in- 904 IPv4 packet to the tunnel decapsulator, causing injection of an 905 encapsulted IPv6 packet to the configured tunnel interface unless the 906 decapsulation checks are able to discard packets injected in such a 907 manner. 909 Therefore, this memo specifies that the decapsulators make these 910 steps to mitigate this threat: 912 - IPv4 source address of the packet MUST be the same as configured 913 for the tunnel end-point, 915 - Independent of any IPv4 ingress filtering the administrator may 916 have configured, the implementation MAY perform IPv4 ingress 917 filtering to check that the IPv4 packets are received from an 918 expected interface (but as this may cause some problems, it may 919 be disabled by default), 921 - IPv6 packets with several, obviously invalid IPv6 source 922 addresses received from the tunnel MUST be discarded (see 923 Section 3.6 for details), and 925 - IPv6 ingress filtering should be performed (typically requiring 926 configuration from the operator), to check that the tunneled 927 IPv6 packets are received from an expected interface. 929 Especially the first verification is vital: to avoid this check, the 930 attacker must be able to know the source of the tunnel (ranging from 931 difficult to predictable) and be able to spoof it (easier). 933 If the remainder threats of tunnel source verification are considered 934 to be significant, a tunneling scheme with authentication should be 935 used instead, for example IPsec [RFC2401] (preferable) or Generic 936 Routing Encapsulation with a pre-configured secret key [RFC2890]. As 937 the configured tunnels are set up more or less manually, setting up 938 the keying material is probably not a problem. However, setting up 939 secure IPsec IPv6-in-IPv4 tunnels is described in another document 940 [V64IPSEC]. 942 If the tunneling is done inside an administrative domain, proper 943 ingress filtering at the edge of the domain can also eliminate the 944 threat from outside of the domain. Therefore shorter tunnels are 945 preferable to longer ones, possibly spanning the whole Internet. 947 Additionally, an implementation MUST treat interfaces to different 948 links as separate, e.g., to ensure that Neighbor Discovery packets 949 arriving on one link does not effect other links. This is especially 950 important for tunnel links. 952 When dropping packets due to failing to match the allowed IPv4 source 953 addresses for a tunnel the node should not "acknowledge" the 954 existence of a tunnel, otherwise this could be used to probe the 955 acceptable tunnel endpoint addresses. For that reason, the 956 specification says that such packets MUST be discarded, and an ICMP 957 error message SHOULD NOT be generated, unless the implementation 958 normally sends ICMP destination unreachable messages for unknown 959 protocols; in such a case, the same code MAY be sent. As should be 960 obvious, the not returning the same ICMP code if an error is returned 961 for other protocols may hint that the IPv6 stack (or the protocol 41 962 tunneling processing) has been enabled -- the behaviour should be 963 consistent on how the implementation otherwise behaves to be 964 transparent to probing. 966 6. Acknowledgments 968 We would like to thank the members of the IPv6 working group, the 969 Next Generation Transition (ngtrans) working group, and the v6ops 970 working group for their many contributions and extensive review of 971 this document. Special thanks are due to Jim Bound, Ross Callon, Bob 972 Hinden, Bill Manning, John Moy, Mohan Parthasarathy, Pekka Savola, 973 Fred Templin, Chirayu Patel, and Tim Chown for many helpful 974 suggestions. Pekka Savola helped in editing the final revisions of 975 the specification. 977 7. References 979 7.1. Normative References 981 [RFC791] J. Postel, "Internet Protocol", RFC 791, September 1981. 