idnits 2.17.1 draft-ietf-taps-transports-07.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 : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (October 07, 2015) is 3123 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Unused Reference: 'RFC6455' is defined on line 2286, but no explicit reference was found in the text -- Obsolete informational reference (is this intentional?): RFC 793 (Obsoleted by RFC 9293) -- Obsolete informational reference (is this intentional?): RFC 896 (Obsoleted by RFC 7805) -- Obsolete informational reference (is this intentional?): RFC 1716 (Obsoleted by RFC 1812) -- Obsolete informational reference (is this intentional?): RFC 1981 (Obsoleted by RFC 8201) -- Obsolete informational reference (is this intentional?): RFC 2460 (Obsoleted by RFC 8200) -- Obsolete informational reference (is this intentional?): RFC 2461 (Obsoleted by RFC 4861) -- Obsolete informational reference (is this intentional?): RFC 2617 (Obsoleted by RFC 7235, RFC 7615, RFC 7616, RFC 7617) -- Obsolete informational reference (is this intentional?): RFC 3205 (Obsoleted by RFC 9205) -- Obsolete informational reference (is this intentional?): RFC 3450 (Obsoleted by RFC 5775) -- Obsolete informational reference (is this intentional?): RFC 3452 (Obsoleted by RFC 5052, RFC 5445) -- Obsolete informational reference (is this intentional?): RFC 3926 (Obsoleted by RFC 6726) -- Obsolete informational reference (is this intentional?): RFC 4614 (Obsoleted by RFC 7414) -- Obsolete informational reference (is this intentional?): RFC 4960 (Obsoleted by RFC 9260) -- Obsolete informational reference (is this intentional?): RFC 5246 (Obsoleted by RFC 8446) -- Obsolete informational reference (is this intentional?): RFC 5672 (Obsoleted by RFC 6376) -- Obsolete informational reference (is this intentional?): RFC 6093 (Obsoleted by RFC 9293) -- Obsolete informational reference (is this intentional?): RFC 6347 (Obsoleted by RFC 9147) -- Obsolete informational reference (is this intentional?): RFC 6824 (Obsoleted by RFC 8684) -- Obsolete informational reference (is this intentional?): RFC 7053 (Obsoleted by RFC 9260) -- Obsolete informational reference (is this intentional?): RFC 7230 (Obsoleted by RFC 9110, RFC 9112) -- Obsolete informational reference (is this intentional?): RFC 7231 (Obsoleted by RFC 9110) -- Obsolete informational reference (is this intentional?): RFC 7232 (Obsoleted by RFC 9110) -- Obsolete informational reference (is this intentional?): RFC 7233 (Obsoleted by RFC 9110) -- Obsolete informational reference (is this intentional?): RFC 7234 (Obsoleted by RFC 9111) -- Obsolete informational reference (is this intentional?): RFC 7235 (Obsoleted by RFC 9110) -- Obsolete informational reference (is this intentional?): RFC 7525 (Obsoleted by RFC 9325) -- Obsolete informational reference (is this intentional?): RFC 7540 (Obsoleted by RFC 9113) == Outdated reference: A later version (-19) exists of draft-ietf-tsvwg-rfc5405bis-05 == Outdated reference: A later version (-08) exists of draft-ietf-aqm-ecn-benefits-06 == Outdated reference: A later version (-13) exists of draft-ietf-tsvwg-sctp-ndata-04 == Outdated reference: A later version (-23) exists of draft-ietf-tsvwg-natsupp-08 Summary: 0 errors (**), 0 flaws (~~), 6 warnings (==), 28 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group G. Fairhurst, Ed. 3 Internet-Draft University of Aberdeen 4 Intended status: Informational B. Trammell, Ed. 5 Expires: April 9, 2016 M. Kuehlewind, Ed. 6 ETH Zurich 7 October 07, 2015 9 Services provided by IETF transport protocols and congestion control 10 mechanisms 11 draft-ietf-taps-transports-07 13 Abstract 15 This document describes services provided by existing IETF protocols 16 and congestion control mechanisms. It is designed to help 17 application and network stack programmers and to inform the work of 18 the IETF TAPS Working Group. 20 Status of This Memo 22 This Internet-Draft is submitted in full conformance with the 23 provisions of BCP 78 and BCP 79. 25 Internet-Drafts are working documents of the Internet Engineering 26 Task Force (IETF). Note that other groups may also distribute 27 working documents as Internet-Drafts. The list of current Internet- 28 Drafts is at http://datatracker.ietf.org/drafts/current/. 30 Internet-Drafts are draft documents valid for a maximum of six months 31 and may be updated, replaced, or obsoleted by other documents at any 32 time. It is inappropriate to use Internet-Drafts as reference 33 material or to cite them other than as "work in progress." 35 This Internet-Draft will expire on April 9, 2016. 37 Copyright Notice 39 Copyright (c) 2015 IETF Trust and the persons identified as the 40 document authors. All rights reserved. 42 This document is subject to BCP 78 and the IETF Trust's Legal 43 Provisions Relating to IETF Documents 44 (http://trustee.ietf.org/license-info) in effect on the date of 45 publication of this document. Please review these documents 46 carefully, as they describe your rights and restrictions with respect 47 to this document. Code Components extracted from this document must 48 include Simplified BSD License text as described in Section 4.e of 49 the Trust Legal Provisions and are provided without warranty as 50 described in the Simplified BSD License. 52 Table of Contents 54 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 55 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 4 56 3. Existing Transport Protocols . . . . . . . . . . . . . . . . 5 57 3.1. Transport Control Protocol (TCP) . . . . . . . . . . . . 5 58 3.1.1. Protocol Description . . . . . . . . . . . . . . . . 5 59 3.1.2. Interface description . . . . . . . . . . . . . . . . 6 60 3.1.3. Transport Features . . . . . . . . . . . . . . . . . 7 61 3.2. Multipath TCP (MPTCP) . . . . . . . . . . . . . . . . . . 8 62 3.2.1. Protocol Description . . . . . . . . . . . . . . . . 8 63 3.2.2. Interface Description . . . . . . . . . . . . . . . . 8 64 3.2.3. Transport features . . . . . . . . . . . . . . . . . 8 65 3.3. Stream Control Transmission Protocol (SCTP) . . . . . . . 9 66 3.3.1. Protocol Description . . . . . . . . . . . . . . . . 9 67 3.3.2. Interface Description . . . . . . . . . . . . . . . . 12 68 3.3.3. Transport Features . . . . . . . . . . . . . . . . . 14 69 3.4. User Datagram Protocol (UDP) . . . . . . . . . . . . . . 15 70 3.4.1. Protocol Description . . . . . . . . . . . . . . . . 15 71 3.4.2. Interface Description . . . . . . . . . . . . . . . . 16 72 3.4.3. Transport Features . . . . . . . . . . . . . . . . . 16 73 3.5. Lightweight User Datagram Protocol (UDP-Lite) . . . . . . 17 74 3.5.1. Protocol Description . . . . . . . . . . . . . . . . 17 75 3.5.2. Interface Description . . . . . . . . . . . . . . . . 18 76 3.5.3. Transport Features . . . . . . . . . . . . . . . . . 18 77 3.6. Datagram Congestion Control Protocol (DCCP) . . . . . . . 19 78 3.6.1. Protocol Description . . . . . . . . . . . . . . . . 19 79 3.6.2. Interface Description . . . . . . . . . . . . . . . . 20 80 3.6.3. Transport Features . . . . . . . . . . . . . . . . . 21 81 3.7. Lightweight User Datagram Protocol (UDP-Lite) . . . . . . 21 82 3.7.1. Protocol Description . . . . . . . . . . . . . . . . 21 83 3.7.2. Interface Description . . . . . . . . . . . . . . . . 22 84 3.7.3. Transport Features . . . . . . . . . . . . . . . . . 22 85 3.8. Internet Control Message Protocol (ICMP) . . . . . . . . 23 86 3.8.1. Protocol Description . . . . . . . . . . . . . . . . 23 87 3.8.2. Interface Description . . . . . . . . . . . . . . . . 24 88 3.8.3. Transport Features . . . . . . . . . . . . . . . . . 24 89 3.9. Realtime Transport Protocol (RTP) . . . . . . . . . . . . 25 90 3.9.1. Protocol Description . . . . . . . . . . . . . . . . 25 91 3.9.2. Interface Description . . . . . . . . . . . . . . . . 26 92 3.9.3. Transport Features . . . . . . . . . . . . . . . . . 26 93 3.10. File Delivery over Unidirectional Transport/Asynchronous 94 Layered Coding Reliable Multicast (FLUTE/ALC) . . . . . . 26 95 3.10.1. Protocol Description . . . . . . . . . . . . . . . . 27 96 3.10.2. Interface Description . . . . . . . . . . . . . . . 29 97 3.10.3. Transport Features . . . . . . . . . . . . . . . . . 29 98 3.11. NACK-Oriented Reliable Multicast (NORM) . . . . . . . . . 30 99 3.11.1. Protocol Description . . . . . . . . . . . . . . . . 30 100 3.11.2. Interface Description . . . . . . . . . . . . . . . 31 101 3.11.3. Transport Features . . . . . . . . . . . . . . . . . 32 102 3.12. Transport Layer Security (TLS) and Datagram TLS (DTLS) as 103 a pseudotransport . . . . . . . . . . . . . . . . . . . . 32 104 3.12.1. Protocol Description . . . . . . . . . . . . . . . . 33 105 3.12.2. Interface Description . . . . . . . . . . . . . . . 34 106 3.12.3. Transport Features . . . . . . . . . . . . . . . . . 34 107 3.13. Hypertext Transport Protocol (HTTP) over TCP as a 108 pseudotransport . . . . . . . . . . . . . . . . . . . . . 35 109 3.13.1. Protocol Description . . . . . . . . . . . . . . . . 36 110 3.13.2. Interface Description . . . . . . . . . . . . . . . 37 111 3.13.3. Transport features . . . . . . . . . . . . . . . . . 37 112 4. Transport Service Features . . . . . . . . . . . . . . . . . 38 113 4.1. Complete Protocol Feature Matrix . . . . . . . . . . . . 40 114 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 42 115 6. Security Considerations . . . . . . . . . . . . . . . . . . . 42 116 7. Contributors . . . . . . . . . . . . . . . . . . . . . . . . 42 117 8. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 43 118 9. Informative References . . . . . . . . . . . . . . . . . . . 43 119 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 52 121 1. Introduction 123 Most Internet applications make use of the Transport Services 124 provided by TCP (a reliable, in-order stream protocol) or UDP (an 125 unreliable datagram protocol). We use the term "Transport Service" 126 to mean the end-to-end service provided to an application by the 127 transport layer. That service can only be provided correctly if 128 information about the intended usage is supplied from the 129 application. The application may determine this information at 130 design time, compile time, or run time, and may include guidance on 131 whether a feature is required, a preference by the application, or 132 something in between. Examples of features of Transport Services are 133 reliable delivery, ordered delivery, content privacy to in-path 134 devices, and integrity protection. 136 The IETF has defined a wide variety of transport protocols beyond TCP 137 and UDP, including SCTP, DCCP, MP-TCP, and UDP-Lite. Transport 138 services may be provided directly by these transport protocols, or 139 layered on top of them using protocols such as WebSockets (which runs 140 over TCP), RTP (over TCP or UDP) or WebRTC data channels (which run 141 over SCTP over DTLS over UDP or TCP). Services built on top of UDP 142 or UDP-Lite typically also need to specify additional mechanisms, 143 including a congestion control mechanism (such as NewReno, TFRC or 144 LEDBAT). This extends the set of available Transport Services beyond 145 those provided to applications by TCP and UDP. 147 [GF: Ledbat is a mechanism, not protocol - hence use the work 148 "support" in para below.] 150 Transport protocols can also be differentiated by the features of the 151 services they provide: for instance, SCTP offers a message-based 152 service providing full or partial reliability and allowing to 153 minimize the head of line blocking due to the support of unordered 154 and unordered message delivery within multiple streams, UDP-Lite and 155 DCCP provide partial integrity protection, and LEDBAT can support 156 low-priority "scavenger" communication. 158 2. Terminology 160 The following terms are defined throughout this document, and in 161 subsequent documents produced by TAPS describing the composition and 162 decomposition of transport services. 164 [EDITOR'S NOTE: we may want to add definitions for the different 165 kinds of interfaces that are important here.] 167 [GF: Interfaces may be covered by Micahel Welzl's companion 168 document?] 170 Transport Service Feature: a specific end-to-end feature that a 171 transport service provides to its clients. Examples include 172 confidentiality, reliable delivery, ordered delivery, message- 173 versus-stream orientation, etc. 175 Transport Service: a set of transport service features, without an 176 association to any given framing protocol, which provides a 177 complete service to an application. 179 Transport Protocol: an implementation that provides one or more 180 different transport services using a specific framing and header 181 format on the wire. 183 Transport Protocol Component: an implementation of a transport 184 service feature within a protocol. 186 Transport Service Instance: an arrangement of transport protocols 187 with a selected set of features and configuration parameters that 188 implements a single transport service, e.g. a protocol stack (RTP 189 over UDP). 191 Application: an entity that uses the transport layer for end-to-end 192 delivery data across the network (this may also be an upper layer 193 protocol or tunnel encapsulation). 195 3. Existing Transport Protocols 197 This section provides a list of known IETF transport protocol and 198 transport protocol frameworks. 200 3.1. Transport Control Protocol (TCP) 202 TCP is an IETF standards track transport protocol. [RFC0793] 203 introduces TCP as follows: "The Transmission Control Protocol (TCP) 204 is intended for use as a highly reliable host-to-host protocol 205 between hosts in packet-switched computer communication networks, and 206 in interconnected systems of such networks." Since its introduction, 207 TCP has become the default connection-oriented, stream-based 208 transport protocol in the Internet. It is widely implemented by 209 endpoints and widely used by common applications. 