idnits 2.17.1 draft-ietf-6man-mtu-option-04.txt: -(786): Line appears to be too long, but this could be caused by non-ascii characters in UTF-8 encoding 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: ---------------------------------------------------------------------------- == There are 2 instances of lines with non-ascii characters in the document. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (23 October 2020) is 1252 days in the past. Is this intentional? Checking references for intended status: Experimental ---------------------------------------------------------------------------- -- Obsolete informational reference (is this intentional?): RFC 1063 (Obsoleted by RFC 1191) -- Obsolete informational reference (is this intentional?): RFC 2460 (Obsoleted by RFC 8200) Summary: 0 errors (**), 0 flaws (~~), 2 warnings (==), 3 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group R. Hinden 3 Internet-Draft Check Point Software 4 Intended status: Experimental G. Fairhurst 5 Expires: 26 April 2021 University of Aberdeen 6 23 October 2020 8 IPv6 Minimum Path MTU Hop-by-Hop Option 9 draft-ietf-6man-mtu-option-04 11 Abstract 13 This document specifies a new Hop-by-Hop IPv6 option that is used to 14 record the minimum Path MTU along the forward path between a source 15 host to a destination host. This collects a minimum Path MTU 16 recorded along the path to the destination. The value can then be 17 communicated back to the source using the return Path MTU field in 18 the option. 20 This Hop-by-Hop option is intended to be used in environments like 21 Data Centers and on paths between Data Centers, to allow them to 22 better take advantage of paths able to support a large Path MTU. The 23 method could also be useful in other environments, including the 24 general Internet. 26 Status of This Memo 28 This Internet-Draft is submitted in full conformance with the 29 provisions of BCP 78 and BCP 79. 31 Internet-Drafts are working documents of the Internet Engineering 32 Task Force (IETF). Note that other groups may also distribute 33 working documents as Internet-Drafts. The list of current Internet- 34 Drafts is at https://datatracker.ietf.org/drafts/current/. 36 Internet-Drafts are draft documents valid for a maximum of six months 37 and may be updated, replaced, or obsoleted by other documents at any 38 time. It is inappropriate to use Internet-Drafts as reference 39 material or to cite them other than as "work in progress." 41 This Internet-Draft will expire on 26 April 2021. 43 Copyright Notice 45 Copyright (c) 2020 IETF Trust and the persons identified as the 46 document authors. All rights reserved. 48 This document is subject to BCP 78 and the IETF Trust's Legal 49 Provisions Relating to IETF Documents (https://trustee.ietf.org/ 50 license-info) in effect on the date of publication of this document. 51 Please review these documents carefully, as they describe your rights 52 and restrictions with respect to this document. Code Components 53 extracted from this document must include Simplified BSD License text 54 as described in Section 4.e of the Trust Legal Provisions and are 55 provided without warranty as described in the Simplified BSD License. 57 Table of Contents 59 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 60 1.1. Example Operation . . . . . . . . . . . . . . . . . . . . 3 61 1.2. Use of the IPv6 Hop-by-Hop Options Header . . . . . . . . 4 62 2. Motivation and Problem Solved . . . . . . . . . . . . . . . . 5 63 3. Requirements Language . . . . . . . . . . . . . . . . . . . . 6 64 4. Applicability Statements . . . . . . . . . . . . . . . . . . 6 65 5. IPv6 Minimum Path MTU Hop-by-Hop Option . . . . . . . . . . . 6 66 6. Router, Host, and Transport Behaviors . . . . . . . . . . . . 7 67 6.1. Router Behavior . . . . . . . . . . . . . . . . . . . . . 7 68 6.2. Host Behavior . . . . . . . . . . . . . . . . . . . . . . 8 69 6.3. Transport Behavior . . . . . . . . . . . . . . . . . . . 8 70 6.3.1. Including the Option in an Outgoing Packet . . . . . 8 71 6.3.2. Validation by the Upper Layer Protocol . . . . . . . 10 72 6.3.3. Receiving the Option . . . . . . . . . . . . . . . . 10 73 6.3.4. Using the Rtn-PMTU Field . . . . . . . . . . . . . . 11 74 6.3.5. Detection of Dropping Packets that include the 75 Option . . . . . . . . . . . . . . . . . . . . . . . 12 76 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 12 77 8. Security Considerations . . . . . . . . . . . . . . . . . . . 13 78 8.1. Network Layer Host Processing . . . . . . . . . . . . . . 13 79 8.2. Validating use of the Option Data . . . . . . . . . . . . 13 80 8.3. Direct use of the Rtn-PMTU Value . . . . . . . . . . . . 14 81 8.4. Using the Rtn-PMTU Value as a Hint for Probing . . . . . 