idnits 2.17.1 draft-templin-intarea-seal-50.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 1 instance of lines with non-RFC6890-compliant IPv4 addresses in the document. If these are example addresses, they should be changed. == The 'Obsoletes: ' line in the draft header should list only the _numbers_ of the RFCs which will be obsoleted by this document (if approved); it should not include the word 'RFC' in the list. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (October 09, 2012) is 4210 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Unused Reference: 'RFC3971' is defined on line 1674, but no explicit reference was found in the text == Unused Reference: 'RFC4861' is defined on line 1681, but no explicit reference was found in the text == Unused Reference: 'MTUDWG' is defined on line 1728, but no explicit reference was found in the text == Unused Reference: 'RFC1063' is defined on line 1737, but no explicit reference was found in the text == Unused Reference: 'RFC1146' is defined on line 1744, but no explicit reference was found in the text == Unused Reference: 'RFC2675' is defined on line 1765, but no explicit reference was found in the text == Unused Reference: 'RFC2780' is defined on line 1772, but no explicit reference was found in the text == Unused Reference: 'RFC4191' is defined on line 1795, but no explicit reference was found in the text == Unused Reference: 'RFC4987' is defined on line 1816, but no explicit reference was found in the text == Unused Reference: 'RFC5226' is defined on line 1819, but no explicit reference was found in the text == Unused Reference: 'RFC5246' is defined on line 1823, but no explicit reference was found in the text == Unused Reference: 'RFC5445' is defined on line 1829, but no explicit reference was found in the text == Unused Reference: 'RFC6335' is defined on line 1845, but no explicit reference was found in the text == Unused Reference: 'TCP-IP' is defined on line 1865, but no explicit reference was found in the text ** Obsolete normative reference: RFC 2460 (Obsoleted by RFC 8200) == Outdated reference: A later version (-09) exists of draft-generic-6man-tunfrag-06 == Outdated reference: A later version (-12) exists of draft-ietf-6man-udpzero-06 == Outdated reference: A later version (-07) exists of draft-ietf-intarea-ipv4-id-update-05 == Outdated reference: A later version (-40) exists of draft-templin-intarea-vet-34 == Outdated reference: A later version (-16) exists of draft-templin-ironbis-11 -- Obsolete informational reference (is this intentional?): RFC 1063 (Obsoleted by RFC 1191) -- Obsolete informational reference (is this intentional?): RFC 1146 (Obsoleted by RFC 6247) -- Obsolete informational reference (is this intentional?): RFC 1981 (Obsoleted by RFC 8201) -- Obsolete informational reference (is this intentional?): RFC 5226 (Obsoleted by RFC 8126) -- Obsolete informational reference (is this intentional?): RFC 5246 (Obsoleted by RFC 8446) -- Obsolete informational reference (is this intentional?): RFC 6434 (Obsoleted by RFC 8504) Summary: 1 error (**), 0 flaws (~~), 22 warnings (==), 7 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group F. Templin, Ed. 3 Internet-Draft Boeing Research & Technology 4 Obsoletes: rfc5320 (if approved) October 09, 2012 5 Intended status: Informational 6 Expires: April 12, 2013 8 The Subnetwork Encapsulation and Adaptation Layer (SEAL) 9 draft-templin-intarea-seal-50.txt 11 Abstract 13 For the purpose of this document, a subnetwork is defined as a 14 virtual topology configured over a connected IP network routing 15 region and bounded by encapsulating border nodes. These virtual 16 topologies are manifested by tunnels that may span multiple IP and/or 17 sub-IP layer forwarding hops, where they may incur packet 18 duplication, packet reordering, source address spoofing and traversal 19 of links with diverse Maximum Transmission Units (MTUs). This 20 document specifies a Subnetwork Encapsulation and Adaptation Layer 21 (SEAL) that addresses these issues. 23 Status of this Memo 25 This Internet-Draft is submitted in full conformance with the 26 provisions of BCP 78 and BCP 79. 28 Internet-Drafts are working documents of the Internet Engineering 29 Task Force (IETF). Note that other groups may also distribute 30 working documents as Internet-Drafts. The list of current Internet- 31 Drafts is at http://datatracker.ietf.org/drafts/current/. 33 Internet-Drafts are draft documents valid for a maximum of six months 34 and may be updated, replaced, or obsoleted by other documents at any 35 time. It is inappropriate to use Internet-Drafts as reference 36 material or to cite them other than as "work in progress." 38 This Internet-Draft will expire on April 12, 2013. 40 Copyright Notice 42 Copyright (c) 2012 IETF Trust and the persons identified as the 43 document authors. All rights reserved. 45 This document is subject to BCP 78 and the IETF Trust's Legal 46 Provisions Relating to IETF Documents 47 (http://trustee.ietf.org/license-info) in effect on the date of 48 publication of this document. Please review these documents 49 carefully, as they describe your rights and restrictions with respect 50 to this document. Code Components extracted from this document must 51 include Simplified BSD License text as described in Section 4.e of 52 the Trust Legal Provisions and are provided without warranty as 53 described in the Simplified BSD License. 55 Table of Contents 57 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 58 1.1. Motivation . . . . . . . . . . . . . . . . . . . . . . . . 4 59 1.2. Approach . . . . . . . . . . . . . . . . . . . . . . . . . 6 60 1.3. Differences with RFC5320 . . . . . . . . . . . . . . . . . 7 61 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 8 62 3. Requirements . . . . . . . . . . . . . . . . . . . . . . . . . 10 63 4. Applicability Statement . . . . . . . . . . . . . . . . . . . 10 64 5. SEAL Specification . . . . . . . . . . . . . . . . . . . . . . 11 65 5.1. SEAL Tunnel Model . . . . . . . . . . . . . . . . . . . . 11 66 5.2. SEAL Model of Operation . . . . . . . . . . . . . . . . . 11 67 5.3. SEAL Header and Trailer Format . . . . . . . . . . . . . . 13 68 5.4. ITE Specification . . . . . . . . . . . . . . . . . . . . 15 69 5.4.1. Tunnel Interface MTU . . . . . . . . . . . . . . . . . 15 70 5.4.2. Tunnel Neighbor Soft State . . . . . . . . . . . . . . 16 71 5.4.3. SEAL Layer Pre-Processing . . . . . . . . . . . . . . 17 72 5.4.4. SEAL Encapsulation and Segmentation . . . . . . . . . 19 73 5.4.5. Outer Encapsulation . . . . . . . . . . . . . . . . . 20 74 5.4.6. Path Probing and ETE Reachability Verification . . . . 21 75 5.4.7. Processing ICMP Messages . . . . . . . . . . . . . . . 21 76 5.4.8. IPv4 Middlebox Reassembly Testing . . . . . . . . . . 22 77 5.4.9. Stateful MTU Determination . . . . . . . . . . . . . . 24 78 5.4.10. Detecting Path MTU Changes . . . . . . . . . . . . . . 24 79 5.5. ETE Specification . . . . . . . . . . . . . . . . . . . . 24 80 5.5.1. Minimum Reassembly Buffer Requirements . . . . . . . . 24 81 5.5.2. Tunnel Neighbor Soft State . . . . . . . . . . . . . . 25 82 5.5.3. IP-Layer Reassembly . . . . . . . . . . . . . . . . . 25 83 5.5.4. Decapsulation and Re-Encapsulation . . . . . . . . . . 26 84 5.6. The SEAL Control Message Protocol (SCMP) . . . . . . . . . 27 85 5.6.1. Generating SCMP Error Messages . . . . . . . . . . . . 28 86 5.6.2. Processing SCMP Error Messages . . . . . . . . . . . . 30 87 6. Link Requirements . . . . . . . . . . . . . . . . . . . . . . 32 88 7. End System Requirements . . . . . . . . . . . . . . . . . . . 32 89 8. Router Requirements . . . . . . . . . . . . . . . . . . . . . 32 90 9. Nested Encapsulation Considerations . . . . . . . . . . . . . 33 91 10. Reliability Considerations . . . . . . . . . . . . . . . . . . 33 92 11. Integrity Considerations . . . . . . . . . . . . . . . . . . . 34 93 12. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 34 94 13. Security Considerations . . . . . . . . . . . . . . . . . . . 34 95 14. Related Work . . . . . . . . . . . . . . . . . . . . . . . . . 35 96 15. Implementation Status . . . . . . . . . . . . . . . . . . . . 36 97 16. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 36 98 17. References . . . . . . . . . . . . . . . . . . . . . . . . . . 36 99 17.1. Normative References . . . . . . . . . . . . . . . . . . . 36 100 17.2. Informative References . . . . . . . . . . . . . . . . . . 37 101 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . . 41 103 1. Introduction 105 As Internet technology and communication has grown and matured, many 106 techniques have developed that use virtual topologies (manifested by 107 tunnels of one form or another) over an actual network that supports 108 the Internet Protocol (IP) [RFC0791][RFC2460]. Those virtual 109 topologies have elements that appear as one hop in the virtual 110 topology, but are actually multiple IP or sub-IP layer hops. These 111 multiple hops often have quite diverse properties that are often not 112 even visible to the endpoints of the virtual hop. This introduces 113 failure modes that are not dealt with well in current approaches. 115 The use of IP encapsulation (also known as "tunneling") has long been 116 considered as the means for creating such virtual topologies. 117 However, the encapsulation headers often include insufficiently 118 provisioned per-packet identification values. IP encapsulation also 119 allows an attacker to produce encapsulated packets with spoofed 120 source addresses even if the source address in the encapsulating 121 header cannot be spoofed. A denial-of-service vector that is not 122 possible in non-tunneled subnetworks is therefore presented. 124 Additionally, the insertion of an outer IP header reduces the 125 effective path MTU visible to the inner network layer. When IPv6 is 126 used as the encapsulation protocol, original sources expect to be 127 informed of the MTU limitation through IPv6 Path MTU discovery 128 (PMTUD) [RFC1981]. When IPv4 is used, this reduced MTU can be 129 accommodated through the use of IPv4 fragmentation, but unmitigated 130 in-the-network fragmentation has been found to be harmful through 131 operational experience and studies conducted over the course of many 132 years [FRAG][FOLK][RFC4963]. Additionally, classical IPv4 PMTUD 133 [RFC1191] has known operational issues that are exacerbated by in- 134 the-network tunnels [RFC2923][RFC4459]. 136 The following subsections present further details on the motivation 137 and approach for addressing these issues. 139 1.1. Motivation 141 Before discussing the approach, it is necessary to first understand 142 the problems. In both the Internet and private-use networks today, 143 IP is ubiquitously deployed as the Layer 3 protocol. The primary 144 functions of IP are to provide for routing, addressing, and a 145 fragmentation and reassembly capability used to accommodate links 146 with diverse MTUs. While it is well known that the IP address space 147 is rapidly becoming depleted, there is a lesser-known but growing 148 consensus that other IP protocol limitations have already or may soon 149 become problematic. 151 First, the Internet historically provided no means for discerning 152 whether the source addresses of IP packets are authentic. This 153 shortcoming is being addressed more and more through the deployment 154 of site border router ingress filters [RFC2827], however the use of 155 encapsulation provides a vector for an attacker to circumvent 156 filtering for the encapsulated packet even if filtering is correctly 157 applied to the encapsulation header. Secondly, the IP header does 158 not include a well-behaved identification value unless the source has 159 included a fragment header for IPv6 or unless the source permits 160 fragmentation for IPv4. These limitations preclude an efficient 161 means for routers to detect duplicate packets and packets that have 162 been re-ordered within the subnetwork. 