211 3.1.1. Protocol Description 213 TCP is a connection-oriented protocol, providing a three way 214 handshake to allow a client and server to set up a connection and 215 negotiate features, and mechanisms for orderly completion and 216 immediate teardown of a connection. TCP is defined by a family of 217 RFCs [RFC4614]. 219 TCP provides multiplexing to multiple sockets on each host using port 220 numbers.] A similar approach is adopted by other IETF-defined 221 transports. An active TCP session is identified by its four-tuple of 222 local and remote IP addresses and local port and remote port numbers. 223 The destination port during connection setup is often used to 224 indicate the requested service. 226 TCP partitions a continuous stream of bytes into segments, sized to 227 fit in IP packets. ICMP-based PathMTU discovery [RFC1191][RFC1981] 228 as well as Packetization Layer Path MTU Discovery (PMTUD) [RFC4821] 229 are supported. 231 Each byte in the stream is identified by a sequence number. The 232 sequence number is used to order segments on receipt, to identify 233 segments in acknowledgments, and to detect unacknowledged segments 234 for retransmission. This is the basis of the reliable, ordered 235 delivery of data in a TCP stream. TCP Selective Acknowledgment 236 [RFC2018] extends this mechanism by making it possible to identify 237 missing segments more precisely, reducing spurious retransmission. 239 Receiver flow control is provided by a sliding window: limiting the 240 amount of unacknowledged data that can be outstanding at a given 241 time. The window scale option [RFC7323] allows a receiver to use 242 windows greater than 64KB. 244 All TCP senders provide Congestion Control [RFC5681]: This uses a 245 separate window, where each time congestion is detected, this 246 congestion window is reduced. Most of the used congestion control 247 mechanisms use one of three mechanisms to detect congestion: A 248 retransmission timer (with exponential back-up), detection of loss 249 (interpreted as a congestion signal), or Explicit Congestion 250 Notification (ECN) [RFC3168] to provide early signaling (see 251 [I-D.ietf-aqm-ecn-benefits]). In addition, a congestion control 252 mechanism may react to changes in delay as an early indication for 253 congestion. 255 A TCP protocol instance can be extended [RFC4614] and tuned. Some 256 features are sender-side only, requiring no negotiation with the 257 receiver; some are receiver-side only, some are explicitly negotiated 258 during connection setup. 260 By default, TCP segment partitioning uses Nagle's algorithm [RFC0896] 261 to buffer data at the sender into large segments, potentially 262 incurring sender-side buffering delay; this algorithm can be disabled 263 by the sender to transmit more immediately, e.g., to reduce latency 264 for interactive sessions. 266 TCP provides a push and a urgent function to enable data to be 267 directly accessed by the receiver wihout having to wait for in-order 268 delivery of the data. However, [RFC6093] does not recommend the use 269 of the urgent flag due to the range of TCP implementations that 270 process TCP urgent indications differently. 272 A checksum provides an Integrity Check and is mandatory across the 273 entire packet. This check protects from delivery of corrupted data 274 and miselivery of packets to the wrong endpoint. This check is 275 relatively weak, applications that require end to end integrity of 276 data are recommended to include a stronger integrity check of their 277 payload data. The TCP checksum does not support partial corruption 278 protection (as in DCCP/UDP-Lite). 280 TCP only supports unicast connections. 282 3.1.2. Interface description 284 A User/TCP Interface is defined in [RFC0793] providing six user 285 commands: Open, Send, Receive, Close, Status. This interface does 286 not describe configuration of TCP options or parameters beside use of 287 the PUSH and URGENT flags. 289 [RFC1122] describes extensions of the TCP/application layer interface 290 for 1) reporting soft errors such as reception fo ICMP error 291 messages, extensive retransmission or urgent pointer advance, 2) 292 providing a possibility to specify the Type-of-Service (TOS) for 293 segments, 3) providing a fush call to empty the TCP send queue, and 294 4) multihoming support. 296 In API implementations derived from the BSD Sockets API, TCP sockets 297 are created using the "SOCK_STREAM" socket type as described in the 298 IEEE Portable Operating System Interface (POSIX) Base Specifications 299 [POSIX]. The features used by a protocol instance may be set and 300 tuned via this API. However, there is no RFC that documents this 301 interface. 303 3.1.3. Transport Features 305 The transport features provided by TCP are: 307 [EDITOR'S NOTE: expand each of these slightly] 309 o unicast transport 311 o connection setup with feature negotiation and application-to-port 312 mapping, implemented using SYN segments and the TCP option field 313 to negotiate features. 315 o port multiplexing: each TCP session is uniquely identified by a 316 combination of the ports and IP address fields. 318 o Uni-or bidirectional communication 320 o stream-oriented delivery in a single stream 322 o fully reliable delivery, implemented using ACKs sent from the 323 receiver to confirm delivery. 325 o error detection: a segment checksum verifies delivery to the 326 correct endpoint and integrity of the data and options. 328 o segmentation: packets are fragmented to a negotiated maximum 329 segment size, further constrained by the effective MTU from PMTUD. 331 o data bundling, an optional mechanism that uses Nagle's algorithm 332 to coalesce data sent within the same RTT into full-sized 333 segments. 335 o flow control using a window-based mechanism, where the receiver 336 advertises the window that it is willing to buffer. 338 o congestion control: a window-based method that uses AIMD to 339 control the sending rate and to conservatively choose a rate after 340 congestion is detected. 342 3.2. Multipath TCP (MPTCP) 344 Multipath TCP [RFC6824] is an extension for TCP to support multi- 345 homing. It is designed to be as transparent as possible to middle- 346 boxes. It does so by establishing regular TCP flows between a pair 347 of source/destination endpoints, and multiplexing the application's 348 stream over these flows. 350 3.2.1. Protocol Description 352 MPTCP uses TCP options for its control plane. They are used to 353 signal multipath capabilities, as well as to negotiate data sequence 354 numbers, and advertise other available IP addresses and establish new 355 sessions between pairs of endpoints. 357 3.2.2. Interface Description 359 By default, MPTCP exposes the same interface as TCP to the 360 application. [RFC6897] however describes a richer API for MPTCP- 361 aware applications. 363 This Basic API describes how an application can 365 o enable or disable MPTCP; 367 o bind a socket to one or more selected local endpoints; 369 o query local and remote endpoint addresses; 371 o get a unique connection identifier (similar to an address-port 372 pair for TCP). 374 The document also recommends the use of extensions defined for SCTP 375 [RFC6458] (see next section) to support multihoming. 377 3.2.3. Transport features 379 As an extension to TCP, MPTCP provides mostly the same features. By 380 establishing multiple sessions between available endpoints, it can 381 additionally provide soft failover solutions should one of the paths 382 become unusable. In addition, by multiplexing one byte stream over 383 separate paths, it can achieve a higher throughput than TCP in 384 certain situations (note however that coupled congestion control 385 [RFC6356] might limit this benefit to maintain fairness to other 386 flows at the bottleneck). When aggregating capacity over multiple 387 paths, and depending on the way packets are scheduled on each TCP 388 subflow, an additional delay and higher jitter might be observed 389 observed before in-order delivery of data to the applications. 391 The transport features provided by MPTCP in addition to TCP therefore 392 are: 394 o congestion control with load balancing over mutiple connections. 396 o endpoint multiplexing of a single byte stream (higher throughput). 398 o address family multiplexing: sub-flows can be started over IPv4 or 399 IPv6 for the same session. 401 o resilience to network failure and/or handover. 403 [AUTHOR'S NOTE: it is unclear whether MPTCP has to provide data 404 bundling.] 406 3.3. Stream Control Transmission Protocol (SCTP) 408 SCTP is a message-oriented standards track transport protocol. The 409 base protocol is specified in [RFC4960]. It supports multi-homing to 410 handle path failures. It also optionally supports path failover to 411 provide resilliance to path failures. An SCTP association has 412 multiple unidirectional streams in each direction and provides in- 413 sequence delivery of user messages only within each stream. This 414 allows it to minimize head of line blocking. SCTP is extensible and 415 the currently defined extensions include mechanisms for dynamic re- 416 configurations of streams [RFC6525] and IP-addresses [RFC5061]. 417 Furthermore, the extension specified in [RFC3758] introduces the 418 concept of partial reliability for user messages. 420 SCTP was originally developed for transporting telephony signalling 421 messages and is deployed in telephony signalling networks, especially 422 in mobile telephony networks. It can also be used for other 423 services, for example in the WebRTC framework for data channels and 424 is therefore deployed in all WEB-browsers supporting WebRTC. 426 3.3.1. Protocol Description 428 SCTP is a connection-oriented protocol using a four way handshake to 429 establish an SCTP association and a three way message exchange to 430 gracefully shut it down. It uses the same port number concept as 431 DCCP, TCP, UDP, and UDP-Lite, and only supports unicast. 433 SCTP uses the 32-bit CRC32c for protecting SCTP packets against bit 434 errors and miselivery of packets to the wrong endpoint. This is 435 stronger than the 16-bit checksums used by TCP or UDP. However, a 436 partial checksum coverage, as provided by DCCP or UDP-Lite is not 437 supported. 439 SCTP has been designed with extensibility in mind. Each SCTP packet 440 starts with a single common header containing the port numbers, a 441 verification tag and the CRC32c checksum. This common header is 442 followed by a sequence of chunks. Each chunk consists of a type 443 field, flags, a length field and a value. [RFC4960] defines how a 444 receiver processes chunks with an unknown chunk type. The support of 445 extensions can be negotiated during the SCTP handshake. 447 SCTP provides a message-oriented service. Multiple small user 448 messages can be bundled into a single SCTP packet to improve the 449 efficiency. For example, this bundling may be done by delaying user 450 messages at the sender similar to the Nagle algorithm used by TCP. 451 User messages which would result in IP packets larger than the MTU 452 will be fragmented at the sender and reassembled at the receiver. 453 There is no protocol limit on the user message size. ICMP-based path 454 MTU discovery as specified for IPv4 in [RFC1191] and for IPv6 in 455 [RFC1981] as well as packetization layer path MTU discovery as 456 specified in [RFC4821] with probe packets using the padding chunks 457 defined the [RFC4820] are supported. 459 [RFC4960] specifies a TCP friendly congestion control to protect the 460 network against overload. SCTP also uses a sliding window flow 461 control to protect receivers against overflow. Similar to TCP, SCTP 462 also supports delaying acknowledgements. [RFC7053] provides a way 463 for the sender of user messages to request the immediate sending of 464 the corresponding acknowledgements. 466 Each SCTP association has between 1 and 65536 uni-directional streams 467 in each direction. The number of streams can be different in each 468 direction. Every user-message is sent on a particular stream. User 469 messages can be sent un-ordered or ordered upon request by the upper 470 layer. Un-ordered messages can be delivered as soon as they are 471 completely received. Ordered messages sent on the same stream are 472 delivered at the receiver in the same order as sent by the sender. 473 For user messages not requiring fragmentation, this minimises head of 474 line blocking. 476 The base protocol defined in [RFC4960] does not allow interleaving of 477 user-messages, which results in sending a large message on one stream 478 can block the sending of user messages on other streams. 479 [I-D.ietf-tsvwg-sctp-ndata] overcomes this limitation. Furthermore, 480 [I-D.ietf-tsvwg-sctp-ndata] specifies multiple algorithms for the 481 sender side selection of which streams to send data from supporting a 482 variety of scheduling algorithms including priority based methods. 483 The stream re-configuration extension defined in [RFC6525] allows 484 streams to be reset during the lifetime of an association and to 485 increase the number of streams, if the number of streams negotiated 486 in the SCTP handshake becomes insufficient. 488 Each user message sent is either delivered to the receiver or, in 489 case of excessive retransmissions, the association is terminated in a 490 non-graceful way [RFC4960], similar to TCP behaviour. In addition to 491 this reliable transfer, the partial reliability extension [RFC3758] 492 allows a sender to abandon user messages. The application can 493 specify the policy for abandoning user messages. Examples for these 494 policies defined in [RFC3758] and [RFC7496] are: 496 o Limiting the time a user message is dealt with by the sender. 