idnits 2.17.1 draft-ietf-behave-rfc3489bis-07.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** It looks like you're using RFC 3978 boilerplate. You should update this to the boilerplate described in the IETF Trust License Policy document (see https://trustee.ietf.org/license-info), which is required now. -- Found old boilerplate from RFC 3978, Section 5.1 on line 22. -- Found old boilerplate from RFC 3978, Section 5.5, updated by RFC 4748 on line 2083. -- Found old boilerplate from RFC 3979, Section 5, paragraph 1 on line 2094. -- Found old boilerplate from RFC 3979, Section 5, paragraph 2 on line 2101. -- Found old boilerplate from RFC 3979, Section 5, paragraph 3 on line 2107. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- == There are 1 instance of lines with private range IPv4 addresses in the document. If these are generic example addresses, they should be changed to use any of the ranges defined in RFC 6890 (or successor): 192.0.2.x, 198.51.100.x or 203.0.113.x. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust Copyright Line does not match the current year == The document seems to lack the recommended RFC 2119 boilerplate, even if it appears to use RFC 2119 keywords. (The document does seem to have the reference to RFC 2119 which the ID-Checklist requires). -- The document seems to lack a disclaimer for pre-RFC5378 work, but may have content which was first submitted before 10 November 2008. If you have contacted all the original authors and they are all willing to grant the BCP78 rights to the IETF Trust, then this is fine, and you can ignore this comment. If not, you may need to add the pre-RFC5378 disclaimer. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (July 8, 2007) is 6130 days in the past. Is this intentional? -- Found something which looks like a code comment -- if you have code sections in the document, please surround them with '' and '' lines. Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) ** Obsolete normative reference: RFC 2818 (Obsoleted by RFC 9110) ** Obsolete normative reference: RFC 2617 (Obsoleted by RFC 7235, RFC 7615, RFC 7616, RFC 7617) ** Obsolete normative reference: RFC 2988 (Obsoleted by RFC 6298) -- Possible downref: Non-RFC (?) normative reference: ref. 'ITU.V42.1994' -- Obsolete informational reference (is this intentional?): RFC 2616 (Obsoleted by RFC 7230, RFC 7231, RFC 7232, RFC 7233, RFC 7234, RFC 7235) == Outdated reference: A later version (-19) exists of draft-ietf-mmusic-ice-16 -- Obsolete informational reference (is this intentional?): RFC 3489 (Obsoleted by RFC 5389) == Outdated reference: A later version (-16) exists of draft-ietf-behave-turn-03 == Outdated reference: A later version (-20) exists of draft-ietf-sip-outbound-09 == Outdated reference: A later version (-08) exists of draft-ietf-behave-nat-behavior-discovery-00 == Outdated reference: A later version (-16) exists of draft-ietf-mmusic-ice-tcp-03 -- Obsolete informational reference (is this intentional?): RFC 2434 (Obsoleted by RFC 5226) Summary: 4 errors (**), 0 flaws (~~), 8 warnings (==), 12 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 BEHAVE Working Group J. Rosenberg 3 Internet-Draft Cisco 4 Obsoletes: 3489 (if approved) C. Huitema 5 Intended status: Standards Track Microsoft 6 Expires: January 9, 2008 R. Mahy 7 Plantronics 8 P. Matthews 9 Avaya 10 D. Wing 11 Cisco 12 July 8, 2007 14 Session Traversal Utilities for (NAT) (STUN) 15 draft-ietf-behave-rfc3489bis-07 17 Status of this Memo 19 By submitting this Internet-Draft, each author represents that any 20 applicable patent or other IPR claims of which he or she is aware 21 have been or will be disclosed, and any of which he or she becomes 22 aware will be disclosed, in accordance with Section 6 of BCP 79. 24 Internet-Drafts are working documents of the Internet Engineering 25 Task Force (IETF), its areas, and its working groups. Note that 26 other groups may also distribute working documents as Internet- 27 Drafts. 29 Internet-Drafts are draft documents valid for a maximum of six months 30 and may be updated, replaced, or obsoleted by other documents at any 31 time. It is inappropriate to use Internet-Drafts as reference 32 material or to cite them other than as "work in progress." 