14 82 8.5. Impact of Middleboxes . . . . . . . . . . . . . . . . . . 15 83 9. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 15 84 10. Change log [RFC Editor: Please remove] . . . . . . . . . . . 15 85 11. References . . . . . . . . . . . . . . . . . . . . . . . . . 17 86 11.1. Normative References . . . . . . . . . . . . . . . . . . 17 87 11.2. Informative References . . . . . . . . . . . . . . . . . 17 88 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 18 90 1. Introduction 92 This draft proposes a new IPv6 Hop-by-Hop Option to be used to record 93 the minimum of the Maximum Transmission Unit (MTU) along the forward 94 path between the source and destination hosts. The source host 95 creates a packet with this option and fills the Min-PMTU field with 96 the value of the MTU for the outbound link that will be used to 97 forward the packet towards the destination host. 99 At each subsequent hop where the option is processed, the router 100 compares the value of the Min-PMTU Field in the option and the MTU of 101 its outgoing link. If the MTU of the link is less than the Min-PMTU, 102 it rewrites the value in the option data with the smaller value. 103 When the packet arrives at the destination host, the host can send 104 the value of the minimum reported MTU for the path back to the source 105 host using the Rtn-PMTU field in the option. The source host can 106 then use this value as an input to the method that sets the Path MTU 107 (PMTU) used by upper layer protocols. 109 1.1. Example Operation 111 The figure below illustrates the operation of the method. In this 112 case, the path between the source and destination hosts comprises 113 three links, the sender has a link MTU of size MTU-S, the link 114 between routers R1 and R2 has an MTU of size 9000 bytes, and the 115 final link to the destination has an MTU of size MTU-D. 117 +--------+ +----+ +----+ +-------+ 118 | | | | | | | | 119 | Sender +---------+ R1 +--------+ R2 +-------- + Dest. | 120 | | | | | | | | 121 +--------+ MTU-S +----+ 9000B +----+ MTU-D +-------+ 123 Three scenarios are described: 125 * Scenario 1, considers all links to have an 9000 byte MTU and the 126 method is supported by both routers. The PMTU is therefore 9000 127 bytes. 129 * Scenario 2, considers the link to the destination host (MTU-D) to 130 have an MTU of 1500 bytes. This is the smallest MTU, router R2 131 updates the Min-PMTU to 1500 bytes and the method correctly 132 updates the PMTU to 1500 bytes. Had there been another smaller 133 MTU at a link further along the path that also supports the 134 method, the lower MTU would also have been detected. 136 * Scenario 3, considers the case where the router preceding the 137 smallest link (R2) does not support the method, and the link to 138 the destination host (MTU-D) has an MTU of 1500 bytes. Therefore, 139 router R2 does not update the Min-PMTU to 1500 bytes. The method 140 then fails to detect the actual PMTU. 142 In Scenarios 2 and 3, a lower PMTU would also fail to be detected in 143 the case where PMTUD had been used and an ICMPv6 Packet to Big (PTB) 144 message had not been delivered to the sender [RFC8201]. 146 These scenarios are summarized in the table below. 148 +-+-----+-----+----+----+----------+-----------------------+ 149 | |MTU-S|MTU-D| R1 | R2 | Rec PMTU | Note | 150 +-+-----+-----+----+----+----------+-----------------------+ 151 |1|9000B|9000B| H | H | 9000 B | Endpoints attempt to | 152 | | | | | | use an 9000 B PMTU. | 153 +-+-----+-----+----+----+----------+-----------------------+ 154 |2|9000B|1500B| H | H | 1500 B | Endpoints attempt to | 155 | | | | | | | use a 1500 B PMTU. | 156 +-+-----+-----+----+----+----------+-----------------------+ 157 |3|9000B|1500B| H | - | 9000 B | Endpoints attempt to | 158 | | | | | | | use an 9000 B PMTU, | 159 | | | | | | | but need to implement | 160 | | | | | | | a method to fall back | 161 | | | | | | | to discover and use a | 162 | | | | | | | 1500 B PMTU. | 163 +-+-----+-----+----+----+----------+-----------------------+ 165 1.2. Use of the IPv6 Hop-by-Hop Options Header 167 IPv6 as specified in [RFC8200] allows nodes to optionally process 168 Hop-by-Hop headers. Specifically from Section 4: 170 * The Hop-by-Hop Options header is not inserted or deleted, but may 171 be examined or processed by any node along a packet's delivery 172 path, until the packet reaches the node (or each of the set of 173 nodes, in the case of multicast) identified in the Destination 174 Address field of the IPv6 header. The Hop-by-Hop Options header, 175 when present, must immediately follow the IPv6 header. Its 176 presence is indicated by the value zero in the Next Header field 177 of the IPv6 header. 