498 o Limiting the number of retransmissions for each fragment of a user 499 message. If the number of retransmissions is limited to 0, one 500 gets a service similar to UDP. 502 o Abandoning messages of lower priority in case of a send buffer 503 shortage. 505 SCTP supports multi-homing. Each SCTP endpoint uses a list of IP- 506 addresses and a single port number. These addresses can be any 507 mixture of IPv4 and IPv6 addresses. These addresses are negotiated 508 during the handshake and the address re-configuration extension 509 specified in [RFC5061] in combination with [RFC4895] can be used to 510 change these addresses in an authenticated way during the livetime of 511 an SCTP association. This allows for transport layer mobility. 512 Multiple addresses are used for improved resilience. If a remote 513 address becomes unreachable, the traffic is switched over to a 514 reachable one, if one exists. Each SCTP end-point supervises 515 continuously the reachability of all peer addresses using a heartbeat 516 mechanism. 518 For securing user messages, the use of TLS over SCTP has been 519 specified in [RFC3436]. However, this solution does not support all 520 services provided by SCTP (for example un-ordered delivery or partial 521 reliability), and therefore the use of DTLS over SCTP has been 522 specified in [RFC6083] to overcome these limitations. When using 523 DTLS over SCTP, the application can use almost all services provided 524 by SCTP. 526 [I-D.ietf-tsvwg-natsupp] defines methods for endpoints and 527 middleboxes to provide support NAT for SCTP over IPv4. For legacy 528 NAT traversal, [RFC6951] defines the UDP encapsulation of SCTP- 529 packets. Alternatively, SCTP packets can be encapsulated in DTLS 530 packets as specified in [I-D.ietf-tsvwg-sctp-dtls-encaps]. The 531 latter encapsulation is used within the WebRTC context. 533 SCTP has a well-defined API, described in the next subsection. 535 3.3.2. Interface Description 537 [RFC4960] defines an abstract API for the base protocol. This API 538 describes the following functions callable by the upper layer of 539 SCTP: Initialize, Associate, Send, Receive, Receive Unsent Message, 540 Receive Unacknowledged Message, Shutdown, Abort, SetPrimary, Status, 541 Change Heartbeat, Request Heartbeat, Get SRTT Report, Set Failure 542 Threshold, Set Protocol Parameters, and Destroy. The following 543 notifications are provided by the SCTP stack to the upper layer: 544 COMMUNICATION UP, DATA ARRIVE, SHUTDOWN COMPLETE, COMMUNICATION LOST, 545 COMMUNICATION ERROR, RESTART, SEND FAILURE, NETWORK STATUS CHANGE. 547 An extension to the BSD Sockets API is defined in [RFC6458] and 548 covers: 550 o the base protocol defined in [RFC4960]. The API allows to control 551 the local addresses and port numbers and the primary path. 552 Furthermore the application has fine control about parameters like 553 retransmission thresholds, the path supervision parameters, the 554 delayed acknowledgement timeout, and the fragmentation point. The 555 API provides a mechanism to allow the SCTP stack to notify the 556 application about event if the application has requested them. 557 These notifications provide Information about status changes of 558 the association and each of the peer addresses. In case of send 559 failures that application can also be notified and user messages 560 can be returned to the application. When sending user messages, 561 the stream id, a payload protocol identifier, an indication 562 whether ordered delivery is requested or not. These parameters 563 can also be provided on message reception. Additionally a context 564 can be provided when sending, which can be use in case of send 565 failures. The sending of arbitrary large user messages is 566 supported. 568 o the SCTP Partial Reliability extension defined in [RFC3758] to 569 specify for a user message the PR-SCTP policy and the policy 570 specific parameter. 572 o the SCTP Authentication extension defined in [RFC4895] allowing to 573 manage the shared keys, the HMAC to use, set the chunk types which 574 are only accepted in an authenticated way, and get the list of 575 chunks which are accepted by the local and remote end point in an 576 authenticated way. 578 o the SCTP Dynamic Address Reconfiguration extension defined in 579 [RFC5061]. It allows to manually add and delete local addresses 580 for SCTP associations and the enabling of automatic address 581 addition and deletion. Furthermore the peer can be given a hint 582 for choosing its primary path. 584 For the following SCTP protocol extensions the BSD Sockets API 585 extension is defined in the document specifying the protocol 586 extensions: 588 o the SCTP Stream Reconfiguration extension defined in [RFC6525]. 589 The API allows to trigger the reset operation for incoming and 590 outgoing streams and the whole association. It provides also a 591 way to notify the association about the corresponding events. 592 Furthermore the application can increase the number of streams. 594 o the UDP Encapsulation of SCTP packets extension defined in 595 [RFC6951]. The API allows the management of the remote UDP 596 encapsulation port. 598 o the SCTP SACK-IMMEDIATELY extension defined in [RFC7053]. The API 599 allows the sender of a user message to request the receiver to 600 send the corresponding acknowledgement immediately. 602 o the additional PR-SCTP policies defined in [RFC7496]. The API 603 allows to enable/disable the PR-SCTP extension, choose the PR-SCTP 604 policies defined in the document and provide statistical 605 information about abandoned messages. 607 Future documents describing SCTP protocol extensions are expected to 608 describe the corresponding BSD Sockets API extension in a "Socket API 609 Considerations" section. 611 The SCTP socket API supports two kinds of sockets: 613 o one-to-one style sockets (by using the socket type "SOCK_STREAM"). 615 o one-to-many style socket (by using the socket type 616 "SOCK_SEQPACKET"). 618 One-to-one style sockets are similar to TCP sockets, there is a 1:1 619 relationship between the sockets and the SCTP associations (except 620 for listening sockets). One-to-many style SCTP sockets are similar 621 to unconnected UDP sockets, where there is a 1:n relationship between 622 the sockets and the SCTP associations. 624 The SCTP stack can provide information to the applications about 625 state changes of the individual paths and the association whenever 626 they occur. These events are delivered similar to user messages but 627 are specifically marked as notifications. 629 New functions have been introduced to support the use of multiple 630 local and remote addresses. Additional SCTP-specific send and 631 receive calls have been defined to permit SCTP-specific information 632 to be snet without using ancillary data in the form of additional 633 cmsgs. These functions provide support for detecting partial 634 delivery of user messages and notifications. 636 The SCTP socket API allows a fine-grained control of the protocol 637 behaviour through an extensive set of socket options. 639 The SCTP kernel implementations of FreeBSD, Linux and Solaris follow 640 mostly the specified extension to the BSD Sockets API for the base 641 protocol and the corresponding supported protocol extensions. 643 3.3.3. Transport Features 645 The transport features provided by SCTP are: 647 [GF: This needs to be harmonised with the components for TCP] 649 o unicast. 651 o connection setup with feature negotiation and application-to-port 652 mapping. 654 o port multiplexing. 656 o message-oriented delivery. 658 o fully reliable or partially reliable delivery. 660 o ordered and unordered delivery within a stream. 662 o support for multiple concurrent streams. 664 o support for stream scheduling prioritization. 666 o flow control. 668 o congestion control. 670 o user message bundling. 672 o user message fragmentation and reassembly. 674 o strong error/misdelivery detection (CRC32c). 676 o transport layer multihoming for resilience. 678 o transport layer mobility. 680 3.4. User Datagram Protocol (UDP) 682 The User Datagram Protocol (UDP) [RFC0768] [RFC2460] is an IETF 683 standards track transport protocol. It provides a unidirectional, 684 datagram protocol that preserves message boundaries. It provides 685 none of the following transport features: error correction, 686 congestion control, or flow control. It can be used to send 687 broadcast datagrams (IPv4) or multicast datagrams (IPv4 and IPv6), in 688 addition to unicast (and anycast) datagrams. IETF guidance on the 689 use of UDP is provided in[I-D.ietf-tsvwg-rfc5405bis]. UDP is widely 690 implemented and widely used by common applications, including DNS. 692 3.4.1. Protocol Description 694 UDP is a connection-less protocol that maintains message boundaries, 695 with no connection setup or feature negotiation. The protocol uses 696 independent messages, ordinarily called datagrams. Each stream of 697 messages is independently managed, therefore retransmission does not 698 hold back data sent using other logical streams. It provides 699 detection of payload errors and misdelivery of packets to the wrong 700 endpoint, either of which result in discard of received datagrams. 702 It is possible to create IPv4 UDP datagrams with no checksum, and 703 while this is generally discouraged [RFC1122] 704 [I-D.ietf-tsvwg-rfc5405bis], certain special cases permit its use. 705 These datagrams relie on the IPv4 header checksum to protect from 706 misdelivery to the wrong endpoint. IPv6 does not by permit UDP 707 datagrams with no checksum, although in certain cases this rule may 708 be relaxed [RFC6935]. The checksum support considerations for 709 omitting the checksum are defined in [RFC6936]. Note that due to the 710 relatively weak form of checksum used by UDP, applications that 711 require end to end integrity of data are recommended to include a 712 stronger integrity check of their payload data. 714 It does not provide reliability and does not provide retransmission. 715 This implies messages may be re-ordered, lost, or duplicated in 716 transit. 718 A receiving application that is unable to run sufficiently fast, or 719 frequently, may miss messages since there is no flow control. The 720 lack of congestion handling implies UDP traffic may experience loss 721 when using an overlaoded path and may cause the loss of messages from 722 other protocols (e.g., TCP) when sharing the same network path. 724 [GF: This para isn't needed": Messages with payload errors are 725 ordinarily detected by an invalid end- to-end checksum and are 726 discarded before being delivered to an application. UDP-Lite (see 727 [RFC3828], and below) provides the ability for portions of the 728 message contents to be exempt from checksum coverage.] 730 On transmission, UDP encapsulates each datagram into an IP packet, 731 which may in turn be fragmented by IP and are reassembled before 732 delivery to the UDP receiver. 734 Applications that need to provide fragmentation or that have other 735 requirements such as receiver flow control, congestion control, 736 PathMTU discovery/PLPMTUD, support for ECN, etc need these to be 737 provided by protocols operating over UDP [I-D.ietf-tsvwg-rfc5405bis]. 739 3.4.2. Interface Description 741 [RFC0768] describes basic requirements for an API for UDP. Guidance 742 on use of common APIs is provided in [I-D.ietf-tsvwg-rfc5405bis]. 744 A UDP endpoint consists of a tuple of (IP address, port number). 745 Demultiplexing using multiple abstract endpoints (sockets) on the 746 same IP address are supported. The same socket may be used by a 747 single server to interact with multiple clients (note: this behavior 748 differs from TCP, which uses a pair of tuples to identify a 749 connection). Multiple server instances (processes) that bind the 750 same socket can cooperate to service multiple clients- the socket 751 implementation arranges to not duplicate the same received unicast 752 message to multiple server processes. 754 Many operating systems also allow a UDP socket to be "connected", 755 i.e., to bind a UDP socket to a specific (remote) UDP endpoint. 756 Unlike TCP's connect primitive, for UDP, this is only a local 757 operation that serves to simplify the local send/receive functions 758 and to filter the traffic for the specified addresses and ports 759 [I-D.ietf-tsvwg-rfc5405bis]. 761 3.4.3. Transport Features 763 The transport features provided by UDP are: 765 o unicast. 767 o multicast, anycast, or IPv4 broadcast. 769 o port multiplexing. A receiving port can be configured to receive 770 datagrams from multiple senders. 772 o message-oriented delivery. 774 o unidirectional or bidirectional. Transmission in each direction 775 is independent. 777 o non-reliable delivery. 779 o non-ordered delivery. 781 o IPv6 jumbograms. 783 o error and misdelivery detection (checksum). 785 o optional checksum. All or none of the payload data is protected. 787 3.5. Lightweight User Datagram Protocol (UDP-Lite) 789 The Lightweight User Datagram Protocol (UDP-Lite) [RFC3828] is an 790 IETF standards track transport protocol. It provides a 791 unidirectional, datagram protocol that preserves message boundaries. 792 IETF guidance on the use of UDP-Lite is provided in 793 [I-D.ietf-tsvwg-rfc5405bis]. 795 3.5.1. Protocol Description 797 UDP-Lite is a connection-less datagram protocol, with no connection 798 setup or feature negotiation. The protocol use messages, rather than 799 a byte-stream. Each stream of messages is independently managed, 800 therefore retransmission does not hold back data sent using other 801 logical streams. 803 It provides multiplexing to multiple sockets on each host using port 804 numbers, and its operation follows that for UDP. An active UDP-Lite 805 session is identified by its four-tuple of local and remote IP 806 addresses and local port and remote port numbers. 