34 The list of current Internet-Drafts can be accessed at 35 http://www.ietf.org/ietf/1id-abstracts.txt. 37 The list of Internet-Draft Shadow Directories can be accessed at 38 http://www.ietf.org/shadow.html. 40 This Internet-Draft will expire on January 9, 2008. 42 Copyright Notice 44 Copyright (C) The IETF Trust (2007). 46 Abstract 48 Session Traversal Utilities for NAT (STUN) is a protocol that serves 49 as a tool for other protocols in dealing with NAT traversal. It can 50 be used by an endpoint to determine the IP address and port allocated 51 to it by a NAT. It can also be used to check connectivity between 52 two endpoints, and as a keep-alive protocol to maintain NAT bindings. 53 STUN works with many existing NATs, and does not require any special 54 behavior from them. 56 STUN is not a NAT traversal solution by itself. Rather, it is a tool 57 to be used in the context of a NAT traversal solution. This is an 58 important change from the previous version of this specification (RFC 59 3489), which presented STUN as a complete solution. 61 This document obsoletes RFC 3489. 63 Table of Contents 65 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 66 2. Evolution from RFC 3489 . . . . . . . . . . . . . . . . . . . 4 67 3. Overview of Operation . . . . . . . . . . . . . . . . . . . . 5 68 4. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 8 69 5. Definitions . . . . . . . . . . . . . . . . . . . . . . . . . 9 70 6. STUN Message Structure . . . . . . . . . . . . . . . . . . . . 10 71 7. Base Protocol Procedures . . . . . . . . . . . . . . . . . . . 12 72 7.1. Forming a Request or an Indication . . . . . . . . . . . 12 73 7.2. Sending the Request or Indication . . . . . . . . . . . . 13 74 7.2.1. Sending over UDP . . . . . . . . . . . . . . . . . . . 13 75 7.2.2. Sending over TCP or TLS-over-TCP . . . . . . . . . . . 14 76 7.3. Receiving a STUN Message . . . . . . . . . . . . . . . . 15 77 7.3.1. Processing a Request . . . . . . . . . . . . . . . . . 16 78 7.3.1.1. Forming a Success or Error Response . . . . . . . 17 79 7.3.1.2. Sending the Success or Error Response . . . . . . 17 80 7.3.2. Processing an Indication . . . . . . . . . . . . . . . 18 81 7.3.3. Processing a Success Response . . . . . . . . . . . . 18 82 7.3.4. Processing an Error Response . . . . . . . . . . . . . 18 83 8. FINGERPRINT Mechanism . . . . . . . . . . . . . . . . . . . . 19 84 9. DNS Discovery of a Server . . . . . . . . . . . . . . . . . . 19 85 10. Authentication and Message-Integrity Mechanisms . . . . . . . 20 86 10.1. Short-Term Credential Mechanism . . . . . . . . . . . . . 21 87 10.1.1. Forming a Request or Indication . . . . . . . . . . . 21 88 10.1.2. Receiving a Request or Indication . . . . . . . . . . 21 89 10.1.3. Receiving a Response . . . . . . . . . . . . . . . . . 22 90 10.2. Long-term Credential Mechanism . . . . . . . . . . . . . 23 91 10.2.1. Forming a Request . . . . . . . . . . . . . . . . . . 24 92 10.2.1.1. First Request . . . . . . . . . . . . . . . . . . 24 93 10.2.1.2. Subsequent Requests . . . . . . . . . . . . . . . 24 94 10.2.2. Receiving a Request . . . . . . . . . . . . . . . . . 24 95 10.2.3. Receiving a Response . . . . . . . . . . . . . . . . . 25 97 11. ALTERNATE-SERVER Mechanism . . . . . . . . . . . . . . . . . . 26 98 12. Backwards Compatibility with RFC 3489 . . . . . . . . . . . . 26 99 12.1. Changes to Client Processing . . . . . . . . . . . . . . 27 100 12.2. Changes to Server Processing . . . . . . . . . . . . . . 27 101 13. STUN Usages . . . . . . . . . . . . . . . . . . . . . . . . . 28 102 14. STUN Attributes . . . . . . . . . . . . . . . . . . . . . . . 29 103 14.1. MAPPED-ADDRESS . . . . . . . . . . . . . . . . . . . . . 30 104 14.2. XOR-MAPPED-ADDRESS . . . . . . . . . . . . . . . . . . . 31 105 14.3. USERNAME . . . . . . . . . . . . . . . . . . . . . . . . 32 106 14.4. MESSAGE-INTEGRITY . . . . . . . . . . . . . . . . . . . . 32 107 14.5. FINGERPRINT . . . . . . . . . . . . . . . . . . . . . . . 33 108 14.6. ERROR-CODE . . . . . . . . . . . . . . . . . . . . . . . 33 109 14.7. REALM . . . . . . . . . . . . . . . . . . . . . . . . . . 34 110 14.8. NONCE . . . . . . . . . . . . . . . . . . . . . . . . . . 35 111 14.9. UNKNOWN-ATTRIBUTES . . . . . . . . . . . . . . . . . . . 35 112 14.10. SERVER . . . . . . . . . . . . . . . . . . . . . . . . . 35 113 14.11. ALTERNATE-SERVER . . . . . . . . . . . . . . . . . . . . 36 114 15. Security Considerations . . . . . . . . . . . . . . . . . . . 