164 For IPv4 encapsulation, when fragmentation is permitted the header 165 includes a 16-bit Identification field, meaning that at most 2^16 166 unique packets with the same (source, destination, protocol)-tuple 167 can be active in the Internet at the same time 168 [I-D.ietf-intarea-ipv4-id-update]. (When middleboxes such as Network 169 Address Translators (NATs) re-write the Identification field to 170 random values, the number of unique packets is even further reduced.) 171 Due to the escalating deployment of high-speed links, however, these 172 numbers have become too small by several orders of magnitude for high 173 data rate packet sources such as tunnel endpoints [RFC4963]. 175 Furthermore, there are many well-known limitations pertaining to IPv4 176 fragmentation and reassembly - even to the point that it has been 177 deemed "harmful" in both classic and modern-day studies (see above). 178 In particular, IPv4 fragmentation raises issues ranging from minor 179 annoyances (e.g., in-the-network router fragmentation [RFC1981]) to 180 the potential for major integrity issues (e.g., mis-association of 181 the fragments of multiple IP packets during reassembly [RFC4963]). 183 As a result of these perceived limitations, a fragmentation-avoiding 184 technique for discovering the MTU of the forward path from a source 185 to a destination node was devised through the deliberations of the 186 Path MTU Discovery Working Group (PMTUDWG) during the late 1980's 187 through early 1990's which resulted in the publication of [RFC1191]. 188 In this negative feedback-based method, the source node provides 189 explicit instructions to routers in the path to discard the packet 190 and return an ICMP error message if an MTU restriction is 191 encountered. However, this approach has several serious shortcomings 192 that lead to an overall "brittleness" [RFC2923]. 194 In particular, site border routers in the Internet have been known to 195 discard ICMP error messages coming from the outside world. This is 196 due in large part to the fact that malicious spoofing of error 197 messages in the Internet is trivial since there is no way to 198 authenticate the source of the messages [RFC5927]. Furthermore, when 199 a source node that requires ICMP error message feedback when a packet 200 is dropped due to an MTU restriction does not receive the messages, a 201 path MTU-related black hole occurs. This means that the source will 202 continue to send packets that are too large and never receive an 203 indication from the network that they are being discarded. This 204 behavior has been confirmed through documented studies showing clear 205 evidence of PMTUD failures for both IPv4 and IPv6 in the Internet 206 today [TBIT][WAND][SIGCOMM]. 208 The issues with both IP fragmentation and this "classical" PMTUD 209 method are exacerbated further when IP tunneling is used [RFC4459]. 210 For example, an ingress tunnel endpoint (ITE) may be required to 211 forward encapsulated packets into the subnetwork on behalf of 212 hundreds, thousands, or even more original sources. If the ITE 213 allows IP fragmentation on the encapsulated packets, persistent 214 fragmentation could lead to undetected data corruption due to 215 Identification field wrapping and/or reassembly congestion at the 216 ETE. If the ITE instead uses classical IP PMTUD it must rely on ICMP 217 error messages coming from the subnetwork that may be suspect, 218 subject to loss due to filtering middleboxes, or insufficiently 219 provisioned for translation into error messages to be returned to the 220 original sources. 222 Although recent works have led to the development of a positive 223 feedback-based end-to-end MTU determination scheme [RFC4821], they do 224 not excuse tunnels from accounting for the encapsulation overhead 225 they add to packets. Moreover, in current practice existing 226 tunneling protocols mask the MTU issues by selecting a "lowest common 227 denominator" MTU that may be much smaller than necessary for most 228 paths and difficult to change at a later date. Therefore, a new 229 approach to accommodate tunnels over links with diverse MTUs is 230 necessary. 232 1.2. Approach 234 For the purpose of this document, a subnetwork is defined as a 235 virtual topology configured over a connected network routing region 236 and bounded by encapsulating border nodes. Example connected network 237 routing regions include Mobile Ad hoc Networks (MANETs), enterprise 238 networks and the global public Internet itself. Subnetwork border 239 nodes forward unicast and multicast packets over the virtual topology 240 across multiple IP and/or sub-IP layer forwarding hops that may 241 introduce packet duplication and/or traverse links with diverse 242 Maximum Transmission Units (MTUs). 244 This document introduces a Subnetwork Encapsulation and Adaptation 245 Layer (SEAL) for tunneling inner network layer protocol packets over 246 IP subnetworks that connect Ingress and Egress Tunnel Endpoints 247 (ITEs/ETEs) of border nodes. It provides a modular specification 248 designed to be tailored to specific associated tunneling protocols. 249 (A transport-mode of operation is also possible, but out of scope for 250 this document.) 252 SEAL provides a mid-layer encapsulation that accommodates links with 253 diverse MTUs, and allows routers in the subnetwork to perform 254 efficient duplicate packet and packet reordering detection. The 255 encapsulation further ensures data origin authentication, packet 256 header integrity and anti-replay in environments in which these 257 functions are necessary. 259 SEAL treats tunnels that traverse the subnetwork as ordinary links 260 that must support network layer services. Moreover, SEAL provides 261 dynamic mechanisms to ensure a maximal path MTU over the tunnel. 262 This is in contrast to static approaches which avoid MTU issues by 263 selecting a lowest common denominator MTU value that may be overly 264 conservative for the vast majority of tunnel paths and difficult to 265 change even when larger MTUs become available. 267 The following sections provide the SEAL normative specifications, 268 while the appendices present non-normative additional considerations. 270 1.3. Differences with RFC5320 272 This specification of SEAL is descended from an experimental 273 independent RFC publication of the same name [RFC5320]. However, 274 this specification introduces a number of important differences from 275 the earlier publication. 277 First, this specification includes a protocol version field in the 278 SEAL header whereas [RFC5320] does not, and therefore cannot be 279 updated by future revisions. This specification therefore obsoletes 280 (i.e., and does not update) [RFC5320]. 282 Secondly, [RFC5320] forms a 32-bit Identification value by 283 concatenating the 16-bit IPv4 Identification field with a 16-bit 284 Identification "extension" field in the SEAL header. This means that 285 [RFC5320] can only operate over IPv4 networks (since IPv6 headers do 286 not include a 16-bit version number) and that the SEAL Identification 287 value can be corrupted if the Identification in the outer IPv4 header 288 is rewritten. In contrast, this specification includes a 32-bit 289 Identification value that is independent of any identification fields 290 found in the inner or outer IP headers, and is therefore compatible 291 with any inner and outer IP protocol version combinations. 293 Additionally, the SEAL segmentation and reassembly procedures defined 294 in [RFC5320] differ significantly from those found in this 295 specification. In particular, this specification defines a 6-bit 296 Offset field that allows for smaller segment sizes when SEAL 297 segmentation is necessary (e.g., in order to observe the IPv4 minimum 298 MTU of 68 bytes). In contrast, [RFC5320] includes a 3-bit Segment 299 field and performs reassembly through concatenation of consecutive 300 segments. 302 The SEAL header in this specification also includes an optional 303 Integrity Check Vector (ICV) that can be used to digitally sign the 304 SEAL header and the leading portion of the encapsulated inner packet. 305 This allows for a lightweight integrity check and a loose data origin 306 authentication capability. The header further includes new control 307 bits as well as a link identification and encapsulation level field 308 for additional control capabilities. 310 Finally, this version of SEAL includes a new messaging protocol known 311 as the SEAL Control Message Protocol (SCMP), whereas [RFC5320] 312 performs signalling through the use of SEAL-encapsulated ICMP 313 messages. The use of SCMP allows SEAL-specific departures from ICMP, 314 as well as a control messaging capability that extends to other 315 specifications, including Virtual Enterprise Traversal (VET) 316 [I-D.templin-intarea-vet]. 318 2. Terminology 320 The following terms are defined within the scope of this document: 322 subnetwork 323 a virtual topology configured over a connected network routing 324 region and bounded by encapsulating border nodes. 326 IP 327 used to generically refer to either Internet Protocol (IP) 328 version, i.e., IPv4 or IPv6. 330 Ingress Tunnel Endpoint (ITE) 331 a virtual interface over which an encapsulating border node (host 332 or router) sends encapsulated packets into the subnetwork. 334 Egress Tunnel Endpoint (ETE) 335 a virtual interface over which an encapsulating border node (host 336 or router) receives encapsulated packets from the subnetwork. 338 SEAL Path 339 a subnetwork path from an ITE to an ETE beginning with an 340 underlying link of the ITE as the first hop. Note that, if the 341 ITE's interface connection to the underlying link assigns multiple 342 IP addresses, each address represents a separate SEAL path. 344 inner packet 345 an unencapsulated network layer protocol packet (e.g., IPv4 346 [RFC0791], OSI/CLNP [RFC0994], IPv6 [RFC2460], etc.) before any 347 outer encapsulations are added. Internet protocol numbers that 348 identify inner packets are found in the IANA Internet Protocol 349 registry [RFC3232]. SEAL protocol packets that incur an 350 additional layer of SEAL encapsulation are also considered inner 351 packets. 353 outer IP packet 354 a packet resulting from adding an outer IP header (and possibly 355 other outer headers) to a SEAL-encapsulated inner packet. 357 packet-in-error 358 the leading portion of an invoking data packet encapsulated in the 359 body of an error control message (e.g., an ICMPv4 [RFC0792] error 360 message, an ICMPv6 [RFC4443] error message, etc.). 362 Packet Too Big (PTB) message 363 a control plane message indicating an MTU restriction (e.g., an 364 ICMPv6 "Packet Too Big" message [RFC4443], an ICMPv4 365 "Fragmentation Needed" message [RFC0792], etc.). 367 Don't Fragment (DF) bit 368 a bit that indicates whether the packet may be fragmented by the 369 network. The DF bit is explicitly included in the IPv4 header 370 [RFC0791] and may be set to '0' to allow fragmentation or '1' to 371 disallow further in-network fragmentation. The bit is absent from 372 the IPv6 header [RFC2460], but implicitly set to '1' becauuse 373 fragmentation can occur only at IPv6 sources. 375 The following abbreviations correspond to terms used within this 376 document and/or elsewhere in common Internetworking nomenclature: 378 HLEN - the length of the SEAL header plus outer headers 380 ICV - Integrity Check Vector 382 MTU - Maximum Transmission Unit 384 SCMP - the SEAL Control Message Protocol 386 SDU - SCMP Destination Unreachable message 388 SPP - SCMP Parameter Problem message 389 SPTB - SCMP Packet Too Big message 391 SEAL - Subnetwork Encapsulation and Adaptation Layer 393 TE - Tunnel Endpoint (i.e., either ingress or egress) 395 VET - Virtual Enterprise Traversal 397 3. Requirements 399 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 400 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 401 document are to be interpreted as described in [RFC2119]. When used 402 in lower case (e.g., must, must not, etc.), these words MUST NOT be 403 interpreted as described in [RFC2119], but are rather interpreted as 404 they would be in common English. 