808 UDP-Lite changes the semantics of the UDP "payload length" field to 809 that of a "checksum coverage length" field, and is identified by a 810 different IP protocol/next-header value. Otherwise, UDP-Lite is 811 semantically identical to UDP. Applications using UDP-Lite therefore 812 can not make assumptions regarding the correctness of the data 813 received in the insensitive part of the UDP-Lite payload. 815 As for UDP, mechanisms for receiver flow control, congestion control, 816 PMTU or PLPMTU discovery, support for ECN, etc need to be provided by 817 upper layer protocols [I-D.ietf-tsvwg-rfc5405bis]. 819 Examples of use include a class of applications that can derive 820 benefit from having partially-damaged payloads delivered, rather than 821 discarded. One use is to support error tolerate payload corruption 822 when used over paths that include error-prone links, another 823 application is when header integrity checks are required, but payload 824 integrity is provided by some other mechanism (e.g., [RFC6936]. 826 A UDP-Lite service may support IPv4 broadcast, multicast, anycast and 827 unicast, and IPv6 multicast, anycast and unicast. 829 3.5.2. Interface Description 831 There is no current API specified in the RFC Series, but guidance on 832 use of common APIs is provided in [I-D.ietf-tsvwg-rfc5405bis]. 834 The interface of UDP-Lite differs from that of UDP by the addition of 835 a single (socket) option that communicates a checksum coverage length 836 value: at the sender, this specifies the intended checksum coverage, 837 with the remaining unprotected part of the payload called the "error- 838 insensitive part". The checksum coverage may also be made visible to 839 the application via the UDP-Lite MIB module [RFC5097]. 841 3.5.3. Transport Features 843 The transport features provided by UDP-Lite are: 845 o unicast. 847 o multicast, anycast, or IPv4 broadcast. 849 o port multiplexing (as for UDP). 851 o message-oriented delivery (as for UDP). 853 o non-reliable delivery (as for UDP). 855 o non-ordered delivery (as for UDP). 857 o error and misdelivery detection (checksum). 859 o partialor full integrity protection. The checksum coverage field 860 indicates the size of the payload data covered by the checksum. 862 3.6. Datagram Congestion Control Protocol (DCCP) 864 Datagram Congestion Control Protocol (DCCP) [RFC4340] is an IETF 865 standards track bidirectional transport protocol that provides 866 unicast connections of congestion-controlled messages without 867 providing reliability. 869 The DCCP Problem Statement describes the goals that DCCP sought to 870 address [RFC4336]. It is suitable for applications that transfer 871 fairly large amounts of data and that can benefit from control over 872 the trade off between timeliness and reliability [RFC4336]. 874 It offers low overhead, and many characteristics common to UDP, but 875 can avoid "Re-inventing the wheel" each time a new multimedia 876 application emerges. Specifically it includes core functions 877 (feature negotiation, path state management, RTT calculation, PMTUD, 878 etc): This allows applications to use a compatible method defining 879 how they send packets and where suitable to choose common algorithms 880 to manage their functions. Examples of suitable applications include 881 interactive applications, streaming media or on-line games [RFC4336]. 883 3.6.1. Protocol Description 885 DCCP is a connection-oriented datagram protocol, providing a three 886 way handshake to allow a client and server to set up a connection, 887 and mechanisms for orderly completion and immediate teardown of a 888 connection. The protocol is defined by a family of RFCs. 890 It provides multiplexing to multiple sockets at each endpoint using 891 port numbers. An active DCCP session is identified by its four-tuple 892 of local and remote IP addresses and local port and remote port 893 numbers. At connection setup, DCCP also exchanges the service code 894 [RFC5595], a mechanism that allows transport instantiations to 895 indicate the service treatment that is expected from the network. 897 The protocol segments data into messages, typically sized to fit in 898 IP packets, but which may be fragmented providing they are less than 899 the maximum packet size. A DCCP interface allows applications to 900 request fragmentation for packets larger than PMTU, but not larger 901 than the maximum packet size allowed by the current congestion 902 control mechanism (CCMPS) [RFC4340]. 904 Each message is identified by a sequence number. The sequence number 905 is used to identify segments in acknowledgments, to detect 906 unacknowledged segments, to measure RTT, etc. The protocol may 907 support ordered or unordered delivery of data, and does not itself 908 provide retransmission. DCCP supports reduced checksum coverage, a 909 partial integrity mechanisms similar to UDP-lIte. There is also a 910 Data Checksum option that when enabled, contains a strong CRC, to 911 enable endpoints to detect application data corruption. 913 Receiver flow control is supported: limiting the amount of 914 unacknowledged data that can be outstanding at a given time. 916 A DCCP protocol instance can be extended [RFC4340] and tuned using 917 features. Some features are sender-side only, requiring no 918 negotiation with the receiver; some are receiver-side only, some are 919 explicitly negotiated during connection setup. 921 A DCCP service is unicast. 923 DCCP supports negotiation of the congestion control profile, to 924 provide Plug and Play congestion control mechanisms. Examples of 925 specified profiles include [RFC4341] [RFC4342] [RFC5662]. All IETF- 926 defined methods provide Congestion Control. 928 DCCP use a Connect packet to initiate a session, and permits half- 929 connections that allow each client to choose the features it wishes 930 to support. Simultaneous open [RFC5596], as in TCP, can enable 931 interoperability in the presence of middleboxes. The Connect packet 932 includes a Service Code field [RFC5595] designed to allow middle 933 boxes and endpoints to identify the characteristics required by a 934 session. 936 A lightweight UDP-based encapsulation (DCCP-UDP) has been defined 937 [RFC6773] that permits DCCP to be used over paths where it is not 938 natively supported. Support in NAPT/NATs is defined in [RFC4340] and 939 [RFC5595]. 941 Upper layer protocols specified on top of DCCP include: DTLS 942 [RFC5595], RTP [RFC5672], ICE/SDP [RFC6773]. 944 A common packet format has allowed tools to evolve that can read and 945 interpret DCCP packets (e.g. Wireshark). 947 3.6.2. Interface Description 949 API characteristics include: - Datagram transmission. - Notification 950 of the current maximum packet size. - Send and reception of zero- 951 length payloads. - Slow Receiver flow control at a receiver. - 952 Detect a Slow receiver at the sender. 954 There is no current API curremntly specified in the RFC Series. 956 3.6.3. Transport Features 958 The transport features provided by DCCP are: 960 o unicast. 962 o connection setup with feature negotiation and application-to-port 963 mapping. 965 o Service Codes. Identifies the upper layer service to the endpoint 966 and network. 968 o port multiplexing. 970 o message-oriented delivery. 972 o non-reliable delivery. 974 o ordered delivery. 976 o flow control. The slow receiver function allows a receiver to 977 control the rate of the sender. 979 o drop notification. Allows a receiver to notify which datagrams 980 were not delivered to the peer upper layer protocol. 982 o timestamps. 984 o partial and full integrity protection (with optional strong 985 integrity check). 987 3.7. Lightweight User Datagram Protocol (UDP-Lite) 989 The Lightweight User Datagram Protocol (UDP-Lite) [RFC3828] is an 990 IETF standards track transport protocol. It provides a 991 unidirectional, datagram protocol that preserves message boundaries. 992 IETF guidance on the use of UDP-Lite is provided in 993 [I-D.ietf-tsvwg-rfc5405bis]. 995 3.7.1. Protocol Description 997 UDP-Lite is a connection-less datagram protocol, with no connection 998 setup or feature negotiation. The protocol use messages, rather than 999 a byte-stream. Each stream of messages is independently managed, 1000 therefore retransmission does not hold back data sent using other 1001 logical streams. 1003 It provides multiplexing to multiple sockets on each host using port 1004 numbers, and its operation follows that for UDP. An active UDP-Lite 1005 session is identified by its four-tuple of local and remote IP 1006 addresses and local port and remote port numbers. 1008 UDP-Lite changes the semantics of the UDP "payload length" field to 1009 that of a "checksum coverage length" field, and is identified by a 1010 different IP protocol/next-header value. Otherwise, UDP-Lite is 1011 semantically identical to UDP. Applications using UDP-Lite therefore 1012 can not make assumptions regarding the correctness of the data 1013 received in the insensitive part of the UDP-Lite payload. 1015 As for UDP, mechanisms for receiver flow control, congestion control, 1016 PMTU or PLPMTU discovery, support for ECN, etc need to be provided by 1017 upper layer protocols [I-D.ietf-tsvwg-rfc5405bis]. 1019 Examples of use include a class of applications that can derive 1020 benefit from having partially-damaged payloads delivered, rather than 1021 discarded. One use is to support error tolerate payload corruption 1022 when used over paths that include error-prone links, another 1023 application is when header integrity checks are required, but payload 1024 integrity is provided by some other mechanism (e.g., [RFC6936]. 1026 A UDP-Lite service may support IPv4 broadcast, multicast, anycast and 1027 unicast, and IPv6 multicast, anycast and unicast. 1029 3.7.2. Interface Description 1031 There is no current API specified in the RFC Series, but guidance on 1032 use of common APIs is provided in [I-D.ietf-tsvwg-rfc5405bis]. 1034 The interface of UDP-Lite differs from that of UDP by the addition of 1035 a single (socket) option that communicates a checksum coverage length 1036 value: at the sender, this specifies the intended checksum coverage, 1037 with the remaining unprotected part of the payload called the "error- 1038 insensitive part". The checksum coverage may also be made visible to 1039 the application via the UDP-Lite MIB module [RFC5097]. 1041 3.7.3. Transport Features 1043 The transport features provided by UDP-Lite are: 1045 o unicast 1047 o multicast, anycast, or IPv4 broadcast. 1049 o port multiplexing (as for UDP). 1051 o message-oriented delivery (as for UDP). 1053 o non-reliable delivery(as for UDP). 1055 o non-ordered delivery (as for UDP). 1057 o partial or full integrity protection. 1059 3.8. Internet Control Message Protocol (ICMP) 1061 The Internet Control Message Protocol (ICMP) [RFC0792] for IPv4 and 1062 [RFC4433] for IPv6 are IETF standards track protocols. 1064 It provides a conection-less unidirectional protocol that delivers 1065 individual messages. It provides none of the following transport 1066 features: error correction, congestion control, or flow control. 1067 Some messages may be sent as broadcast datagrams (IPv4) or multicast 1068 datagrams (IPv4 and IPv6), in addition to unicast (and anycast) 1069 datagrams. 1071 3.8.1. Protocol Description 1073 ICMP is a conection-less unidirectional protocol that delivers 1074 individual messages. The protocol uses independent messages, 1075 ordinarily called datagrams. Each message is required to carry a 1076 checksum as an integrity check and to protect from misdelivery to the 1077 wrong endpoint. 1079 ICMP messages typically relay diagnostic information from an endpoint 1080 [RFC1122] or network device [RFC1716] addressed to the sender of a 1081 flow. This usually contains the network protocol header of a packet 1082 that encountered the reported issue. Some formats of messages may 1083 also carry other payload data. Each message carries an integrity 1084 check calculated in the same way as UDP. 1086 The RFC series defines additional IPv6 message formats to support a 1087 range of uses. In the case of IPv6 the protocol incorporates 1088 neighbour discovery [RFC2461] [RFC3971]} (provided by ARP for IPv4) 1089 and the Multicast Listener Discovery (MLD) [RFC2710] group management 1090 functions (provided by IGMP for IPv4). 1092 Reliable transmission can not be assumed. A receiving application 1093 that is unable to run sufficiently fast, or frequently, may miss 1094 messages since there is no flow or congestion control. In addition 1095 some network devices rate-limit ICMP messages. 1097 Transport Protocols and upper layer protocols can use ICMP messages 1098 to help them take appropriate decisions when network or endpoint 1099 errors are reported. For example to implement, ICMP-based PathMTU 1100 discovery [RFC1191][RFC1981] or assist in Packetization Layer Path 1101 MTU Discovery (PMTUD) [RFC4821]. Such reactions to received messages 1102 needs to protects from off-path data injection 1103 [I-D.ietf-tsvwg-rfc5405bis], avoiding an application receiving 1104 packets that were created by an unauthorized third party. An 1105 application therefore needs to ensure that aLL messaged are 1106 appropriately validated, by checking the payload of the messages to 1107 ensure these are received in response to actually transmitted traffic 1108 (e.g., a reported error condition that corresponds to a UDP datagram 1109 or TCP segment was actually sent by the application). This requires 1110 context [RFC6056], such as local state about communication instances 1111 to each destination (e.g., in the TCP, DCCP, or SCTP protocols). 1112 This state is not always maintained by UDP-based applications 1113 [I-D.ietf-tsvwg-rfc5405bis]. 