179 * NOTE: While [RFC2460] required that all nodes must examine and 180 process the Hop-by-Hop Options header, it is now expected that 181 nodes along a packet's delivery path only examine and process the 182 Hop-by-Hop Options header if explicitly configured to do so. 184 The Hop-by-Hop Option defined in this document is designed to take 185 advantage of this property of how Hop-by-Hop options are processed. 186 Nodes that do not support this Option SHOULD ignore them. This can 187 mean that the Min-PMTU value does not account for all links along a 188 path. 190 2. Motivation and Problem Solved 192 The current state of Path MTU Discovery on the Internet is 193 problematic. The mechanisms defined in [RFC8201] are known to not 194 work well in all environments. This fails to work in various cases, 195 including when nodes in the middle of the network do not send ICMP 196 PTB messages, or rate-limited messages to the point of not making 197 them a useful mechanism, or do not have a return path to the source 198 host. 200 This results in many transport connections being configured to use 201 smaller packets (e.g., 1280 bytes) by default and makes it difficult 202 to take advantage of paths with a larger PMTU where they do exist. 203 Applications that can gain benefit from sending large packets are 204 forced to use IPv6 Fragmentation [RFC8200], which can reduce the 205 reliability of Internet communication [RFC8900]. 207 Transport encapsulations and network-layer tunnels further reduce the 208 the payload size available for a transport to use. Also, some use- 209 cases increase packet overhead, for example, Network Virtualization 210 Using Generic Routing Encapsulation (NVGRE) [RFC7637] encapsulates L2 211 packets in an outer IP header and does not allow IP Fragmentation. 213 Sending small packets can limit performance, e.g., when packet 214 processing is limited by the packet rate. The potential of multi- 215 gigabit Ethernet will not be realized if the packet size is limited 216 to 1280 bytes, because this exceeds the packet per second rate that 217 most nodes can process. For example, the packet per second rate 218 required to reach wire speed on a 10G Ethernet link with 1280 byte 219 packets is about 977K packets per second (pps), vs. 139K pps for 9000 220 byte packets. A significant difference. 222 The purpose of the this draft is to improve the situation by defining 223 a mechanism that does not rely on reception of ICMPv6 Packet Too Big 224 messages from nodes in the middle of the network. Instead, this 225 provides information to the destination host about the minimum Path 226 MTU, and sends this information back to the source host. This is 227 expected to work better than the current RFC8201-based mechanisms. 229 3. Requirements Language 231 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 232 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 233 "OPTIONAL" in this document are to be interpreted as described in BCP 234 14 [RFC2119] [RFC8174] when, and only when, they appear in all 235 capitals, as shown here. 237 4. Applicability Statements 239 This Hop-by-Hop Option header is intended to be used in environments 240 such as Data Centers and on paths between Data Centers, to allow a 241 host to better take advantage of a path that is able to support a 242 large PMTU. 244 The design of the option is sufficiently simple that it could be 245 executed on a router's fast path. A strong pull from router vendors 246 customers will be required to create critical mass for this to 247 happen. This could initially be the case for connections within and 248 between Data Centers. 250 The method could also be useful in other environments, including the 251 general Internet, if and when this Hop-by-Hop Option is supported on 252 these paths. 254 5. IPv6 Minimum Path MTU Hop-by-Hop Option 256 The Minimum Path MTU Hop-by-Hop Option has the following format: 258 Option Option Option 259 Type Data Len Data 260 +--------+--------+--------+--------+---------+-------+-+ 261 |BBCTTTTT|00000100| Min-PMTU | Rtn-PMTU |R| 262 +--------+--------+--------+--------+---------+-------+-+ 264 Option Type (see Section 4.2 of [RFC8200]): 266 BB 00 Skip over this option and continue processing. 268 C 1 Option data can change en route to the packet's final 269 destination. 271 TTTTT 10000 Option Type assigned from IANA [IANA-HBH]. 273 Length: 4 The size of the each value field in Option Data 274 field supports PMTU values from 0 to 65,535 octets. 276 Min-PMTU: n 16-bits. The minimum MTU recorded along the path 277 in octets, reflecting the smallest link MTU that 278 the packet experienced along the path. 279 A value less than the IPv6 minimum link 280 MTU [RFC8200] should be ignored. 