406 4. Applicability Statement 408 SEAL was originally motivated by the specific case of subnetwork 409 abstraction for Mobile Ad hoc Networks (MANETs), however the domain 410 of applicability also extends to subnetwork abstractions over 411 enterprise networks, ISP networks, SOHO networks, the global public 412 Internet itself, and any other connected network routing region. 414 SEAL provides a network sublayer for encapsulation of an inner 415 network layer packet within outer encapsulating headers. SEAL can 416 also be used as a sublayer within a transport layer protocol data 417 payload, where transport layer encapsulation is typically used for 418 Network Address Translator (NAT) traversal as well as operation over 419 subnetworks that give preferential treatment to certain "core" 420 Internet protocols (e.g., TCP, UDP, etc.). The SEAL header is 421 processed the same as for IPv6 extension headers, i.e., it is not 422 part of the outer IP header but rather allows for the creation of an 423 arbitrarily extensible chain of headers in the same way that IPv6 424 does. 426 To accommodate MTU diversity, the Ingress Tunnel Endpoint (ITE) may 427 need to perform any necessary fragmentation which the Egress Tunnel 428 Endpoint (ETE) must reassemble. The ETE further acts as a passive 429 observer that informs the ITE of any packet size limitations. This 430 allows the ITE to return appropriate PMTUD feedback even if the 431 network path between the ITE and ETE filters ICMP messages. 433 SEAL further provides mechanisms to ensure data origin 434 authentication, packet header integrity, and anti-replay. The SEAL 435 framework is therefore similar to the IP Security (IPsec) 436 Authentication Header (AH) [RFC4301][RFC4302], however it provides 437 only minimal hop-by-hop authenticating services while leaving full 438 data integrity, authentication and confidentiality services as an 439 end-to-end consideration. While SEAL performs data origin 440 authentication, the origin site must also perform the necessary 441 ingress filtering in order to provide full source address 442 verification [I-D.ietf-savi-framework]. 444 In many aspects, SEAL also very closely resembles the Generic Routing 445 Encapsulation (GRE) framework [RFC1701]. SEAL can therefore be 446 applied in the same use cases that are traditionally addressed by 447 GRE, but goes beyond GRE to also provide additional capabilities 448 (e.,g., path MTU accommodation, data origin authentication, etc.) as 449 described in this document. 451 5. SEAL Specification 453 The following sections specify the operation of SEAL: 455 5.1. SEAL Tunnel Model 457 SEAL is an encapsulation sublayer used within point-to-point and non- 458 broadcast, multiple access (NBMA) tunnels. Each SEAL path is 459 configured over one or more underlying interfaces attached to 460 subnetwork links. The SEAL tunnel connects an ITE to one or more ETE 461 "neighbors" via encapsulation across an underlying subnetwork, where 462 the tunnel neighbor relationship may be either unidirectional or 463 bidirectional. 465 A unidirectional tunnel neighbor relationship allows the near end ITE 466 to send data packets forward to the far end ETE, while the ETE only 467 returns control messages when necessary. A bidirectional tunnel 468 neighbor relationship is one over which both TEs can exchange both 469 data and control messages. 471 Implications of the SEAL unidirectional and bidirectional models are 472 the same as discussed in [I-D.templin-intarea-vet]. 474 5.2. SEAL Model of Operation 476 SEAL-enabled ITEs encapsulate each inner packet in a SEAL header and 477 any outer header encapsulations as shown in Figure 1: 479 +--------------------+ 480 ~ outer IP header ~ 481 +--------------------+ 482 ~ other outer hdrs ~ 483 +--------------------+ 484 ~ SEAL Header ~ 485 +--------------------+ +--------------------+ 486 | | --> | | 487 ~ Inner ~ --> ~ Inner ~ 488 ~ Packet ~ --> ~ Packet ~ 489 | | --> | | 490 +--------------------+ +----------+---------+ 492 Figure 1: SEAL Encapsulation 494 The ITE inserts the SEAL header according to the specific tunneling 495 protocol. For simple encapsulation of an inner network layer packet 496 within an outer IP header, the ITE inserts the SEAL header following 497 the outer IP header and before the inner packet as: IP/SEAL/{inner 498 packet}. 500 For encapsulations over transports such as UDP, the ITE inserts the 501 SEAL header following the outer transport layer header and before the 502 inner packet, e.g., as IP/UDP/SEAL/{inner packet}. In that case, the 503 UDP header is seen as an "other outer header" as depicted in Figure 1 504 and the outer IP and transport layer headers are together seen as the 505 outer encapsulation headers. 507 SEAL supports both "nested" tunneling and "re-encapsulating" 508 tunneling. Nested tunneling occurs when a first tunnel is 509 encapsulated within a second tunnel, which may then further be 510 encapsulated within additional tunnels. Nested tunneling can be 511 useful, and stands in contrast to "recursive" tunneling which is an 512 anomalous condition incurred due to misconfiguration or a routing 513 loop. Considerations for nested tunneling and avoiding recursive 514 tunneling are discussed in Section 4 of [RFC2473]. 516 Re-encapsulating tunneling occurs when a packet arrives at a first 517 ETE, which then acts as an ITE to re-encapsulate and forward the 518 packet to a second ETE connected to the same subnetwork. In that 519 case each ITE/ETE transition represents a segment of a bridged path 520 between the ITE nearest the source and the ETE nearest the 521 destination. Combinations of nested and re-encapsulating tunneling 522 are also naturally supported by SEAL. 524 The SEAL ITE considers each underlying interface as the ingress 525 attachment point to a SEAL path to the ETE. The ITE therefore may 526 experience different path MTUs on different SEAL paths. 528 Finally, the SEAL ITE ensures that the inner network layer protocol 529 will see a minimum MTU of 1500 bytes over each SEAL path regardless 530 of the outer network layer protocol version, i.e., even if a small 531 amount of fragmentation and reassembly are necessary. This is 532 necessary to avoid path MTU "black holes" for the minimum MTU 533 configured by the vast majority of links in the Internet. 535 5.3. SEAL Header and Trailer Format 537 The SEAL header is formatted as follows: 539 0 1 2 3 540 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 541 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 542 |VER|C|A|R|L|I|V|X|M| Offset | NEXTHDR | LINK_ID |LEVEL| 543 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 544 | Identification (optional) | 545 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 546 | Integrity Check Vector (ICV) (optional) | 547 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 549 Figure 2: SEAL Header Format 551 VER (2) 552 a 2-bit version field. This document specifies Version 0 of the 553 SEAL protocol, i.e., the VER field encodes the value 0. 555 C (1) 556 the "Control/Data" bit. Set to 1 by the ITE in SEAL Control 557 Message Protocol (SCMP) control messages, and set to 0 in ordinary 558 data packets. 560 A (1) 561 the "Acknowledgement Requested" bit. Set to 1 by the ITE in SEAL 562 data packets for which it wishes to receive an explicit 563 acknowledgement from the ETE. 565 R (1) 566 the "Redirects Permitted" bit (reserved for use by VET: 567 [I-D.templin-intarea-vet]). 569 L (1) 570 the "Rate Limit" bit. 572 I (1) 573 the "Identification Included" bit. 575 V (1) 576 the "ICV included" bit. 578 X (1) a 1-bit reserved field. 580 M (1) the "More Segments" bit. Set to 1 in a non-final segment and 581 set to 0 in the final segment of the SEAL packet. 583 Offset (6) a 6-bit Offset field. Set to 0 in the first segment of a 584 segmented SEAL packet. Set to an integral number of 32 byte 585 blocks in subsequent segments (e.g., an Offset of 10 indicates a 586 block that begins at the 320th byte in the packet). 588 NEXTHDR (8) an 8-bit field that encodes the next header Internet 589 Protocol number the same as for the IPv4 protocol and IPv6 next 590 header fields. 592 LINK_ID (5) 593 a 5-bit link identification value, set to a unique value by the 594 ITE for each SEAL path over which it will send encapsulated 595 packets to the ETE (up to 32 SEAL paths per ETE are therefore 596 supported). Note that, if the ITE's interface connection to the 597 underlying link assigns multiple IP addresses, each address 598 represents a separate SEAL path that must be assigned a separate 599 LINK_ID. 601 LEVEL (3) 602 a 3-bit nesting level; use to limit the number of tunnel nesting 603 levels. Set to an integer value up to 7 in the innermost SEAL 604 encapsulation, and decremented by 1 for each successive additional 605 SEAL encapsulation nesting level. Up to 8 levels of nesting are 606 therefore supported. 608 Identification (32) 609 an optional 32-bit per-packet identification field; present when 610 I==1. Set to a 32-bit value (beginning with 0) that is 611 monotonically-incremented for each SEAL packet transmitted to this 612 ETE. 614 Integrity Check Vector (ICV) (32) 615 an optional 32-bit header integrity check value; present when 616 V==1. Covers the leading 128 bytes of the packet beginning with 617 the SEAL header. The value 128 is chosen so that at least the 618 SEAL header as well as the inner packet network and transport 619 layer headers are covered by the integrity check. 621 5.4. ITE Specification 623 5.4.1. Tunnel Interface MTU 625 The tunnel interface must present a constant MTU value to the inner 626 network layer as the size for admission of inner packets into the 627 interface. Since NBMA tunnel virtual interfaces may support a large 628 set of SEAL paths that accept widely varying maximum packet sizes, 629 however, a number of factors should be taken into consideration when 630 selecting a tunnel interface MTU. 632 Due to the ubiquitous deployment of standard Ethernet and similar 633 networking gear, the nominal Internet cell size has become 1500 634 bytes; this is the de facto size that end systems have come to expect 635 will either be delivered by the network without loss due to an MTU 636 restriction on the path or a suitable ICMP Packet Too Big (PTB) 637 message returned. When large packets sent by end systems incur 638 additional encapsulation at an ITE, however, they may be dropped 639 silently within the tunnel since the network may not always deliver 640 the necessary PTBs [RFC2923]. The ITE SHOULD therefore set a tunnel 641 interface MTU of at least 1500 bytes. 643 The inner network layer protocol consults the tunnel interface MTU 644 when admitting a packet into the interface. For non-SEAL inner IPv4 645 packets with the IPv4 Don't Fragment (DF) cleared (i.e, DF==0), if 646 the packet is larger than the tunnel interface MTU the inner IPv4 647 layer uses IPv4 fragmentation to break the packet into fragments no 648 larger than the tunnel interface MTU. The ITE then admits each 649 fragment into the interface as an independent packet. 651 For all other inner packets, the inner network layer admits the 652 packet if it is no larger than the tunnel interface MTU; otherwise, 653 it drops the packet and sends a PTB error message to the source with 654 the MTU value set to the tunnel interface MTU. The message contains 655 as much of the invoking packet as possible without the entire message 656 exceeding the network layer MINMTU size. 658 The ITE can alternatively set an indefinite MTU on the tunnel 659 interface such that all inner packets are admitted into the interface 660 regardless of their size. For ITEs that host applications that use 661 the tunnel interface directly, this option must be carefully 662 coordinated with protocol stack upper layers since some upper layer 663 protocols (e.g., TCP) derive their packet sizing parameters from the 664 MTU of the outgoing interface and as such may select too large an 665 initial size. This is not a problem for upper layers that use 666 conservative initial maximum segment size estimates and/or when the 667 tunnel interface can reduce the upper layer's maximum segment size, 668 e.g., by reducing the size advertised in the MSS option of outgoing 669 TCP messages (sometimes known as "MSS clamping"). 