36 115 15.1. Attacks against the Protocol . . . . . . . . . . . . . . 36 116 15.1.1. Outside Attacks . . . . . . . . . . . . . . . . . . . 36 117 15.1.2. Inside Attacks . . . . . . . . . . . . . . . . . . . . 36 118 15.2. Attacks Affecting the Usage . . . . . . . . . . . . . . . 36 119 15.2.1. Attack I: DDoS Against a Target . . . . . . . . . . . 37 120 15.2.2. Attack II: Silencing a Client . . . . . . . . . . . . 37 121 15.2.3. Attack III: Assuming the Identity of a Client . . . . 37 122 15.2.4. Attack IV: Eavesdropping . . . . . . . . . . . . . . . 38 123 15.3. Hash Agility Plan . . . . . . . . . . . . . . . . . . . . 38 124 16. IAB Considerations . . . . . . . . . . . . . . . . . . . . . . 38 125 17. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 39 126 17.1. STUN Methods Registry . . . . . . . . . . . . . . . . . . 39 127 17.2. STUN Attribute Registry . . . . . . . . . . . . . . . . . 39 128 17.3. STUN Error Code Registry . . . . . . . . . . . . . . . . 40 129 18. Changes Since RFC 3489 . . . . . . . . . . . . . . . . . . . . 40 130 19. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 42 131 20. References . . . . . . . . . . . . . . . . . . . . . . . . . . 42 132 20.1. Normative References . . . . . . . . . . . . . . . . . . 42 133 20.2. Informational References . . . . . . . . . . . . . . . . 43 134 Appendix A. C Snippet to Determine STUN Message Types . . . . . . 44 135 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 44 136 Intellectual Property and Copyright Statements . . . . . . . . . . 46 138 1. Introduction 140 The protocol defined in this specification, Session Traversal 141 Utilities for NAT, provides a toolkit of functions for dealing with 142 NATs. It provides a means for an endpoint to determine the IP 143 address and port allocated by a NAT that corresponds to its private 144 IP address and port. It also provides a way for an endpoint to keep 145 a NAT binding alive. With some extensions, the protocol can be used 146 to do connectivity checks between two endpoints 147 [I-D.ietf-mmusic-ice], or to relay packets between two endpoints 148 [I-D.ietf-behave-turn]. 150 In keeping with its toolkit nature, this specification defines an 151 extensible packet format, defines operation over several transport 152 protocols, and provides for two forms of authentication. 154 STUN is intended to be used in context of one or more NAT traversal 155 solutions. These solutions are known as STUN usages. Each usage 156 describes how STUN is utilized to achieve the NAT traversal solution. 157 Typically, a usage indicates when STUN messages get sent, which 158 optional attributes to include, what server is used, and what 159 authentication mechanism is to be used. Interactive Connectivity 160 Establishment (ICE) [I-D.ietf-mmusic-ice] is one usage of ICE. SIP 161 Outbound [I-D.ietf-sip-outbound] is another usage of ICE. In some 162 cases, a usage will require extensions to STUN. A STUN extension can 163 be in the form of new methods, attributes, or error response codes. 164 More information on STUN usages can be found in Section 13. 166 2. Evolution from RFC 3489 168 STUN was originally defined in RFC 3489 [RFC3489]. That 169 specification, sometimes referred to as "classic STUN", represented 170 itself as a complete solution to the NAT traversal problem. In that 171 solution, a client would discover whether it was behind a NAT, 172 determine its NAT type, discover its IP address and port on the 173 public side of the outermost NAT, and then utilize that IP address 174 and port within the body of protocols, such as the Session Initiation 175 Protocol (SIP) [RFC3261]. However, experience since the publication 176 of RFC 3489 has found that classic STUN simply does not work 177 sufficiently well to be a deployable solution. The address and port 178 learned through classic STUN are sometimes usable for communications 179 with a peer, and sometimes not. Classic STUN provided no way to 180 discover whether it would, in fact, work or not, and it provided no 181 remedy in cases where it did not. Furthermore, classic STUN's 182 algorithm for classification of NAT types was found to be faulty, as 183 many NATs did not fit cleanly into the types defined there. Classic 184 STUN also had security vulnerabilities which required an extremely 185 complicated mechanism to address, and despite the complexity of the 186 mechanism, were not fully remedied. 