idnits 2.17.1 draft-ietf-tsvwg-datagram-plpmtud-22.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- -- The abstract seems to indicate that this document updates RFC8201, but the header doesn't have an 'Updates:' line to match this. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year (Using the creation date from RFC4821, updated by this document, for RFC5378 checks: 2003-10-21) -- The document seems to lack a disclaimer for pre-RFC5378 work, but may have content which was first submitted before 10 November 2008. If you have contacted all the original authors and they are all willing to grant the BCP78 rights to the IETF Trust, then this is fine, and you can ignore this comment. If not, you may need to add the pre-RFC5378 disclaimer. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (10 June 2020) is 1388 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) ** Obsolete normative reference: RFC 4960 (Obsoleted by RFC 9260) == Outdated reference: A later version (-13) exists of draft-ietf-intarea-tunnels-10 == Outdated reference: A later version (-34) exists of draft-ietf-quic-transport-27 Summary: 1 error (**), 0 flaws (~~), 3 warnings (==), 3 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Internet Engineering Task Force G. Fairhurst 3 Internet-Draft T. Jones 4 Updates: 4821, 4960, 6951, 8085, 8261 (if University of Aberdeen 5 approved) M. Tuexen 6 Intended status: Standards Track I. Ruengeler 7 Expires: 12 December 2020 T. Voelker 8 Muenster University of Applied Sciences 9 10 June 2020 11 Packetization Layer Path MTU Discovery for Datagram Transports 12 draft-ietf-tsvwg-datagram-plpmtud-22 14 Abstract 16 This document describes a robust method for Path MTU Discovery 17 (PMTUD) for datagram Packetization Layers (PLs). It describes an 18 extension to RFC 1191 and RFC 8201, which specifies ICMP-based Path 19 MTU Discovery for IPv4 and IPv6. The method allows a PL, or a 20 datagram application that uses a PL, to discover whether a network 21 path can support the current size of datagram. This can be used to 22 detect and reduce the message size when a sender encounters a packet 23 black hole (where packets are discarded). The method can probe a 24 network path with progressively larger packets to discover whether 25 the maximum packet size can be increased. This allows a sender to 26 determine an appropriate packet size, providing functionality for 27 datagram transports that is equivalent to the Packetization Layer 28 PMTUD specification for TCP, specified in RFC 4821. 30 This document updates RFC 4821 to specify the PLPMTUD method for 31 datagram PLs. It also updates RFC 8085 to refer to the method 32 specified in this document instead of the method in RFC 4821 for use 33 with UDP datagrams. Section 7.3 of RFC 4960 recommends an endpoint 34 apply the techniques in RFC 4821 on a per-destination-address basis. 35 RFC 4960, RFC 6951, and RFC 8261 are updated to recommend that SCTP, 36 SCTP encapsulated in UDP and SCTP encapsulated in DTLS use the method 37 specified in this document instead of the method in RFC 4821. 39 The document also provides implementation notes for incorporating 40 Datagram PMTUD into IETF datagram transports or applications that use 41 datagram transports. 43 When published, this specification updates RFC 4960, RFC 4821, RFC 44 8085 and RFC 8261. 46 Status of This Memo 48 This Internet-Draft is submitted in full conformance with the 49 provisions of BCP 78 and BCP 79. 51 Internet-Drafts are working documents of the Internet Engineering 52 Task Force (IETF). Note that other groups may also distribute 53 working documents as Internet-Drafts. The list of current Internet- 54 Drafts is at https://datatracker.ietf.org/drafts/current/. 56 Internet-Drafts are draft documents valid for a maximum of six months 57 and may be updated, replaced, or obsoleted by other documents at any 58 time. It is inappropriate to use Internet-Drafts as reference 59 material or to cite them other than as "work in progress." 61 This Internet-Draft will expire on 12 December 2020. 63 Copyright Notice 65 Copyright (c) 2020 IETF Trust and the persons identified as the 66 document authors. All rights reserved. 68 This document is subject to BCP 78 and the IETF Trust's Legal 69 Provisions Relating to IETF Documents (https://trustee.ietf.org/ 70 license-info) in effect on the date of publication of this document. 71 Please review these documents carefully, as they describe your rights 72 and restrictions with respect to this document. Code Components 73 extracted from this document must include Simplified BSD License text 74 as described in Section 4.e of the Trust Legal Provisions and are 75 provided without warranty as described in the Simplified BSD License. 77 Table of Contents 79 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 4 80 1.1. Classical Path MTU Discovery . . . . . . . . . . . . . . 4 81 1.2. Packetization Layer Path MTU Discovery . . . . . . . . . 6 82 1.3. Path MTU Discovery for Datagram Services . . . . . . . . 7 83 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 8 84 3. Features Required to Provide Datagram PLPMTUD . . . . . . . . 11 85 4. DPLPMTUD Mechanisms . . . . . . . . . . . . . . . . . . . . . 14 86 4.1. PLPMTU Probe Packets . . . . . . . . . . . . . . . . . . 14 87 4.2. Confirmation of Probed Packet Size . . . . . . . . . . . 15 88 4.3. Black Hole Detection and Reducing the PLPMTU . . . . . . 15 89 4.4. The Maximum Packet Size (MPS) . . . . . . . . . . . . . . 17 90 4.5. Disabling the Effect of PMTUD . . . . . . . . . . . . . . 18 91 4.6. Response to PTB Messages . . . . . . . . . . . . . . . . 18 92 4.6.1. Validation of PTB Messages . . . . . . . . . . . . . 18 93 4.6.2. Use of PTB Messages . . . . . . . . . . . . . . . . . 19 95 5. Datagram Packetization Layer PMTUD . . . . . . . . . . . . . 20 96 5.1. DPLPMTUD Components . . . . . . . . . . . . . . . . . . . 21 97 5.1.1. Timers . . . . . . . . . . . . . . . . . . . . . . . 21 98 5.1.2. Constants . . . . . . . . . . . . . . . . . . . . . . 22 99 5.1.3. Variables . . . . . . . . . . . . . . . . . . . . . . 23 100 5.1.4. Overview of DPLPMTUD Phases . . . . . . . . . . . . . 24 101 5.2. State Machine . . . . . . . . . . . . . . . . . . . . . . 26 102 5.3. Search to Increase the PLPMTU . . . . . . . . . . . . . . 29 103 5.3.1. Probing for a larger PLPMTU . . . . . . . . . . . . . 29 104 5.3.2. Selection of Probe Sizes . . . . . . . . . . . . . . 30 105 5.3.3. Resilience to Inconsistent Path Information . . . . . 30 106 5.4. Robustness to Inconsistent Paths . . . . . . . . . . . . 31 107 6. Specification of Protocol-Specific Methods . . . . . . . . . 31 108 6.1. Application support for DPLPMTUD with UDP or UDP-Lite . . 31 109 6.1.1. Application Request . . . . . . . . . . . . . . . . . 32 110 6.1.2. Application Response . . . . . . . . . . . . . . . . 32 111 6.1.3. Sending Application Probe Packets . . . . . . . . . . 32 112 6.1.4. Initial Connectivity . . . . . . . . . . . . . . . . 32 113 6.1.5. Validating the Path . . . . . . . . . . . . . . . . . 32 114 6.1.6. Handling of PTB Messages . . . . . . . . . . . . . . 32 115 6.2. DPLPMTUD for SCTP . . . . . . . . . . . . . . . . . . . . 33 116 6.2.1. SCTP/IPv4 and SCTP/IPv6 . . . . . . . . . . . . . . . 33 117 6.2.1.1. Initial Connectivity . . . . . . . . . . . . . . 33 118 6.2.1.2. Sending SCTP Probe Packets . . . . . . . . . . . 33 119 6.2.1.3. Validating the Path with SCTP . . . . . . . . . . 34 120 6.2.1.4. PTB Message Handling by SCTP . . . . . . . . . . 34 121 6.2.2. DPLPMTUD for SCTP/UDP . . . . . . . . . . . . . . . . 34 122 6.2.2.1. Initial Connectivity . . . . . . . . . . . . . . 35 123 6.2.2.2. Sending SCTP/UDP Probe Packets . . . . . . . . . 35 124 6.2.2.3. Validating the Path with SCTP/UDP . . . . . . . . 35 125 6.2.2.4. Handling of PTB Messages by SCTP/UDP . . . . . . 35 126 6.2.3. DPLPMTUD for SCTP/DTLS . . . . . . . . . . . . . . . 35 127 6.2.3.1. Initial Connectivity . . . . . . . . . . . . . . 35 128 6.2.3.2. Sending SCTP/DTLS Probe Packets . . . . . . . . . 36 129 6.2.3.3. Validating the Path with SCTP/DTLS . . . . . . . 36 130 6.2.3.4. Handling of PTB Messages by SCTP/DTLS . . . . . . 36 131 6.3. DPLPMTUD for QUIC . . . . . . . . . . . . . . . . . . . . 36 132 7. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 36 133 8. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 36 134 9. Security Considerations . . . . . . . . . . . . . . . . . . . 37 135 10. References . . . . . . . . . . . . . . . . . . . . . . . . . 38 136 10.1. Normative References . . . . . . . . . . . . . . . . . . 38 137 10.2. Informative References . . . . . . . . . . . . . . . . . 39 138 Appendix A. Revision Notes . . . . . . . . . . . . . . . . . . . 41 139 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 46 141 1. Introduction 143 The IETF has specified datagram transport using UDP, SCTP, and DCCP, 144 as well as protocols layered on top of these transports (e.g., SCTP/ 145 UDP, DCCP/UDP, QUIC/UDP), and direct datagram transport over the IP 146 network layer. This document describes a robust method for Path MTU 147 Discovery (PMTUD) that can be used with these transport protocols (or 148 the applications that use their transport service) to discover an 149 appropriate size of packet to use across an Internet path. 151 1.1. Classical Path MTU Discovery 153 Classical Path Maximum Transmission Unit Discovery (PMTUD) can be 154 used with any transport that is able to process ICMP Packet Too Big 155 (PTB) messages (e.g., [RFC1191] and [RFC8201]). In this document, 156 the term PTB message is applied to both IPv4 ICMP Unreachable 157 messages (type 3) that carry the error Fragmentation Needed (Type 3, 158 Code 4) [RFC0792] and ICMPv6 Packet Too Big messages (Type 2) 159 [RFC4443]. When a sender receives a PTB message, it reduces the 160 effective MTU to the value reported as the Link MTU in the PTB 161 message. A method from time-to-time increases the packet size in 162 attempt to discover an increase in the supported PMTU. The packets 163 sent with a size larger than the current effective PMTU are known as 164 probe packets. 166 Packets not intended as probe packets are either fragmented to the 167 current effective PMTU, or the attempt to send fails with an error 168 code. Applications can be provided with a primitive to let them read 169 the Maximum Packet Size (MPS), derived from the current effective 170 PMTU. 172 Classical PMTUD is subject to protocol failures. One failure arises 173 when traffic using a packet size larger than the actual PMTU is 174 black-holed (all datagrams larger than the actual PMTU, are 175 discarded). This could arise when the PTB messages are not delivered 176 back to the sender for some reason (see for example [RFC2923]). 178 Examples where PTB messages are not delivered include: 180 * The generation of ICMP messages is usually rate limited. This 181 could result in no PTB messages being generated to the sender (see 182 section 2.4 of [RFC4443]) 184 * ICMP messages can be filtered by middleboxes (including firewalls) 185 [RFC4890]. A firewall could be configured with a policy to block 186 incoming ICMP messages, which would prevent reception of PTB 187 messages to a sending endpoint behind this firewall. 189 * When the router issuing the ICMP message drops a tunneled packet, 190 the resulting ICMP message will be directed to the tunnel ingress. 191 This tunnel endpoint is responsible for forwarding the ICMP 192 message and also processing the quoted packet within the payload 193 field to remove the effect of the tunnel, and return a correctly 194 formatted ICMP message to the sender [I-D.ietf-intarea-tunnels]. 195 Failure to do this prevents the PTB message reaching the original 196 sender. 198 * Asymmetry in forwarding can result in there being no return route 199 to the original sender, which would prevent an ICMP message being 200 delivered to the sender. This issue can also arise when policy- 201 based routing is used, Equal Cost Multipath (ECMP) routing is 202 used, or a middlebox acts as an application load balancer. An 203 example is where the path towards the server is chosen by ECMP 204 routing depending on bytes in the IP payload. In this case, when 205 a packet sent by the server encounters a problem after the ECMP 206 router, then any resulting ICMP message also needs to be directed 207 by the ECMP router towards the original sender. 