282 Rtn-PMTU: n 15-bits. The returned Path MTU field, carrying the 15 283 most significant bits of the latest received Min-PMTU 284 field for the forward path. The value zero means that 285 no Reported MTU is being returned. 287 R n 1-bit. R-Flag. Set by the source to signal that 288 the destination host should include the received 289 Rtn-PMTU field updated by the reported Min-PMTU value. 291 NOTE: The encoding of the final two octets (Rtn-PMTU and R-Flag) 292 could be implemented by a mask of the latest received Min-PMTU value 293 with 0xFFFE, discarding the right-most bit and then performing a 294 logical 'OR' with the R-Flag value of the sender. 296 6. Router, Host, and Transport Behaviors 298 6.1. Router Behavior 300 Routers that are not configured to support Hop-by-Hop Options SHOULD 301 ignore this option and SHOULD forward the packet. 303 Routers that support Hop-by-Hop Options, but that are not configured 304 to support this option SHOULD ignore the option and SHOULD forward 305 the packet. 307 Routers that recognize this option SHOULD compare the value of the 308 Min-PMTU field with the MTU configured for the outgoing link. If the 309 MTU of the outgoing link is less than the Min-PMTU, the router 310 rewrites the Min-PMTU in the Option to use the smaller value. 312 A router MUST ignore and MUST NOT change the Rtn-PMTU field or the 313 R-Flag in the option. 315 Discussion: 317 * The design of this option makes it feasible to be implemented 318 within the fast path of a router, because the processing 319 requirements are minimal. 321 6.2. Host Behavior 323 When requested to send an IPv6 packet with the Minimum Path MTU 324 option, the source host includes the option in an outgoing packet. 325 The source host SHOULD fill the Min-PMTU field with the MTU 326 configured for the link over which it will send the packet on the 327 next hop towards the destination host. If this value is not updated, 328 the field MUST be set to zero. 330 The source host SHOULD set the Rtn-PMTU field to the cached value of 331 the reported Min-PMTU value for the flow ( see Section 6.3.3). If 332 this value is not set, for example, because there is no cached 333 reported Min-PMTU value, the field MUST be set to zero. 335 The source host MAY request the destination host to return the 336 reported Min-PMTU value by setting the R-Flag in the option of an 337 outgoing packet. 339 6.3. Transport Behavior 341 6.3.1. Including the Option in an Outgoing Packet 343 The upper layer protocol can request the Minimum Path MTU option is 344 included in an outgoing IPv6 packet. This option does not need to be 345 included in all packets belonging to a flow. A transport protocol 346 (or upper layer protocol) can include this option only on specific 347 packets used to test the path. 349 When it includes the option, the host supplies the previously cached 350 value of the received Minimum Path MTU for the flow to set the Rtn- 351 PMTU field (see Section 6.3.3). If a valid cached received Minimum 352 Path MTU is not available, the Rtn-PMTU field value MUST be set to 353 zero. 355 The source host MAY request the destination host to send a packet 356 carrying the option by setting the R-Flag. The R-Flag SHOULD NOT be 357 set when the Minimum Path MTU Option was sent solely to feedback the 358 return Path MTU. 360 NOTE: Including this option in a large packet (e.g., one larger than 361 the present PMTU) is not likely to be useful, since the large packet 362 would itself be dropped by any link along the path with a smaller 363 MTU, preventing the Min-PMTU information from reaching the 364 destination host. 366 Discussion: 368 * In the case of TCP, the option could be included in packets 369 carrying a SYN segment as part of the connection set up, or can 370 periodically be sent in packets carrying other segments. 371 Including this packet in a SYN could increase the probability that 372 the SYN segment is lost when routers on the path drop packets with 373 this option (see Section 6.3.5). NOTE: A TCP connection can also 374 negotiate the Maximum Segment Size (MSS), which acts as an upper 375 limit to the packet size that can be sent by a TCP sender. 377 * The use with datagram transport protocols (e.g., UDP) is harder to 378 characterize because applications using datagram transports range 379 from very short-lived (low data-volume applications) exchanges, to 380 longer (bulk) exchanges of packets between the source and 381 destination hosts [RFC8085]. 383 * Simple-exchange protocols (i.e., low data-volume applications 384 [RFC8085] that only send one or a few packets per transaction, 385 might assume that the PMTU is symmetrical. That is, the PMTU is 386 the same in both directions, or at least not smaller for the 387 return path. This optimization does not hold when the paths are 388 not symmetric. 