983 [RFC1191] Mogul, J., and S. Deering., "Path MTU Discovery", RFC 1191, 984 November 1990. 986 [RFC1981] McCann, J., S. Deering, and J. Mogul. "Path MTU Discovery 987 for IP version 6", RFC 1981, August 1996. 989 [RFC2119] S. Bradner, "Key words for use in RFCs to Indicate 990 Requirement Levels", RFC 2119, March 1997. 992 [RFC2460] Deering, S., and Hinden, R. "Internet Protocol, Version 6 993 (IPv6) Specification", RFC 2460, December 1998. 995 [RFC2463] A. Conta, S. Deering, "Internet Control Message Protocol 996 (ICMPv6) for the Internet Protocol Version 6 (IPv6) 997 Specification", RFC 2463, December 1998. 999 7.2. Informative References 1001 [ASSIGNED] IANA, "Assigned numbers online database", 1002 http://www.iana.org/numbers.html 1004 [DNSOPV6] Durand, A., Ihren, J., and Savola P., "Operational 1005 Considerations and Issues with IPv6 DNS", draft-ietf-dnsop- 1006 ipv6-dns-issues-07.txt, work-in-progress, May 2004. 1008 [KM97] Kent, C., and J. Mogul, "Fragmentation Considered Harmful". 1009 In Proc. SIGCOMM '87 Workshop on Frontiers in Computer 1010 Communications Technology. August 1987. 1012 [V6SEC] P. Savola, "IPv6 Transition/Co-existence Security 1013 Considerations", draft-savola-v6ops-security-overview- 1014 02.txt, work-in-progress, February 2004. 1016 [V64IPSEC] Graveman, R., et al., "Using IPsec to Secure IPv6-over-IPv4 1017 Tunnels", draft-tschofenig-v6ops-secure-tunnels-00.txt, 1018 work-in-progress, June 2004. 1020 [RFC1122] Braden, R., "Requirements for Internet Hosts - Communication 1021 Layers", STD 3, RFC 1122, October 1989. 1023 [RFC1435] S. Knowles, "IESG Advice from Experience with Path MTU 1024 Discovery", RFC 1435, March 1993. 1026 [RFC1812] F. Baker, "Requirements for IP Version 4 Routers", RFC 1812, 1027 June 1995. 1029 [RFC2401] Kent, S., Atkinson, R., "Security Architecture for the 1030 Internet Protocol", RFC 2401, November 1998. 1032 [RFC2461] Narten, T., Nordmark, E., and Simpson, W. "Neighbor 1033 Discovery for IP Version 6 (IPv6)", RFC 2461, December 1998. 1035 [RFC2462] Thomson, S., and Narten, T. "IPv6 Stateless Address 1036 Autoconfiguration," RFC 2462, December 1998. 1038 [RFC2667] D. Thaler, "IP Tunnel MIB", RFC 2667, August 1999. 1040 [RFC2827] Ferguson, P., and Senie, D., "Network Ingress Filtering: 1041 Defeating Denial of Service Attacks which employ IP Source 1042 Address Spoofing", RFC 2827, May 2000. 1044 [RFC2890] Dommety, G., "Key and Sequence Number Extensions to GRE", 1045 RFC 2890, September 2000. 1047 [RFC2923] K. Lahey, "TCP Problems with Path MTU Discovery", RFC 2923, 1048 September 2000. 1050 [RFC2983] D. Black, "Differentiated Services and Tunnels", RFC 2983, 1051 October 2000. 1053 [RFC3056] B. Carpenter, and K. Moore, "Connection of IPv6 Domains via 1054 IPv4 Clouds", RFC 3056, February 2001. 1056 [RFC3168] K. Ramakrishnan, S. Floyd, D. Black, "The Addition of 1057 Explicit Congestion Notification (ECN) to IP", RFC 3168, 1058 September 2001. 1060 [RFC3232] Reynolds, J., "Assigned Numbers: RFC 1700 is Replaced by an 1061 On-line Database", RFC 3232, January 2002. 1063 [RFC3484] R. Draves, "Default Address Selection for IPv6", RFC 3484, 1064 February 2003. 1066 [RFC3493] Gilligan, R., et al, "Basic Socket Interface Extensions for 1067 IPv6", RFC 3493, February 2003. 1069 [RFC3513] Hinden, R., and S. Deering, "IP Version 6 Addressing 1070 Architecture", RFC 3513, April 2003. 1072 [RFC3596] Thomson, S., C. Huitema, V. Ksinant, and M. Souissi, "DNS 1073 Extensions to support IP version 6", RFC 3596, October 2003. 1075 [RFC3704] Baker, F., and Savola P., "Ingress Filtering for Multihomed 1076 Networks", RFC 3704, BCP 84, March 2004. 