671 In light of the above considerations, the ITE should configure an 672 indefinite MTU on tunnel *router* interfaces so that SEAL performs 673 all subnetwork adaptation from within the interface as specified in 674 Section 5.4.3. The ITE can instead set a smaller MTU on tunnel 675 *host* interfaces (e.g., the smallest MTU among all of the underlying 676 links minus the size of the encapsulation headers) but SHOULD NOT set 677 an MTU smaller than 1500 bytes. 679 5.4.2. Tunnel Neighbor Soft State 681 The tunnel virtual interface maintains a number of soft state 682 variables for each ETE and for each SEAL path. 684 When per-packet identification is required, the ITE maintains a per 685 ETE window of Identification values for the packets it has recently 686 sent to this ETE. The ITE then sets a variable "USE_ID" to TRUE, and 687 includes an Identification in each packet it sends to this ETE; 688 otherwise, it sets USE_ID to FALSE. 690 When data origin authentication and integrity checking is required, 691 the ITE also maintains a per ETE integrity check vector (ICV) 692 calculation algorithm and a symmetric secret key to calculate the ICV 693 in each packet it will send to this ETE. The ITE then sets a 694 variable "USE_ICV" to TRUE, and includes an ICV in each packet it 695 sends to this ETE; otherwise, it sets USE_ICV to FALSE. 697 For IPv4 SEAL paths, the ITE further maintains a variable 698 "RATE_LIMIT" initialized to FALSE. If the SEAL path subsequently 699 exhibits unavoidable IPv4 fragmentation the ETE sets RATE_LIMIT to 700 TRUE. 702 For each SEAL path, the ITE must also account for encapsulation 703 header lengths. The ITE therefore maintains the per SEAL path 704 constant values "SHLEN" set to the length of the SEAL header, "THLEN" 705 set to the length of the outer encapsulating transport layer headers 706 (or 0 if outer transport layer encapsulation is not used), "IHLEN" 707 set to the length of the outer IP layer header, and "HLEN" set to 708 (SHLEN+THLEN+IHLEN). (The ITE must include the length of the 709 uncompressed headers even if header compression is enabled when 710 calculating these lengths.) In addition, the ITE maintains a per 711 SEAL path variable "PATH_MTU" initialized to the maximum of 1500 712 bytes and the MTU of the underlying link minus HLEN. (Thereafter, 713 the ITE must not reduce PATH_MTU to a value smaller than 1500 bytes.) 715 The ITE may instead maintain the packet sizing variables and 716 constants as per ETE (rather than per SEAL path) values. In that 717 case, the values reflect the lowest-common-denominator size across 718 all of the SEAL paths associated with this ETE. 720 5.4.3. SEAL Layer Pre-Processing 722 The SEAL layer is logically positioned between the inner and outer 723 network protocol layers, where the inner layer is seen as the (true) 724 network layer and the outer layer is seen as the (virtual) data link 725 layer. Each packet to be processed by the SEAL layer is either 726 admitted into the tunnel interface by the inner network layer 727 protocol as described in Section 5.4.1 or is undergoing re- 728 encapsulation from within the tunnel interface. The SEAL layer sees 729 the former class of packets as inner packets that include inner 730 network and transport layer headers, and sees the latter class of 731 packets as transitional SEAL packets that include the outer and SEAL 732 layer headers that were inserted by the previous hop SEAL ITE. For 733 these transitional packets, the SEAL layer re-encapsulates the packet 734 with new outer and SEAL layer headers when it forwards the packet to 735 the next hop SEAL ITE. 737 In both cases, the ITE sets a variable 'MINMTU' to the minimum MTU 738 for the SEAL path over which encapsulated packets will travel. For 739 IPv6 paths the ITE sets MINMTU=1280 (see: [RFC2460]) and for IPv4 740 paths the ITE sets MINMTU=576 even though the true MINMTU for IPv4 is 741 only 68 bytes (see: [RFC0791]). The ITE can also set MINMTU to a 742 larger value (but no larger than 1500 bytes) if there is reason to 743 believe that the path MTU is larger. If this value proves too large, 744 the ITE will receive PTB message feedback either from the ETE or from 745 a router on the path and will be able to reduce its MINMTU to a 746 smaller value. 748 We now discuss the SEAL layer pre-processing actions for these two 749 classes of packets. 751 5.4.3.1. Inner Packet Pre-Processing 753 For each inner packet admitted into the tunnel interface, if the 754 packet is itself a SEAL packet (i.e., one with the port number for 755 SEAL in the transport layer header or one with the protocol number 756 for SEAL in the IP layer header) and the LEVEL field of the SEAL 757 header contains the value 0, the ITE silently discards the packet. 759 Otherwise, for non-SEAL IPv4 inner packets with DF==0 in the IP 760 header and IPv6 inner packets with a fragment header and with (MF=0; 761 Offset=0), if the packet is larger than (MINMTU-HLEN) the ITE uses IP 762 fragmentation to fragment the packet into N roughly equal-length 763 pieces, where N is minimized and each fragment is significantly 764 smaller than (MINMTU-HLEN) to allow for additional encapsulations in 765 the path. The ITE then submits each fragment for SEAL encapsulation 766 as specified in Section 5.4.4. 768 For all other inner packets: 770 o if the packet is no larger than larger than (MINMTU-HLEN), the ITE 771 submits it for SEAL encapsulation as specified in Section 5.4.4 773 o if the packet is larger than (MINMTU-HLEN) but no larger than 1500 774 bytes, the ITE submits the packet for SEAL segmentation as 775 specifed in Section 5.4.4. For IPv6 packets, the ITE also sends 776 an ICMPv6 PTB message with the MTU field set to (MIN(MINMTU,1280)- 777 HLEN) and the Code field set to 1 toward the original source 778 subject to rate limiting (see: [I-D.generic-6man-tunfrag]). This 779 PTB message is advisory in nature and does not represent a packet 780 loss; instead, it instructs the original IPv6 host to begin 781 fragmenting future packets that it will send to this destination. 783 o if the packet is larger than 1500 bytes but no larger than 784 PATH_MTU for the corresponding SEAL path the ITE submits it for 785 SEAL encapsulation as specified in Section 5.4.4. 787 o if the packet is larger than PATH_MTU, the ITE drops the packet 788 and sends an ordinary PTB message appropriate to the inner 789 protocol version with the MTU field set to PATH_MTU. (For IPv4 790 SEAL packets with DF==0, the ITE should set DF=1 and re-calculate 791 the IPv4 header checksum before generating the PTB message in 792 order to avoid bogon filters.) After sending the PTB message, the 793 ITE discards the inner packet. 795 5.4.3.2. Transitional SEAL Packet Pre-Processing 797 For each transitional packet that is processed by the SEAL layer from 798 within the tunnel interface, the ITE sets aside the SEAL 799 encapsulation headers that were added by the previous hop and 800 examines the inner packet as follows: 802 o if the packet is no larger than larger than (MINMTU-HLEN), the ITE 803 submits it for SEAL encapsulation as specified in Section 5.4.4. 805 o if the packet is larger than (MINMTU-HLEN) but no larger than 1500 806 bytes, the ITE submits the packet for SEAL segmentation as 807 specifed in Section 5.4.4. The ITE also sends an SCMP PTB (SPTB) 808 message toward the previous hop SEAL ITE with the MTU field set to 809 (MINMTU-HLEN) subject to rate limiting (see: Section 5.6.1.1). 810 This SPTB message is advisory in nature and does not represent a 811 packet loss; instead, it instructs the previous hop ITE to begin 812 sending smaller SEAL segments. 814 o if the packet is larger than 1500 bytes but no larger than 815 PATH_MTU for the corresponding SEAL path the ITE submits it for 816 SEAL encapsulation as specified in Section 5.4.4. 818 o if the packet is larger than PATH_MTU, the ITE drops the packet 819 and sends an SPTB message to the previous hop subject to rate 820 limiting (see: Section 5.6.1.1) with the MTU field set to 821 PATH_MTU. After sending the SPTB message, the ITE discards the 822 packet. 824 5.4.4. SEAL Encapsulation and Segmentation 826 For each inner packet/fragment submitted for SEAL encapsulation, the 827 ITE next encapsulates the packet in a SEAL header formatted as 828 specified in Section 5.3. The SEAL header includes an Identification 829 field when USE_ID is TRUE, followed by an ICV field when USE_ICV is 830 TRUE. 832 The ITE next sets C=0 in the SEAL header. The ITE also sets A=1 if 833 necessary for ETE reachability determination (see: Section 5.4.6) or 834 for stateful MTU determination (see Section 5.4.9). Otherwise, the 835 ITE sets A=0. Next, when RATE_LIMIT is TRUE the ITE sets L=1; 836 otherwise, it sets L=0. The ITE also sets X=0. 838 The ITE then sets LINK_ID to the value assigned to the underlying 839 SEAL path, and sets NEXTHDR to the protocol number corresponding to 840 the address family of the encapsulated inner packet. For example, 841 the ITE sets NEXTHDR to the value '4' for encapsulated IPv4 packets 842 [RFC2003], '41' for encapsulated IPv6 packets [RFC2473][RFC4213], 843 '80' for encapsulated OSI/CLNP packets [RFC1070], etc. 845 Next, if the inner packet is not itself a SEAL packet the ITE sets 846 LEVEL to an integer value between 0 and 7 as a specification of the 847 number of additional layers of nested SEAL encapsulations permitted. 848 If the inner packet is a SEAL packet that is undergoing nested 849 encapsulation, the ITE instead sets LEVEL to the value that appears 850 in the inner packet's SEAL header minus 1. If the inner packet is 851 undergoing SEAL re-encapsulation, the ITE instead copies the LEVEL 852 value from the SEAL header of the packet to be re-encapsulated. 854 Next, if the inner packet is no larger than (MINMTU-HLEN) or larger 855 than 1500, the ITE sets (M=0; Offset=0). Otherwise, the ITE breaks 856 the inner packet into a N roughly equal-length non-overlapping 857 segmnets (where N is minimized and each fragment is significantly 858 smaller than (MINMTU-HLEN) to allow for additional encapsulations in 859 the path) then appends a clone of the SEAL header from the first 860 segment onto the head of each additional segment. The ITE then sets 861 (M=1; Offset=0) in the first segment, sets (M=0/1; Offset=i) in the 862 second segment, sets (M=0/1; Offset=j) in the third segment (if 863 needed), etc., then finally sets (M=0; Offset=k) in the final segment 864 (where i, j, k, etc. are the number of 32 byte blocks that preceded 865 this segment). 867 When USE_ID is FALSE, the ITE next sets I=0. Otherwise, the ITE sets 868 I=1 and writes a monotonically-incrementing integer value for this 869 ETE in the Identification field beginning with 0 in the first packet 870 transmitted. (For SEAL packets that have been split into multiple 871 pieces, the ITE writes the same Identification value in each piece.) 872 The monotonically-incrementing requirement is to satisfy ETEs that 873 use this value for anti-replay purposes. The value is incremented 874 modulo 2^32, i.e., it wraps back to 0 when the previous value was 875 (2^32 - 1). 877 When USE_ICV is FALSE, the ITE next sets V=0. Otherwise, the ITE 878 sets V=1 and calculates the packet header ICV value using an 879 algorithm agreed on by the ITE and ETE. When data origin 880 authentication is required, the algorithm uses a symmetric secret key 881 so that the ETE can verify that the ICV was generated by the ITE. 882 Beginning with the SEAL header, the ITE calculates the ICV over the 883 leading 128 bytes of the packet (or up to the end of the packet if 884 there are fewer than 128 bytes) and places result in the ICV field. 