390 * The use of this option with DNS and DNSSEC over UDP ought to work 391 for paths where the PMTU is symmetric. The DNS server will learn 392 the PMTU from the DNS query messages. If the Rtn-PMTU value is 393 smaller, then a large DNSSEC response might be dropped and the 394 known problems with PMTUD will then occur. DNS and DNSSEC over 395 transport protocols that can carry the PMTU ought to work. 397 * Applications that use Anycast should include this option in all 398 packets, because the actual destination host will vary due to the 399 nature of Anycast. 401 6.3.2. Validation by the Upper Layer Protocol 403 An upper layer protocol (e.g., transport endpoint) using this option 404 needs to provide protection from data injection attacks by off-path 405 devices [RFC8085]. This requires a method to assure that the 406 information in the Option Data is provided by a node on the path. 407 For example, a TCP connection or UDP application that maintains the 408 related state and uses a randomized ephemeral port would provide this 409 basic validation to protect from off-path data injection. IPsec 410 [RFC4301] and TLS [RFC8446] provide greater assurance. 412 The Upper Layer discards any received packet when the packet 413 validation fails. When packet validation fails, the Upper Layer MUST 414 also discard the associated Option Data from the minimum Path MTU 415 option without further processing. 417 6.3.3. Receiving the Option 419 An upper layer protocol that receives a Minimum Path MTU Option 420 included with a valid packet caches the value of the last received 421 Min-PMTU. This value is specific to the instance of the upper layer 422 protocol (i.e., matching the IPv6 flow ID, port-fields in UDP or the 423 SPI in IPsec [RFC4301], etc), not to the pair of source and 424 destination addresses, because network devices can make forwarding 425 decisions that impact the PMTU of a flow based on the presence and 426 value of the packet's upper layer fields. 428 For a connection-oriented upper layer protocol, caching of the 429 received Min-PMTU could be implemented by saving the value in the 430 connection context at the transport layer. A connection-less upper 431 layer (e.g., one using UDP), requires the upper layer protocol to 432 cache the value for each flow it uses. 434 A destination host that receives a Minimum Path MTU Option with the 435 R-Flag SHOULD include the Minimum Path MTU option in the next 436 outgoing IPv6 packet for the corresponding flow. 438 A simple mechanism could only include this option (with the Rtn-PMTU 439 field set) the first time this option is received or when it notifies 440 a change in the Minimum Path MTU. This limits the number of packets 441 including the option packets that are sent. However, this does not 442 provide robustness to packet loss or recovery after a sender looses 443 state. 445 Path characteristics can change and the actual PMTU could increase or 446 decrease over time. For instance, following a path change when 447 packets are then forwarded over a link with a different MTU than that 448 previously used. To bound the delay in discovering a change in the 449 actual PMTU, a sender with a link MTU larger than the current PMTU 450 SHOULD periodically send the Minimum Path MTU Option with the R-bit 451 set. DPLPMTUD provides recommendations concerning how this could be 452 implemented (see Section 5.3 of [RFC8899]). Since the option 453 consumes less capacity than a full-sized probe packet, there can be 454 advantage in using this to detect a change in the path 455 characteristics. 457 Discussion: 459 * Some upper layer protocols send packets less frequently than 460 packets that the host receives packets. This provides less 461 frequent feedback of the received Rtn-PMTU value. However, a host 462 always sends the most recent Rtn-PMTU value. 464 6.3.4. Using the Rtn-PMTU Field 466 The Rtn-PMTU field provides an indication of the PMTU from on-path 467 routers. It does not necessarily reflect the actual PMTU between the 468 sender and destination. Care therefore needs to be exercised in 469 using the Rtn-PMTU value. Specifically: 471 * The actual PMTU can be lower than the Rtn-PMTU value because Min- 472 PMTU field was not updated by a router on the path that did not 473 process the option. 475 * The actual PMTU may be lower than the Rtn-PMTU value because the 476 there is a layer 2 device with a lower MTU that does not perform 477 IPv6 forwarding. 