188 For these reasons, this specification obsoletes RFC 3489, and instead 189 describes STUN as a tool that is utilized as part of a complete NAT 190 traversal solution. ICE is a complete NAT traversal solutions for 191 protocols based on the offer/answer [RFC3264] methodology, such as 192 SIP. SIP Outbound is a complete solution for traversal of SIP 193 signaling, and it uses STUN in a very different way. Though it is 194 possible that a protocol may be able to use STUN by itself (classic 195 STUN) as a traversal solution, such usage is not described here and 196 is strongly discouraged for the reasons described above. 198 The on-the-wire protocol described here is changed only slightly from 199 classic STUN. The protocol now runs over TCP in addition to UDP. 200 Extensibility was added to the protocol in a more structured way. A 201 magic-cookie mechanism for demultiplexing STUN with application 202 protocols was added by stealing 32 bits from the 128 bit transaction 203 ID defined in RFC 3489, allowing the change to be backwards 204 compatible. Mapped addresses are encoded using a new exclusive-or 205 format. There are other, more minor changes. See Section 18 for a 206 more complete listing. 208 Due to the change in scope, STUN has also been renamed from "Simple 209 Traversal of UDP Through NAT" to "Session Traversal Utilities for 210 NAT". The acronym remains STUN, which is all anyone ever remembers 211 anyway. 213 3. Overview of Operation 215 This section is descriptive only. 217 /--------\ 218 // STUN \\ 219 | Agent | 220 \\ (server) // 221 \--------/ 223 +----------------+ Public Internet 224 ................| NAT 2 |....................... 225 +----------------+ 227 +----------------+ Private NET 2 228 ................| NAT 1 |....................... 229 +----------------+ 231 /--------\ 232 // STUN \\ 233 | Agent | 234 \\ (client) // Private NET 1 235 \--------/ 237 Figure 1: One possible STUN Configuration 239 One possible STUN configuration is shown in Figure 1. In this 240 configuration, there are two entities (called STUN agents) that 241 implement the STUN protocol. The lower agent in the figure is 242 connected to private network 1. This network connects to private 243 network 2 through NAT 1. Private network 2 connects to the public 244 Internet through NAT 2. The upper agent in the figure resides on the 245 public Internet. 247 STUN is a client-server protocol. It supports two types of 248 transactions. One is a request/response transaction in which a 249 client sends a request to a server, and the server returns a 250 response. The second is an indication transaction in which a client 251 sends an indication to the server and the server does not respond. 252 Both types of transactions include a transaction ID, which is a 253 randomly selected 96-bit number. For request/response transactions, 254 this transaction ID allows the client to associate the response with 255 the request that generated it; for indications, this simply serves as 256 a debugging aid. 258 All STUN messages start with a fixed header that includes a method, a 259 class, and the transaction ID. The method indicates which of the 260 various requests or indications this is; this specification defines 261 just one method, Binding, but other methods are expected to be 262 defined in other documents. The class indicates whether this is a 263 request, a (success) response, an error response, or an indication. 264 Following the fixed header comes zero or more attributes, which are 265 type-length-value extensions that convey additional information for 266 the specific message. 268 This document defines a single method called Binding. The Binding 269 method can be used either in request/response transactions or in 270 indication transactions. When used in request/response transactions, 271 the Binding method can be used to determine the particular "binding" 272 a NAT has allocated to a STUN client. When used in either request/ 273 response or in indication transactions, the Binding method can also 274 be used to keep these "bindings" alive. 