209 * There are additional cases where the next hop destination fails to 210 receive a packet because of its size. This could be due to 211 misconfiguration of the layer 2 path between nodes, for instance 212 the MTU configured in a layer 2 switch, or misconfiguration of the 213 Maximum Receive Unit (MRU). If a packet is dropped by the link, 214 this will not cause a PTB message to be sent to the original 215 sender. 217 Another failure could result if a node that is not on the network 218 path sends a PTB message that attempts to force a sender to change 219 the effective PMTU [RFC8201]. A sender can protect itself from 220 reacting to such messages by utilizing the quoted packet within a PTB 221 message payload to validate that the received PTB message was 222 generated in response to a packet that had actually originated from 223 the sender. However, there are situations where a sender would be 224 unable to provide this validation. Examples where validation of the 225 PTB message is not possible include: 227 * When a router issuing the ICMP message implements RFC792 228 [RFC0792], it is only required to include the first 64 bits of the 229 IP payload of the packet within the quoted payload. There could 230 be insufficient bytes remaining for the sender to interpret the 231 quoted transport information. 233 Note: The recommendation in RFC1812 [RFC1812] is that IPv4 routers 234 return a quoted packet with as much of the original datagram as 235 possible without the length of the ICMP datagram exceeding 576 236 bytes. IPv6 routers include as much of the invoking packet as 237 possible without the ICMPv6 packet exceeding 1280 bytes [RFC4443]. 239 * The use of tunnels/encryption can reduce the size of the quoted 240 packet returned to the original source address, increasing the 241 risk that there could be insufficient bytes remaining for the 242 sender to interpret the quoted transport information. 244 * Even when the PTB message includes sufficient bytes of the quoted 245 packet, the network layer could lack sufficient context to 246 validate the message, because validation depends on information 247 about the active transport flows at an endpoint node (e.g., the 248 socket/address pairs being used, and other protocol header 249 information). 251 * When a packet is encapsulated/tunneled over an encrypted 252 transport, the tunnel/encapsulation ingress might have 253 insufficient context, or computational power, to reconstruct the 254 transport header that would be needed to perform validation. 256 * When an ICMP message is generated by a router in a network segment 257 that has inserted a header into a packet, the quoted packet could 258 contain additional protocol header information that was not 259 included in the original sent packet, and which the PL sender does 260 not process or may not know how to process. This could disrupt 261 the ability of the sender to validate this PTB message. 263 * A Network Address Translation (NAT) device that translates a 264 packet header, ought to also translate ICMP messages and update 265 the ICMP quoted packet [RFC5508] in that message. If this is not 266 correctly translated then the sender would not be able to 267 associate the message with the PL that originated the packet, and 268 hence this ICMP message cannot be validated. 270 1.2. Packetization Layer Path MTU Discovery 272 The term Packetization Layer (PL) has been introduced to describe the 273 layer that is responsible for placing data blocks into the payload of 274 IP packets and selecting an appropriate MPS. This function is often 275 performed by a transport protocol (e.g., DCCP, RTP, SCTP, QUIC), but 276 can also be performed by other encapsulation methods working above 277 the transport layer. 279 In contrast to PMTUD, Packetization Layer Path MTU Discovery 280 (PLPMTUD) [RFC4821] introduced a method that does not rely upon 281 reception and validation of PTB messages. It is therefore more 282 robust than Classical PMTUD. This has become the recommended 283 approach for implementing discovery of the PMTU [BCP145]. 285 It uses a general strategy where the PL sends probe packets to search 286 for the largest size of unfragmented datagram that can be sent over a 287 network path. Probe packets are sent to explore using a larger 288 packet size. If a probe packet is successfully delivered (as 289 determined by the PL), then the PLPMTU is raised to the size of the 290 successful probe. If a black hole is detected (e.g., where packets 291 of size PLPMTU are consistently not received), the method reduces the 292 PLPMTU. 294 Datagram PLPMTUD introduces flexibility in implementation. At one 295 extreme, it can be configured to only perform Black Hole Detection 296 and recovery with increased robustness compared to Classical PMTUD. 297 At the other extreme, all PTB processing can be disabled, and PLPMTUD 298 replaces Classical PMTUD. 300 PLPMTUD can also include additional consistency checks without 301 increasing the risk that data is lost when probing to discover the 302 Path MTU. For example, information available at the PL, or higher 303 layers, enables received PTB messages to be validated before being 304 utilized. 306 1.3. Path MTU Discovery for Datagram Services 308 Section 5 of this document presents a set of algorithms for datagram 309 protocols to discover the largest size of unfragmented datagram that 310 can be sent over a network path. The method relies upon features of 311 the PL described in Section 3 and applies to transport protocols 312 operating over IPv4 and IPv6. It does not require cooperation from 313 the lower layers, although it can utilize PTB messages when these 314 received messages are made available to the PL. 316 The message size guidelines in section 3.2 of the UDP Usage 317 Guidelines [BCP145] state "an application SHOULD either use the Path 318 MTU information provided by the IP layer or implement Path MTU 319 Discovery (PMTUD)", but does not provide a mechanism for discovering 320 the largest size of unfragmented datagram that can be used on a 321 network path. The present document updates RFC 8085 to specify this 322 method in place of PLPMTUD [RFC4821] and provides a mechanism for 323 sharing the discovered largest size as the MPS (see Section 4.4). 325 Section 10.2 of [RFC4821] recommended a PLPMTUD probing method for 326 the Stream Control Transport Protocol (SCTP). SCTP utilizes probe 327 packets consisting of a minimal sized HEARTBEAT chunk bundled with a 328 PAD chunk as defined in [RFC4820]. However, RFC 4821 did not provide 329 a complete specification. The present document replaces that 330 description by providing a complete specification. 332 The Datagram Congestion Control Protocol (DCCP) [RFC4340] requires 333 implementations to support Classical PMTUD and states that a DCCP 334 sender "MUST maintain the MPS allowed for each active DCCP session". 335 It also defines the current congestion control MPS (CCMPS) supported 336 by a network path. This recommends use of PMTUD, and suggests use of 337 control packets (DCCP-Sync) as path probe packets, because they do 338 not risk application data loss. The method defined in this 339 specification can be used with DCCP. 341 Section 4 and Section 5 define the protocol mechanisms and 342 specification for Datagram Packetization Layer Path MTU Discovery 343 (DPLPMTUD). 345 Section 6 specifies the method for datagram transports and provides 346 information to enable the implementation of PLPMTUD with other 347 datagram transports and applications that use datagram transports. 349 Section 6 also provides updated recommendations for [RFC6951] and 350 [RFC8261]. 352 2. Terminology 354 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 355 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 356 "OPTIONAL" in this document are to be interpreted as described in BCP 357 14 [RFC2119] [RFC8174] when, and only when, they appear in all 358 capitals, as shown here. 360 The following terminology is defined. Relevant terms are directly 361 copied from [RFC4821], and the definitions in [RFC1122]. 363 Acknowledged PL: A PL that includes a mechanism that can confirm 364 successful delivery of datagrams to the remote PL endpoint (e.g., 365 SCTP). Typically, the PL receiver returns acknowledgments 366 corresponding to the received datagrams, which can be utilised to 367 detect black-holing of packets (c.f., Unacknowledged PL). 369 Actual PMTU: The Actual PMTU is the PMTU of a network path between a 370 sender PL and a destination PL, which the DPLPMTUD algorithm seeks 371 to determine. 373 Black Hole: A Black Hole is encountered when a sender is unaware 374 that packets are not being delivered to the destination end point. 375 Two types of Black Hole are relevant to DPLPMTUD: 377 * Packets encounter a packet Black Hole when packets are not 378 delivered to the destination endpoint (e.g., when the sender 379 transmits packets of a particular size with a previously known 380 effective PMTU and they are discarded by the network). 382 * An ICMP Black Hole is encountered when the sender is unaware 383 that packets are not delivered to the destination endpoint 384 because PTB messages are not received by the originating PL 385 sender. 387 Classical Path MTU Discovery: Classical PMTUD is a process described 388 in [RFC1191] and [RFC8201], in which nodes rely on PTB messages to 389 learn the largest size of unfragmented packet that can be used 390 across a network path. 392 Datagram: A datagram is a transport-layer protocol data unit, 393 transmitted in the payload of an IP packet. 395 Effective PMTU: The Effective PMTU is the current estimated value 396 for PMTU that is used by a PMTUD. This is equivalent to the 397 PLPMTU derived by PLPMTUD plus the size of any headers added below 398 the PL, including the IP layer headers. 400 EMTU_S: The Effective MTU for sending (EMTU_S) is defined in 401 [RFC1122] as "the maximum IP datagram size that may be sent, for a 402 particular combination of IP source and destination addresses...". 404 EMTU_R: The Effective MTU for receiving (EMTU_R) is designated in 405 [RFC1122] as "the largest datagram size that can be reassembled". 407 Link: A Link is a communication facility or medium over which nodes 408 can communicate at the link layer, i.e., a layer below the IP 409 layer. Examples are Ethernet LANs and Internet (or higher) layer 410 tunnels. 412 Link MTU: The Link Maximum Transmission Unit (MTU) is the size in 413 bytes of the largest IP packet, including the IP header and 414 payload, that can be transmitted over a link. Note that this 415 could more properly be called the IP MTU, to be consistent with 416 how other standards organizations use the acronym. This includes 417 the IP header, but excludes link layer headers and other framing 418 that is not part of IP or the IP payload. Other standards 419 organizations generally define the link MTU to include the link 420 layer headers. This specification continues the requirement in 421 [RFC4821], that states "All links MUST enforce their MTU: links 422 that might non- deterministically deliver packets that are larger 423 than their rated MTU MUST consistently discard such packets." 425 MAX_PLPMTU: The MAX_PLPMTU is the largest size of PLPMTU that 426 DPLPMTUD will attempt to use (see the constants defined in 427 Section 5.1.2). 429 MIN_PLPMTU: The MIN_PLPMTU is the smallest size of PLPMTU that 430 DPLPMTUD will attempt to use (see the constants defined in 431 Section 5.1.2). 433 MPS: The Maximum Packet Size (MPS) is the largest size of 434 application data block that can be sent across a network path by a 435 PL using a single Datagram (see Section 4.4). 437 MSL: Maximum Segment Lifetime (MSL) The maximum delay a packet is 438 expected to experience across a path, taken as 2 minutes [BCP145]. 