1115 Any response to ICMP error messages ought to be robust to temporary 1116 routing failures (sometimes called "soft errors"), e.g., transient 1117 ICMP "unreachable" messages ought to not normally cause a 1118 communication abort [RFC5461] [I-D.ietf-tsvwg-rfc5405bis]. 1120 3.8.2. Interface Description 1122 ICMP processing is integrated into many connection-oriented 1123 transports, but like other functions needs to be provided by an 1124 upper-layer protocol when using UDP and UDP-Lite. On some stacks, a 1125 bound socket also allows a UDP application to be notified when ICMP 1126 error messages are received for its transmissions 1127 [I-D.ietf-tsvwg-rfc5405bis]. 1129 3.8.3. Transport Features 1131 The transport features provided by ICMP are: 1133 o unidirectional. 1135 o multicast, anycast and IP4 broadcast. 1137 o message-oriented delivery. 1139 o non-reliable delivery. 1141 o non-ordered delivery. 1143 o error and misdelivery detection (checksum). 1145 3.9. Realtime Transport Protocol (RTP) 1147 RTP provides an end-to-end network transport service, suitable for 1148 applications transmitting real-time data, such as audio, video or 1149 data, over multicast or unicast network services, including TCP, UDP, 1150 UDP-Lite, or DCCP. 1152 [EDITOR'S NOTE: Varun Singh signed up as contributor for this 1153 section. Given the complexity of RTP, suggest to have an abbreviated 1154 section here contrasting RTP with other transports, and focusing on 1155 those features that are RTP-unique. Gorry Fairhurst contributed this 1156 stub section] 1158 3.9.1. Protocol Description 1160 The RTP standard [RFC3550] defines a pair of protocols, RTP and the 1161 Real Time Control Protocol, RTCP. The transport does not provide 1162 connection setup, but relies on out-of-band techniques or associated 1163 control protocols to setup, negotiate parameters or tear-down a 1164 session. 1166 An RTP sender encapsulates audio/video data into RTP packets to 1167 transport media streams. The RFC-series specifies RTP media formats 1168 allow packets to carry a wide range of media, and specifies a wide 1169 range of mulriplexing, error control and other support mechanisms. 1171 If a frame of media data is large, it will be fragment this into 1172 several RTP packets. If small, several frames may be bundled into a 1173 single RTP packet. RTP may runs over a congestion-controlled or non- 1174 congestion-controlled transport protocol. 1176 An RTP receiver collects RTP packets from network, validates them for 1177 correctness, and sends them to the media decoder input-queue. 1178 Missing packet detection is performed by the channel decoder. The 1179 play-out buffer is ordered by time stamp and is used to reorder 1180 packets. Damaged frames may be repaired before the media payloads 1181 are decompressed to display or store the data. 1183 RTCP is an associated control protocol that works with RTP. Both the 1184 RTP sender and receiver can send RTCP report packets. This is used 1185 to periodically send control information and report performance. 1186 Based on received RTCP feedback, an RTP sender can adjust the 1187 transmission, e.g., perform rate adaptation at the application layer 1188 in the case of congestion. 1190 An RTCP receiver report (RTCP RR) is returned to the sender 1191 periodically to report key parameters (e.g, the fraction of packets 1192 lost in the last reporting interval, the cumulative number of packets 1193 lost, the highest sequence number received, and the inter-arrival 1194 jitter). The RTCP RR packets also contain timing information that 1195 allows the sender to estimate the network round trip time (RTT) to 1196 the receivers. 1198 The interval between reports sent from each receiver tends to be on 1199 the order of a few seconds on average, although this varies with the 1200 session rate, and sub-second reporting intervals are possible for 1201 high rate sessions. The interval is randomised to avoid 1202 synchronization of reports from multiple receivers. 1204 3.9.2. Interface Description 1206 [EDITOR'S NOTE: to do] 1208 3.9.3. Transport Features 1210 The transport features provided by RTP are: 1212 o unicast. 1214 o multicast, anycast or IPv4 broadcast. 1216 o port multiplexing. 1218 o message-oriented delivery. 1220 o associated protocols for connection setup with feature negotiation 1221 and application-to-port mapping. 1223 o support for media types and other extensions. 1225 o segmentation and aggregation. 1227 o performance reporting. 1229 o drop notification. 1231 o timestamps. 1233 3.10. File Delivery over Unidirectional Transport/Asynchronous Layered 1234 Coding Reliable Multicast (FLUTE/ALC) 1236 FLUTE/ALC is an IETF standards track protocol specified in [RFC6726] 1237 and [RFC5775],. ALC provides an underlying reliable transport service 1238 and FLUTE a file-oriented specialization of the ALC service (e.g., to 1239 carry associated metadata). The [RFC6726] and [RFC5775] protocols 1240 are non-backward-compatible updates of the [RFC3926] and [RFC3450] 1241 experimental protocols; these experimental protocols are currently 1242 largely deployed in the 3GPP Multimedia Broadcast and Multicast 1243 Services (MBMS) (see [MBMS], section 7) and similar contexts (e.g., 1244 the Japanese ISDB-Tmm standard). 1246 The FLUTE/ALC protocol has been designed to support massively 1247 scalable reliable bulk data dissemination to receiver groups of 1248 arbitrary size using IP Multicast over any type of delivery network, 1249 including unidirectional networks (e.g., broadcast wireless 1250 channels). However, the FLUTE/ALC protocol also supports point-to- 1251 point unicast transmissions. 1253 FLUTE/ALC bulk data dissemination has been designed for discrete file 1254 or memory-based "objects". Transmissions happen either in push mode, 1255 where content is sent once, or in on-demand mode, where content is 1256 continuously sent during periods of time that can largely exceed the 1257 average time required to download the session objects (see [RFC5651], 1258 section 4.2). 1260 Altough FLUTE/ALC is not well adapted to byte- and message-streaming, 1261 there is an exception: FLUTE/ALC is used to carry 3GPP Dynamic 1262 Adaptive Streaming over HTTP (DASH) when scalability is a requirement 1263 (see [MBMS], section 5.6). In that case, each Audio/Video segment is 1264 transmitted as a distinct FLUTE/ALC object in push mode. FLUTE/ALC 1265 uses packet erasure coding (also known as Application-Level Forward 1266 Erasure Correction, or AL-FEC) in a proactive way. The goal of using 1267 AL-FEC is both to increase the robustness in front of packet erasures 1268 and to improve the efficiency of the on-demand service. FLUTE/ALC 1269 transmissions can be governed by a congestion control mechanism such 1270 as the "Wave and Equation Based Rate Control" (WEBRC) [RFC3738] when 1271 FLUTE/ALC is used in a layered transmission manner, with several 1272 session channels over which ALC packets are sent. However many 1273 FLUTE/ALC deployments involve only Constant Bit Rate (CBR) channels 1274 with no competing flows, for which a sender-based rate control 1275 mechanism is sufficient. In any case, FLUTE/ALC's reliability, 1276 delivery mode, congestion control, and flow/rate control mechanisms 1277 are distinct components that can be separately controlled to meet 1278 different application needs. 1280 3.10.1. Protocol Description 1282 The FLUTE/ALC protocol works on top of UDP (though it could work on 1283 top of any datagram delivery transport protocol), without requiring 1284 any connectivity from receivers to the sender. Purely unidirectional 1285 networks are therefore supported by FLUTE/ALC. This guarantees 1286 scalability to an unlimited number of receivers in a session, since 1287 the sender behaves exactly the same regardness of the number of 1288 receivers. 1290 FLUTE/ALC supports the transfer of bulk objects such as file or in- 1291 memory content, using either a push or an on-demand mode. in push 1292 mode, content is sent once to the receivers, while in on-demand mode, 1293 content is sent continuously during periods of time that can greatly 1294 exceed the average time required to download the session objects. 1296 This enables receivers to join a session asynchronously, at their own 1297 discretion, receive the content and leave the session. In this case, 1298 data content is typically sent continuously, in loops (also known as 1299 "carousels"). FLUTE/ALC also supports the transfer of an object 1300 stream, with loose real-time constraints. This is particularly 1301 useful to carry 3GPP DASH when scalability is a requirement and 1302 unicast transmissions over HTTP cannot be used ([MBMS], section 5.6). 1303 In this case, packets are sent in sequence using push mode. FLUTE/ 1304 ALC is not well adapted to byte- and message-streaming and other 1305 solutions could be preferred (e.g., FECFRAME [RFC6363] with real-time 1306 flows). 1308 The FLUTE file delivery instantiation of ALC provides a metadata 1309 delivery service. Each object of the FLUTE/ALC session is described 1310 in a dedicated entry of a File Delivery Table (FDT), using an XML 1311 format (see [RFC6726], section 3.2). This metadata can include, but 1312 is not restricted to, a URI attribute (to identify and locate the 1313 object), a media type attribute, a size attribute, an encoding 1314 attribute, or a message digest attribute. Since the set of objects 1315 sent within a session can be dynamic, with new objects being added 1316 and old ones removed, several instances of the FDT can be sent and a 1317 mechanism is provided to identify a new FDT Instance. 1319 To provide robustness against packet loss and improve the efficiency 1320 of the on-demand mode, FLUTE/ALC relies on packet erasure coding (AL- 1321 FEC). AL-FEC encoding is proactive (since there is no feedback and 1322 therefore no (N)ACK-based retransmission) and ALC packets containing 1323 repair data are sent along with ALC packets containing source data. 1324 Several FEC Schemes have been standardized; FLUTE/ALC does not 1325 mandate the use of any particular one. Several strategies concerning 1326 the transmission order of ALC source and repair packets are possible, 1327 in particular in on-demand mode where it can deeply impact the 1328 service provided (e.g., to favor the recovery of objects in sequence, 1329 or at the other extreme, to favor the recovery of all objects in 1330 parallel), and FLUTE/ALC does not mandate nor recommend the use of 1331 any particular one. 1333 A FLUTE/ALC session is composed of one or more channels, associated 1334 to different destination unicast and/or multicast IP addresses. ALC 1335 packets are sent in those channels at a certain transmission rate, 1336 with a rate that often differs depending on the channel. FLUTE/ALC 1337 does not mandate nor recommend any strategy to select which ALC 1338 packet to send on which channel. FLUTE/ALC can use a multiple rate 1339 congestion control building block (e.g., WEBRC) to provide congestion 1340 control that is feedback free, where receivers adjust their reception 1341 rates individually by joining and leaving channels associated with 1342 the session. To that purpose, the ALC header provides a specific 1343 field to carry congestion control specific information. However 1344 FLUTE/ALC does not mandate the use of a particular congestion control 1345 mechanism although WEBRC is mandatory to support in case of Internet 1346 ([RFC6726], section 1.1.4). FLUTE/ALC is often used over a network 1347 path with pre-provisoned capacity [RFC5404] whete theres are no flows 1348 competing for capacity. In this case, a sender-based rate control 1349 mechanism and a single channel is sufficient. 1351 [RFC6584] provides per-packet authentication, integrity, and anti- 1352 replay protection in the context of the ALC and NORM protocols. 1353 Several mechanisms are proposed that seamlessly integrate into these 1354 protocols using the ALC and NORM header extension mechanisms. 1356 3.10.2. Interface Description 1358 The FLUTE/ALC specification does not describe a specific application 1359 programming interface (API) to control protocol operation. 1360 Open source reference implementations of FLUTE/ALC are available at 1361 http://planete-bcast.inrialpes.fr/ (no longer maintained) and 1362 http://mad.cs.tut.fi/ (no longer maintained), and these 1363 implementations specify and document their own APIs. Commercial 1364 versions are also available, some derived from the above 1365 implementations, with their own API. 1367 3.10.3. Transport Features 1369 The transport features provided by FLUTE/ALC are: 1371 o unicast 1373 o multicast, anycast or IPv4 broadcast. 1375 o per-object dynamic meta-data delivery. 1377 o push delivery or on-demand delivery service. 1379 o fully reliable or partially reliable delivery (of file or in- 1380 memory objects). 1382 o ordered or unordered delivery (of file or in-memory objects). 1384 o per-packet authentication, integrity, and anti-replay services. 1386 o proactive packet erasure coding (AL-FEC) to recover from packet 1387 erasures and improve the on-demand delivery service, 1389 o error detection (through UDP and lower level checksums). 1391 o congestion control for layered flows (e.g., with WEBRC). 1393 o rate control transmission in a given channel. 1395 3.11. NACK-Oriented Reliable Multicast (NORM) 1397 NORM is an IETF standards track protocol specified in [RFC5740]. The 1398 protocol was designed to support reliable bulk data dissemination to 1399 receiver groups using IP Multicast but also provides for point-to- 1400 point unicast operation. Its support for bulk data dissemination 1401 includes discrete file or computer memory-based "objects" as well as 1402 byte- and message-streaming. NORM is designed to incorporate packet 1403 erasure coding as an inherent part of its selective ARQ in response 1404 to receiver negative acknowledgements. The packet erasure coding can 1405 also be proactively applied for forward protection from packet loss. 1406 NORM transmissions are governed by the TCP-friendly congestion 1407 control. NORM's reliability, congestion control, and flow control 1408 mechanism are distinct components and can be separately controlled to 1409 meet different application needs. 