479 * The actual PMTU may be larger than the Rtn-PMTU value because of a 480 corrupted, delayed or mis-ordered response. A source host SHOULD 481 ignore a Rtn-PMTU value larger than the MTU configured for the 482 outgoing link. 484 Using the method has the potential to complete discovery of the 485 correct value in a single round trip time, even over paths that have 486 successive links each configured with a lower MTU. 488 To avoid unintentional dropping of packets that exceed the actual 489 PMTU (e.g., Scenario 3 in Section 1.1), the source host can delay 490 increasing the PMTU until a probe packet with the size of the Rtn- 491 PMTU value has been successfully acknowledged by the upper layer, 492 confirming that the path supports the larger PMTU. This probing 493 increases robustness, but adds one additional path round trip time 494 before the PMTU is updated. This use resembles that of PTB messages 495 in section 4.6 of DPLPMTUD [RFC8899] (with the important difference 496 that a PTB message can only seek to lower the PMTU, whereas this 497 option could trigger a probe packet to seek to increase the PMTU.) 499 Section 5.2 of [RFC8201] provides guidance on the caching of PMTU 500 information and also the relation to IPv6 flow labels. 501 Implementations should consider the impact of Equal Cost Multipath 502 (ECMP) [RFC6438]. Specifically, whether a PMTU ought be maintained 503 for each transport endpoint, or for each network address. 505 6.3.5. Detection of Dropping Packets that include the Option 507 There is evidence that some middleboxes drop packets that include 508 Hop-by-Hop options. For example, a firewall might drop a packet that 509 carries an unknown extension header or option. This practice is 510 expected to decrease as an option becomes more widely used. It could 511 result in generation of an ICMPv6 message indicating the problem. 512 This could be used to (temporarily) suspend use of this option. 514 A middlebox that silently discards a packet with this option results 515 in dropping of any packet using the option. This dropping be avoided 516 by appropriate configuration in a controlled environment, such as 517 within a data centre, but needs to be considered for Internet usage. 518 Section 6.2 recommends that this option is not used on packets where 519 loss might adversely impact performance. 521 7. IANA Considerations 523 No IANA actions are requested in this document. 525 IANA has assigned and registered a new IPv6 Hop-by-Hop Option type 526 from the "Destination Options and Hop-by-Hop Options" registry 527 [IANA-HBH]. This assignment is shown in Section 5. 529 8. Security Considerations 531 This section discusses the security considerations. It first reviews 532 host processing when receiving this option at the network layer. It 533 then considers two ways in which the Option Data can be processed, 534 followed by two approaches for using the Option Data. Finally, it 535 discusses middlebox implications related to use in the general 536 Internet. 538 8.1. Network Layer Host Processing 540 A malicious attacker can forge a packet directed at a host that 541 carries the minimum Path MTU option. By design, the fields of this 542 IP option can be modified by the network. 544 Reception of this packet will incur receive processing as the network 545 stack parses the packet before the packet is delivered to the upper 546 layer protocol. This network layer option processing is normally 547 completed before any upper layer protocol delivery checks are 548 performed. 550 The network layer does not normally have sufficient information to 551 validate that the packet carrying an option originated from the 552 destination (or an on-path node). It also does not typically have 553 sufficient context to demultiplex the packet to identify the related 554 transport flow. This can mean that any changes resulting from 555 reception of the option apply to all flows between a pair of 556 endpoints. 558 These considerations are no different to other uses of Hop-by-Hop 559 options, and this is the use case for PMTUD. The following section 560 describes a mitigation for this attack. 562 8.2. Validating use of the Option Data 564 Transport protocols should be designed to provide protection from 565 data injection attacks by off-path devices and mechanisms should be 566 described in the Security Considerations for each transport 567 specification (see Section 5.1 of the UDP Guidelines [RFC8085]). For 568 example, a TCP or UDP application that maintains the related state 569 and uses a randomized ephemeral port would provide basic protection. 570 TLS [RFC8446] or IPsec [RFC4301] provide cryptographic 571 authentication. An upper layer protocol that validates each received 572 packet discards any packet when this validation fails. In this case, 573 the host MUST also discard the associated Option Data from the 574 minimum Path MTU option without further processing (Section 6.3). 