885 (For SEAL packets that have been split into two pieces, each piece 886 calculates its own ICV value.) 888 The ITE then adds the outer encapsulating headers as specified in 889 Section 5.4.5. 891 5.4.5. Outer Encapsulation 893 Following SEAL encapsulation, the ITE next encapsulates each segment 894 in the requisite outer transport (when necessary) and IP layer 895 headers. When a transport layer header such as UDP or TCP is 896 included, the ITE writes the port number for SEAL in the transport 897 destination service port field. 899 When UDP encapsulation is used, the ITE sets the UDP checksum field 900 to zero for IPv4 packets and also sets the UDP checksum field to zero 901 for IPv6 packets if the encapsulation will include an ICV. Further 902 considerations for setting the UDP checksum field for IPv6 packets 903 are discussed in [I-D.ietf-6man-udpzero]. 905 The ITE then sets the outer IP layer headers the same as specified 906 for ordinary IP encapsulation (e.g., [RFC1070][RFC2003], [RFC2473], 907 [RFC4213], etc.) except that for ordinary SEAL packets the ITE copies 908 the "TTL/Hop Limit", "Type of Service/Traffic Class" and "Congestion 909 Experienced" values in the inner network layer header into the 910 corresponding fields in the outer IP header. For transitional SEAL 911 packets undergoing re-encapsulation, the ITE instead copies the "TTL/ 912 Hop Limit", "Type of Service/Traffic Class" and "Congestion 913 Experienced" values in the outer IP header of the received packet 914 into the corresponding fields in the outer IP header of the packet to 915 be forwarded (i.e., the values are transferred between outer headers 916 and *not* copied from the inner network layer header). 918 The ITE also sets the IP protocol number to the appropriate value for 919 the first protocol layer within the encapsulation (e.g., UDP, TCP, 920 SEAL, etc.). When IPv6 is used as the outer IP protocol, the ITE 921 then sets the flow label value in the outer IPv6 header the same as 922 described in [RFC6438]. When IPv4 is used as the outer IP protocol, 923 the ITE instead sets DF=0 in the IPv4 header to allow the packet to 924 be fragmented if it encounters a restricting link. (For IPv6 SEAL 925 paths, the DF bit is implicitly set to 1.) 927 The ITE finally sends each outer packet via the underlying link 928 corresponding to LINK_ID. For IPv4 SEAL paths with RATE_LIMIT=TRUE, 929 the ITE sends the packet subject to rate limiting so that the IPv4 930 Identification value is not repeated within the IPv4 Maximum Segment 931 Lifetime (i.e., 120 seconds) [RFC1122]. 933 5.4.6. Path Probing and ETE Reachability Verification 935 All SEAL data packets sent by the ITE are considered implicit probes. 936 SEAL data packets will elicit an SCMP message from the ETE if it 937 needs to acknowledge a probe and/or report an error condition. SEAL 938 data packets may also be dropped by either the ETE or a router on the 939 path, which may or may not result in an ICMP message being returned 940 to the ITE. 942 The ITE processes ICMP messages as specified in Section 5.4.7. 944 The ITE processes SCMP messages as specified in Section 5.6.2. 946 5.4.7. Processing ICMP Messages 948 When the ITE sends SEAL packets, it may receive ICMP error messages 949 [RFC0792][RFC4443] from an ordinary router within the subnetwork. 950 Each ICMP message includes an outer IP header, followed by an ICMP 951 header, followed by a portion of the SEAL data packet that generated 952 the error (also known as the "packet-in-error") beginning with the 953 outer IP header. 955 The ITE should process ICMPv4 Protocol Unreachable messages and 956 ICMPv6 Parameter Problem messages with Code "Unrecognized Next Header 957 type encountered" as a hint that the IP destination address does not 958 implement SEAL. The ITE can optionally ignore ICMP messages that do 959 not include sufficient information in the packet-in-error, or process 960 them as a hint that the SEAL path may be failing. 962 For other ICMP messages, the ITE should use any outer header 963 information available as a first-pass authentication filter (e.g., to 964 determine if the source of the message is within the same 965 administrative domain as the ITE) and discards the message if first 966 pass filtering fails. 968 Next, the ITE examines the packet-in-error beginning with the SEAL 969 header. If the value in the Identification field (if present) is not 970 within the window of packets the ITE has recently sent to this ETE, 971 or if the value in the SEAL header ICV field (if present) is 972 incorrect, the ITE discards the message. 974 Next, if the received ICMP message is a PTB the ITE sets the 975 temporary variable "PMTU" for this SEAL path to the MTU value in the 976 PTB message. If PMTU==0, the ITE consults a plateau table (e.g., as 977 described in [RFC1191]) to determine PMTU based on the length field 978 in the outer IP header of the packet-in-error. For example, if the 979 ITE receives a PTB message with MTU==0 and length 4KB, it can set 980 PMTU=2KB. If the ITE subsequently receives a PTB message with MTU==0 981 and length 2KB, it can set PMTU=1792, etc. to a minimum value of 982 PMTU=(1500+HLEN). If the ITE is performing stateful MTU 983 determination for this SEAL path (see Section 5.4.9), the ITE next 984 sets PATH_MTU=MAX((PMTU-HLEN), 1500). 986 If the ICMP message was not discarded, the ITE then transcribes it 987 into a message to return to the previous hop. If the inner packet 988 was a SEAL data packet, the ITE transcribes the ICMP message into an 989 SCMP message. Otherwise, the ITE transcribes the ICMP message into a 990 message appropriate for the inner protocol version. 992 To transcribe the message, the ITE extracts the inner packet from 993 within the ICMP message packet-in-error field and uses it to generate 994 a new message corresponding to the type of the received ICMP message. 995 For SCMP messages, the ITE generates the message the same as 996 described for ETE generation of SCMP messages in Section 5.6.1. For 997 (S)PTB messages, the ITE writes (PMTU-HLEN) in the MTU field. 999 The ITE finally forwards the transcribed message to the previous hop 1000 toward the inner source address. 1002 5.4.8. IPv4 Middlebox Reassembly Testing 1004 The ITE can perform a qualification exchange to ensure that the 1005 subnetwork correctly delivers fragments to the ETE. This procedure 1006 can be used, e.g., to determine whether there are middleboxes on the 1007 path that violate the [RFC1812], Section 5.2.6 requirement that: "A 1008 router MUST NOT reassemble any datagram before forwarding it". 1010 The ITE should use knowledge of its topological arrangement as an aid 1011 in determining when middlebox reassembly testing is necessary. For 1012 example, if the ITE is aware that the ETE is located somewhere in the 1013 public Internet, middlebox reassembly testing should not be 1014 necessary. If the ITE is aware that the ETE is located behind a NAT 1015 or a firewall, however, then middlebox reassembly testing is 1016 recommended. 1018 The ITE can perform a middlebox reassembly test by selecting a data 1019 packet to be used as a probe. While performing the test with real 1020 data packets, the ITE should select only inner packets that are no 1021 larger than (1500-HLEN) bytes for testing purposes. The ITE can also 1022 construct a dummy probe packet instead of using ordinary SEAL data 1023 packets. 1025 To generate a dummy probe packet, the ITE creates a packet buffer 1026 beginning with the same outer headers, SEAL header and inner network 1027 layer header that would appear in an ordinary data packet, then pads 1028 the packet with random data to a length that is at least 128 bytes 1029 but no longer than (1500-HLEN) bytes. The ITE then writes the value 1030 '0' in the inner network layer TTL (for IPv4) or Hop Limit (for IPv6) 1031 field. 1033 The ITE then sets C=0 in the SEAL header of the probe packet and sets 1034 the NEXTHDR field to the inner network layer protocol type. (The ITE 1035 may also set A=1 if it requires a positive acknowledgement; 1036 otherwise, it sets A=0.) Next, the ITE sets LINK_ID and LEVEL to the 1037 appropriate values for this SEAL path, sets Identification and I=1 1038 (when USE_ID is TRUE), then finally calculates the ICV and sets 1039 V=1(when USE_ICV is TRUE). 1041 The ITE then encapsulates the probe packet in the appropriate outer 1042 headers, splits it into two outer IPv4 fragments, then sends both 1043 fragments over the same SEAL path. 1045 The ITE should send a series of probe packets (e.g., 3-5 probes with 1046 1sec intervals between tests) instead of a single isolated probe in 1047 case of packet loss. If the ETE returns an SCMP PTB message with MTU 1048 != 0, then the SEAL path correctly supports fragmentation; otherwise, 1049 the ITE enables stateful MTU determination for this SEAL path as 1050 specified in Section 5.4.9. 1052 (Examples of middleboxes that may perform reassembly include stateful 1053 NATs and firewalls. Such devices could still allow for stateless MTU 1054 determination if they gather the fragments of a fragmented IPv4 SEAL 1055 data packet for packet analysis purposes but then forward the 1056 fragments on to the final destination rather than forwarding the 1057 reassembled packet.) 1059 5.4.9. Stateful MTU Determination 1061 SEAL supports a stateless MTU determination capability, however the 1062 ITE may in some instances wish to impose a stateful MTU limit on a 1063 particular SEAL path. For example, when the ETE is situated behind a 1064 middlebox that performs IPv4 reassembly (see: Section 5.4.8) it is 1065 imperative that fragmentation be avoided. In other instances (e.g., 1066 when the SEAL path includes performance-constrained links), the ITE 1067 may deem it necessary to cache a conservative static MTU in order to 1068 avoid sending large packets that would only be dropped due to an MTU 1069 restriction somewhere on the path. 1071 To determine a static MTU value, the ITE sends a series of dummy 1072 probe packets of various sizes to the ETE with A=1 in the SEAL header 1073 and DF=1 in the outer IP header. The ITE then caches the size 'S' of 1074 the largest packet for which it receives a probe reply from the ETE 1075 by setting PATH_MTU=MAX((S-HLEN), 1500) for this SEAL path. 1077 For example, the ITE could send probe packets of 4KB, followed by 1078 2KB, followed by 1792 bytes, etc. While probing, the ITE processes 1079 any ICMP PTB message it receives as a potential indication of probe 1080 failure then discards the message. 1082 5.4.10. Detecting Path MTU Changes 1084 When stateful MTU determination is used, the ITE SHOULD periodically 1085 reset PATH_MTU and/or re-probe the path to determine whether PATH_MTU 1086 has increased. If the path still has a too-small MTU, the ITE will 1087 receive a PTB message that reports a smaller size. 1089 For IPv4 SEAL paths, when the path correctly implements fragmentation 1090 and RATE_LIMIT is TRUE, the ITE can periodically reset 1091 RATE_LIMIT=FALSE to determine whether the path still requires rate 1092 limiting. If the ITE receives an SPTB message it should again set 1093 RATE_LIMIT=TRUE. 1095 5.5. ETE Specification 1097 5.5.1. Minimum Reassembly Buffer Requirements 1099 For IPv6, the ETE must configure a minimum reassembly buffer size of 1100 1500 bytes for the reassembly of outer IPv6 packets (see: [RFC2460]. 1101 For IPv4, the ETE must also configure a minimum reassembly buffer 1102 size of 1500 bytes for the reassembly of outer IPv4 packets, i.e., 1103 even though the true minimum reassembly size for IPv4 is only 576 1104 bytes [RFC1122]. 1106 In addition to this outer reassembly buffer requirement, the ETE must 1107 further configure a minimum SEAL reassembly buffer size of (1500 + 1108 HLEN) bytes for the reassembly of segmented SEAL packets (see: 1109 Section 5.5.4). 