1411 3.11.1. Protocol Description 1413 [EDITOR'S NOTE: needs to be more clear about the application of FEC 1414 and packet erasure coding; expand ARQ.] 1416 The NORM protocol is encapsulated in UDP datagrams and thus provides 1417 multiplexing for multiple sockets on hosts using port numbers. For 1418 purposes of loosely coordinated IP Multicast, NORM is not strictly 1419 connection-oriented although per-sender state is maintained by 1420 receivers for protocol operation. [RFC5740] does not specify a 1421 handshake protocol for connection establishment and separate session 1422 initiation can be used to coordinate port numbers. However, in-band 1423 "client-server" style connection establishment can be accomplished 1424 with the NORM congestion control signaling messages using port 1425 binding techniques like those for TCP client-server connections. 1427 NORM supports bulk "objects" such as file or in-memory content but 1428 also can treat a stream of data as a logical bulk object for purposes 1429 of packet erasure coding. In the case of stream transport, NORM can 1430 support either byte streams or message streams where application- 1431 defined message boundary information is carried in the NORM protocol 1432 messages. This allows the receiver(s) to join/re-join and recover 1433 message boundaries mid-stream as needed. Application content is 1434 carried and identified by the NORM protocol with encoding symbol 1435 identifiers depending upon the Forward Error Correction (FEC) Scheme 1436 [RFC3452] configured. NORM uses NACK-based selective ARQ to reliably 1437 deliver the application content to the receiver(s). NORM proactively 1438 measures round-trip timing information to scale ARQ timers 1439 appropriately and to support congestion control. For multicast 1440 operation, timer-based feedback suppression is uses to achieve group 1441 size scaling with low feedback traffic levels. The feedback 1442 suppression is not applied for unicast operation. 1444 NORM uses rate-based congestion control based upon the TCP-Friendly 1445 Rate Control (TFRC) [RFC4324] principles that are also used in DCCP 1446 [RFC4340]. NORM uses control messages to measure RTT and collect 1447 congestion event (e..g, loss event, ECN event, etc) information from 1448 the receiver(s) to support dynamic rate control adjustment. The TCP- 1449 Friendly Multicast Congestion Control (TFMCC) [RFC4654] used provides 1450 some extra features to support multicast but is functionally 1451 equivalent to TFRC in the unicast case. 1453 NORM's reliability mechanism is decoupled from congestion control. 1454 This allows alternative arrangements of transport services to be 1455 invoked. For example, fixed-rate reliable delivery can be supported 1456 or unreliable (but optionally "better than best effort" via packet 1457 erasure coding) delivery with rate-control per TFRC can be achieved. 1458 Additionally, alternative congestion control techniques may be 1459 applied. For example, TFRC rate control with congestion event 1460 detection based on ECN for links with high packet loss (e.g., 1461 wireless) has been implemented and demonstrated with NORM. 1463 While NORM is NACK-based for reliability transfer, it also supports a 1464 positive acknowledgment (ACK) mechanism that can be used for receiver 1465 flow control. Again, since this mechanism is decoupled from the 1466 reliability and congestion control, applications that have different 1467 needs in this aspect can use the protocol differently. One example 1468 is the use of NORM for quasi-reliable delivery where timely delivery 1469 of newer content may be favored over completely reliable delivery of 1470 older content within buffering and RTT constraints. 1472 3.11.2. Interface Description 1474 The NORM specification does not describe a specific application 1475 programming interface (API) to control protocol operation. A freely- 1476 available, open source reference implementation of NORM is available 1477 at https://www.nrl.navy.mil/itd/ncs/products/norm, and a documented 1478 API is provided for this implementation. While a sockets-like API is 1479 not currently documented, the existing API supports the necessary 1480 functions for that to be implemented. 1482 3.11.3. Transport Features 1484 The transport features provided by NORM are: 1486 o unicast or multicast. 1488 o stream-oriented delivery in a single stream. 1490 o object-oriented delivery of discrete data or file items. 1492 o reliable delivery. 1494 o unordered unidirectional delivery (of in-memory data or file bulk 1495 content objects). 1497 o error detection (UDP checksum). 1499 o segmentation. 1501 o data bundling (Nagle's algorithm). 1503 o flow control (timer-based and/or ack-based). 1505 o congestion control. 1507 o packet erasure coding (both proactively and as part of ARQ). 1509 3.12. Transport Layer Security (TLS) and Datagram TLS (DTLS) as a 1510 pseudotransport 1512 Transport Layer Security (TLS) and Datagram TLS (DTLS) are IETF 1513 protocols that provide several security-related features to 1514 applications. TLS is designed to run on top of a reliable streaming 1515 transport protocol (usually TCP), while DTLS is designed to run on 1516 top of a best-effort datagram protocol (UDP or DCCP [RFC5238]). At 1517 the time of writing, the current version of TLS is 1.2; it is defined 1518 in [RFC5246]. DTLS provides nearly identical functionality to 1519 applications; it is defined in [RFC6347] and its current version is 1520 also 1.2. The TLS protocol evolved from the Secure Sockets Layer 1521 (SSL) protocols developed in the mid 90s to support protection of 1522 HTTP traffic. 1524 While older versions of TLS and DTLS are still in use, they provide 1525 weaker security guarantees. [RFC7457] outlines important attacks on 1526 TLS and DTLS. [RFC7525] is a Best Current Practices (BCP) document 1527 that describes secure configurations for TLS and DTLS to counter 1528 these attacks. The recommendations are applicable for the vast 1529 majority of use cases. 1531 [NOTE: The Logjam authors (weakdh.org) give (inconclusive) evidence 1532 that one of the recommendations of [RFC7525], namely the use of 1533 DHE-1024 as a fallback, may not be sufficient in all cases to counter 1534 an attacker with the resources of a nation-state. It is unclear at 1535 this time if the RFC is going to be updated as a result, or whether 1536 there will be an RFC7525bis.] 1538 3.12.1. Protocol Description 1540 Both TLS and DTLS provide the same security features and can thus be 1541 discussed together. The features they provide are: 1543 o Confidentiality 1545 o Data integrity 1547 o Peer authentication (optional) 1549 o Perfect forward secrecy (optional) 1551 The authentication of the peer entity can be omitted; a common web 1552 use case is where the server is authenticated and the client is not. 1553 TLS also provides a completely anonymous operation mode in which 1554 neither peer's identity is authenticated. It is important to note 1555 that TLS itself does not specify how a peering entity's identity 1556 should be interpreted. For example, in the common use case of 1557 authentication by means of an X.509 certificate, it is the 1558 application's decision whether the certificate of the peering entity 1559 is acceptable for authorization decisions. Perfect forward secrecy, 1560 if enabled and supported by the selected algorithms, ensures that 1561 traffic encrypted and captured during a session at time t0 cannot be 1562 later decrypted at time t1 (t1 > t0), even if the long-term secrets 1563 of the communicating peers are later compromised. 1565 As DTLS is generally used over an unreliable datagram transport such 1566 as UDP, applications will need to tolerate loss, re-ordered, or 1567 duplicated datagrams. Like TLS, DTLS conveys application data in a 1568 sequence of independent records. However, because records are mapped 1569 to unreliable datagrams, there are several features unique to DTLS 1570 that are not applicable to TLS: 1572 o Record replay detection (optional). 1574 o Record size negotiation (estimates of PMTU and record size 1575 expansion factor). 1577 o Coveyance of IP don't fragment (DF) bit settings by application. 1579 o An anti-DoS stateless cookie mechanism (optional). 1581 Generally, DTLS follows the TLS design as closely as possible. To 1582 operate over datagrams, DTLS includes a sequence number and limited 1583 forms of retransmission and fragmentation for its internal 1584 operations. The sequence number may be used for detecting replayed 1585 information, according to the windowing procedure described in 1586 Section 4.1.2.6 of [RFC6347]. Note also that DTLS forbids the use of 1587 stream ciphers, which are essentially incompatible when operating on 1588 independent encrypted records. 1590 3.12.2. Interface Description 1592 TLS is commonly invoked using an API provided by packages such as 1593 OpenSSL, wolfSSL, or GnuTLS. Using such APIs entails the 1594 manipulation of several important abstractions, which fall into the 1595 following categories: long-term keys and algorithms, session state, 1596 and communications/connections. There may also be special APIs 1597 required to deal with time and/or random numbers, both of which are 1598 needed by a variety of encryption algorithms and protocols. 1600 Considerable care is required in the use of TLS APIs in order to 1601 create a secure application. The programmer should have at least a 1602 basic understanding of encryption and digital signature algorithms 1603 and their strengths, public key infrastructure (including X.509 1604 certificates and certificate revocation), and the sockets API. See 1605 [RFC7525] and [RFC7457], as mentioned above. 1607 As an example, in the case of OpenSSL, the primary abstractions are 1608 the library itself and method (protocol), session, context, cipher 1609 and connection. After initializing the library and setting the 1610 method, a cipher suite is chosen and used to configure a context 1611 object. Session objects may then be minted according to the 1612 parameters present in a context object and associated with individual 1613 connections. Depending on how precisely the programmer wishes to 1614 select different algorithmic or protocol options, various levels of 1615 details may be required. 1617 3.12.3. Transport Features 1619 Both TLS and DTLS employ a layered architecture. The lower layer is 1620 commonly called the record protocol. It is responsible for: 1622 o message fragmentation 1624 o authentication and integrity via message authentication codes 1625 (MAC) 1627 o data encryption 1629 o scheduling transmission using the underlying transport protocol 1631 DTLS augments the TLS record protocol with: 1633 o ordering and replay protection, implemented using sequence 1634 numbers. 1636 Several protocols are layered on top of the record protocol. These 1637 include the handshake, alert, and change cipher spec protocols. 1638 There is also the data protocol, used to carry application traffic. 1639 The handshake protocol is used to establish cryptographic and 1640 compression parameters when a connection is first set up. In DTLS, 1641 this protocol also has a basic fragmentation and retransmission 1642 capability and a cookie-like mechanism to resist DoS attacks. (TLS 1643 compression is not recommended at present). The alert protocol is 1644 used to inform the peer of various conditions, most of which are 1645 terminal for the connection. The change cipher spec protocol is used 1646 to synchronize changes in cryptographic parameters for each peer. 1648 3.13. Hypertext Transport Protocol (HTTP) over TCP as a pseudotransport 1650 Hypertext Transfer Protocol (HTTP) is an application-level protocol 1651 widely used on the Internet. Version 1.1 of the protocol is 1652 specified in [RFC7230] [RFC7231] [RFC7232] [RFC7233] [RFC7234] 1653 [RFC7235], and version 2 in [RFC7540]. Furthermore, HTTP is used as 1654 a substrate for other application-layer protocols. There are various 1655 reasons for this practice listed in [RFC3205]; these include being a 1656 well-known and well-understood protocol, reusability of existing 1657 servers and client libraries, easy use of existing security 1658 mechanisms such as HTTP digest authentication [RFC2617] and TLS 1659 [RFC5246], the ability of HTTP to traverse firewalls which makes it 1660 work with a lot of infrastructure, and cases where a application 1661 server often needs to support HTTP anyway. 1663 Depending on application's needs, the use of HTTP as a substrate 1664 protocol may add complexity and overhead in comparison to a special- 1665 purpose protocol (e.g. HTTP headers, suitability of the HTTP 1666 security model etc.). [RFC3205] address this issues and provides 1667 some guidelines and concerns about the use of HTTP standard port 80 1668 and 443, the use of HTTP URL scheme and interaction with existing 1669 firewalls, proxies and NATs. 1671 Though not strictly bound to TCP, HTTP is almost exclusively run over 1672 TCP, and therefore inherits its properties when used in this way. 1674 3.13.1. Protocol Description 1676 Hypertext Transfer Protocol (HTTP) is a request/response protocol. A 1677 client sends a request containing a request method, URI and protocol 1678 version followed by a MIME-like message (see [RFC7231] for the 1679 differences between an HTTP object and a MIME message), containing 1680 information about the client and request modifiers. The message can 1681 contain a message body carrying application data as well. The server 1682 responds with a status or error code followed by a MIME-like message 1683 containing information about the server and information about carried 1684 data and it can include a message body. It is possible to specify a 1685 data format for the message body using MIME media types [RFC2045]. 