440 Packet: A Packet is the IP header(s) and any extension headers/ 441 options plus the IP payload. 443 Packetization Layer (PL): The PL is a layer of the network stack 444 that places data into packets and performs transport protocol 445 functions. Examples of a PL include: TCP, SCTP, SCTP over UDP, 446 SCTP over DTLS, or QUIC. 448 Path: The Path is the set of links and routers traversed by a packet 449 between a source node and a destination node by a particular flow. 451 Path MTU (PMTU): The Path MTU (PMTU) is the minimum of the Link MTU 452 of all the links forming a network path between a source node and 453 a destination node, as used by PMTUD. 455 PTB: In this document, the term PTB message is applied to both IPv4 456 ICMP Unreachable messages (type 3) that carry the error 457 Fragmentation Needed (Type 3, Code 4) [RFC0792] and ICMPv6 Packet 458 Too Big messages (Type 2) [RFC4443]. 460 PTB_SIZE: The PTB_SIZE is a value reported in a validated PTB 461 message that indicates next hop link MTU of a router along the 462 path. 464 PL_PTB_SIZE: The size reported in a validated PTB message, reduced 465 by the size of all headers added by layers below the PL. 467 PLPMTU: The Packetization Layer PMTU is an estimate of the largest 468 size of PL datagram that can be sent by a path, controled by 469 PLPMTUD. 471 PLPMTUD: Packetization Layer Path MTU Discovery (PLPMTUD), the 472 method described in this document for datagram PLs, which is an 473 extension to Classical PMTU Discovery. 475 Probe packet: A probe packet is a datagram sent with a purposely 476 chosen size (typically the current PLPMTU or larger) to detect if 477 packets of this size can be successfully sent end-to-end across 478 the network path. 480 Unacknowledged PL: A PL that does not itself provide a mechanism to 481 confirm delivery of datagrams to the remote PL endpoint (e.g., 482 UDP), and therefore requires DPLPMTUD to provide a mechanism to 483 detect black-holing of packets (c.f., Acknowledged PL). 485 3. Features Required to Provide Datagram PLPMTUD 487 The principles expressed in [RFC4821] apply to the use of the 488 technique with any PL. TCP PLPMTUD has been defined using standard 489 TCP protocol mechanisms. Unlike TCP, a datagram PL requires 490 additional mechanisms and considerations to implement PLPMTUD. 492 The requirements for datagram PLPMTUD are: 494 1. Managing the PLPMTU: For datagram PLs, the PLPMTU is managed by 495 DPLPMTUD. A PL MUST NOT send a datagram (other than a probe 496 packet) with a size at the PL that is larger than the current 497 PLPMTU. 499 2. Probe packets: The network interface below PL is REQUIRED to 500 provide a way to transmit a probe packet that is larger than the 501 PLPMTU. In IPv4, a probe packet MUST be sent with the Don't 502 Fragment (DF) bit set in the IP header, and without network layer 503 endpoint fragmentation. In IPv6, a probe packet is always sent 504 without source fragmentation (as specified in section 5.4 of 505 [RFC8201]). 507 3. Reception feedback: The destination PL endpoint is REQUIRED to 508 provide a feedback method that indicates to the DPLPMTUD sender 509 when a probe packet has been received by the destination PL 510 endpoint. Section 6 provides examples of how a PL can provide 511 this acknowledgment of received probe packets. 513 4. Probe loss recovery: It is RECOMMENDED to use probe packets that 514 do not carry any user data that would require retransmission if 515 lost. Most datagram transports permit this. If a probe packet 516 contains user data requiring retransmission in case of loss, the 517 PL (or layers above) are REQUIRED to arrange any retransmission/ 518 repair of any resulting loss. The PL is REQUIRED to be robust in 519 the case where probe packets are lost due to other reasons 520 (including link transmission error, congestion). 522 5. PMTU parameters: A DPLPMTUD sender is RECOMMENDED to utilize 523 information about the maximum size of packet that can be 524 transmitted by the sender on the local link (e.g., the local Link 525 MTU). A PL sender MAY utilize similar information about the 526 maximum size of network layer packet that a receiver can accept 527 when this is supplied (note this could be less than EMTU_R). 528 This avoids implementations trying to send probe packets that can 529 not be transferred by the local link. Too high of a value could 530 reduce the efficiency of the search algorithm. Some applications 531 also have a maximum transport protocol data unit (PDU) size, in 532 which case there is no benefit from probing for a size larger 533 than this (unless a transport allows multiplexing multiple 534 applications PDUs into the same datagram). 536 6. Processing PTB messages: A DPLPMTUD sender MAY optionally utilize 537 PTB messages received from the network layer to help identify 538 when a network path does not support the current size of probe 539 packet. Any received PTB message MUST be validated before it is 540 used to update the PLPMTU discovery information [RFC8201]. This 541 validation confirms that the PTB message was sent in response to 542 a packet originating by the sender, and needs to be performed 543 before the PLPMTU discovery method reacts to the PTB message. A 544 PTB message MUST NOT be used to increase the PLPMTU [RFC8201], 545 but could trigger a probe to test for a larger PLPMTU. A valid 546 PTB_SIZE is converted to a PL_PTB_SIZE before it is to be used in 547 the DPLPMTUD state machine. A PL_PTB_SIZE that is greater than 548 that currently probed SHOULD be ignored. (This PTB message ought 549 to be discarded without further processing, but could be utilized 550 as an input that enables a resilience mode). 552 7. Probing and congestion control: A PL MAY use a congestion 553 controller to decide when to send a probe packet. If 554 transmission of probe packets is limited by the congestion 555 controller, this could result in transmission of probe packets 556 being delayed or suspended during congestion. When the 557 transmission of probe packets is not controlled by the congestion 558 controller, the interval between probe packets MUST be at least 559 one RTT. Loss of a probe packet SHOULD NOT be treated as an 560 indication of congestion and SHOULD NOT trigger a congestion 561 control reaction [RFC4821], because this could result in 562 unnecessary reduction of the sending rate. An update to the 563 PLPMTU (or MPS) MUST NOT increase the congestion window measured 564 in bytes [RFC4821]. Therefore, an increase in the packet size 565 does not cause an increase in the data rate in bytes per second. 566 A PL that maintains the congestion window in terms of a limit to 567 the number of outstanding fixed size packets SHOULD adapt this 568 limit to compensate for the size of the actual packets. The 569 transmission of probe packets can interact with the operation of 570 a PL that performs burst mitigation or pacing and could need 571 transmission of probe packets to be regulated by these methods. 573 8. Probing and flow control: Flow control at the PL concerns the 574 end-to-end flow of data using the PL service. Flow control 575 SHOULD NOT apply to DPLPMTU when probe packets use a design that 576 does not carry user data to the remote application. 578 9. Shared PLPMTU state: The PMTU value calculated from the PLPMTU 579 MAY also be stored with the corresponding entry associated with 580 the destination in the IP layer cache, and used by other PL 581 instances. The specification of PLPMTUD [RFC4821] states: "If 582 PLPMTUD updates the MTU for a particular path, all Packetization 583 Layer sessions that share the path representation (as described 584 in Section 5.2 of [RFC4821]) SHOULD be notified to make use of 585 the new MTU". Such methods MUST be robust to the wide variety of 586 underlying network forwarding behaviors. Section 5.2 of 587 [RFC8201] provides guidance on the caching of PMTU information 588 and also the relation to IPv6 flow labels. 590 In addition, the following principles are stated for design of a 591 DPLPMTUD method: 593 * A PL MAY be designed to segment data blocks larger than the MPS 594 into multiple datagrams. However, not all datagram PLs support 595 segmentation of data blocks. It is RECOMMENDED that methods avoid 596 forcing an application to use an arbitrary small MPS for 597 transmission while the method is searching for the currently 598 supported PLPMTU. A reduced MPS can adversely impact the 599 performance of an application. 601 * To assist applications in choosing a suitable data block size, the 602 PL is RECOMMENDED to provide a primitive that returns the MPS 603 derived from the PLPMTU to the higher layer using the PL. The 604 value of the MPS can change following a change in the path, or 605 loss of probe packets. 607 * Path validation: It is RECOMMENDED that methods are robust to path 608 changes that could have occurred since the path characteristics 609 were last confirmed, and to the possibility of inconsistent path 610 information being received. 612 * Datagram reordering: A method is REQUIRED to be robust to the 613 possibility that a flow encounters reordering, or the traffic 614 (including probe packets) is divided over more than one network 615 path. 617 * Datagram delay and duplication: The feedback mechanism is REQUIRED 618 to be robust to the possibility that packets could be 619 significantly delayed or duplicated along a network path. 621 * When to probe: It is RECOMMENDED that methods determine whether 622 the path has changed since it last measured the path. This can 623 help determine when to probe the path again. 625 4. DPLPMTUD Mechanisms 627 This section lists the protocol mechanisms used in this 628 specification. 630 4.1. PLPMTU Probe Packets 632 The DPLPMTUD method relies upon the PL sender being able to generate 633 probe packets with a specific size. TCP is able to generate these 634 probe packets by choosing to appropriately segment data being sent 635 [RFC4821]. In contrast, a datagram PL that constructs a probe packet 636 has to either request an application to send a data block that is 637 larger than that generated by an application, or to utilize padding 638 functions to extend a datagram beyond the size of the application 639 data block. Protocols that permit exchange of control messages 640 (without an application data block) can generate a probe packet by 641 extending a control message with padding data. The total size of a 642 probe packet includes all headers and padding added to the payload 643 data being sent (e.g., including protocol option fields, security- 644 related fields such as an Authenticated Encryption with Associated 645 Data (AEAD) tag and TLS record layer padding). 647 A receiver is REQUIRED to be able to distinguish an in-band data 648 block from any added padding. This is needed to ensure that any 649 added padding is not passed on to an application at the receiver. 651 This results in three possible ways that a sender can create a probe 652 packet: 654 Probing using padding data: A probe packet that contains only 655 control information together with any padding, which is needed to 656 be inflated to the size of the probe packet. Since these probe 657 packets do not carry an application-supplied data block, they do 658 not typically require retransmission, although they do still 659 consume network capacity and incur endpoint processing. 661 Probing using application data and padding data: A probe packet that 662 contains a data block supplied by an application that is combined 663 with padding to inflate the length of the datagram to the size of 664 the probe packet. 666 Probing using application data: A probe packet that contains a data 667 block supplied by an application that matches the size of the 668 probe packet. This method requests the application to issue a 669 data block of the desired probe size. 671 A PL that uses a probe packet carrying application data and needs 672 protection from the loss of this probe packet could perform 673 transport-layer retransmission/repair of the data block (e.g., by 674 retransmission after loss is detected or by duplicating the data 675 block in a datagram without the padding data). This retransmitted 676 data block might possibly need to be sent using a smaller PLPMTU, 677 which could force the PL to to use a smaller packet size to traverse 678 the end-to-end path. (This could utilize endpoint network-layer 679 fragmentation or a PL that can re-segment the data block into 680 multiple datagrams). 