576 A network node on the path has visibility of all packets it forwards. 577 By observing the network packet payload, the node might be able to 578 construct a packet that might be validated by the destination host. 579 Such a node would also be able to drop or limit the flow in other 580 ways that could be potentially more disruptive. Authenticating the 581 packet, for example, using IPsec [RFC4301] or TLS [RFC8446] mitigates 582 this attack. 584 8.3. Direct use of the Rtn-PMTU Value 586 The simplest way to utilize the Rtn-PMTU value is to directly use 587 this to update the PMTU. This approach results in a set of security 588 issues when the option carries malicious data: 590 * A direct update of the PMTU using the Rtn-PMTU value could result 591 in an attacker inflating or reducing the size of the host PMTU for 592 the destination. Forcing a reduction in the PMTU can decrease the 593 efficiency of network use, might increase the number of packets/ 594 fragments required to send the same volume of payload data, and 595 prevents sending an unfragmented datagram larger than the PMTU. 596 Increasing the PMTU can result in black-holing (see Section 1.1 of 597 [RFC8899]) when the source sends packets larger than the actual 598 PMTU. This persists until the PMTU is next updated. 600 * The method can be used to solicit a response from the destination 601 host. A malicious attacker could forge a packet that cause the 602 sender to add the option to a packet sent to the source. A forged 603 value of Rtn-PMTU in the Option Data might also impact the remote 604 endpoint, as described in the previous bullet. This persists 605 until a valid minimum Path MTU option is received. This attack 606 could be mitigated by limiting the sending of the minimum Path MTU 607 option in reply to incoming packets that carry the option. 609 8.4. Using the Rtn-PMTU Value as a Hint for Probing 611 Another way to utilize the Rtn-PMTU value is to indirectly trigger a 612 probe to determine if the path supports a PMTU of size Rtn-PMTU. 613 This approach needs context for the flow, and hence assumes an upper 614 layer protocol that validates the packet that carries the option 615 Section 8.2. This is the case when used in combination with DPLPMTUD 616 [RFC8899]. A set of security considerations result when an option 617 carries malicious data: 619 * If the forged packet carries a validated option with a non-zero 620 Rtn-PMTU field, the upper layer protocol could utilize the 621 information in the Rtn-PMTU field. A Rtn-PMTU larger than the 622 current PMTU can trigger a probe for a new size. 624 * If the forged packet carries a non-zero Min-PMTU field, the upper 625 layer protocol would change the cached information about the path 626 from the source. The cached information at the destination host 627 will be overwritten when the host receives another packet that 628 includes a minimum Path MTU option corresponding to the flow. 630 * Processing of the option could cause a destination host to add the 631 minimum Path MTU option to a packet sent to the source host. This 632 option will carry a Rtn-PMTU value that could have been updated by 633 the forged packet. The impact of the source host receiving this 634 resembles that discussed previously. 636 8.5. Impact of Middleboxes 638 There is evidence that some middleboxes drop packets that include 639 Hop-by-Hop options. For example, a firewall might drop a packet that 640 carries an unknown extension header or option. This practice is 641 expected to decrease as the option becomes more widely used. Methods 642 to address this are discussed in Section 6.3.5. 644 When a forged packet cause a packet to be sent including the minimum 645 Path MTU option, and the return path does not forward packets with 646 this option, the packet will be dropped Section 6.3.5. This attack 647 is mitigated by validating the option data before use and by limiting 648 the rate of responses generated. An upper layer could further 649 mitigate the impact by responding to a R-Flag by including the option 650 in a packet that does not carry application data. 652 9. Acknowledgments 654 A somewhat similar mechanism was proposed for IPv4 in 1988 in 655 [RFC1063] by Jeff Mogul, C. Kent, Craig Partridge, and Keith 656 McCloghire. It was later obsoleted in 1990 by [RFC1191] the current 657 deployed approach to Path MTU Discovery. 659 Helpful comments were received from Tom Herbert, Tom Jones, Fred 660 Templin, Ole Troan, [Your name here], and other members of the 6MAN 661 working group. 663 10. Change log [RFC Editor: Please remove] 665 draft-ietf-6man-mtu-option-04, 2020-Oct-23 667 * Fixes for typos. 