276 In the Binding request/response transaction, a Binding Request is 277 sent from a STUN client to a STUN server. When the Binding Request 278 arrives at the STUN server, it may have passed through one or more 279 NATs between the STUN client and the STUN server (in Figure 1, there 280 were two such NATs). As the Binding Request message passes through a 281 NAT, the NAT will modify the source transport address (that is, the 282 source IP address and the source port) of the packet. As a result, 283 the source transport address of the request received by the server 284 will be the public IP address and port created by the NAT closest to 285 the server. This is called a reflexive transport address. The STUN 286 server copies that source transport address into an XOR-MAPPED- 287 ADDRESS attribute in the STUN Binding Response and sends the Binding 288 Response back to the the STUN client. As this packet passes back 289 through a NAT, the NAT will modify the destination transport address, 290 but the transport address in the XOR-MAPPED-ADDRESS attribute within 291 the body of the STUN response will remain untouched. In this way, 292 the client can learn its reflexive transport address allocated by the 293 outermost NAT with respect to the STUN server. 295 In some usages, STUN must be multiplexed with other protocols (e.g., 296 [I-D.ietf-mmusic-ice], [I-D.ietf-sip-outbound]). In these usages, 297 there must be a way to inspect a packet and determine if it is a STUN 298 packet or not. STUN provides two fields in the STUN header with 299 fixed values that can be used for this purpose. If this is not 300 sufficient, then STUN packets can also contain a FINGERPRINT value 301 which can further be used to distinguish the packets. 303 STUN has optional mechanisms for providing authentication and message 304 integrity when required. These mechanisms revolve around the use of 305 a username, password, and message-integrity value. Two of these 306 mechanisms, the long-term credential mechanism and the short-term 307 credential mechanism, are defined in this specification. Each usage 308 specifies the mechanisms allowed with that usage. 310 In the short-term credential mechanism, the client and the server 311 exchange a username and password through some out-of-band method 312 prior to the STUN exchange. For example, in the ICE usage 313 [I-D.ietf-mmusic-ice] the two endpoints use out-of-band signaling to 314 exchange a username and password. The client then includes the 315 username and a message-integrity value in the request message, where 316 the message-integrity value is computed as a cryptographic hash of 317 the message contents and the password. If the server replies with a 318 success response, then the response will include a message-integrity 319 value (computed using the same username and password), but the 320 username is not included. Error responses are not message-integrity 321 protected. 323 In the long-term credential mechanism, the client and server share a 324 pre-provisioned username and password and perform a digest challenge/ 325 response exchange inspired by (but differing in details) to the one 326 defined for HTTP [RFC2617]. Initially, the client sends a request 327 message (e.g., a Binding Request) without any username or message- 328 integrity value included. The server replies with an error response 329 indicating that the request must be authenticated. This error 330 response includes a realm value and a nonce value. The client then 331 uses the realm value to help it select a username and password (for 332 example, the client might have a number of username and password 333 combinations stored, each one keyed by a different realm value). The 334 client then retries the request, this time including the realm, the 335 username, the nonce, and a message-integrity value in the request, 336 where the message-integrity value is computed as a cryptographic hash 337 of the message contents and the password. The nonce is provided by 338 the server and merely echoed by the client into the request. It is 339 chosen by the server such that it encodes information about the 340 client, the time-of-day, or other parameters. In this way, if an 341 attacker should attempt to replay the request, the server would find 342 the nonce invalid, and then reject the request. If the server 343 replies with a success response, then the response will include a 344 message-integrity value (computed using the same username and 345 password), but realm, username, and nonce are not included. Error 346 responses are not message-integrity protected. If the client has 347 further requests to send, it can try to reuse the same username, 348 realm, and nonce values. If the server does not accept them, it will 349 reply with an error response giving a realm and nonce value again. 