682 DPLPMTUD MAY choose to use only one of these methods to simplify the 683 implementation. 685 Probe messages sent by a PL MUST contain enough information to 686 uniquely identify the probe within Maximum Segment Lifetime (e.g., 687 including a unique identifier from the PL or the DPLPMTUD 688 implementation), while being robust to reordering and replay of probe 689 response and PTB messages. 691 4.2. Confirmation of Probed Packet Size 693 The PL needs a method to determine (confirm) when probe packets have 694 been successfully received end-to-end across a network path. 696 Transport protocols can include end-to-end methods that detect and 697 report reception of specific datagrams that they send (e.g., DCCP, 698 SCTP, and QUIC provide keep-alive/heartbeat features). When 699 supported, this mechanism MAY also be used by DPLPMTUD to acknowledge 700 reception of a probe packet. 702 A PL that does not acknowledge data reception (e.g., UDP and UDP- 703 Lite) is unable itself to detect when the packets that it sends are 704 discarded because their size is greater than the actual PMTU. These 705 PLs need to rely on an application protocol to detect this loss. 707 Section 6 specifies this function for a set of IETF-specified 708 protocols. 710 4.3. Black Hole Detection and Reducing the PLPMTU 712 The description that follows uses the set of constants defined in 713 Section 5.1.2 and variables defined in Section 5.1.3. 715 Black Hole Detection is triggered by an indication that the network 716 path could be unable to support the current PLPMTU size. 718 There are three indicators that can detect black holes: 720 * A validated PTB message can be received that indicates a 721 PL_PTB_SIZE less than the current PLPMTU. A DPLPMTUD method MUST 722 NOT rely solely on this method. 724 * A PL can use the DPLPMTUD probing mechanism to periodically 725 generate probe packets of the size of the current PLPMTU (e.g., 726 using the confirmation timer Section 5.1.1). A timer tracks 727 whether acknowledgments are received. Successive loss of probes 728 is an indication that the current path no longer supports the 729 PLPMTU (e.g., when the number of probe packets sent without 730 receiving an acknowledgment, PROBE_COUNT, becomes greater than 731 MAX_PROBES). 733 * A PL can utilize an event that indicates the network path no 734 longer sustains the sender's PLPMTU size. This could use a 735 mechanism implemented within the PL to detect excessive loss of 736 data sent with a specific packet size and then conclude that this 737 excessive loss could be a result of an invalid PLPMTU (as in 738 PLPMTUD for TCP [RFC4821]). 740 The three methods can result in different transmission patterns for 741 packet probes and are expected to result in different responsiveness 742 following a change in the actual PMTU. 744 A PL MAY inhibit sending probe packets when no application data has 745 been sent since the previous probe packet. A PL that resumes sending 746 user data MAY continue PLPMTU discovery for each path. This allows 747 it to use an up-to-date PLPMTU. However, this could result in 748 additional packets being sent. 750 When the method detects the current PLPMTU is not supported, DPLPMTUD 751 sets a lower PLPMTU, and sets a lower MPS. The PL then confirms that 752 the new PLPMTU can be successfully used across the path. A probe 753 packet could need to have a size less than the size of the data block 754 generated by the application. 756 4.4. The Maximum Packet Size (MPS) 758 The result of probing determines a usable PLPMTU, which is used to 759 set the MPS used by the application. The MPS is smaller than the 760 PLPMTU because it is reduced by the size of PL headers (including the 761 overhead of security-related fields such as an AEAD tag and TLS 762 record layer padding). The relationship between the MPS and the 763 PLPMTUD is illustrated in Figure 1. 765 any additional 766 headers .--- MPS -----. 767 | | | 768 v v v 769 +------------------------------+ 770 | IP | ** | PL | protocol data | 771 +------------------------------+ 773 <----- PLPMTU -----> 774 <---------- PMTU --------------> 776 Figure 1: Relationship between MPS and PLPMTU 778 A PL is unable to send a packet (other than a probe packet) with a 779 size larger than the current PLPMTU at the network layer. To avoid 780 this, a PL MAY be designed to segment data blocks larger than the MPS 781 into multiple datagrams. 783 DPLPMTUD seeks to avoid IP fragmentation. An attempt to send a data 784 block larger than the MPS will therefore fail if a PL is unable to 785 segment data. To determine the largest data block that can be sent, 786 a PL SHOULD provide applications with a primitive that returns the 787 MPS, derived from the current PLPMTU. 789 If DPLPMTUD results in a change to the MPS, the application needs to 790 adapt to the new MPS. A particular case can arise when packets have 791 been sent with a size less than the MPS and the PLPMTU was 792 subsequently reduced. If these packets are lost, the PL MAY segment 793 the data using the new MPS. If a PL is unable to re-segment a 794 previously sent datagram (e.g., [RFC4960]), then the sender either 795 discards the datagram or could perform retransmission using network- 796 layer fragmentation to form multiple IP packets not larger than the 797 PLPMTU. For IPv4, the use of endpoint fragmentation by the sender is 798 preferred over clearing the DF bit in the IPv4 header. Operational 799 experience reveals that IP fragmentation can reduce the reliability 800 of Internet communication [I-D.ietf-intarea-frag-fragile], which may 801 reduce the probability of successful retransmission. 803 4.5. Disabling the Effect of PMTUD 805 A PL implementing this specification MUST suspend network layer 806 processing of outgoing packets that enforces a PMTU 807 [RFC1191][RFC8201] for each flow utilizing DPLPMTUD, and instead use 808 DPLPMTUD to control the size of packets that are sent by a flow. 809 This removes the need for the network layer to drop or fragment sent 810 packets that have a size greater than the PMTU. 812 4.6. Response to PTB Messages 814 This method requires the DPLPMTUD sender to validate any received PTB 815 message before using the PTB information. The response to a PTB 816 message depends on the PL_PTB_SIZE calculated from the PTB_SIZE in 817 the PTB message, the state of the PLPMTUD state machine, and the IP 818 protocol being used. 820 Section 4.6.1 first describes validation for both IPv4 ICMP 821 Unreachable messages (type 3) and ICMPv6 Packet Too Big messages, 822 both of which are referred to as PTB messages in this document. 824 4.6.1. Validation of PTB Messages 826 This section specifies utilization and validation of PTB messages. 828 * A simple implementation MAY ignore received PTB messages and in 829 this case the PLPMTU is not updated when a PTB message is 830 received. 832 * A PL that supports PTB messages MUST validate these messages 833 before they are further processed. 835 A PL that receives a PTB message from a router or middlebox performs 836 ICMP validation (see Section 4 of [RFC8201] and Section 5.2 of 837 [BCP145]). Because DPLPMTUD operates at the PL, the PL needs to 838 check that each received PTB message is received in response to a 839 packet transmitted by the endpoint PL performing DPLPMTUD. 841 The PL MUST check the protocol information in the quoted packet 842 carried in an ICMP PTB message payload to validate the message 843 originated from the sending node. This validation includes 844 determining that the combination of the IP addresses, the protocol, 845 the source port and destination port match those returned in the 846 quoted packet - this is also necessary for the PTB message to be 847 passed to the corresponding PL. 849 The validation SHOULD utilize information that it is not simple for 850 an off-path attacker to determine [BCP145]. For example, it could 851 check the value of a protocol header field known only to the two PL 852 endpoints. A datagram application that uses well-known source and 853 destination ports ought to also rely on other information to complete 854 this validation. 856 These checks are intended to provide protection from packets that 857 originate from a node that is not on the network path. A PTB message 858 that does not complete the validation MUST NOT be further utilized by 859 the DPLPMTUD method, as discussed in the Security Considerations 860 section. 862 Section 4.6.2 describes this processing of PTB messages. 864 4.6.2. Use of PTB Messages 866 PTB messages that have been validated MAY be utilized by the DPLPMTUD 867 algorithm, but MUST NOT be used directly to set the PLPMTU. 869 Before using the size reported in the PTB message it must first be 870 converted to a PL_PTB_SIZE. The PL_PTB_SIZE is smaller than the 871 PTB_SIZE because it is reduced by headers below the PL including any 872 IP options or extensions added to the PL packet. 874 A method that utilizes these PTB messages can improve the speed at 875 which the algorithm detects an appropriate PLPMTU by triggering an 876 immediate probe for the PL_PTB_SIZE (resulting in a network-layer 877 packet of size PTB_SIZE), compared to one that relies solely on 878 probing using a timer-based search algorithm. 880 A set of checks are intended to provide protection from a router that 881 reports an unexpected PTB_SIZE. The PL also needs to check that the 882 indicated PL_PTB_SIZE is less than the size used by probe packets and 883 at least the minimum size accepted. 885 This section provides a summary of how PTB messages can be utilized. 886 (This uses the set of constants defined in Section 5.1.2). This 887 processing depends on the PL_PTB_SIZE and the current value of a set 888 of variables: 890 PL_PTB_SIZE < MIN_PLPMTU 891 * Invalid PL_PTB_SIZE see Section 4.6.1. 893 * PTB message ought to be discarded without further processing 894 (i.e., PLPMTU is not modified). 896 * The information could be utilized as an input that triggers 897 enabling a resilience mode (see Section 5.3.3). 899 MIN_PLPMTU < PL_PTB_SIZE < BASE_PLPMTU 900 * A robust PL MAY enter an error state (see Section 5.2) for an 901 IPv4 path when the PL_PTB_SIZE reported in the PTB message is 902 larger than or equal to 68 bytes [RFC0791] and when this is 903 less than the BASE_PLPMTU. 905 * A robust PL MAY enter an error state (see Section 5.2) for an 906 IPv6 path when the PL_PTB_SIZE reported in the PTB message is 907 larger than or equal to 1280 bytes [RFC8200] and when this is 908 less than the BASE_PLPMTU. 910 BASE_PLPMTU <= PL_PTB_SIZE < PLPMTU 911 * This could be an indication of a black hole. The PLPMTU SHOULD 912 be set to BASE_PLPMTU (the PLPMTU is reduced to the BASE_PLPMTU 913 to avoid unnecessary packet loss when a black hole is 914 encountered). 916 * The PL ought to start a search to quickly discover the new 917 PLPMTU. The PL_PTB_SIZE reported in the PTB message can be 918 used to initialize a search algorithm. 920 PLPMTU < PL_PTB_SIZE < PROBED_SIZE 921 * The PLPMTU continues to be valid, but the size of a packet used 922 to search (PROBED_SIZE) was larger than the actual PMTU. 924 * The PLPMTU is not updated. 926 * The PL can use the reported PL_PTB_SIZE from the PTB message as 927 the next search point when it resumes the search algorithm. 929 PL_PTB_SIZE >= PROBED_SIZE 930 * Inconsistent network signal. 932 * PTB message ought to be discarded without further processing 933 (i.e., PLPMTU is not modified). 