1686 Furthermore, the protocol has numerous additional features; features 1687 relevant to pseudotransport are described below. 1689 Content negotiation, specified in [RFC7231], is a mechanism provided 1690 by HTTP for selecting a representation on a requested resource. The 1691 client and server negotiate acceptable data formats, charsets, data 1692 encoding (e.g. data can be transferred compressed, gzip), etc. HTTP 1693 can accommodate exchange of messages as well as data streaming (using 1694 chunked transfer encoding [RFC7230]). It is also possible to request 1695 a part of a resource using range requests specified in [RFC7233]. 1696 The protocol provides powerful cache control signalling defined in 1697 [RFC7234]. 1699 HTTP 1.1's and HTTP 2.0's persistent connections can be use to 1700 perform multiple request-response transactions during the life-time 1701 of a single HTTP connection. Moreover, HTTP 2.0 connections can 1702 multiplex many request/response pairs in parallel on a single 1703 connection. This reduces connection establishment overhead and the 1704 effect of TCP slow-start on each transaction, important for HTTP's 1705 primary use case. 1707 It is possible to combine HTTP with security mechanisms, like TLS 1708 (denoted by HTTPS), which adds protocol properties provided by such a 1709 mechanism (e.g. authentication, encryption, etc.). TLS's 1710 Application-Layer Protocol Negotiation (ALPN) extension [RFC7301] can 1711 be used for HTTP version negotiation within TLS handshake which 1712 eliminates addition round-trip. Arbitrary cookie strings, included 1713 as part of the MIME headers, are often used as bearer tokens in HTTP. 1715 Application layer protocols using HTTP as substrate may use existing 1716 method and data formats, or specify new methods and data formats. 1717 Furthermore some protocols may not fit a request/response paradigm 1718 and instead rely on HTTP to send messages (e.g. [RFC6546]). Because 1719 HTTP is working in many restricted infrastructures, it is also used 1720 to tunnel other application-layer protocols. 1722 3.13.2. Interface Description 1724 There are many HTTP libraries available exposing different APIs. The 1725 APIs provide a way to specify a request by providing a URI, a method, 1726 request modifiers and optionally a request body. For the response, 1727 callbacks can be registered that will be invoked when the response is 1728 received. If TLS is used, API expose a registration of callbacks in 1729 case a server requests client authentication and when certificate 1730 verification is needed. 1732 World Wide Web Consortium (W3C) standardized the XMLHttpRequest API 1733 [XHR], an API that can be use for sending HTTP/HTTPS requests and 1734 receiving server responses. Besides XML data format, request and 1735 response data format can also be JSON, HTML and plain text. 1736 Specifically JavaScript and XMLHttpRequest are a ubiquitous 1737 programming model for websites, and more general applications, where 1738 native code is less attractive. 1740 Representational State Transfer (REST) [REST] is another example how 1741 applications can use HTTP as transport protocol. REST is an 1742 architecture style for building application on the Internet. It uses 1743 HTTP as a communication protocol. 1745 3.13.3. Transport features 1747 The transport features provided by HTTP, when used as a 1748 pseudotransport, are: 1750 o unicast. 1752 o message and stream-oriented transfer. 1754 o bi- or unidirectional transmission. 1756 o ordered delivery. 1758 o fully reliable delivery. 1760 o object range request. 1762 o message content type negotiation. 1764 o flow control. 1766 HTTPS (HTTP over TLS) additionally provides the following components: 1768 o authentication (of one or both ends of a connection). 1770 o confidentiality. 1772 o integrity protection. 1774 4. Transport Service Features 1776 [EDITOR'S NOTE: This section is still work-in-progress. This list is 1777 probably not complete and/or too detailed.] 1779 The transport protocol components analyzed in this document which can 1780 be used as a basis for defining common transport service features, 1781 normalized and separated into categories, are as follows: 1783 o Control Functions 1785 * Addressing 1787 + unicast 1789 + multicast, anycast and IPv4 broadcast 1791 + use of NAPT-compatible port numbers 1793 * Multihoming support 1795 + multihoming for resilience 1797 + multihoming for mobility 1799 - specify handover latency? 1801 + multihoming for load-balancing 1803 - specify interleaving delay? 1805 * Multiplexing 1807 + application to port mapping 1809 + single vs. multiple streaming 1811 o Delivery 1813 * reliability 1815 + fully reliable delivery 1817 + partially reliable delivery 1818 - packet erasure coding 1820 + unreliable delivery 1822 - drop notification 1824 - Integrity protection 1826 o checksum for error detection 1828 o partial payload checksum protection 1830 o checksum optional 1832 * ordering 1834 + ordered delivery 1836 + unordered delivery 1838 - unordered delivery of in-memory data 1840 * type/framing 1842 + stream-oriented delivery 1844 + message-oriented delivery 1846 + object-oriented delivery of discrete data or file items 1848 - object content type negotiation 1850 + range-based partical object transmission 1852 + file bulk content objects 1854 o Transmission control 1856 * rate control 1858 + timer-based 1860 + ACK-based 1862 * congestion control 1864 * flow control 1865 * segmentation 1867 * data/message bundling (Nagle's algorithm) 1869 * stream scheduling prioritization 1871 o Security 1873 * authentication of one end of a connection 1875 * authentication of both ends of a connection 1877 * confidentiality 1879 * cryptographic integrity protection 1881 A future revision of this document will define transport service 1882 features based upon this list. 1884 [EDITOR'S NOTE: this section will drawn from the candidate features 1885 provided by protocol components in the previous section - please 1886 discuss on taps@ietf.org list] 1888 4.1. Complete Protocol Feature Matrix 1890 [EDITOR'S NOTE: Dave Thaler has signed up as a contributor for this 1891 section. Michael Welzl also has a beginning of a matrix which could 1892 be useful here.] 1894 [EDITOR'S NOTE: The below is a strawman proposal below by Gorry 1895 Fairhurst for initial discussion] 1897 The table below summarises protocol mechanisms that have been 1898 standardised. It does not make an assessment on whether specific 1899 implementations are fully compliant to these specifications. 1901 +-----------------+---------+---------+---------+---------+---------+ 1902 | Mechanism | UDP | UDP-L | DCCP | SCTP | TCP | 1903 +-----------------+---------+---------+---------+---------+---------+ 1904 | Unicast | Yes | Yes | Yes | Yes | Yes | 1905 | | | | | | | 1906 | Mcast/IPv4Bcast | Yes(2) | Yes | No | No | No | 1907 | | | | | | | 1908 | Port Mux | Yes | Yes | Yes | Yes | Yes | 1909 | | | | | | | 1910 | Mode | Dgram | Dgram | Dgram | Dgram | Stream | 1911 | | | | | | | 1912 | Connected | No | No | Yes | Yes | Yes | 1913 | | | | | | | 1914 | Data bundling | No | No | No | Yes | Yes | 1915 | | | | | | | 1916 | Feature Nego | No | No | Yes | Yes | Yes | 1917 | | | | | | | 1918 | Options | No | No | Support | Support | Support | 1919 | | | | | | | 1920 | Data priority | * | * | * | Yes | No | 1921 | | | | | | | 1922 | Data bundling | No | No | No | Yes | Yes | 1923 | | | | | | | 1924 | Reliability | None | None | None | Select | Full | 1925 | | | | | | | 1926 | Ordered deliv | No | No | No | Stream | Yes | 1927 | | | | | | | 1928 | Corruption Tol. | No | Support | Support | No | No | 1929 | | | | | | | 1930 | Flow Control | No | No | Support | Yes | Yes | 1931 | | | | | | | 1932 | PMTU/PLPMTU | (1) | (1) | Yes | Yes | Yes | 1933 | | | | | | | 1934 | Cong Control | (1) | (1) | Yes | Yes | Yes | 1935 | | | | | | | 1936 | ECN Support | (1) | (1) | Yes | TBD | Yes | 1937 | | | | | | | 1938 | NAT support | Limited | Limited | Support | TBD | Support | 1939 | | | | | | | 1940 | Security | DTLS | DTLS | DTLS | DTLS | TLS, AO | 1941 | | | | | | | 1942 | UDP encaps | N/A | None | Yes | Yes | None | 1943 | | | | | | | 1944 | RTP support | Support | Support | Support | ? | Support | 1945 +-----------------+---------+---------+---------+---------+---------+ 1947 Note (1): this feature requires support in an upper layer protocol. 1949 Note (2): this feature requires support in an upper layer protocol 1950 when used with IPv6. 1952 5. IANA Considerations 1954 This document has no considerations for IANA. 1956 6. Security Considerations 1958 This document surveys existing transport protocols and protocols 1959 providing transport-like services. Confidentiality, integrity, and 1960 authenticity are among the features provided by those services. This 1961 document does not specify any new components or mechanisms for 1962 providing these features. Each RFC listed in this document discusses 1963 the security considerations of the specification it contains. 1965 7. Contributors 1967 [Editor's Note: turn this into a real contributors section with 1968 addresses once we figure out how to trick the toolchain into doing 1969 so] 1971 o Section 3.2 on MPTCP was contributed by Simone Ferlin-Oliviera 1972 (ferlin@simula.no) and Olivier Mehani 1973 (olivier.mehani@nicta.com.au) 1975 o Section 3.4 on UDP was contributed by Kevin Fall (kfall@kfall.com) 1977 o Section 3.3 on SCTP was contributed by Michael Tuexen (tuexen@fh- 1978 muenster.de) 1980 o Section 3.10 on FLUTE/ALC was contributed by Vincent Roca 1981 (vincent.roca@inria.fr) 1983 o Section 3.11 on NORM was contributed by Brian Adamson 1984 (brian.adamson@nrl.navy.mil) 1986 o Section 3.12 on TLS and DTLS was contributed by Ralph Holz 1987 (ralph.holz@nicta.com.au) and Olivier Mehani 1988 (olivier.mehani@nicta.com.au) 1990 o Section 3.13 on HTTP was contributed by Dragana Damjanovic 1991 (ddamjanovic@mozilla.com) 1993 8. Acknowledgments 1995 Thanks to Karen Nielsen, Joe Touch, and Michael Welzl for the 1996 comments, feedback, and discussion. This work is partially supported 1997 by the European Commission under grant agreements FP7-ICT-318627 1998 mPlane and from the Horizon 2020 research and innovation program 1999 under grant agreement No. 644334 (NEAT); support does not imply 2000 endorsement. 2002 9. Informative References 2004 [RFC0768] Postel, J., "User Datagram Protocol", STD 6, RFC 768, DOI 2005 10.17487/RFC0768, August 1980, 2006 . 2008 [RFC0792] Postel, J., "Internet Control Message Protocol", STD 5, 2009 RFC 792, DOI 10.17487/RFC0792, September 1981, 2010 . 2012 [RFC0793] Postel, J., "Transmission Control Protocol", STD 7, RFC 2013 793, DOI 10.17487/RFC0793, September 1981, 2014 . 2016 [RFC0896] Nagle, J., "Congestion Control in IP/TCP Internetworks", 2017 RFC 896, DOI 10.17487/RFC0896, January 1984, 2018 . 2020 [RFC1122] Braden, R., Ed., "Requirements for Internet Hosts - 2021 Communication Layers", STD 3, RFC 1122, DOI 10.17487/ 2022 RFC1122, October 1989, 2023 . 2025 [RFC1191] Mogul, J. and S. Deering, "Path MTU discovery", RFC 1191, 2026 DOI 10.17487/RFC1191, November 1990, 2027 . 2029 [RFC1716] Almquist, P. and F. Kastenholz, "Towards Requirements for 2030 IP Routers", RFC 1716, DOI 10.17487/RFC1716, November 2031 1994, . 2033 [RFC1981] McCann, J., Deering, S., and J. Mogul, "Path MTU Discovery 2034 for IP version 6", RFC 1981, DOI 10.17487/RFC1981, August 2035 1996, . 2037 [RFC2018] Mathis, M., Mahdavi, J., Floyd, S., and A. Romanow, "TCP 2038 Selective Acknowledgment Options", RFC 2018, DOI 10.17487/ 2039 RFC2018, October 1996, 2040 . 2042 [RFC2045] Freed, N. and N. Borenstein, "Multipurpose Internet Mail 2043 Extensions (MIME) Part One: Format of Internet Message 2044 Bodies", RFC 2045, DOI 10.17487/RFC2045, November 1996, 2045 . 2047 [RFC2460] Deering, S. and R. Hinden, "Internet Protocol, Version 6 2048 (IPv6) Specification", RFC 2460, DOI 10.17487/RFC2460, 2049 December 1998, . 2051 [RFC2461] Narten, T., Nordmark, E., and W. Simpson, "Neighbor 2052 Discovery for IP Version 6 (IPv6)", RFC 2461, DOI 2053 10.17487/RFC2461, December 1998, 2054 . 2056 [RFC2617] Franks, J., Hallam-Baker, P., Hostetler, J., Lawrence, S., 2057 Leach, P., Luotonen, A., and L. Stewart, "HTTP 2058 Authentication: Basic and Digest Access Authentication", 2059 RFC 2617, DOI 10.17487/RFC2617, June 1999, 2060 . 2062 [RFC2710] Deering, S., Fenner, W., and B. Haberman, "Multicast 2063 Listener Discovery (MLD) for IPv6", RFC 2710, DOI 2064 10.17487/RFC2710, October 1999, 2065 . 2067 [RFC3168] Ramakrishnan, K., Floyd, S., and D. Black, "The Addition 2068 of Explicit Congestion Notification (ECN) to IP", RFC 2069 3168, DOI 10.17487/RFC3168, September 2001, 2070 . 2072 [RFC3205] Moore, K., "On the use of HTTP as a Substrate", BCP 56, 2073 RFC 3205, DOI 10.17487/RFC3205, February 2002, 2074 . 2076 [RFC3436] Jungmaier, A., Rescorla, E., and M. Tuexen, "Transport 2077 Layer Security over Stream Control Transmission Protocol", 2078 RFC 3436, DOI 10.17487/RFC3436, December 2002, 2079 . 2081 [RFC3450] Luby, M., Gemmell, J., Vicisano, L., Rizzo, L., and J. 2082 Crowcroft, "Asynchronous Layered Coding (ALC) Protocol 2083 Instantiation", RFC 3450, DOI 10.17487/RFC3450, December 2084 2002, . 2086 [RFC3452] Luby, M., Vicisano, L., Gemmell, J., Rizzo, L., Handley, 2087 M., and J. Crowcroft, "Forward Error Correction (FEC) 2088 Building Block", RFC 3452, DOI 10.17487/RFC3452, December 2089 2002, . 