1111 5.5.2. Tunnel Neighbor Soft State 1113 When data origin authentication and integrity checking is required, 1114 the ETE maintains a per-ITE ICV calculation algorithm and a symmetric 1115 secret key to verify the ICV. When per-packet identification is 1116 required, the ETE also maintains a window of Identification values 1117 for the packets it has recently received from this ITE. 1119 When the tunnel neighbor relationship is bidirectional, the ETE 1120 further maintains a per SEAL path mapping of outer IP and transport 1121 layer addresses to the LINK_ID that appears in packets received from 1122 the ITE. 1124 5.5.3. IP-Layer Reassembly 1126 The ETE reassembles fragmented IP packets that are explcitly 1127 addressed to itself. For IP fragments that are received via a SEAL 1128 tunnel, the ETE SHOULD maintain conservative reassembly cache high- 1129 and low-water marks. When the size of the reassembly cache exceeds 1130 this high-water mark, the ETE SHOULD actively discard stale 1131 incomplete reassemblies (e.g., using an Active Queue Management (AQM) 1132 strategy) until the size falls below the low-water mark. The ETE 1133 SHOULD also actively discard any pending reassemblies that clearly 1134 have no opportunity for completion, e.g., when a considerable number 1135 of new fragments have arrived before a fragment that completes a 1136 pending reassembly arrives. 1138 The ETE processes non-SEAL IP packets as specified in the normative 1139 references, i.e., it performs any necessary IP reassembly then 1140 discards the packet if it is larger than the reassembly buffer size 1141 or delivers the (fully-reassembled) packet to the appropriate upper 1142 layer protocol module. 1144 For SEAL packets, the ITE performs any necessary IP reassembly then 1145 submits the packet for SEAL decapsulation as specified in Section 1146 5.5.4. (Note that if the packet is larger than the reassembly buffer 1147 size, the ITE still returns the leading portion of the (partially) 1148 reassembled packet.) 1150 5.5.4. Decapsulation and Re-Encapsulation 1152 For each SEAL packet accepted for decapsulation, when I==1 the ETE 1153 first examines the Identification field. If the Identification is 1154 not within the window of acceptable values for this ITE, the ETE 1155 silently discards the packet. 1157 Next, if V==1 the ETE verifies the ICV value (with the ICV field 1158 itself reset to 0) and silently discards the packet if the value is 1159 incorrect. 1161 Next, if the packet arrived as multiple IPv4 fragments and L ==0, the 1162 ETE sends an SPTB message back to the ITE with MTU set to the size of 1163 the largest fragment received minus HLEN (see: Section 5.6.1.1). 1165 Next, if the packet arrived as multiple IP fragments and the inner 1166 packet is larger than 1500 bytes, the ETE silently discards the 1167 packet; otherwise, it continues to process the packet. 1169 Next, if there is an incorrect value in a SEAL header field (e.g., an 1170 incorrect "VER" field value), the ETE discards the packet. If the 1171 SEAL header has C==0, the ETE also returns an SCMP "Parameter 1172 Problem" (SPP) message (see Section 5.6.1.2). 1174 Next, if the SEAL header has C==1, the ETE processes the packet as an 1175 SCMP packet as specified in Section 5.6.2. Otherwise, the ETE 1176 continues to process the packet as a SEAL data packet. 1178 Next, if the SEAL header has (M==1 || Offset!==0) the ETE checks to 1179 see if the other segments of this already-segmented SEAL packet have 1180 arrived, i.e., by looking for additional segments that have the same 1181 outer IP source address, destination address, source transport port 1182 number (if present) and SEAL Identification value. If the other 1183 segments have already arrived, the ETE discards the SEAL header and 1184 other outer headers from the non-initial segments and appends them 1185 onto the end of the first segment. Otherwise, the ETE caches the 1186 segment for at most 60 seconds while awaiting the arrival of its 1187 partners. During this process, the ETE discards any segments that 1188 are overlapping with respect to segments that have already been 1189 received. 1191 Next, if the SEAL header in the (reassembled) packet has A==1, the 1192 ETE sends an SPTB message back to the ITE with MTU=0 (see: Section 1193 5.6.1.1). 1195 Finally, the ETE discards the outer headers and processes the inner 1196 packet according to the header type indicated in the SEAL NEXTHDR 1197 field. If the inner (TTL / Hop Limit) field encodes the value 0, the 1198 ETE silently discards the packet. Otherwise, if the next hop toward 1199 the inner destination address is via a different interface than the 1200 SEAL packet arrived on, the ETE discards the SEAL header and delivers 1201 the inner packet either to the local host or to the next hop 1202 interface if the packet is not destined to the local host. 1204 If the next hop is on the same interface the SEAL packet arrived on, 1205 however, the ETE submits the packet for SEAL re-encapsulation 1206 beginning with the specification in Section 5.4.3 above and without 1207 decrementing the value in the inner (TTL / Hop Limit) field. In this 1208 process, the packet remains within the tunnel (i.e., it does not exit 1209 and then re-enter the tunnel); hence, the packet is not discarded if 1210 the LEVEL field in the SEAL header contains the value 0. 1212 5.6. The SEAL Control Message Protocol (SCMP) 1214 SEAL provides a companion SEAL Control Message Protocol (SCMP) that 1215 uses the same message types and formats as for the Internet Control 1216 Message Protocol for IPv6 (ICMPv6) [RFC4443]. As for ICMPv6, each 1217 SCMP message includes a 32-bit header and a variable-length body. 1218 The ITE encapsulates the SCMP message in a SEAL header and outer 1219 headers as shown in Figure 3: 1221 +--------------------+ 1222 ~ outer IP header ~ 1223 +--------------------+ 1224 ~ other outer hdrs ~ 1225 +--------------------+ 1226 ~ SEAL Header ~ 1227 +--------------------+ +--------------------+ 1228 | SCMP message header| --> | SCMP message header| 1229 +--------------------+ +--------------------+ 1230 | | --> | | 1231 ~ SCMP message body ~ --> ~ SCMP message body ~ 1232 | | --> | | 1233 +--------------------+ +--------------------+ 1235 SCMP Message SCMP Packet 1236 before encapsulation after encapsulation 1238 Figure 3: SCMP Message Encapsulation 1240 The following sections specify the generation, processing and 1241 relaying of SCMP messages. 1243 5.6.1. Generating SCMP Error Messages 1245 ETEs generate SCMP error messages in response to receiving certain 1246 SEAL data packets using the format shown in Figure 4: 1248 0 1 2 3 1249 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 1250 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1251 | Type | Code | Checksum | 1252 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1253 | Type-Specific Data | 1254 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1255 | As much of the invoking SEAL data packet as possible | 1256 ~ (beginning with the SEAL header) without the SCMP ~ 1257 | packet exceeding MINMTU bytes (*) | 1259 (*) also known as the "packet-in-error" 1261 Figure 4: SCMP Error Message Format 1263 The error message includes the 32-bit SCMP message header, followed 1264 by a 32-bit Type-Specific Data field, followed by the leading portion 1265 of the invoking SEAL data packet beginning with the SEAL header as 1266 the "packet-in-error". The packet-in-error includes as much of the 1267 invoking packet as possible extending to a length that would not 1268 cause the entire SCMP packet following outer encapsulation to exceed 1269 MINMTU bytes. 1271 When the ETE processes a SEAL data packet for which the 1272 Identification and ICV values are correct but an error must be 1273 returned, it prepares an SCMP error message as shown in Figure 4. 1274 The ETE sets the Type and Code fields to the same values that would 1275 appear in the corresponding ICMPv6 message [RFC4443], but calculates 1276 the Checksum beginning with the SCMP message header using the 1277 algorithm specified for ICMPv4 in [RFC0792]. 1279 The ETE next encapsulates the SCMP message in the requisite SEAL and 1280 outer headers as shown in Figure 3. During encapsulation, the ETE 1281 sets the outer destination address/port numbers of the SCMP packet to 1282 the values associated with the ITE and sets the outer source address/ 1283 port numbers to its own outer address/port numbers. 1285 The ETE then sets (C=1; A=0; L=0; X=0; M=0; Offset=0) in the SEAL 1286 header, then sets I, V, NEXTHDR and LEVEL to the same values that 1287 appeared in the SEAL header of the data packet. If the neighbor 1288 relationship between the ITE and ETE is unidirectional, the ETE next 1289 sets the LINK_ID field to the same value that appeared in the SEAL 1290 header of the data packet. Otherwise, the ETE sets the LINK_ID field 1291 to the value it would use in sending a SEAL packet to this ITE. 1293 When I==1, the ETE next sets the Identification field to an 1294 appropriate value for the ITE. If the neighbor relationship between 1295 the ITE and ETE is unidirectional, the ETE sets the Identification 1296 field to the same value that appeared in the SEAL header of the data 1297 packet. Otherwise, the ETE sets the Identification field to the 1298 value it would use in sending the next SEAL packet to this ITE. 1300 When V==1, the ETE then calculates and sets the ICV field the same as 1301 specified for SEAL data packet encapsulation in Section 5.4.4. 1303 Finally, the ETE sends the resulting SCMP packet to the ITE the same 1304 as specified for SEAL data packets in Section 5.4.5. 1306 The following sections describe additional considerations for various 1307 SCMP error messages: 1309 5.6.1.1. Generating SCMP Packet Too Big (SPTB) Messages 1311 An ITE generates an SCMP "Packet Too Big" (SPTB) message when it 1312 receives a SEAL data packet that is larger than (MINMTU-HLEN) but no 1313 larger than 1500 bytes. The ITE sends the SPTB message toward the 1314 previous hop SEAL ITE with the MTU field set to (MINMTU-HLEN) subject 1315 to rate limiting. In the case of nested tunneling, the previous hop 1316 SEAL ITE is determined by examining the source addressing information 1317 in the outer encapsulations of the SEAL data packet submitted for 1318 nested encapsulation. In the case of re-encapsulating tunneling, the 1319 SEAL node acts as an ETE with respect to the previous hop ITE and as 1320 an ITE with respect to the next hop ETE. The node must therefore 1321 retain the outer encapsulations that were used in the previous 1322 segment of this (segmented) SEAL tunnel (i.e., as it transitions from 1323 ETE on the previous hop segment to ITE on the next hop segment) so 1324 that it will have the source addressing information for the previous 1325 hop ITE. 1327 An ETE generates an SPTB message when it receives a SEAL data packet 1328 that arrived as multiple outer IPv4 fragments and for which L==0. 1329 The ETE prepares the SPTB message the same as for the corresponding 1330 ICMPv6 PTB message, and writes the length of the largest outer IP 1331 fragment received minus HLEN in the MTU field of the message. 1333 The ETE also generates an SPTB message when it accepts a SEAL 1334 protocol data packet with A==1 in the SEAL header. The ETE prepares 1335 the SPTB message the same as above, except that it writes the value 0 1336 in the MTU field. 1338 5.6.1.2. Generating Other SCMP Error Messages 1340 An ETE generates an SCMP "Destination Unreachable" (SDU) message 1341 under the same circumstances that an IPv6 system would generate an 1342 ICMPv6 Destination Unreachable message. 1344 An ETE generates an SCMP "Parameter Problem" (SPP) message when it 1345 receives a SEAL packet with an incorrect value in the SEAL header. 1347 TEs generate other SCMP message types using methods and procedures 1348 specified in other documents. For example, SCMP message types used 1349 for tunnel neighbor coordinations are specified in VET 1350 [I-D.templin-intarea-vet]. 