935 * The information could be utilized as an input to trigger 936 enabling a resilience mode. 938 5. Datagram Packetization Layer PMTUD 940 This section specifies Datagram PLPMTUD (DPLPMTUD). The method can 941 be introduced at various points (as indicated with * in the figure 942 below) in the IP protocol stack to discover the PLPMTU so that an 943 application can utilize an appropriate MPS for the current network 944 path. 946 DPLPMTUD SHOULD only be performed at one layer between a pair of 947 endpoints. Therefore, an upper PL or application should avoid using 948 DPLPMTUD when this is already enabled in a lower layer. A PL MUST 949 adjust the MPS indicated by DPLPMTUD to account for any additional 950 overhead introduced by the PL. 952 +----------------------+ 953 | Application* | 954 +-----+------------+---+ 955 | | 956 +---+--+ +--+--+ 957 | QUIC*| |SCTP*| 958 +---+--+ +-+-+-+ 959 | | | 960 +---+ +----+ | 961 | | | 962 +-+--+-+ | 963 | UDP | | 964 +---+--+ | 965 | | 966 +-----------+-------+--+ 967 | Network Interface | 968 +----------------------+ 970 Figure 2: Examples where DPLPMTUD can be implemented 972 The central idea of DPLPMTUD is probing by a sender. Probe packets 973 are sent to find the maximum size of user message that can be 974 completely transferred across the network path from the sender to the 975 destination. 977 The following sections identify the components needed for 978 implementation, provides an overview of the phases of operation, and 979 specifies the state machine and search algorithm. 981 5.1. DPLPMTUD Components 983 This section describes the timers, constants, and variables of 984 DPLPMTUD. 986 5.1.1. Timers 988 The method utilizes up to three timers: 990 PROBE_TIMER: The PROBE_TIMER is configured to expire after a period 991 longer than the maximum time to receive an acknowledgment to a 992 probe packet. This value MUST NOT be smaller than 1 second, and 993 SHOULD be larger than 15 seconds. Guidance on selection of the 994 timer value are provided in Section 3.1.1 of the UDP Usage 995 Guidelines [BCP145]. 997 PMTU_RAISE_TIMER: The PMTU_RAISE_TIMER is configured to the period a 998 sender will continue to use the current PLPMTU, after which it re- 999 enters the Search phase. This timer has a period of 600 seconds, 1000 as recommended by PLPMTUD [RFC4821]. 1002 DPLPMTUD MAY inhibit sending probe packets when no application 1003 data has been sent since the previous probe packet. A PL 1004 preferring to use an up-to-date PMTU once user data is sent again, 1005 can choose to continue PMTU discovery for each path. However, 1006 this will result in sending additional packets. 1008 CONFIRMATION_TIMER: When an acknowledged PL is used, this timer MUST 1009 NOT be used. For other PLs, the CONFIRMATION_TIMER is configured 1010 to the period a PL sender waits before confirming the current 1011 PLPMTU is still supported. This is less than the PMTU_RAISE_TIMER 1012 and used to decrease the PLPMTU (e.g., when a black hole is 1013 encountered). Confirmation needs to be frequent enough when data 1014 is flowing that the sending PL does not black hole extensive 1015 amounts of traffic. Guidance on selection of the timer value are 1016 provided in Section 3.1.1 of the UDP Usage Guidelines [BCP145]. 1018 DPLPMTUD MAY inhibit sending probe packets when no application 1019 data has been sent since the previous probe packet. A PL 1020 preferring to use an up-to-date PMTU once user data is sent again, 1021 can choose to continue PMTU discovery for each path. However, 1022 this could result in sending additional packets. 1024 DPLPMTD specifies various timers, however an implementation could 1025 choose to realise these timer functions using a single timer. 1027 5.1.2. Constants 1029 The following constants are defined: 1031 MAX_PROBES: The MAX_PROBES is the maximum value of the PROBE_COUNT 1032 counter (see Section 5.1.3). MAX_PROBES represents the limit for 1033 the number of consecutive probe attempts of any size. Search 1034 algorithms benefit from a MAX_PROBES value greater than 1 because 1035 this can provide robustness to isolated packet loss. The default 1036 value of MAX_PROBES is 3. 1038 MIN_PLPMTU: The MIN_PLPMTU is the smallest size of PLPMTU that 1039 DPLPMTUD will attempt to use. An endpoint could need to be 1040 configure the MIN_PLPMTU to provide space for extension headers 1041 and other encapsulations at layers below the PL. This value can 1042 be interface and path dependent. For IPv6, this size is greater 1043 than or equal to the size at the PL that results in an 1280 byte 1044 IPv6 packet, as specified in [RFC8200]. For IPv4, this size is 1045 greater than or equal to the size at the PL that results in an 68 1046 byte IPv4 packet. Note: An IPv4 router is required to be able to 1047 forward a datagram of 68 bytes without further fragmentation. 1048 This is the combined size of an IPv4 header and the minimum 1049 fragment size of 8 bytes. In addition, receivers are required to 1050 be able to reassemble fragmented datagrams at least up to 576 1051 bytes, as stated in section 3.3.3 of [RFC1122]. 1053 MAX_PLPMTU: The MAX_PLPMTU is the largest size of PLPMTU. This has 1054 to be less than or equal to the maximum size of the PL packet that 1055 can be sent on the outgoing interface (constrained by the local 1056 interface MTU). When known, this also ought to be less than the 1057 maximum size of PL packet that can be received by the remote 1058 endpoint (constrained by EMTU_R). It can be limited by the design 1059 or configuration of the PL being used. An application, or PL, MAY 1060 choose a smaller MAX_PLPMTU when there is no need to send packets 1061 larger than a specific size. 1063 BASE_PLPMTU: The BASE_PLPMTU is a configured size expected to work 1064 for most paths. The size is equal to or larger than the 1065 MIN_PLPMTU and smaller than the MAX_PLPMTU. For most PLs a 1066 suitable BASE_PLPMTU will be larger than 1200 bytes. When using 1067 IPv4, there is no currently equivalent size specified and a 1068 default BASE_PLPMTU of 1200 bytes is RECOMMENDED. 1070 5.1.3. Variables 1072 This method utilizes a set of variables: 1074 PROBED_SIZE: The PROBED_SIZE is the size of the current probe packet 1075 as determined at the PL. This is a tentative value for the 1076 PLPMTU, which is awaiting confirmation by an acknowledgment. 1078 PROBE_COUNT: The PROBE_COUNT is a count of the number of successive 1079 unsuccessful probe packets that have been sent. Each time a probe 1080 packet is acknowledged, the value is set to zero. (Some probe 1081 loss is expected while searching, therefore loss of a single probe 1082 is not an indication of a PMTU problem.) 1084 The figure below illustrates the relationship between the packet size 1085 constants and variables at a point of time when the DPLPMTUD 1086 algorithm performs path probing to increase the size of the PLPMTU. 1087 A probe packet has been sent of size PROBED_SIZE. Once this is 1088 acknowledged, the PLPMTU will raise to PROBED_SIZE allowing the 1089 DPLPMTUD algorithm to further increase PROBED_SIZE toward sending a 1090 probe with the size of the actual PMTU. 1092 MIN_PLPMTU MAX_PLPMTU 1093 <-------------------------------------------> 1094 | | | 1095 v | | 1096 BASE_PLPMTU | v 1097 | PROBED_SIZE 1098 v 1099 PLPMTU 1101 Figure 3: Relationships between packet size constants and variables 1103 5.1.4. Overview of DPLPMTUD Phases 1105 This section provides a high-level informative view of the DPLPMTUD 1106 method, by describing the movement of the method through several 1107 phases of operation. More detail is available in the state machine 1108 Section 5.2. 1110 +------+ 1111 +------->| Base |-----------------+ Connectivity 1112 | +------+ | or BASE_PLPMTU 1113 | | | confirmation failed 1114 | | v 1115 | | Connectivity +-------+ 1116 | | and BASE_PLPMTU | Error | 1117 | | confirmed +-------+ 1118 | | | Consistent 1119 | v | connectivity 1120 Black Hole | +--------+ | and BASE_PLPMTU 1121 detected | | Search |<---------------+ confirmed 1122 | +--------+ 1123 | ^ | 1124 | | | 1125 | Raise | | Search 1126 | timer | | algorithm 1127 | expired | | completed 1128 | | | 1129 | | v 1130 | +-----------------+ 1131 +---| Search Complete | 1132 +-----------------+ 1134 Figure 4: DPLPMTUD Phases 1136 Base: The Base Phase confirms connectivity to the remote peer using 1137 packets of the BASE_PLPMTU. The confirmation of connectivity is 1138 implicit for a connection-oriented PL (where it can be performed 1139 in a PL connection handshake). A connectionless PL sends a probe 1140 packet and uses acknowledgment of this probe packet to confirm 1141 that the remote peer is reachable. 1143 The sender also confirms that BASE_PLPMTU is supported across the 1144 network path. This may be achieved using a PL mechanism (e.g., 1145 using a handshake packet of size BASE_PLPMTU), or by sending a 1146 probe packet of size BASE_PLPMTU and confirming that this is 1147 received. 1149 A probe packet of size BASE_PLPMTU can be sent immediately on the 1150 initial entry to the Base Phase (following a connectivity check). 1151 A PL that does not wish to support a path with a PLPMTU less than 1152 BASE_PLPMTU can simplify the phase into a single step by 1153 performing the connectivity checks with a probe of the BASE_PLPMTU 1154 size. 1156 Once confirmed, DPLPMTUD enters the Search Phase. If the Base 1157 Phase fails to confirm the BASE_PLPMTU, DPLPMTUD enters the Error 1158 Phase. 1160 Search: The Search Phase utilizes a search algorithm to send probe 1161 packets to seek to increase the PLPMTU. The algorithm concludes 1162 when it has found a suitable PLPMTU, by entering the Search 1163 Complete Phase. 1165 A PL could respond to PTB messages using the PTB to advance or 1166 terminate the search, see Section 4.6. 1168 Search Complete: The Search Complete Phase is entered when the 1169 PLPMTU is supported across the network path. A PL can use a 1170 CONFIRMATION_TIMER to periodically repeat a probe packet for the 1171 current PLPMTU size. If the sender is unable to confirm 1172 reachability (e.g., if the CONFIRMATION_TIMER expires) or the PL 1173 signals a lack of reachability, a black hole has been detected and 1174 DPLPMTUD enters the Base phase. 1176 The PMTU_RAISE_TIMER is used to periodically resume the search 1177 phase to discover if the PLPMTU can be raised. Black Hole 1178 Detection causes the sender to enter the Base Phase. 1180 Error: The Error Phase is entered when there is conflicting or 1181 invalid PLPMTU information for the path (e.g., a failure to 1182 support the BASE_PLPMTU) that cause DPLPMTUD to be unable to 1183 progress and the PLPMTU is lowered. 1185 DPLPMTUD remains in the Error Phase until a consistent view of the 1186 path can be discovered and it has also been confirmed that the 1187 path supports the BASE_PLPMTU (or DPLPMTUD is suspended). 1189 A method that only reduces the PLPMTU to a suitable size would be 1190 sufficient to ensure reliable operation, but can be very inefficient 1191 when the actual PMTU changes or when the method (for whatever reason) 1192 makes a suboptimal choice for the PLPMTU. 1194 A full implementation of DPLPMTUD provides an algorithm enabling the 1195 DPLPMTUD sender to increase the PLPMTU following a change in the 1196 characteristics of the path, such as when a link is reconfigured with 1197 a larger MTU, or when there is a change in the set of links traversed 1198 by an end-to-end flow (e.g., after a routing or path fail-over 1199 decision). 1201 5.2. State Machine 1203 A state machine for DPLPMTUD is depicted in Figure 5. If multipath 1204 or multihoming is supported, a state machine is needed for each path. 1206 Note: Not all changes are shown to simplify the diagram. 