351 4. Terminology 353 In this document, the key words "MUST", "MUST NOT", "REQUIRED", 354 "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", 355 and "OPTIONAL" are to be interpreted as described in BCP 14, RFC 2119 356 [RFC2119] and indicate requirement levels for compliant STUN 357 implementations. 359 5. Definitions 361 STUN Agent: An entity that implements the STUN protocol. Agents can 362 act as STUN clients for some transactions and as STUN servers for 363 other transactions. 365 STUN Client: A logical role in the STUN protocol. A STUN client 366 sends STUN requests or STUN indications, and receives STUN 367 responses. The term "STUN client" is also used colloquially to 368 refer to a STUN agent that only acts as a STUN client. 370 STUN Server: A logical role in the STUN protocol. A STUN server 371 receives STUN requests or STUN indications and sends STUN 372 responses. The term "STUN server" is also used colloquially to 373 refer to a STUN agent that only acts as a STUN server. 375 Transport Address: The combination of an IP address and port number 376 (such as a UDP or TCP port number). 378 Reflexive Transport Address: A transport address learned by a client 379 that identifies that client as seen by another host on an IP 380 network, typically a STUN server. When there is an intervening 381 NAT between the client and the other host, the reflexive transport 382 address represents the mapped address allocated to the client on 383 the public side of the NAT. Reflexive transport addresses are 384 learned from the mapped address attribute (MAPPED-ADDRESS or XOR- 385 MAPPED-ADDRESS) in STUN responses. 387 Mapped Address: Same meaning as Reflexive Address. This term is 388 retained only for for historic reasons and due to the naming of 389 the MAPPED-ADDRESS and XOR-MAPPED-ADDRESS attributes. 391 Long Term Credential: A username and associated password that 392 represent a shared secret between client and server. Long term 393 credentials are generally granted to the client when a subscriber 394 enrolls in a service and persist until the subscriber leaves the 395 service or explicitly changes the credential. 397 Long Term Password: The password from a long term credential. 399 Short Term Credential: A temporary username and associated password 400 which represent a shared secret between client and server. Short 401 term credentials are obtained through some kind of protocol 402 mechanism between the client server, preceding the STUN exchange. 403 A short term credential has an explicit temporal scope, which may 404 be based on a specific amount of time (such as 5 minutes) or on an 405 event (such as termination of a SIP dialog). The specific scope 406 of a short term credential is defined by the application usage. 408 Short Term Password: The password component of a short term 409 credential. 411 STUN Indication: A STUN message that does not receive a response 413 Attribute: The STUN term for a Type-Length-Value (TLV) object that 414 can be added to a STUN message. Attributes are divided into two 415 types: comprehension-required and comprehension-optional. STUN 416 agents can safely ignore comprehension-optional attributes they 417 don't understand, but cannot successfully process a message if it 418 contains comprehension-required attributes that are not 419 understood. 421 RTO: Retransmission TimeOut 423 6. STUN Message Structure 425 STUN messages are encoded in binary using network-oriented format 426 (most significant byte or octet first, also commonly known as big- 427 endian). The transmission order is described in detail in Appendix B 428 of RFC791 [RFC0791]. Unless otherwise noted, numeric constants are 429 in decimal (base 10). 431 All STUN messages MUST start with a 20-byte header followed by zero 432 or more Attributes. The STUN header contains a STUN message type, 433 magic cookie, transaction ID, and message length. 