2091 [RFC3550] Schulzrinne, H., Casner, S., Frederick, R., and V. 2092 Jacobson, "RTP: A Transport Protocol for Real-Time 2093 Applications", STD 64, RFC 3550, DOI 10.17487/RFC3550, 2094 July 2003, . 2096 [RFC3738] Luby, M. and V. Goyal, "Wave and Equation Based Rate 2097 Control (WEBRC) Building Block", RFC 3738, DOI 10.17487/ 2098 RFC3738, April 2004, 2099 . 2101 [RFC3758] Stewart, R., Ramalho, M., Xie, Q., Tuexen, M., and P. 2102 Conrad, "Stream Control Transmission Protocol (SCTP) 2103 Partial Reliability Extension", RFC 3758, DOI 10.17487/ 2104 RFC3758, May 2004, 2105 . 2107 [RFC3828] Larzon, L-A., Degermark, M., Pink, S., Jonsson, L-E., Ed., 2108 and G. Fairhurst, Ed., "The Lightweight User Datagram 2109 Protocol (UDP-Lite)", RFC 3828, DOI 10.17487/RFC3828, July 2110 2004, . 2112 [RFC3926] Paila, T., Luby, M., Lehtonen, R., Roca, V., and R. Walsh, 2113 "FLUTE - File Delivery over Unidirectional Transport", RFC 2114 3926, DOI 10.17487/RFC3926, October 2004, 2115 . 2117 [RFC3971] Arkko, J., Ed., Kempf, J., Zill, B., and P. Nikander, 2118 "SEcure Neighbor Discovery (SEND)", RFC 3971, DOI 2119 10.17487/RFC3971, March 2005, 2120 . 2122 [RFC4324] Royer, D., Babics, G., and S. Mansour, "Calendar Access 2123 Protocol (CAP)", RFC 4324, DOI 10.17487/RFC4324, December 2124 2005, . 2126 [RFC4336] Floyd, S., Handley, M., and E. Kohler, "Problem Statement 2127 for the Datagram Congestion Control Protocol (DCCP)", RFC 2128 4336, DOI 10.17487/RFC4336, March 2006, 2129 . 2131 [RFC4340] Kohler, E., Handley, M., and S. Floyd, "Datagram 2132 Congestion Control Protocol (DCCP)", RFC 4340, DOI 2133 10.17487/RFC4340, March 2006, 2134 . 2136 [RFC4341] Floyd, S. and E. Kohler, "Profile for Datagram Congestion 2137 Control Protocol (DCCP) Congestion Control ID 2: TCP-like 2138 Congestion Control", RFC 4341, DOI 10.17487/RFC4341, March 2139 2006, . 2141 [RFC4342] Floyd, S., Kohler, E., and J. Padhye, "Profile for 2142 Datagram Congestion Control Protocol (DCCP) Congestion 2143 Control ID 3: TCP-Friendly Rate Control (TFRC)", RFC 4342, 2144 DOI 10.17487/RFC4342, March 2006, 2145 . 2147 [RFC4433] Kulkarni, M., Patel, A., and K. Leung, "Mobile IPv4 2148 Dynamic Home Agent (HA) Assignment", RFC 4433, DOI 2149 10.17487/RFC4433, March 2006, 2150 . 2152 [RFC4614] Duke, M., Braden, R., Eddy, W., and E. Blanton, "A Roadmap 2153 for Transmission Control Protocol (TCP) Specification 2154 Documents", RFC 4614, DOI 10.17487/RFC4614, September 2155 2006, . 2157 [RFC4654] Widmer, J. and M. Handley, "TCP-Friendly Multicast 2158 Congestion Control (TFMCC): Protocol Specification", RFC 2159 4654, DOI 10.17487/RFC4654, August 2006, 2160 . 2162 [RFC4820] Tuexen, M., Stewart, R., and P. Lei, "Padding Chunk and 2163 Parameter for the Stream Control Transmission Protocol 2164 (SCTP)", RFC 4820, DOI 10.17487/RFC4820, March 2007, 2165 . 2167 [RFC4821] Mathis, M. and J. Heffner, "Packetization Layer Path MTU 2168 Discovery", RFC 4821, DOI 10.17487/RFC4821, March 2007, 2169 . 2171 [RFC4895] Tuexen, M., Stewart, R., Lei, P., and E. Rescorla, 2172 "Authenticated Chunks for the Stream Control Transmission 2173 Protocol (SCTP)", RFC 4895, DOI 10.17487/RFC4895, August 2174 2007, . 2176 [RFC4960] Stewart, R., Ed., "Stream Control Transmission Protocol", 2177 RFC 4960, DOI 10.17487/RFC4960, September 2007, 2178 . 2180 [RFC5061] Stewart, R., Xie, Q., Tuexen, M., Maruyama, S., and M. 2181 Kozuka, "Stream Control Transmission Protocol (SCTP) 2182 Dynamic Address Reconfiguration", RFC 5061, DOI 10.17487/ 2183 RFC5061, September 2007, 2184 . 2186 [RFC5097] Renker, G. and G. Fairhurst, "MIB for the UDP-Lite 2187 protocol", RFC 5097, DOI 10.17487/RFC5097, January 2008, 2188 . 2190 [RFC5246] Dierks, T. and E. Rescorla, "The Transport Layer Security 2191 (TLS) Protocol Version 1.2", RFC 5246, DOI 10.17487/ 2192 RFC5246, August 2008, 2193 . 2195 [RFC5238] Phelan, T., "Datagram Transport Layer Security (DTLS) over 2196 the Datagram Congestion Control Protocol (DCCP)", RFC 2197 5238, DOI 10.17487/RFC5238, May 2008, 2198 . 2200 [RFC5404] Westerlund, M. and I. Johansson, "RTP Payload Format for 2201 G.719", RFC 5404, DOI 10.17487/RFC5404, January 2009, 2202 . 2204 [RFC5461] Gont, F., "TCP's Reaction to Soft Errors", RFC 5461, DOI 2205 10.17487/RFC5461, February 2009, 2206 . 2208 [RFC5595] Fairhurst, G., "The Datagram Congestion Control Protocol 2209 (DCCP) Service Codes", RFC 5595, DOI 10.17487/RFC5595, 2210 September 2009, . 2212 [RFC5596] Fairhurst, G., "Datagram Congestion Control Protocol 2213 (DCCP) Simultaneous-Open Technique to Facilitate NAT/ 2214 Middlebox Traversal", RFC 5596, DOI 10.17487/RFC5596, 2215 September 2009, . 2217 [RFC5651] Luby, M., Watson, M., and L. Vicisano, "Layered Coding 2218 Transport (LCT) Building Block", RFC 5651, DOI 10.17487/ 2219 RFC5651, October 2009, 2220 . 2222 [RFC5662] Shepler, S., Ed., Eisler, M., Ed., and D. Noveck, Ed., 2223 "Network File System (NFS) Version 4 Minor Version 1 2224 External Data Representation Standard (XDR) Description", 2225 RFC 5662, DOI 10.17487/RFC5662, January 2010, 2226 . 2228 [RFC5672] Crocker, D., Ed., "RFC 4871 DomainKeys Identified Mail 2229 (DKIM) Signatures -- Update", RFC 5672, DOI 10.17487/ 2230 RFC5672, August 2009, 2231 . 2233 [RFC5740] Adamson, B., Bormann, C., Handley, M., and J. Macker, 2234 "NACK-Oriented Reliable Multicast (NORM) Transport 2235 Protocol", RFC 5740, DOI 10.17487/RFC5740, November 2009, 2236 . 2238 [RFC5775] Luby, M., Watson, M., and L. Vicisano, "Asynchronous 2239 Layered Coding (ALC) Protocol Instantiation", RFC 5775, 2240 DOI 10.17487/RFC5775, April 2010, 2241 . 2243 [RFC5681] Allman, M., Paxson, V., and E. Blanton, "TCP Congestion 2244 Control", RFC 5681, DOI 10.17487/RFC5681, September 2009, 2245 . 2247 [RFC6056] Larsen, M. and F. Gont, "Recommendations for Transport- 2248 Protocol Port Randomization", BCP 156, RFC 6056, DOI 2249 10.17487/RFC6056, January 2011, 2250 . 2252 [RFC6083] Tuexen, M., Seggelmann, R., and E. Rescorla, "Datagram 2253 Transport Layer Security (DTLS) for Stream Control 2254 Transmission Protocol (SCTP)", RFC 6083, DOI 10.17487/ 2255 RFC6083, January 2011, 2256 . 2258 [RFC6093] Gont, F. and A. Yourtchenko, "On the Implementation of the 2259 TCP Urgent Mechanism", RFC 6093, DOI 10.17487/RFC6093, 2260 January 2011, . 2262 [RFC6525] Stewart, R., Tuexen, M., and P. Lei, "Stream Control 2263 Transmission Protocol (SCTP) Stream Reconfiguration", RFC 2264 6525, DOI 10.17487/RFC6525, February 2012, 2265 . 2267 [RFC6546] Trammell, B., "Transport of Real-time Inter-network 2268 Defense (RID) Messages over HTTP/TLS", RFC 6546, DOI 2269 10.17487/RFC6546, April 2012, 2270 . 2272 [RFC6347] Rescorla, E. and N. Modadugu, "Datagram Transport Layer 2273 Security Version 1.2", RFC 6347, DOI 10.17487/RFC6347, 2274 January 2012, . 2276 [RFC6356] Raiciu, C., Handley, M., and D. Wischik, "Coupled 2277 Congestion Control for Multipath Transport Protocols", RFC 2278 6356, DOI 10.17487/RFC6356, October 2011, 2279 . 2281 [RFC6363] Watson, M., Begen, A., and V. Roca, "Forward Error 2282 Correction (FEC) Framework", RFC 6363, DOI 10.17487/ 2283 RFC6363, October 2011, 2284 . 2286 [RFC6455] Fette, I. and A. Melnikov, "The WebSocket Protocol", RFC 2287 6455, DOI 10.17487/RFC6455, December 2011, 2288 . 2290 [RFC6458] Stewart, R., Tuexen, M., Poon, K., Lei, P., and V. 2291 Yasevich, "Sockets API Extensions for the Stream Control 2292 Transmission Protocol (SCTP)", RFC 6458, DOI 10.17487/ 2293 RFC6458, December 2011, 2294 . 2296 [RFC6584] Roca, V., "Simple Authentication Schemes for the 2297 Asynchronous Layered Coding (ALC) and NACK-Oriented 2298 Reliable Multicast (NORM) Protocols", RFC 6584, DOI 2299 10.17487/RFC6584, April 2012, 2300 . 2302 [RFC6726] Paila, T., Walsh, R., Luby, M., Roca, V., and R. Lehtonen, 2303 "FLUTE - File Delivery over Unidirectional Transport", RFC 2304 6726, DOI 10.17487/RFC6726, November 2012, 2305 . 2307 [RFC6773] Phelan, T., Fairhurst, G., and C. Perkins, "DCCP-UDP: A 2308 Datagram Congestion Control Protocol UDP Encapsulation for 2309 NAT Traversal", RFC 6773, DOI 10.17487/RFC6773, November 2310 2012, . 2312 [RFC6824] Ford, A., Raiciu, C., Handley, M., and O. Bonaventure, 2313 "TCP Extensions for Multipath Operation with Multiple 2314 Addresses", RFC 6824, DOI 10.17487/RFC6824, January 2013, 2315 . 2317 [RFC6897] Scharf, M. and A. Ford, "Multipath TCP (MPTCP) Application 2318 Interface Considerations", RFC 6897, DOI 10.17487/RFC6897, 2319 March 2013, . 2321 [RFC6935] Eubanks, M., Chimento, P., and M. Westerlund, "IPv6 and 2322 UDP Checksums for Tunneled Packets", RFC 6935, DOI 2323 10.17487/RFC6935, April 2013, 2324 . 2326 [RFC6936] Fairhurst, G. and M. Westerlund, "Applicability Statement 2327 for the Use of IPv6 UDP Datagrams with Zero Checksums", 2328 RFC 6936, DOI 10.17487/RFC6936, April 2013, 2329 . 2331 [RFC6951] Tuexen, M. and R. Stewart, "UDP Encapsulation of Stream 2332 Control Transmission Protocol (SCTP) Packets for End-Host 2333 to End-Host Communication", RFC 6951, DOI 10.17487/ 2334 RFC6951, May 2013, 2335 . 2337 [RFC7053] Tuexen, M., Ruengeler, I., and R. Stewart, "SACK- 2338 IMMEDIATELY Extension for the Stream Control Transmission 2339 Protocol", RFC 7053, DOI 10.17487/RFC7053, November 2013, 2340 . 2342 [RFC7230] Fielding, R., Ed. and J. Reschke, Ed., "Hypertext Transfer 2343 Protocol (HTTP/1.1): Message Syntax and Routing", RFC 2344 7230, DOI 10.17487/RFC7230, June 2014, 2345 . 2347 [RFC7231] Fielding, R., Ed. and J. Reschke, Ed., "Hypertext Transfer 2348 Protocol (HTTP/1.1): Semantics and Content", RFC 7231, DOI 2349 10.17487/RFC7231, June 2014, 2350 . 2352 [RFC7232] Fielding, R., Ed. and J. Reschke, Ed., "Hypertext Transfer 2353 Protocol (HTTP/1.1): Conditional Requests", RFC 7232, DOI 2354 10.17487/RFC7232, June 2014, 2355 . 2357 [RFC7233] Fielding, R., Ed., Lafon, Y., Ed., and J. Reschke, Ed., 2358 "Hypertext Transfer Protocol (HTTP/1.1): Range Requests", 2359 RFC 7233, DOI 10.17487/RFC7233, June 2014, 2360 . 2362 [RFC7234] Fielding, R., Ed., Nottingham, M., Ed., and J. Reschke, 2363 Ed., "Hypertext Transfer Protocol (HTTP/1.1): Caching", 2364 RFC 7234, DOI 10.17487/RFC7234, June 2014, 2365 . 2367 [RFC7235] Fielding, R., Ed. and J. Reschke, Ed., "Hypertext Transfer 2368 Protocol (HTTP/1.1): Authentication", RFC 7235, DOI 2369 10.17487/RFC7235, June 2014, 2370 . 2372 [RFC7301] Friedl, S., Popov, A., Langley, A., and E. Stephan, 2373 "Transport Layer Security (TLS) Application-Layer Protocol 2374 Negotiation Extension", RFC 7301, DOI 10.17487/RFC7301, 2375 July 2014, . 2377 [RFC7323] Borman, D., Braden, B., Jacobson, V., and R. 2378 Scheffenegger, Ed., "TCP Extensions for High Performance", 2379 RFC 7323, DOI 10.17487/RFC7323, September 2014, 2380 . 2382 [RFC7457] Sheffer, Y., Holz, R., and P. Saint-Andre, "Summarizing 2383 Known Attacks on Transport Layer Security (TLS) and 2384 Datagram TLS (DTLS)", RFC 7457, DOI 10.17487/RFC7457, 2385 February 2015, . 2387 [RFC7496] Tuexen, M., Seggelmann, R., Stewart, R., and S. Loreto, 2388 "Additional Policies for the Partially Reliable Stream 2389 Control Transmission Protocol Extension", RFC 7496, DOI 2390 10.17487/RFC7496, April 2015, 2391 . 2393 [RFC7525] Sheffer, Y., Holz, R., and P. Saint-Andre, 2394 "Recommendations for Secure Use of Transport Layer 2395 Security (TLS) and Datagram Transport Layer Security 2396 (DTLS)", BCP 195, RFC 7525, DOI 10.17487/RFC7525, May 2397 2015, . 2399 [RFC7540] Belshe, M., Peon, R., and M. Thomson, Ed., "Hypertext 2400 Transfer Protocol Version 2 (HTTP/2)", RFC 7540, DOI 2401 10.17487/RFC7540, May 2015, 2402 . 2404 [I-D.ietf-tsvwg-rfc5405bis] 2405 Eggert, L., Fairhurst, G., and G. Shepherd, "UDP Usage 2406 Guidelines", draft-ietf-tsvwg-rfc5405bis-05 (work in 2407 progress), August 2015. 2409 [I-D.ietf-aqm-ecn-benefits] 2410 Fairhurst, G. and M. Welzl, "The Benefits of using 2411 Explicit Congestion Notification (ECN)", draft-ietf-aqm- 2412 ecn-benefits-06 (work in progress), July 2015. 2414 [I-D.ietf-tsvwg-sctp-dtls-encaps] 2415 Tuexen, M., Stewart, R., Jesup, R., and S. Loreto, "DTLS 2416 Encapsulation of SCTP Packets", draft-ietf-tsvwg-sctp- 2417 dtls-encaps-09 (work in progress), January 2015. 2419 [I-D.ietf-tsvwg-sctp-ndata] 2420 Stewart, R., Tuexen, M., Loreto, S., and R. Seggelmann, 2421 "Stream Schedulers and User Message Interleaving for the 2422 Stream Control Transmission Protocol", draft-ietf-tsvwg- 2423 sctp-ndata-04 (work in progress), July 2015. 2425 [I-D.ietf-tsvwg-natsupp] 2426 Stewart, R., Tuexen, M., and I. Ruengeler, "Stream Control 2427 Transmission Protocol (SCTP) Network Address Translation 2428 Support", draft-ietf-tsvwg-natsupp-08 (work in progress), 2429 July 2015. 2431 [XHR] van Kesteren, A., Aubourg, J., Song, J., and H. Steen, 2432 "XMLHttpRequest working draft 2433 (http://www.w3.org/TR/XMLHttpRequest/)", 2000. 2435 [REST] Fielding, R., "Architectural Styles and the Design of 2436 Network-based Software Architectures, Ph. D. (UC Irvine), 2437 Chapter 5: Representational State Transfer", 2000. 2439 [POSIX] 1-2008, IEEE., "IEEE Standard for Information Technology 2440 -- Portable Operating System Interface (POSIX) Base 2441 Specifications, Issue 7", n.d.. 2443 [MBMS] 3GPP TSG WS S4, ., "3GPP TS 26.346: Multimedia Broadcast/ 2444 Multicast Service (MBMS); Protocols and codecs, release 13 2445 (http://www.3gpp.org/DynaReport/26346.htm).", 2015. 2447 Authors' Addresses 2449 Godred Fairhurst (editor) 2450 University of Aberdeen 2451 School of Engineering, Fraser Noble Building 2452 Aberdeen AB24 3UE 2454 Email: gorry@erg.abdn.ac.uk 2455 Brian Trammell (editor) 2456 ETH Zurich 2457 Gloriastrasse 35 2458 8092 Zurich 2459 Switzerland 2461 Email: ietf@trammell.ch 2463 Mirja Kuehlewind (editor) 2464 ETH Zurich 2465 Gloriastrasse 35 2466 8092 Zurich 2467 Switzerland 2469 Email: mirja.kuehlewind@tik.ee.ethz.ch