1208 | | 1209 | Start | PL indicates loss 1210 | | of connectivity 1211 v v 1212 +---------------+ +---------------+ 1213 | DISABLED | | ERROR | 1214 +---------------+ PROBE_TIMER expiry: +---------------+ 1215 | PL indicates PROBE_COUNT = MAX_PROBES or ^ | 1216 | connectivity PTB: PL_PTB_SIZE < BASE_PLPMTU | | 1217 +--------------------+ +---------------+ | 1218 | | | 1219 v | BASE_PLPMTU Probe | 1220 +---------------+ acked | 1221 | BASE |--------------------->+ 1222 +---------------+ | 1223 ^ | ^ ^ | 1224 Black hole detected | | | | Black hole detected | 1225 +--------------------+ | | +--------------------+ | 1226 | +----+ | | 1227 | PROBE_TIMER expiry: | | 1228 | PROBE_COUNT < MAX_PROBES | | 1229 | | | 1230 | PMTU_RAISE_TIMER expiry | | 1231 | +-----------------------------------------+ | | 1232 | | | | | 1233 | | v | v 1234 +---------------+ +---------------+ 1235 |SEARCH_COMPLETE| | SEARCHING | 1236 +---------------+ +---------------+ 1237 | ^ ^ | | ^ 1238 | | | | | | 1239 | | +-----------------------------------------+ | | 1240 | | MAX_PLPMTU Probe acked or | | 1241 | | PROBE_TIMER expiry: PROBE_COUNT = MAX_PROBES or | | 1242 +----+ PTB: PL_PTB_SIZE = PLPMTU +----+ 1243 CONFIRMATION_TIMER expiry: PROBE_TIMER expiry: 1244 PROBE_COUNT < MAX_PROBES or PROBE_COUNT < MAX_PROBES or 1245 PLPMTU Probe acked Probe acked or PTB: 1246 PLPMTU < PL_PTB_SIZE < PROBED_SIZE 1248 Figure 5: State machine for Datagram PLPMTUD 1250 The following states are defined: 1252 DISABLED: The DISABLED state is the initial state before probing has 1253 started. It is also entered from any other state, when the PL 1254 indicates loss of connectivity. This state is left once the PL 1255 indicates connectivity to the remote PL. When transitioning to 1256 the BASE state, a probe packet of size BASE_PLPMTU can be sent 1257 immediately. 1259 BASE: The BASE state is used to confirm that the BASE_PLPMTU size is 1260 supported by the network path and is designed to allow an 1261 application to continue working when there are transient 1262 reductions in the actual PMTU. It also seeks to avoid long 1263 periods when a sender searching for a larger PLPMTU is unaware 1264 that packets are not being delivered due to a packet or ICMP Black 1265 Hole. 1267 On entry, the PROBED_SIZE is set to the BASE_PLPMTU size and the 1268 PROBE_COUNT is set to zero. 1270 Each time a probe packet is sent, the PROBE_TIMER is started. The 1271 state is exited when the probe packet is acknowledged, and the PL 1272 sender enters the SEARCHING state. 1274 The state is also left when the PROBE_COUNT reaches MAX_PROBES or 1275 a received PTB message is validated. This causes the PL sender to 1276 enter the ERROR state. 1278 SEARCHING: The SEARCHING state is the main probing state. This 1279 state is entered when probing for the BASE_PLPMTU completes. 1281 Each time a probe packet is acknowledged, the PROBE_COUNT is set 1282 to zero, the PLPMTU is set to the PROBED_SIZE and then the 1283 PROBED_SIZE is increased using the search algorithm (as described 1284 in Section 5.3. 1286 When a probe packet is sent and not acknowledged within the period 1287 of the PROBE_TIMER, the PROBE_COUNT is incremented and a new probe 1288 packet is transmitted. 1290 The state is exited to enter SEARCH_COMPLETE when the PROBE_COUNT 1291 reaches MAX_PROBES, a validated PTB is received that corresponds 1292 to the last successfully probed size (PL_PTB_SIZE = PLPMTU), or a 1293 probe of size MAX_PLPMTU is acknowledged (PLPMTU = MAX_PLPMTU). 1295 When a black hole is detected in the SEARCHING state, this causes 1296 the PL sender to enter the BASE state. 1298 SEARCH_COMPLETE: The SEARCH_COMPLETE state indicates that a search 1299 has completed. This is the normal maintenance state, where the PL 1300 is not probing to update the PLPMTU. DPLPMTUD remains in this 1301 state until either the PMTU_RAISE_TIMER expires or a black hole is 1302 detected. 1304 When DPLPMTUD uses an unacknowledged PL and is in the 1305 SEARCH_COMPLETE state, a CONFIRMATION_TIMER periodically resets 1306 the PROBE_COUNT and schedules a probe packet with the size of the 1307 PLPMTU. If MAX_PROBES successive PLPMTUD sized probes fail to be 1308 acknowledged the method enters the BASE state. When used with an 1309 acknowledged PL (e.g., SCTP), DPLPMTUD SHOULD NOT continue to 1310 generate PLPMTU probes in this state. 1312 ERROR: The ERROR state represents the case where either the network 1313 path is not known to support a PLPMTU of at least the BASE_PLPMTU 1314 size or when there is contradictory information about the network 1315 path that would otherwise result in excessive variation in the MPS 1316 signaled to the higher layer. The state implements a method to 1317 mitigate oscillation in the state-event engine. It signals a 1318 conservative value of the MPS to the higher layer by the PL. The 1319 state is exited when packet probes no longer detect the error. 1320 The PL sender then enters the SEARCHING state. 1322 Implementations are permitted to enable endpoint fragmentation if 1323 the DPLPMTUD is unable to validate MIN_PLPMTU within PROBE_COUNT 1324 probes. If DPLPMTUD is unable to validate MIN_PLPMTU the 1325 implementation will transition to the DISABLED state. 1327 Note: MIN_PLPMTU could be identical to BASE_PLPMTU, simplifying 1328 the actions in this state. 1330 5.3. Search to Increase the PLPMTU 1332 This section describes the algorithms used by DPLPMTUD to search for 1333 a larger PLPMTU. 1335 5.3.1. Probing for a larger PLPMTU 1337 Implementations use a search algorithm across the search range to 1338 determine whether a larger PLPMTU can be supported across a network 1339 path. 1341 The method discovers the search range by confirming the minimum 1342 PLPMTU and then using the probe method to select a PROBED_SIZE less 1343 than or equal to MAX_PLPMTU. MAX_PLPMTU is the minimum of the local 1344 MTU and EMTU_R (when this is learned from the remote endpoint). The 1345 MAX_PLPMTU MAY be reduced by an application that sets a maximum to 1346 the size of datagrams it will send. 1348 The PROBE_COUNT is initialized to zero when the first probe with a 1349 size greater than or equal to PLPMTUD is sent. Each probe packet 1350 successfully sent to the remote peer is confirmed by acknowledgment 1351 at the PL, see Section 4.1. 1353 Each time a probe packet is sent to the destination, the PROBE_TIMER 1354 is started. The timer is canceled when the PL receives 1355 acknowledgment that the probe packet has been successfully sent 1356 across the path Section 4.1. This confirms that the PROBED_SIZE is 1357 supported, and the PROBED_SIZE value is then assigned to the PLPMTU. 1358 The search algorithm can continue to send subsequent probe packets of 1359 an increasing size. 1361 If the timer expires before a probe packet is acknowledged, the probe 1362 has failed to confirm the PROBED_SIZE. Each time the PROBE_TIMER 1363 expires, the PROBE_COUNT is incremented, the PROBE_TIMER is 1364 reinitialized, and a new probe of the same size or any other size 1365 (determined by the search algorithm) can be sent. The maximum number 1366 of consecutive failed probes is configured (MAX_PROBES). If the 1367 value of the PROBE_COUNT reaches MAX_PROBES, probing will stop, and 1368 the PL sender enters the SEARCH_COMPLETE state. 1370 5.3.2. Selection of Probe Sizes 1372 The search algorithm determines a minimum useful gain in PLPMTU. It 1373 would not be constructive for a PL sender to attempt to probe for all 1374 sizes. This would incur unnecessary load on the path. 1375 Implementations SHOULD select the set of probe packet sizes to 1376 maximize the gain in PLPMTU from each search step. 1378 Implementations could optimize the search procedure by selecting step 1379 sizes from a table of common PMTU sizes. When selecting the 1380 appropriate next size to search, an implementer ought to also 1381 consider that there can be common sizes of MPS that applications seek 1382 to use, and their could be common sizes of MTU used within the 1383 network. 1385 5.3.3. Resilience to Inconsistent Path Information 1387 A decision to increase the PLPMTU needs to be resilient to the 1388 possibility that information learned about the network path is 1389 inconsistent. A path is inconsistent when, for example, probe 1390 packets are lost due to other reasons (i.e., not packet size) or due 1391 to frequent path changes. Frequent path changes could occur by 1392 unexpected "flapping" - where some packets from a flow pass along one 1393 path, but other packets follow a different path with different 1394 properties. 1396 A PL sender is able to detect inconsistency from the sequence of 1397 PLPMTU probes that are acknowledged or the sequence of PTB messages 1398 that it receives. When inconsistent path information is detected, a 1399 PL sender could use an alternate search mode that clamps the offered 1400 MPS to a smaller value for a period of time. This avoids unnecessary 1401 loss of packets. 1403 5.4. Robustness to Inconsistent Paths 1405 Some paths could be unable to sustain packets of the BASE_PLPMTU 1406 size. The Error State could be implemented to provide rubustness to 1407 such paths. This allows fallback to a smaller than desired PLPMTU, 1408 rather than suffer connectivity failure. This could utilize methods 1409 such as endpoint IP fragmentation to enable the PL sender to 1410 communicate using packets smaller than the BASE_PLPMTU. 1412 6. Specification of Protocol-Specific Methods 1414 DPLPMTUD requires protocol-specific details to be specified for each 1415 PL that is used. 1417 The first subsection provides guidance on how to implement the 1418 DPLPMTUD method as a part of an application using UDP or UDP-Lite. 1419 The guidance also applies to other datagram services that do not 1420 include a specific transport protocol (such as a tunnel 1421 encapsulation). The following subsections describe how DPLPMTUD can 1422 be implemented as a part of the transport service, allowing 1423 applications using the service to benefit from discovery of the 1424 PLPMTU without themselves needing to implement this method when using 1425 SCTP and QUIC. 1427 6.1. Application support for DPLPMTUD with UDP or UDP-Lite 1429 The current specifications of UDP [RFC0768] and UDP-Lite [RFC3828] do 1430 not define a method in the RFC-series that supports PLPMTUD. In 1431 particular, the UDP transport does not provide the transport features 1432 needed to implement datagram PLPMTUD. 1434 The DPLPMTUD method can be implemented as a part of an application 1435 built directly or indirectly on UDP or UDP-Lite, but relies on 1436 higher-layer protocol features to implement the method [BCP145]. 1438 Some primitives used by DPLPMTUD might not be available via the 1439 Datagram API (e.g., the ability to access the PLPMTU from the IP 1440 layer cache, or interpret received PTB messages). 1442 In addition, it is recommended that PMTU discovery is not performed 1443 by multiple protocol layers. An application SHOULD avoid using 1444 DPLPMTUD when the underlying transport system provides this 1445 capability. A common method for managing the PLPMTU has benefits, 1446 both in the ability to share state between different processes and 1447 opportunities to coordinate probing for different PL instances. 1449 6.1.1. Application Request 1451 An application needs an application-layer protocol mechanism (such as 1452 a message acknowledgment method) that solicits a response from a 1453 destination endpoint. The method SHOULD allow the sender to check 1454 the value returned in the response to provide additional protection 1455 from off-path insertion of data [BCP145]. Suitable methods include a 1456 parameter known only to the two endpoints, such as a session ID or 1457 initialized sequence number. 1459 6.1.2. Application Response 1461 An application needs an application-layer protocol mechanism to 1462 communicate the response from the destination endpoint. This 1463 response could indicate successful reception of the probe across the 1464 path, but could also indicate that some (or all packets) have failed 1465 to reach the destination. 1467 6.1.3. Sending Application Probe Packets 1469 A probe packet can carry an application data block, but the 1470 successful transmission of this data is at risk when used for 1471 probing. Some applications might prefer to use a probe packet that 1472 does not carry an application data block to avoid disruption to data 1473 transfer. 1475 6.1.4. Initial Connectivity 1477 An application that does not have other higher-layer information 1478 confirming connectivity with the remote peer SHOULD implement a 1479 connectivity mechanism using acknowledged probe packets before 1480 entering the BASE state. 