1078 8. Authors' Addresses 1080 Erik Nordmark 1081 Sun Microsystems Laboratories 1082 180, avenue de l'Europe 1083 38334 SAINT ISMIER Cedex, France 1084 Tel : +33 (0)4 76 18 88 03 1085 Fax : +33 (0)4 76 18 88 88 1086 Email : erik.nordmark@sun.com 1088 Robert E. Gilligan 1089 Intransa, Inc. 1090 2870 Zanker Rd., Suite 100 1091 San Jose, CA 95134 1093 Tel : +1 408 678 8600 1094 Fax : +1 408 678 8800 1095 Email : gilligan@intransa.com, gilligan@leaf.com 1097 9. Changes from RFC 2893 1099 The motivation for the bulk of these changes are to simplify the 1100 document to only contain the mechanisms of wide-spread use. 1102 RFC 2893 contains a mechanism called automatic tunneling. But a much 1103 more general mechanism is specified in RFC 3056 [RFC3056] which gives 1104 each node with a (global) IPv4 address a /48 IPv6 prefix i.e., enough 1105 for a whole site. 1107 The following changes have been performed since RFC 2893: 1109 - Removed references to A6 and retained AAAA. 1111 - Removed automatic tunneling and use of IPv4-compatible 1112 addresses. 1114 - Removed default Configured Tunnel using IPv4 "Anycast Address" 1116 - Removed Source Address Selection section since this is now 1117 covered by another document ([RFC3484]). 1119 - Removed brief mention of 6over4. 1121 - Split into normative and non-normative references and other 1122 reference cleanup. 1124 - Dropped "or equal" in if (IPv4 path MTU - 20) is less than or 1125 equal to 1280 1127 - Dropped this: However, IPv6 may be used in some environments 1128 where interoperability with IPv4 is not required. IPv6 nodes 1129 that are designed to be used in such environments need not use 1130 or even implement these mechanisms. 1132 - Described Static MTU and Dynamic MTU cases separately; clarified 1133 that the dynamic path MTU mechanism is OPTIONAL but if it is 1134 implemented it should follow the rules in section 3.2.2. 1136 - Specified Static MTU to default to a MTU of 1280 to 1480 bytes, 1137 and that this may be configurable. Discussed the issues with 1138 using Static MTU at more length. 1140 - Specified minimal rules for IPv4 reassembly and IPv6 MRU to 1141 enhance interoperability and to minimize blacholes. 1143 - Restated the "currently underway" language about Type-of- 1144 Service, and loosely point at [RFC2983] and [RFC3168]. 1146 - Fixed reference to Assigned Numbers to be to online version 1147 (with proper pointer to "Assigned Numbers is obsolete" RFC). 1149 - Clarified text about ingress filtering e.g. that it applies to 1150 packet delivered to transport protocols on the decapsulator as 1151 well as packets being forwarded by the decapsulator, and how the 1152 decapsulator's checks help when IPv4 and IPv6 ingress filtering 1153 is in place. 1155 - Removed unidirectional tunneling; assume all tunnels are 1156 bidirectional. 1158 - Removed the guidelines for advertising addresses in DNS as 1159 slightly out of scope, referring to another document for the 1160 details. 1162 - Removed the SHOULD requirement that the link-local addresses 1163 should be formed based on IPv4 addresses. 1165 - Added a SHOULD for implementing a knob to be able to set the 1166 source address of the tunnel, and add discussion why this is 1167 useful. 1169 - Added stronger wording for source address checks: both IPv4 and 1170 IPv6 source addresses MUST be checked, and RPF-like ingress 1171 filtering is optional. 1173 - Rewrote security considerations to be more precise about the 1174 threats of tunneling. 1176 - Added a note that using TTL=255 when encapsulating might be 1177 useful for decapsulation security checks later on. 1179 - Added more discussion in Section 3.2 why using an "infinite" 1180 IPv6 MTU leads to likely interoperability problems. 1182 - Added an explicit requirement that if both MTU determination 1183 methods are used, choosing one should be possible on a per- 1184 tunnel basis. 1186 - Clarified that ICMPv4 error handling is only applicable to 1187 dynamic MTU determination. 1189 - Removed/clarified DNS record filtering; an API is a SHOULD and 1190 if it does not exist, MUST NOT filter anything. Refer more 1191 strongly to RFC3484 on the DNS record ordering/processing. 