669 draft-ietf-6man-mtu-option-03, 2020-Sept-14 671 * Rewrite to make text and terminology more consistent. 673 * Added the notion of validating the packet before use of the HBH 674 option data. 675 * Method aligned with the way common APIs send/receive HBH option 676 data. 677 * Added reference to DPLPMTUD and clarified upper layer usage. 678 * Completed security considerations section. 680 draft-ietf-6man-mtu-option-02, 2020-March-9 682 * Editorial changes to make text and terminology more consistent. 683 * Added reference to DPLPMTUD. 685 draft-ietf-6man-mtu-option-01, 2019-September-13 687 * Changes to show IANA assigned code point. 688 * Editorial changes to make text and terminology more consistent. 689 * Added a reference to RFC8200 in Section 2 and a reference to 690 RFC6438 in Section 6.3. 692 draft-ietf-6man-mtu-option-00, 2019-August-9 694 * First 6man w.g. draft version. 695 * Changes to request IANA allocation of code point. 696 * Editorial changes. 698 draft-hinden-6man-mtu-option-02, 2019-July-5 700 * Changed option format to also include the Returned PMTU value and 701 Return flag and made related text changes in Section 6.2 to 702 describe this behavior. 703 * ICMP Packet Too Big messages are no longer used for feedback to 704 the source host. 705 * Added to Acknowledgements Section that a similar mechanism was 706 proposed for IPv4 in 1988 in [RFC1063]. 707 * Editorial changes. 709 draft-hinden-6man-mtu-option-01, 2019-March-05 711 * Changed requested status from Standards Track to Experimental to 712 allow use of experimental option type (11110) to allow for 713 experimentation. Removed request for IANA Option assignment. 714 * Added Section 2 "Motivation and Problem Solved" section to better 715 describe what the purpose of this document is. 716 * Added appendix describing planned experiments and how the results 717 will be measured. 718 * Editorial changes. 720 draft-hinden-6man-mtu-option-00, 2018-Oct-16 721 * Initial draft. 723 11. References 725 11.1. Normative References 727 [IANA-HBH] "Destination Options and Hop-by-Hop Options", 728 . 731 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 732 Requirement Levels", BCP 14, RFC 2119, 733 DOI 10.17487/RFC2119, March 1997, 734 . 736 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 737 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 738 May 2017, . 740 [RFC8200] Deering, S. and R. Hinden, "Internet Protocol, Version 6 741 (IPv6) Specification", STD 86, RFC 8200, 742 DOI 10.17487/RFC8200, July 2017, 743 . 745 [RFC8201] McCann, J., Deering, S., Mogul, J., and R. Hinden, Ed., 746 "Path MTU Discovery for IP version 6", STD 87, RFC 8201, 747 DOI 10.17487/RFC8201, July 2017, 748 . 750 11.2. Informative References 752 [RFC1063] Mogul, J., Kent, C., Partridge, C., and K. McCloghrie, "IP 753 MTU discovery options", RFC 1063, DOI 10.17487/RFC1063, 754 July 1988, . 756 [RFC1191] Mogul, J. and S. Deering, "Path MTU discovery", RFC 1191, 757 DOI 10.17487/RFC1191, November 1990, 758 . 760 [RFC2460] Deering, S. and R. Hinden, "Internet Protocol, Version 6 761 (IPv6) Specification", RFC 2460, DOI 10.17487/RFC2460, 762 December 1998, . 764 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 765 Internet Protocol", RFC 4301, DOI 10.17487/RFC4301, 766 December 2005, . 768 [RFC6438] Carpenter, B. and S. Amante, "Using the IPv6 Flow Label 769 for Equal Cost Multipath Routing and Link Aggregation in 770 Tunnels", RFC 6438, DOI 10.17487/RFC6438, November 2011, 771 . 773 [RFC7637] Garg, P., Ed. and Y. Wang, Ed., "NVGRE: Network 774 Virtualization Using Generic Routing Encapsulation", 775 RFC 7637, DOI 10.17487/RFC7637, September 2015, 776 . 778 [RFC8085] Eggert, L., Fairhurst, G., and G. Shepherd, "UDP Usage 779 Guidelines", BCP 145, RFC 8085, DOI 10.17487/RFC8085, 780 March 2017, . 782 [RFC8446] Rescorla, E., "The Transport Layer Security (TLS) Protocol 783 Version 1.3", RFC 8446, DOI 10.17487/RFC8446, August 2018, 784 . 786 [RFC8899] Fairhurst, G., Jones, T., Tüxen, M., Rüngeler, I., and T. 787 Völker, "Packetization Layer Path MTU Discovery for 788 Datagram Transports", RFC 8899, DOI 10.17487/RFC8899, 789 September 2020, . 791 [RFC8900] Bonica, R., Baker, F., Huston, G., Hinden, R., Troan, O., 792 and F. Gont, "IP Fragmentation Considered Fragile", 793 BCP 230, RFC 8900, DOI 10.17487/RFC8900, September 2020, 794 . 796 Authors' Addresses 798 Robert M. Hinden 799 Check Point Software 800 959 Skyway Road 801 San Carlos, CA 94070 802 United States of America 804 Email: bob.hinden@gmail.com 806 Godred Fairhurst 807 University of Aberdeen 808 School of Engineering 809 Fraser Noble Building 810 Aberdeen 811 AB24 3UE 812 United Kingdom 814 Email: gorry@erg.abdn.ac.uk