1482 6.1.5. Validating the Path 1484 An application that does not have other higher-layer information 1485 confirming correct delivery of datagrams SHOULD implement the 1486 CONFIRMATION_TIMER to periodically send probe packets while in the 1487 SEARCH_COMPLETE state. 1489 6.1.6. Handling of PTB Messages 1491 An application that is able and wishes to receive PTB messages MUST 1492 perform ICMP validation as specified in Section 5.2 of [BCP145]. 1493 This requires that the application checks each received PTB message 1494 to validate that it was is received in response to transmitted 1495 traffic and that the reported PL_PTB_SIZE is less than the current 1496 probed size (see Section 4.6.2). A validated PTB message MAY be used 1497 as input to the DPLPMTUD algorithm, but MUST NOT be used directly to 1498 set the PLPMTU. 1500 6.2. DPLPMTUD for SCTP 1502 Section 10.2 of [RFC4821] specified a recommended PLPMTUD probing 1503 method for SCTP and Section 7.3 of [RFC4960] recommended an endpoint 1504 apply the techniques in RFC4821 on a per-destination-address basis. 1505 The specification for DPLPMTUD continues the practice of using the PL 1506 to discover the PMTU, but updates, RFC4960 with a recommendation to 1507 use the method specified in this document: The RECOMMENDED method for 1508 generating probes is to add a chunk consisting only of padding to an 1509 SCTP message. The PAD chunk defined in [RFC4820] SHOULD be attached 1510 to a minimum length HEARTBEAT (HB) chunk to build a probe packet. 1511 This enables probing without affecting the transfer of user messages 1512 and without being limited by congestion control or flow control. 1513 This is preferred to using DATA chunks (with padding as required) as 1514 path probes. 1516 Section 6.9 of [RFC4960] describes dividing the user messages into 1517 data chunks sent by the PL when using SCTP. This notes that once an 1518 SCTP message has been sent, it cannot be re-segmented. [RFC4960] 1519 describes the method to retransmit data chunks when the MPS has 1520 reduced, and the use of IP fragmentation for this case. This is 1521 unchanged by this document. 1523 6.2.1. SCTP/IPv4 and SCTP/IPv6 1525 6.2.1.1. Initial Connectivity 1527 The base protocol is specified in [RFC4960]. This provides an 1528 acknowledged PL. A sender can therefore enter the BASE state as soon 1529 as connectivity has been confirmed. 1531 6.2.1.2. Sending SCTP Probe Packets 1533 Probe packets consist of an SCTP common header followed by a 1534 HEARTBEAT chunk and a PAD chunk. The PAD chunk is used to control 1535 the length of the probe packet. The HEARTBEAT chunk is used to 1536 trigger the sending of a HEARTBEAT ACK chunk. The reception of the 1537 HEARTBEAT ACK chunk acknowledges reception of a successful probe. A 1538 successful probe updates the association and path counters, but an 1539 unsuccessful probe is discounted (assumed to be a result of choosing 1540 too large a PLPMTU). 1542 The SCTP sender needs to be able to determine the total size of a 1543 probe packet. The HEARTBEAT chunk could carry a Heartbeat 1544 Information parameter that includes, besides the information 1545 suggested in [RFC4960], the probe size to help an implementation 1546 associate a HEARTBEAT-ACK with the size of probe that was sent. The 1547 sender could also use other methods, such as sending a nonce and 1548 verifying the information returned also contains the corresponding 1549 nonce. The length of the PAD chunk is computed by reducing the 1550 probing size by the size of the SCTP common header and the HEARTBEAT 1551 chunk. The payload of the PAD chunk contains arbitrary data. When 1552 transmitted at the IP layer, the PMTU size also includes the IPv4 or 1553 IPv6 header(s). 1555 Probing can start directly after the PL handshake, this can be done 1556 before data is sent. Assuming this behavior (i.e., the PMTU is 1557 smaller than or equal to the interface MTU), this process will take 1558 several round trip time periods, dependent on the number of DPLPMTUD 1559 probes sent. The Heartbeat timer can be used to implement the 1560 PROBE_TIMER. 1562 6.2.1.3. Validating the Path with SCTP 1564 Since SCTP provides an acknowledged PL, a sender MUST NOT implement 1565 the CONFIRMATION_TIMER while in the SEARCH_COMPLETE state. 1567 6.2.1.4. PTB Message Handling by SCTP 1569 Normal ICMP validation MUST be performed as specified in Appendix C 1570 of [RFC4960]. This requires that the first 8 bytes of the SCTP 1571 common header are quoted in the payload of the PTB message, which can 1572 be the case for ICMPv4 and is normally the case for ICMPv6. 1574 When a PTB message has been validated, the PL_PTB_SIZE calculated 1575 from the PTB_SIZE reported in the PTB message SHOULD be used with the 1576 DPLPMTUD algorithm, providing that the reported PL_PTB_SIZE is less 1577 than the current probe size (see Section 4.6). 1579 6.2.2. DPLPMTUD for SCTP/UDP 1581 The UDP encapsulation of SCTP is specified in [RFC6951]. 1583 This specification updates the reference to RFC 4821 in section 5.6 1584 of RFC 6951 to refer to XXXTHISRFCXXX. RFC 6951 is updated by 1585 addition of the following sentence at the end of section 5.6: "The 1586 RECOMMENDED method for determining the MTU of the path is specified 1587 in XXXTHISRFCXXX". 1589 XXX RFC EDITOR - please replace XXXTHISRFCXXX when published XXX 1591 6.2.2.1. Initial Connectivity 1593 A sender can enter the BASE state as soon as SCTP connectivity has 1594 been confirmed. 1596 6.2.2.2. Sending SCTP/UDP Probe Packets 1598 Packet probing can be performed as specified in Section 6.2.1.2. The 1599 size of the probe packet includes the 8 bytes of UDP Header. This 1600 has to be considered when filling the probe packet with the PAD 1601 chunk. 1603 6.2.2.3. Validating the Path with SCTP/UDP 1605 SCTP provides an acknowledged PL, therefore a sender does not 1606 implement the CONFIRMATION_TIMER while in the SEARCH_COMPLETE state. 1608 6.2.2.4. Handling of PTB Messages by SCTP/UDP 1610 ICMP validation MUST be performed for PTB messages as specified in 1611 Appendix C of [RFC4960]. This requires that the first 8 bytes of the 1612 SCTP common header are contained in the PTB message, which can be the 1613 case for ICMPv4 (but note the UDP header also consumes a part of the 1614 quoted packet header) and is normally the case for ICMPv6. When the 1615 validation is completed, the PL_PTB_SIZE calculated from the PTB_SIZE 1616 in the PTB message SHOULD be used with the DPLPMTUD providing that 1617 the reported PL_PTB_SIZE is less than the current probe size. 1619 6.2.3. DPLPMTUD for SCTP/DTLS 1621 The Datagram Transport Layer Security (DTLS) encapsulation of SCTP is 1622 specified in [RFC8261]. This is used for data channels in WebRTC 1623 implementations. This specification updates the reference to RFC 1624 4821 in section 5 of RFC 8261 to refer to XXXTHISRFCXXX. 1626 XXX RFC EDITOR - please replace XXXTHISRFCXXX when published XXX 1628 6.2.3.1. Initial Connectivity 1630 A sender can enter the BASE state as soon as SCTP connectivity has 1631 been confirmed. 1633 6.2.3.2. Sending SCTP/DTLS Probe Packets 1635 Packet probing can be done, as specified in Section 6.2.1.2. The 1636 maximum payload is reduced by the size of the DTLS headers, which has 1637 to be considered when filling the PAD chunk. The size of the probe 1638 packet includes the DTLS PL headers. This has to be considered when 1639 filling the probe packet with the PAD chunk. 1641 6.2.3.3. Validating the Path with SCTP/DTLS 1643 Since SCTP provides an acknowledged PL, a sender MUST NOT implement 1644 the CONFIRMATION_TIMER while in the SEARCH_COMPLETE state. 1646 6.2.3.4. Handling of PTB Messages by SCTP/DTLS 1648 [RFC4960] does not specify a way to validate SCTP/DTLS ICMP message 1649 payload and neither does this document. This can prevent processing 1650 of PTB messages at the PL. 1652 6.3. DPLPMTUD for QUIC 1654 QUIC [I-D.ietf-quic-transport] is a UDP-based PL that provides 1655 reception feedback. The UDP payload includes a QUIC packet header, a 1656 protected payload, and any authentication fields. It supports 1657 padding and packet coalescence that can be used to construct probe 1658 packets. From the perspective of DPLPMTUD, QUIC can function as an 1659 acknowledged PL. [I-D.ietf-quic-transport] describes the method for 1660 using DPLPMTUD with QUIC packets. 1662 7. Acknowledgments 1664 This work was partially funded by the European Union's Horizon 2020 1665 research and innovation programme under grant agreement No. 644334 1666 (NEAT). The views expressed are solely those of the author(s). 1668 Thanks to all that have commented or contributed, the TSVWG and QUIC 1669 working groups, and Mathew Calder and Julius Flohr for providing 1670 early implementations. 1672 8. IANA Considerations 1674 This memo includes no request to IANA. 1676 If there are no requirements for IANA, the section will be removed 1677 during conversion into an RFC by the RFC Editor. 1679 9. Security Considerations 1681 The security considerations for the use of UDP and SCTP are provided 1682 in the referenced RFCs. 1684 To avoid excessive load, the interval between individual probe 1685 packets MUST be at least one RTT, and the interval between rounds of 1686 probing is determined by the PMTU_RAISE_TIMER. 1688 A PL sender needs to ensure that the method used to confirm reception 1689 of probe packets protects from off-path attackers injecting packets 1690 into the path. This protection is provided in IETF-defined protocols 1691 (e.g., TCP, SCTP) using a randomly-initialized sequence number. A 1692 description of one way to do this when using UDP is provided in 1693 section 5.1 of [BCP145]). 1695 There are cases where ICMP Packet Too Big (PTB) messages are not 1696 delivered due to policy, configuration or equipment design (see 1697 Section 1.1). This method therefore does not rely upon PTB messages 1698 being received, but is able to utilize these when they are received 1699 by the sender. PTB messages could potentially be used to cause a 1700 node to inappropriately reduce the PLPMTU. A node supporting 1701 DPLPMTUD MUST therefore appropriately validate the payload of PTB 1702 messages to ensure these are received in response to transmitted 1703 traffic (i.e., a reported error condition that corresponds to a 1704 datagram actually sent by the path layer, see Section 4.6.1). 1706 An on-path attacker able to create a PTB message could forge PTB 1707 messages that include a valid quoted IP packet. Such an attack could 1708 be used to drive down the PLPMTU. An on-path device could similarly 1709 force a reduction of the PLPMTU by implementing a policy that drops 1710 packets larger than a configured size. There are two ways this 1711 method can be mitigated against such attacks: First, by ensuring that 1712 a PL sender never reduces the PLPMTU below the base size, solely in 1713 response to receiving a PTB message. This is achieved by first 1714 entering the BASE state when such a message is received. Second, the 1715 design does not require processing of PTB messages, a PL sender could 1716 therefore suspend processing of PTB messages (e.g., in a robustness 1717 mode after detecting that subsequent probes actually confirm that a 1718 size larger than the PTB_SIZE is supported by a path). 