1193 - Add a note that the destination IPv4 address could also be a 1194 multicast address. 1196 - Make it RECOMMENDED to provide a toggle to perform strict 1197 ingress filtering on an interface. 1199 - Made a lot of miscellaneous editorial cleanups. 1201 9.1. Changes from draft-ietf-v6ops-mech-v2-00 1203 [[ RFC-Editor note: remove the change history between the drafts 1204 before publication. ]] 1206 - Clarified in section 2.2 that there is no assumption that the 1207 DNS server knows the IPv4/IPv6 capabilities of the requesting 1208 node. 1210 - Clarified in section 2.2 that a filtering resolver might want to 1211 take into account external factors e.g., whether IPv6 interfaces 1212 have been configured on the node. 1214 - Clarified in section 2.3 that part of the motivation for the 1215 section is that this is the opposite of common DNS practices in 1216 IPv4; advertising unreachable IPv4 addresses in the DNS is 1217 common. 1219 - Removed the now artificial separation in a section on "common 1220 tunneling techniques" and "configured tunneling" to make one 1221 section on "configured tunneling". 1223 - Restructured the section on tunnel MTU to make the relationship 1224 between static tunnel MTU and dynamic tunnel MTU more clear. 1225 This includes fixing the unclear language about "must be 1280 1226 but may be configurable". 1228 - Added warning about manually configuring large tunnel MTUs 1229 causing excessive fragmentation. 1231 - Added warning about IPv4 PMTU blackholes when using dynamic MTU. 1233 - Clarified that when decapsulating the receiver must be liberal 1234 and allow for padding of the encapsulated packet. 1236 - Added example that when reflecting ICMPv4 errors as ICMPv6 1237 errors it would be appropriate to use ICMPv6 unreachable type 1238 with code "address unreachable" since an error from inside the 1239 tunnel is in effect a link specific problem from IPv6's 1240 perspective. 1242 - Consolidated the text on ingress filtering and created a 1243 separate section on the threat related to source address 1244 spoofing through open decapsulators. 1246 - Clarified "martian" filtering as follows: 0.0.0.0 should be 1247 0.0.0.0/8, same for 127. (per RFC1812), and elaborated that the 1248 broadcast address check includes both the 255.255.255.255 1249 address and all the broadcast addresses of the decapsulator. 1251 - Clarified that packets which fail the checks (such as verifying 1252 the IPv4 source address, martian, and ingress filtering) on the 1253 decapsulator should be silently dropped. 1255 - Clarified that while source link layer address options and 1256 target link layer address options are ignored in received ND 1257 packets, the ND packets themselves are processed as normal. 1259 9.2. Changes from draft-ietf-v6ops-mech-v2-01 1261 - Removed unidirectional tunnels; assume all the tunnels are 1262 bidirectional. 1264 - Removed the definition of IPv4-compatible IPv6 addresses. 1266 - Removed redundant text in the Hop Limit processing rules. 1268 - Removed the guidelines for advertising addresses in DNS as 1269 slightly out of scope, referring to another document for the 1270 details. 1272 - Removed the SHOULD requirement that the link-local addresses 1273 should be formed based on IPv4 addresses. 1275 - Added more discussion on the ICMPv4/6 Path MTU Discovery and the 1276 required number of packet drops. 1278 - Added a SHOULD for implementing a knob to be able to set the 1279 source address of the tunnel, and add discussion why this is 1280 useful. 1282 - Added stronger wording for source address checks: both IPv4 and 1283 IPv6 source addresses MUST be checked, and RPF-like ingress 1284 filtering is optional. 1286 - Rewrote security considerations to be more precise about the 1287 threats of tunneling. 