1352 5.6.2. Processing SCMP Error Messages 1354 An ITE may receive SCMP messages with C==1 in the SEAL header after 1355 sending packets to an ETE. The ITE first verifies that the outer 1356 addresses of the SCMP packet are correct, and (when I==1) that the 1357 Identification field contains an acceptable value. The ITE next 1358 verifies that the SEAL header fields are set correctly as specified 1359 in Section 5.6.1. When V==1, the ITE then verifies the ICV value. 1360 The ITE next verifies the Checksum value in the SCMP message header. 1361 If any of these values are incorrect, the ITE silently discards the 1362 message; otherwise, it processes the message as follows: 1364 5.6.2.1. Processing SCMP PTB Messages 1366 After an ITE sends a SEAL data packet to an ETE, it may receive an 1367 SPTB message with a packet-in-error containing the leading portion of 1368 the packet (see: Section 5.6.1.1). For IP SPTB messages with MTU==0, 1369 the ITE processes the message as confirmation that the ETE received a 1370 SEAL data packet with A==1 in the SEAL header. The ITE then discards 1371 the message. 1373 For SPTB messages with MTU != 0, the ITE processes the message as an 1374 indication of a packet size limitation as follows. If the inner 1375 packet is itself a SEAL packet, and the inner packet length is less 1376 than 1500, the ITE reduces its MINMTU value for this ITE. If the 1377 inner packet is a non-SEAL IPv4 packet and the inner packet length is 1378 less than 1500, the ITE instead sets RATE_LIMIT=1. For all other 1379 cases, if the inner packet length is larger than 1500 and the MTU 1380 value is not substantially less than 1500 bytes, the value is likely 1381 to reflect the true MTU of the restricting link on the path to the 1382 ETE; otherwise, a router on the path may be generating runt 1383 fragments. 1385 In that case, the ITE can consult a plateau table (e.g., as described 1386 in [RFC1191]) to rewrite the MTU value to a reduced size. For 1387 example, if the ITE receives an IPv4 SPTB message with MTU==256 and 1388 inner packet length 4KB, it can rewrite the MTU to 2KB. If the ITE 1389 subsequently receives an IPv4 SPTB message with MTU==256 and inner 1390 packet length 2KB, it can rewrite the MTU to 1792, etc., to a minimum 1391 of 1500 bytes. If the ITE is performing stateful MTU determination 1392 for this SEAL path, it then writes the new MTU value minus HLEN in 1393 PATH_MTU. 1395 The ITE then checks its forwarding tables to discover the previous 1396 hop toward the source address of the inner packet. If the previous 1397 hop is reached via the same tunnel interface the SPTB message arrived 1398 on, the ITE relays the message to the previous hop. In order to 1399 relay the message, the first writes zero in the Identification and 1400 ICV fields of the SEAL header within the packet-in-error. The ITE 1401 next rewrites the outer SEAL header fields with values corresponding 1402 to the previous hop and recalculates the ICV using the ICV 1403 calculation parameters associated with the previous hop. Next, the 1404 ITE replaces the SPTB's outer headers with headers of the appropriate 1405 protocol version and fills in the header fields as specified in 1406 Section 5.4.5, where the destination address/port correspond to the 1407 previous hop and the source address/port correspond to the ITE. The 1408 ITE then sends the message to the previous hop the same as if it were 1409 issuing a new SPTB message. (Note that, in this process, the values 1410 within the SEAL header of the packet-in-error are meaningless to the 1411 previous hop and therefore cannot be used by the previous hop for 1412 authentication purposes.) 1414 If the previous hop is not reached via the same tunnel interface, the 1415 ITE instead transcribes the message into a format appropriate for the 1416 inner packet (i.e., the same as described for transcribing ICMP 1417 messages in Section 5.4.7) and sends the resulting transcribed 1418 message to the original source. (NB: if the inner packet within the 1419 SPTB message is an IPv4 SEAL packet with DF==0, the ITE should set 1420 DF=1 and re-calculate the IPv4 header checksum while transcribing the 1421 message in order to avoid bogon filters.) The ITE then discards the 1422 SPTB message. 1424 Note that the ITE may receive an SPTB message from another ITE that 1425 is at the head end of a nested level of encapsulation. The ITE has 1426 no security associations with this nested ITE, hence it should 1427 consider this SPTB message the same as if it had received an ICMP PTB 1428 message from an ordinary router on the path to the ETE. That is, the 1429 ITE should examine the packet-in-error field of the SPTB message and 1430 only process the message if it is able to recognize the packet as one 1431 it had previously sent. 1433 5.6.2.2. Processing Other SCMP Error Messages 1435 An ITE may receive an SDU message with an appropriate code under the 1436 same circumstances that an IPv6 node would receive an ICMPv6 1437 Destination Unreachable message. The ITE either transcribes or 1438 relays the message toward the source address of the inner packet 1439 within the packet-in-error the same as specified for SPTB messages in 1440 Section 5.6.2.1. 1442 An ITE may receive an SPP message when the ETE receives a SEAL packet 1443 with an incorrect value in the SEAL header. The ITE should examine 1444 the SEAL header within the packet-in-error to determine whether a 1445 different setting should be used in subsequent packets, but does not 1446 relay the message further. 1448 TEs process other SCMP message types using methods and procedures 1449 specified in other documents. For example, SCMP message types used 1450 for tunnel neighbor coordinations are specified in VET 1451 [I-D.templin-intarea-vet]. 1453 6. Link Requirements 1455 Subnetwork designers are expected to follow the recommendations in 1456 Section 2 of [RFC3819] when configuring link MTUs. 1458 7. End System Requirements 1460 End systems are encouraged to implement end-to-end MTU assurance 1461 (e.g., using Packetization Layer Path MTU Discovery (PLPMTUD) per 1462 [RFC4821]) even if the subnetwork is using SEAL. 1464 When end systems use PLPMTUD, SEAL will ensure that the tunnel 1465 behaves as a link in the path that assures an MTU of at least 1500 1466 bytes while not precluding discovery of larger MTUs. The PMPMTUD 1467 mechanism will therefore be able to function as designed in order to 1468 discover and utilize larger MTUs. 1470 8. Router Requirements 1472 Routers within the subnetwork are expected to observe the standard IP 1473 router requirements, including the implementation of IP fragmentation 1474 and reassembly as well as the generation of ICMP messages 1475 [RFC0792][RFC1122][RFC1812][RFC2460][RFC4443][RFC6434]. 1477 Note that, even when routers support existing requirements for the 1478 generation of ICMP messages, these messages are often filtered and 1479 discarded by middleboxes on the path to the original source of the 1480 message that triggered the ICMP. It is therefore not possible to 1481 assume delivery of ICMP messages even when routers are correctly 1482 implemented. 1484 9. Nested Encapsulation Considerations 1486 SEAL supports nested tunneling for up to 8 layers of encapsulation. 1487 In this model, the SEAL ITE has a tunnel neighbor relationship only 1488 with ETEs at its own nesting level, i.e., it does not have a tunnel 1489 neighbor relationship with other ITEs, nor with ETEs at other nesting 1490 levels. 1492 Therefore, when an ITE 'A' within an inner nesting level needs to 1493 return an error message to an ITE 'B' within an outer nesting level, 1494 it generates an ordinary ICMP error message the same as if it were an 1495 ordinary router within the subnetwork. 'B' can then perform message 1496 validation as specified in Section 5.4.7, but full message origin 1497 authentication is not possible. 1499 Since ordinary ICMP messages are used for coordinations between ITEs 1500 at different nesting levels, nested SEAL encapsulations should only 1501 be used when the ITEs are within a common administrative domain 1502 and/or when there is no ICMP filtering middlebox such as a firewall 1503 or NAT between them. An example would be a recursive nesting of 1504 mobile networks, where the first network receives service from an 1505 ISP, the second network receives service from the first network, the 1506 third network receives service from the second network, etc. 1508 NB: As an alternative, the SCMP protocol could be extended to allow 1509 ITE 'A' to return an SCMP message to ITE 'B' rather than return an 1510 ICMP message. This would conceptually allow the control messages to 1511 pass through firewalls and NATs, however it would give no more 1512 message origin authentication assurance than for ordinary ICMP 1513 messages. It was therefore determined that the complexity of 1514 extending the SCMP protocol was of little value within the context of 1515 the anticipated use cases for nested encapsulations. 1517 10. Reliability Considerations 1519 Although a SEAL tunnel may span an arbitrarily-large subnetwork 1520 expanse, the IP layer sees the tunnel as a simple link that supports 1521 the IP service model. Links with high bit error rates (BERs) (e.g., 1522 IEEE 802.11) use Automatic Repeat-ReQuest (ARQ) mechanisms [RFC3366] 1523 to increase packet delivery ratios, while links with much lower BERs 1524 typically omit such mechanisms. Since SEAL tunnels may traverse 1525 arbitrarily-long paths over links of various types that are already 1526 either performing or omitting ARQ as appropriate, it would therefore 1527 be inefficient to require the tunnel endpoints to also perform ARQ. 1529 11. Integrity Considerations 1531 The SEAL header includes an integrity check field that covers the 1532 SEAL header and at least the inner packet headers. This provides for 1533 header integrity verification on a segment-by-segment basis for a 1534 segmented re-encapsulating tunnel path. 1536 Fragmentation and reassembly schemes must also consider packet- 1537 splicing errors, e.g., when two fragments from the same packet are 1538 concatenated incorrectly, when a fragment from packet X is 1539 reassembled with fragments from packet Y, etc. The primary sources 1540 of such errors include implementation bugs and wrapping IPv4 ID 1541 fields. 1543 In particular, the IPv4 16-bit ID field can wrap with only 64K 1544 packets with the same (src, dst, protocol)-tuple alive in the system 1545 at a given time [RFC4963]. When the IPv4 ID field is re-written by a 1546 middlebox such as a NAT or Firewall, ID field wrapping can occur with 1547 even fewer packets alive in the system. 1549 When outer IPv4 fragmentation is unavoidable, SEAL institutes rate 1550 limiting so that the number of packets admitted into the tunnel by 1551 the ITE does not exceed the number of unique packets that may be 1552 alive within the Internet. 1554 12. IANA Considerations 1556 The IANA is requested to allocate a User Port number for "SEAL" in 1557 the 'port-numbers' registry for the TCP and UDP protocols. 1559 The IANA is further requested to allocate an IP protocol number for 1560 "SEAL" in the "protocol-numbers" registry. 1562 . 1564 13. Security Considerations 1566 SEAL provides a segment-by-segment data origin authentication and 1567 anti-replay service across the (potentially) multiple segments of a 1568 re-encapsulating tunnel. It further provides a segment-by-segment 1569 integrity check of the headers of encapsulated packets, but does not 1570 verify the integrity of the rest of the packet beyond the headers 1571 unless fragmentation is unavoidable. SEAL therefore considers full 1572 message integrity checking, authentication and confidentiality as 1573 end-to-end considerations. 1575 An amplification/reflection/buffer overflow attack is possible when 1576 an attacker sends IP fragments with spoofed source addresses to an 1577 ETE in an attempt to clog the ETE's reassembly buffer and/or cause 1578 the ETE to generate a stream of SCMP messages returned to a victim 1579 ITE. The SCMP message ICV, Identification, as well as the inner 1580 headers of the packet-in-error, provide mitigation for the ETE to 1581 detect and discard SEAL segments with spoofed source addresses. 