1720 Parsing the quoted packet inside a PTB message can introduce addional 1721 per-packet processing at the PL sender. This processing SHOULD be 1722 limited to avoid a denial of service attack when arbitrary headers 1723 are included. Rate-limiting the processing could result in PTB 1724 messages not being received by a PL, however the DPLPMTUD method is 1725 robust to such loss. 1727 The successful processing of an ICMP message can trigger a probe when 1728 the reported PTB size is valid, but this does not directly update the 1729 PLPMTU for the path. This prevents a message attempting to black 1730 hole data by indicating a size larger than supported by the path. 1732 It is possible that the information about a path is not stable. This 1733 could be a result of forwarding across more than one path that has a 1734 different actual PMTU or a single path presents a varying PMTU. The 1735 design of a PLPMTUD implementation SHOULD consider how to mitigate 1736 the effects of varying path information. One possible mitigation is 1737 to provide robustness (see Section 5.4) in the method that avoids 1738 oscillation in the MPS. 1740 DPLPMTUD methods can introduce padding data to inflate the length of 1741 the datagram to the total size required for a probe packet. The 1742 total size of a probe packet includes all headers and padding added 1743 to the payload data being sent (e.g., including security-related 1744 fields such as an AEAD tag and TLS record layer padding). The value 1745 of the padding data does not influence the DPLPMTUD search algorithm, 1746 and therefore needs to be set consistent with the policy of the PL. 1748 If a PL can make use of cryptographic confidentiality or data- 1749 integrity mechanisms, then the design ought to avoid adding anything 1750 (e.g., padding) to DPLPMTUD probe packets that is not also protected 1751 by those cryptographic mechanisms. 1753 10. References 1755 10.1. Normative References 1757 [BCP145] Eggert, L., Fairhurst, G., and G. Shepherd, "UDP Usage 1758 Guidelines", BCP 145, RFC 8085, March 2017. 1760 1762 [RFC0768] Postel, J., "User Datagram Protocol", STD 6, RFC 768, 1763 DOI 10.17487/RFC0768, August 1980, 1764 . 1766 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, 1767 DOI 10.17487/RFC0791, September 1981, 1768 . 1770 [RFC1191] Mogul, J.C. and S.E. Deering, "Path MTU discovery", 1771 RFC 1191, DOI 10.17487/RFC1191, November 1990, 1772 . 1774 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1775 Requirement Levels", BCP 14, RFC 2119, 1776 DOI 10.17487/RFC2119, March 1997, 1777 . 1779 [RFC3828] Larzon, L-A., Degermark, M., Pink, S., Jonsson, L-E., Ed., 1780 and G. Fairhurst, Ed., "The Lightweight User Datagram 1781 Protocol (UDP-Lite)", RFC 3828, DOI 10.17487/RFC3828, July 1782 2004, . 1784 [RFC4820] Tuexen, M., Stewart, R., and P. Lei, "Padding Chunk and 1785 Parameter for the Stream Control Transmission Protocol 1786 (SCTP)", RFC 4820, DOI 10.17487/RFC4820, March 2007, 1787 . 1789 [RFC4960] Stewart, R., Ed., "Stream Control Transmission Protocol", 1790 RFC 4960, DOI 10.17487/RFC4960, September 2007, 1791 . 1793 [RFC6951] Tuexen, M. and R. Stewart, "UDP Encapsulation of Stream 1794 Control Transmission Protocol (SCTP) Packets for End-Host 1795 to End-Host Communication", RFC 6951, 1796 DOI 10.17487/RFC6951, May 2013, 1797 . 1799 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 1800 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 1801 May 2017, . 1803 [RFC8200] Deering, S. and R. Hinden, "Internet Protocol, Version 6 1804 (IPv6) Specification", STD 86, RFC 8200, 1805 DOI 10.17487/RFC8200, July 2017, 1806 . 1808 [RFC8201] McCann, J., Deering, S., Mogul, J., and R. Hinden, Ed., 1809 "Path MTU Discovery for IP version 6", STD 87, RFC 8201, 1810 DOI 10.17487/RFC8201, July 2017, 1811 . 1813 [RFC8261] Tuexen, M., Stewart, R., Jesup, R., and S. Loreto, 1814 "Datagram Transport Layer Security (DTLS) Encapsulation of 1815 SCTP Packets", RFC 8261, DOI 10.17487/RFC8261, November 1816 2017, . 1818 10.2. Informative References 1820 [I-D.ietf-intarea-frag-fragile] 1821 Bonica, R., Baker, F., Huston, G., Hinden, R., Troan, O., 1822 and F. Gont, "IP Fragmentation Considered Fragile", Work 1823 in Progress, Internet-Draft, draft-ietf-intarea-frag- 1824 fragile-17, 30 September 2019, . 1827 [I-D.ietf-intarea-tunnels] 1828 Touch, J. and M. Townsley, "IP Tunnels in the Internet 1829 Architecture", Work in Progress, Internet-Draft, draft- 1830 ietf-intarea-tunnels-10, 12 September 2019, 1831 . 1834 [I-D.ietf-quic-transport] 1835 Iyengar, J. and M. Thomson, "QUIC: A UDP-Based Multiplexed 1836 and Secure Transport", Work in Progress, Internet-Draft, 1837 draft-ietf-quic-transport-27, 21 February 2020, 1838 . 1841 [RFC0792] Postel, J., "Internet Control Message Protocol", STD 5, 1842 RFC 792, DOI 10.17487/RFC0792, September 1981, 1843 . 1845 [RFC1122] Braden, R., Ed., "Requirements for Internet Hosts - 1846 Communication Layers", STD 3, RFC 1122, 1847 DOI 10.17487/RFC1122, October 1989, 1848 . 1850 [RFC1812] Baker, F., Ed., "Requirements for IP Version 4 Routers", 1851 RFC 1812, DOI 10.17487/RFC1812, June 1995, 1852 . 1854 [RFC2923] Lahey, K., "TCP Problems with Path MTU Discovery", 1855 RFC 2923, DOI 10.17487/RFC2923, September 2000, 1856 . 1858 [RFC4340] Kohler, E., Handley, M., and S. Floyd, "Datagram 1859 Congestion Control Protocol (DCCP)", RFC 4340, 1860 DOI 10.17487/RFC4340, March 2006, 1861 . 1863 [RFC4443] Conta, A., Deering, S., and M. Gupta, Ed., "Internet 1864 Control Message Protocol (ICMPv6) for the Internet 1865 Protocol Version 6 (IPv6) Specification", STD 89, 1866 RFC 4443, DOI 10.17487/RFC4443, March 2006, 1867 . 1869 [RFC4821] Mathis, M. and J. Heffner, "Packetization Layer Path MTU 1870 Discovery", RFC 4821, DOI 10.17487/RFC4821, March 2007, 1871 . 1873 [RFC4890] Davies, E. and J. Mohacsi, "Recommendations for Filtering 1874 ICMPv6 Messages in Firewalls", RFC 4890, 1875 DOI 10.17487/RFC4890, May 2007, 1876 . 1878 [RFC5508] Srisuresh, P., Ford, B., Sivakumar, S., and S. Guha, "NAT 1879 Behavioral Requirements for ICMP", BCP 148, RFC 5508, 1880 DOI 10.17487/RFC5508, April 2009, 1881 . 1883 Appendix A. Revision Notes 1885 Note to RFC-Editor: please remove this entire section prior to 1886 publication. 1888 Individual draft -00: 1890 * Comments and corrections are welcome directly to the authors or 1891 via the IETF TSVWG working group mailing list. 1893 * This update is proposed for WG comments. 1895 Individual draft -01: 1897 * Contains the first representation of the algorithm, showing the 1898 states and timers 1900 * This update is proposed for WG comments. 1902 Individual draft -02: 1904 * Contains updated representation of the algorithm, and textual 1905 corrections. 1907 * The text describing when to set the effective PMTU has not yet 1908 been validated by the authors 1910 * To determine security to off-path-attacks: We need to decide 1911 whether a received PTB message SHOULD/MUST be validated? The text 1912 on how to handle a PTB message indicating a link MTU larger than 1913 the probe has yet not been validated by the authors 1915 * No text currently describes how to handle inconsistent results 1916 from arbitrary re-routing along different parallel paths 1918 * This update is proposed for WG comments. 1920 Working Group draft -00: 1922 * This draft follows a successful adoption call for TSVWG 1924 * There is still work to complete, please comment on this draft. 1926 Working Group draft -01: 1928 * This draft includes improved introduction. 1930 * The draft is updated to require ICMP validation prior to accepting 1931 PTB messages - this to be confirmed by WG 1933 * Section added to discuss Selection of Probe Size - methods to be 1934 evaluated and recommendations to be considered 1936 * Section added to align with work proposed in the QUIC WG. 1938 Working Group draft -02: 1940 * The draft was updated based on feedback from the WG, and a 1941 detailed review by Magnus Westerlund. 1943 * The document updates RFC 4821. 1945 * Requirements list updated. 1947 * Added more explicit discussion of a simpler black-hole detection 1948 mode. 1950 * This draft includes reorganisation of the section on IETF 1951 protocols. 1953 * Added more discussion of implementation within an application. 1955 * Added text on flapping paths. 1957 * Replaced 'effective MTU' with new term PLPMTU. 1959 Working Group draft -03: 1961 * Updated figures 1963 * Added more discussion on blackhole detection 1965 * Added figure describing just blackhole detection 1966 * Added figure relating MPS sizes 1968 Working Group draft -04: 1970 * Described phases and named these consistently. 1972 * Corrected transition from confirmation directly to the search 1973 phase (Base has been checked). 1975 * Redrawn state diagrams. 1977 * Renamed BASE_MTU to BASE_PMTU (because it is a base for the PMTU). 1979 * Clarified Error state. 1981 * Clarified suspending DPLPMTUD. 1983 * Verified normative text in requirements section. 1985 * Removed duplicate text. 1987 * Changed all text to refer to /packet probe/probe packet/ 1988 /validation/verification/ added term /Probe Confirmation/ and 1989 clarified BlackHole detection. 1991 Working Group draft -05: 1993 * Updated security considerations. 1995 * Feedback after speaking with Joe Touch helped improve UDP-Options 1996 description. 1998 Working Group draft -06: 2000 * Updated description of ICMP issues in section 1.1 2002 * Update to description of QUIC. 2004 Working group draft -07: 2006 * Moved description of the PTB processing method from the PTB 2007 requirements section. 2009 * Clarified what is performed in the PTB validation check. 2011 * Updated security consideration to explain PTB security without 2012 needing to read the rest of the document. 2014 * Reformatted state machine diagram 2016 Working group draft -08: 2018 * Moved to rfcxml v3+ 2020 * Rendered diagrams to svg in html version. 2022 * Removed Appendix A. Event-driven state changes. 2024 * Removed section on DPLPMTUD with UDP Options. 2026 * Shortened the description of phases. 2028 Working group draft -09: 2030 * Remove final mention of UDP Options 2032 * Add Initial Connectivity sections to each PL 2034 * Add to disable outgoing pmtu enforcement of packets 2036 Working group draft -10: 2038 * Address comments from Lars Eggert 2040 * Reinforce that PROBE_COUNT is successive attempts to probe for any 2041 size 2043 * Redefine MAX_PROBES to 3 2045 * Address PTB_SIZE of 0 or less that MIN_PLPMTU 2047 Working group draft -11: 2049 * Restore a sentence removed in previous rev 2051 * De-acronymise QUIC 2053 * Address some nits 2055 Working group draft -12: 2057 * Add TSVWG, QUIC and implementers to acknowledgments 2059 * Shorten a diagram line. 2061 * Address nits from Julius and Wes. 2063 * Be clearer when talking about IP layer caches 2065 Working group draft -13, -14: 2067 * Updated after WGLC. 2069 Working group draft -15: 2071 * Updated after AD evaluation and prepared for IETF-LC. 2073 Working group draft -16: 2075 * Updated text after SECDIR review. 2077 Working group draft -17: 2079 * Updated text after GENART and IETF-LC. 2081 * Renamed BASE_MTU to BASE_PLPMTU, and MIN and MAX PMTU to PLPMTU 2082 (because these are about a base for the PLPMTU), and ensured 2083 consistent separation of PMTU and PLPMTU. 2085 * Adopted US-style English throughout. 2087 Working group draft -18: 2089 * Updated text and address nits from OPSDIR, ART and IESG reviews. 2091 * Order PTB processing based on PL_PTB_SIZE 2093 Working group draft -19: 2095 * Updated text and address nits based on comments from Tim Chown and 2096 Murray S. Kucherawy. 2098 Working group draft -20: 2100 * Address nits and comments from IESG 2102 * Refer to BCP 145 rather than RFC 8085 in most places. 2104 * Update probing method text for SCTP and QUIC. 2106 Working group draft -21: 2108 * Update QUIC text for skipping into BASE state. 2110 Working group draft -22: 2112 * Add a section reference to MPS 2114 * Clarify MIN_PLPMTU text 2116 * Remove most QUIC text 2118 * Make QUIC reference informative. 2120 Authors' Addresses 2122 Godred Fairhurst 2123 University of Aberdeen 2124 School of Engineering 2125 Fraser Noble Building 2126 Aberdeen 2127 AB24 3UE 2128 United Kingdom 2130 Email: gorry@erg.abdn.ac.uk 2132 Tom Jones 2133 University of Aberdeen 2134 School of Engineering 2135 Fraser Noble Building 2136 Aberdeen 2137 AB24 3UE 2138 United Kingdom 2140 Email: tom@erg.abdn.ac.uk 2142 Michael Tuexen 2143 Muenster University of Applied Sciences 2144 Stegerwaldstrasse 39 2145 48565 Steinfurt 2146 Germany 2148 Email: tuexen@fh-muenster.de 2150 Irene Ruengeler 2151 Muenster University of Applied Sciences 2152 Stegerwaldstrasse 39 2153 48565 Steinfurt 2154 Germany 2156 Email: i.ruengeler@fh-muenster.de 2157 Timo Voelker 2158 Muenster University of Applied Sciences 2159 Stegerwaldstrasse 39 2160 48565 Steinfurt 2161 Germany 2163 Email: timo.voelker@fh-muenster.de