1289 - Added a note that using TTL=255 when encapsulating might be 1290 useful for decapsulation security checks later on. 1292 - Added more discussion in Section 3.2 why using an "infinite" 1293 IPv6 MTU leads to likely interoperability problems. 1295 - Added an explicit requirement that if both MTU determination 1296 methods are used, choosing one should be possible on a per- 1297 tunnel basis. 1299 - Clarified that ICMPv4 error handling is only applicable to 1300 dynamic MTU determination. 1302 - Specified Static MTU to default to a MTU of 1280 to 1480 bytes, 1303 and that this may be configurable. Discussed the issues with 1304 using Static MTU at more length. 1306 - Specified minimal rules for IPv4 reassembly and IPv6 MRU to 1307 enhance interoperability and to minimize blacholes. 1309 - Made a lot of miscellaneous editorial cleanups. 1311 9.3. Changes from draft-ietf-v6ops-mech-v2-02 1313 - Removed DNS record filtering, and made it a SHOULD NOT; 1314 summarized the DNS record ordering. 1316 - Further clarified the usage of IPv6 RPF checks, as they 1317 typically require manual configuration. 1319 - Add a note on link-local address formation strategy to point out 1320 that minimizing the probability of renumbering may be desirable. 1322 - Clarify that with multiple IPv4 addresses, and the link-local 1323 address formed based on one of them, the selection can be done 1324 administratively or by the implementation. 1326 - Refer to separate documents on generic IPv6 security 1327 considerations and IPsec set-up details. 1329 - Clarify that the IPv4 end-point address can be predictable in 1330 some situations. 1332 - Add a note that the destination IPv4 address could also be a 1333 multicast address. 1335 - Make it RECOMMENDED to provide a toggle to perform strict 1336 ingress filtering on an interface. 1338 Intellectual Property Statement 1340 The IETF takes no position regarding the validity or scope of any 1341 Intellectual Property Rights or other rights that might be claimed to 1342 pertain to the implementation or use of the technology described in 1343 this document or the extent to which any license under such rights 1344 might or might not be available; nor does it represent that it has 1345 made any independent effort to identify any such rights. Information 1346 on the procedures with respect to rights in RFC documents can be 1347 found in BCP 78 and BCP 79. 1349 Copies of IPR disclosures made to the IETF Secretariat and any 1350 assurances of licenses to be made available, or the result of an 1351 attempt made to obtain a general license or permission for the use of 1352 such proprietary rights by implementers or users of this 1353 specification can be obtained from the IETF on-line IPR repository at 1354 http://www.ietf.org/ipr. 1356 The IETF invites any interested party to bring to its attention any 1357 copyrights, patents or patent applications, or other proprietary 1358 rights that may cover technology that may be required to implement 1359 this standard. Please address the information to the IETF at ietf- 1360 ipr@ietf.org. 1362 Disclaimer of Validity 1364 This document and the information contained herein are provided on an 1365 "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS 1366 OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY AND THE INTERNET 1367 ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS OR IMPLIED, 1368 INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE 1369 INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED 1370 WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. 1372 Copyright Statement 1374 Copyright (C) The Internet Society (2004). This document is subject 1375 to the rights, licenses and restrictions contained in BCP 78, and 1376 except as set forth therein, the authors retain all their rights. 1378 Acknowledgment 1380 Funding for the RFC Editor function is currently provided by the 1381 Internet Society.