1583 The SEAL header is sent in-the-clear the same as for the outer IP and 1584 other outer headers. In this respect, the threat model is no 1585 different than for IPv6 extension headers. Unlike IPv6 extension 1586 headers, however, the SEAL header can be protected by an integrity 1587 check that also covers the inner packet headers. 1589 Security issues that apply to tunneling in general are discussed in 1590 [RFC6169]. 1592 14. Related Work 1594 Section 3.1.7 of [RFC2764] provides a high-level sketch for 1595 supporting large tunnel MTUs via a tunnel-level segmentation and 1596 reassembly capability to avoid IP level fragmentation. 1598 Section 3 of [RFC4459] describes inner and outer fragmentation at the 1599 tunnel endpoints as alternatives for accommodating the tunnel MTU. 1601 Section 4 of [RFC2460] specifies a method for inserting and 1602 processing extension headers between the base IPv6 header and 1603 transport layer protocol data. The SEAL header is inserted and 1604 processed in exactly the same manner. 1606 IPsec/AH is [RFC4301][RFC4301] is used for full message integrity 1607 verification between tunnel endpoints, whereas SEAL only ensures 1608 integrity for the inner packet headers. The AYIYA proposal 1609 [I-D.massar-v6ops-ayiya] uses similar means for providing message 1610 authentication and integrity. 1612 SEAL, along with the Virtual Enterprise Traversal (VET) 1613 [I-D.templin-intarea-vet] tunnel virtual interface abstraction, are 1614 the functional building blocks for the Internet Routing Overlay 1615 Network (IRON) [I-D.templin-ironbis] and Routing and Addressing in 1616 Networks with Global Enterprise Recursion (RANGER) [RFC5720][RFC6139] 1617 architectures. 1619 The concepts of path MTU determination through the report of 1620 fragmentation and extending the IPv4 Identification field were first 1621 proposed in deliberations of the TCP-IP mailing list and the Path MTU 1622 Discovery Working Group (MTUDWG) during the late 1980's and early 1623 1990's. An historical analysis of the evolution of these concepts, 1624 as well as the development of the eventual PMTUD mechanism, appears 1625 in [RFC5320]. 1627 15. Implementation Status 1629 An early implementation of the first revision of SEAL [RFC5320] is 1630 available at: http://isatap.com/seal. 1632 16. Acknowledgments 1634 The following individuals are acknowledged for helpful comments and 1635 suggestions: Jari Arkko, Fred Baker, Iljitsch van Beijnum, Oliver 1636 Bonaventure, Teco Boot, Bob Braden, Brian Carpenter, Steve Casner, 1637 Ian Chakeres, Noel Chiappa, Remi Denis-Courmont, Remi Despres, Ralph 1638 Droms, Aurnaud Ebalard, Gorry Fairhurst, Washam Fan, Dino Farinacci, 1639 Joel Halpern, Sam Hartman, John Heffner, Thomas Henderson, Bob 1640 Hinden, Christian Huitema, Eliot Lear, Darrel Lewis, Joe Macker, Matt 1641 Mathis, Erik Nordmark, Dan Romascanu, Dave Thaler, Joe Touch, Mark 1642 Townsley, Ole Troan, Margaret Wasserman, Magnus Westerlund, Robin 1643 Whittle, James Woodyatt, and members of the Boeing Research & 1644 Technology NST DC&NT group. 1646 Discussions with colleagues following the publication of [RFC5320] 1647 have provided useful insights that have resulted in significant 1648 improvements to this, the Second Edition of SEAL. 1650 Path MTU determination through the report of fragmentation was first 1651 proposed by Charles Lynn on the TCP-IP mailing list in 1987. 1652 Extending the IP identification field was first proposed by Steve 1653 Deering on the MTUDWG mailing list in 1989. 1655 17. References 1657 17.1. Normative References 1659 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, 1660 September 1981. 1662 [RFC0792] Postel, J., "Internet Control Message Protocol", STD 5, 1663 RFC 792, September 1981. 1665 [RFC1122] Braden, R., "Requirements for Internet Hosts - 1666 Communication Layers", STD 3, RFC 1122, October 1989. 1668 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1669 Requirement Levels", BCP 14, RFC 2119, March 1997. 1671 [RFC2460] Deering, S. and R. Hinden, "Internet Protocol, Version 6 1672 (IPv6) Specification", RFC 2460, December 1998. 1674 [RFC3971] Arkko, J., Kempf, J., Zill, B., and P. Nikander, "SEcure 1675 Neighbor Discovery (SEND)", RFC 3971, March 2005. 1677 [RFC4443] Conta, A., Deering, S., and M. Gupta, "Internet Control 1678 Message Protocol (ICMPv6) for the Internet Protocol 1679 Version 6 (IPv6) Specification", RFC 4443, March 2006. 1681 [RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman, 1682 "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861, 1683 September 2007. 1685 17.2. Informative References 1687 [FOLK] Shannon, C., Moore, D., and k. claffy, "Beyond Folklore: 1688 Observations on Fragmented Traffic", December 2002. 1690 [FRAG] Kent, C. and J. Mogul, "Fragmentation Considered Harmful", 1691 October 1987. 1693 [I-D.generic-6man-tunfrag] 1694 Templin, F., "IPv6 Path MTU Updates", 1695 draft-generic-6man-tunfrag-06 (work in progress), 1696 July 2012. 1698 [I-D.ietf-6man-udpzero] 1699 Fairhurst, G. and M. Westerlund, "IPv6 UDP Checksum 1700 Considerations", draft-ietf-6man-udpzero-06 (work in 1701 progress), June 2012. 1703 [I-D.ietf-intarea-ipv4-id-update] 1704 Touch, J., "Updated Specification of the IPv4 ID Field", 1705 draft-ietf-intarea-ipv4-id-update-05 (work in progress), 1706 May 2012. 1708 [I-D.ietf-savi-framework] 1709 Wu, J., Bi, J., Bagnulo, M., Baker, F., and C. Vogt, 1710 "Source Address Validation Improvement Framework", 1711 draft-ietf-savi-framework-06 (work in progress), 1712 January 2012. 1714 [I-D.massar-v6ops-ayiya] 1715 Massar, J., "AYIYA: Anything In Anything", 1716 draft-massar-v6ops-ayiya-02 (work in progress), July 2004. 1718 [I-D.templin-intarea-vet] 1719 Templin, F., "Virtual Enterprise Traversal (VET)", 1720 draft-templin-intarea-vet-34 (work in progress), 1721 June 2012. 1723 [I-D.templin-ironbis] 1724 Templin, F., "The Internet Routing Overlay Network 1725 (IRON)", draft-templin-ironbis-11 (work in progress), 1726 June 2012. 1728 [MTUDWG] "IETF MTU Discovery Working Group mailing list, 1729 gatekeeper.dec.com/pub/DEC/WRL/mogul/mtudwg-log, November 1730 1989 - February 1995.". 1732 [RFC0994] International Organization for Standardization (ISO) and 1733 American National Standards Institute (ANSI), "Final text 1734 of DIS 8473, Protocol for Providing the Connectionless- 1735 mode Network Service", RFC 994, March 1986. 1737 [RFC1063] Mogul, J., Kent, C., Partridge, C., and K. McCloghrie, "IP 1738 MTU discovery options", RFC 1063, July 1988. 1740 [RFC1070] Hagens, R., Hall, N., and M. Rose, "Use of the Internet as 1741 a subnetwork for experimentation with the OSI network 1742 layer", RFC 1070, February 1989. 1744 [RFC1146] Zweig, J. and C. Partridge, "TCP alternate checksum 1745 options", RFC 1146, March 1990. 1747 [RFC1191] Mogul, J. and S. Deering, "Path MTU discovery", RFC 1191, 1748 November 1990. 1750 [RFC1701] Hanks, S., Li, T., Farinacci, D., and P. Traina, "Generic 1751 Routing Encapsulation (GRE)", RFC 1701, October 1994. 1753 [RFC1812] Baker, F., "Requirements for IP Version 4 Routers", 1754 RFC 1812, June 1995. 1756 [RFC1981] McCann, J., Deering, S., and J. Mogul, "Path MTU Discovery 1757 for IP version 6", RFC 1981, August 1996. 1759 [RFC2003] Perkins, C., "IP Encapsulation within IP", RFC 2003, 1760 October 1996. 1762 [RFC2473] Conta, A. and S. Deering, "Generic Packet Tunneling in 1763 IPv6 Specification", RFC 2473, December 1998. 1765 [RFC2675] Borman, D., Deering, S., and R. Hinden, "IPv6 Jumbograms", 1766 RFC 2675, August 1999. 1768 [RFC2764] Gleeson, B., Heinanen, J., Lin, A., Armitage, G., and A. 1769 Malis, "A Framework for IP Based Virtual Private 1770 Networks", RFC 2764, February 2000. 1772 [RFC2780] Bradner, S. and V. Paxson, "IANA Allocation Guidelines For 1773 Values In the Internet Protocol and Related Headers", 1774 BCP 37, RFC 2780, March 2000. 1776 [RFC2827] Ferguson, P. and D. Senie, "Network Ingress Filtering: 1777 Defeating Denial of Service Attacks which employ IP Source 1778 Address Spoofing", BCP 38, RFC 2827, May 2000. 1780 [RFC2923] Lahey, K., "TCP Problems with Path MTU Discovery", 1781 RFC 2923, September 2000. 1783 [RFC3232] Reynolds, J., "Assigned Numbers: RFC 1700 is Replaced by 1784 an On-line Database", RFC 3232, January 2002. 1786 [RFC3366] Fairhurst, G. and L. Wood, "Advice to link designers on 1787 link Automatic Repeat reQuest (ARQ)", BCP 62, RFC 3366, 1788 August 2002. 1790 [RFC3819] Karn, P., Bormann, C., Fairhurst, G., Grossman, D., 1791 Ludwig, R., Mahdavi, J., Montenegro, G., Touch, J., and L. 1792 Wood, "Advice for Internet Subnetwork Designers", BCP 89, 1793 RFC 3819, July 2004. 1795 [RFC4191] Draves, R. and D. Thaler, "Default Router Preferences and 1796 More-Specific Routes", RFC 4191, November 2005. 1798 [RFC4213] Nordmark, E. and R. Gilligan, "Basic Transition Mechanisms 1799 for IPv6 Hosts and Routers", RFC 4213, October 2005. 1801 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 1802 Internet Protocol", RFC 4301, December 2005. 1804 [RFC4302] Kent, S., "IP Authentication Header", RFC 4302, 1805 December 2005. 1807 [RFC4459] Savola, P., "MTU and Fragmentation Issues with In-the- 1808 Network Tunneling", RFC 4459, April 2006. 1810 [RFC4821] Mathis, M. and J. Heffner, "Packetization Layer Path MTU 1811 Discovery", RFC 4821, March 2007. 1813 [RFC4963] Heffner, J., Mathis, M., and B. Chandler, "IPv4 Reassembly 1814 Errors at High Data Rates", RFC 4963, July 2007. 1816 [RFC4987] Eddy, W., "TCP SYN Flooding Attacks and Common 1817 Mitigations", RFC 4987, August 2007. 1819 [RFC5226] Narten, T. and H. Alvestrand, "Guidelines for Writing an 1820 IANA Considerations Section in RFCs", BCP 26, RFC 5226, 1821 May 2008. 1823 [RFC5246] Dierks, T. and E. Rescorla, "The Transport Layer Security 1824 (TLS) Protocol Version 1.2", RFC 5246, August 2008. 1826 [RFC5320] Templin, F., "The Subnetwork Encapsulation and Adaptation 1827 Layer (SEAL)", RFC 5320, February 2010. 1829 [RFC5445] Watson, M., "Basic Forward Error Correction (FEC) 1830 Schemes", RFC 5445, March 2009. 1832 [RFC5720] Templin, F., "Routing and Addressing in Networks with 1833 Global Enterprise Recursion (RANGER)", RFC 5720, 1834 February 2010. 1836 [RFC5927] Gont, F., "ICMP Attacks against TCP", RFC 5927, July 2010. 1838 [RFC6139] Russert, S., Fleischman, E., and F. Templin, "Routing and 1839 Addressing in Networks with Global Enterprise Recursion 1840 (RANGER) Scenarios", RFC 6139, February 2011. 1842 [RFC6169] Krishnan, S., Thaler, D., and J. Hoagland, "Security 1843 Concerns with IP Tunneling", RFC 6169, April 2011. 1845 [RFC6335] Cotton, M., Eggert, L., Touch, J., Westerlund, M., and S. 1846 Cheshire, "Internet Assigned Numbers Authority (IANA) 1847 Procedures for the Management of the Service Name and 1848 Transport Protocol Port Number Registry", BCP 165, 1849 RFC 6335, August 2011. 1851 [RFC6434] Jankiewicz, E., Loughney, J., and T. Narten, "IPv6 Node 1852 Requirements", RFC 6434, December 2011. 1854 [RFC6438] Carpenter, B. and S. Amante, "Using the IPv6 Flow Label 1855 for Equal Cost Multipath Routing and Link Aggregation in 1856 Tunnels", RFC 6438, November 2011. 1858 [SIGCOMM] Luckie, M. and B. Stasiewicz, "Measuring Path MTU 1859 Discovery Behavior", November 2010. 1861 [TBIT] Medina, A., Allman, M., and S. Floyd, "Measuring 1862 Interactions Between Transport Protocols and Middleboxes", 1863 October 2004. 1865 [TCP-IP] "Archive/Hypermail of Early TCP-IP Mail List, 1866 http://www-mice.cs.ucl.ac.uk/multimedia/misc/tcp_ip/, May 1867 1987 - May 1990.". 1869 [WAND] Luckie, M., Cho, K., and B. Owens, "Inferring and 1870 Debugging Path MTU Discovery Failures", October 2005. 1872 Author's Address 1874 Fred L. Templin (editor) 1875 Boeing Research & Technology 1876 P.O. Box 3707 1877 Seattle, WA 98124 1878 USA 1880 Email: fltemplin@acm.org