435 0 1 2 3 436 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 437 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 438 |0 0| STUN Message Type | Message Length | 439 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 440 | Magic Cookie | 441 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 442 | | 443 | Transaction ID (96 bits) | 444 | | 445 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 447 Figure 2: Format of STUN Message Header 449 The most significant two bits of every STUN message MUST be zeroes. 450 This can be used to differentiate STUN packets from other protocols 451 when STUN is multiplexed with other protocols on the same port. 453 The message type defines the message class (request, success 454 response, failure response, or indication) and the message method 455 (the primary function) of the STUN message. Although there are four 456 message classes, there are only two types of transactions in STUN: 457 request/response transactions (which consist of a request message and 458 a response message), and indication transactions (which consists a 459 single indication message). Response classes are split into error 460 and success responses to aid in quickly processing the STUN message. 462 The message type field is decomposed further into the following 463 structure: 465 +--+--+-+-+-+-+-+-+-+-+-+-+-+-+ 466 |M |M |M|M|M|C|M|M|M|C|M|M|M|M| 467 |11|10|9|8|7|1|6|5|4|0|3|2|1|0| 468 +--+--+-+-+-+-+-+-+-+-+-+-+-+-+ 470 Figure 3: Format of STUN Message Type Field 472 Here the bits in the message type field are shown as most-significant 473 (M11) through least-significant (M0). M11 through M0 represent a 12- 474 bit encoding of the method. C1 and C0 represent a 2 bit encoding of 475 the class. A class of 0b00 is a Request, a class of 0b01 is an 476 indication, a class of 0b10 is a success response, and a class of 477 0b11 is an error response. This specification defines a single 478 method, Binding. The method and class are orthogonal, so that four 479 each method, a request, success response, error response and 480 indication are defined for that method. 482 For example, a Binding Request has class=0b00 (request) and 483 method=0b000000000001 (Binding), and is encoded into the first 16 484 bits as 0x0001. A Binding response has class=0b10 (success response) 485 and method=0b000000000001, and is encoded into the first 16 bits as 486 0x0101. 488 Note: This unfortunate encoding is due to assignment of values in 489 [RFC3489] which did not consider encoding Indications, Success, 490 and Errors using bit fields. 492 The magic cookie field MUST contain the fixed value 0x2112A442 in 493 network byte order. In RFC 3489 [RFC3489], this field was part of 494 the transaction ID; placing the magic cookie in this location allows 495 a server to detect if the client will understand certain attributes 496 that were added in this revised specification. In addition, it aids 497 in distinguishing STUN packets from packets of other protocols when 498 STUN is multiplexed with those other protocols on the same port. 500 The transaction ID is a 96 bit identifier, used to uniquely identify 501 STUN transactions. The transaction ID is chosen by the STUN client. 502 It primarily serves to correlate requests with responses, though it 503 also plays a small role in helping to prevent certain types of 504 attacks. As such, the transaction ID MUST be uniformly and randomly 505 chosen from the interval 0 .. 2**96-1. Resends of the same request 506 reuse the same transaction ID, but the client MUST choose a new 507 transaction ID for new transactions unless the new request is bit- 508 wise identical to the previous request and sent from the same 509 transport address to the same IP address. Success and error 510 responses MUST carry the same transaction ID as their corresponding 511 request. When an agent is acting as a STUN server and STUN client on 512 the same port, the transaction IDs in requests sent by the agent have 513 no relationship to the transaction IDs in requests received by the 514 agent. 516 The message length MUST contain the size, in bytes, of the message 517 not including the 20 byte STUN header. Since all STUN attributes are 518 padded to a multiple of four bytes, the last two bits of this field 519 are always zero. This provides another way to distinguish STUN 520 packets from packets of other protocols. 522 Following the STUN fixed portion of the header are zero or more 523 attributes. Each attribute is TLV (type-length-value) encoded. The 524 details of the encoding, and of the attributes themselves is given in 525 Section 14. 527 7. Base Protocol Procedures 529 This section defines the base procedures of the STUN protocol. It 530 describes how messages are formed, how they are sent, and how they 531 are processed when they are received. It also defines the detailed 532 processing of the Binding method. Other sections in this document 533 describe optional procedures that a usage may elect to use in certain 534 situations. Other documents may define other extensions to STUN, by 535 adding new methods, new attributes, or new error response codes. 537 7.1. Forming a Request or an Indication 539 When formulating a request or indication message, the client MUST 540 follow the rules in Section 6 when creating the header. In addition, 541 the message class MUST be either "Request" or "Indication" (as 542 appropriate), and the method must be either Binding or some method 543 defined in another document. 545 The client then adds any attributes specified by the method or the 546 usage. For example, some usages may specify that the client use an 547 authentication method (Section 10) or the FINGERPRINT attribute 548 (Section 8). 550 For the Binding method with no authentication, no attributes are 551 required unless the usage specifies otherwise. 553 7.2. Sending the Request or Indication 555 The client then sends the request to the server. This document 556 specifies how to send STUN messages over UDP, TCP, or TLS-over-TCP; 557 other transport protocols may be added in the future. The STUN usage 558 must specify which transport protocol is used, and how the client 559 determines the IP address and port of the server. Section 9 560 describes a DNS-based method of determining the IP address and port 561 of a server which a usage may elect to use. 563 At any time, a client MAY have multiple outstanding STUN requests 564 with the same STUN server (that is, multiple transactions in 565 progress, with different transaction ids). 567 7.2.1. Sending over UDP 569 When running STUN over UDP it is possible that the STUN message might 570 be dropped by the network. Reliability of STUN request/response 571 transactions is accomplished through retransmissions of the request 572 message by the client application itself. STUN indications are not 573 retransmitted; thus indication transactions over UDP are not 574 reliable. 576 A client SHOULD retransmit a STUN request message starting with an 577 interval of RTO ("Retransmission TimeOut"), doubling after each 578 retransmission. The RTO is an estimate of the round-trip-time, and 579 is computed as described in RFC 2988 [RFC2988], with two exceptions. 580 First, the initial value for RTO SHOULD be configurable (rather than 581 the 3s recommended in RFC 2988). In fixed- line access links, a 582 value of 100ms is RECOMMENDED. Secondly, the value of RTO MUST NOT 583 be rounded up to the nearest second. Rather, a 1ms accuracy MUST be 584 maintained. As with TCP, the usage of Karn's algorithm is 585 RECOMMENDED. When applied to STUN, it means that RTT estimates 586 SHOULD NOT be computed from STUN transactions which result in the 587 retransmission of a request. 589 The value for RTO SHOULD be cached by an client after the completion 590 of the transaction, and used as the starting value for RTO for the 591 next transaction to the same server (based on equality of IP 592 address). The value SHOULD be considered stale and discarded after 593 10 minutes. 595 Retransmissions continue until a response is received, or until a 596 total of 7 requests have been sent. If, after the last request, a 597 duration equal to 16 times the RTO has passed without a response, the 598 client SHOULD consider the transaction to have failed. A STUN 599 transaction over UDP is also considered failed if there has been a 600 transport failure of some sort, such as a fatal ICMP error. For 601 example, assuming an RTO of 100ms, requests would be sent at times 602 0ms, 100ms, 300ms, 700ms, 1500ms, 3100ms, and 6300ms. If the client 603 has not received a response after 7900ms, the client will consider 604 the transaction to have timed out. 606 7.2.2. Sending over TCP or TLS-over-TCP 608 For TCP and TLS-over-TCP, the client opens a TCP connection to the 609 server. 611 In some usage of STUN, STUN is sent as the only protocol over the TCP 612 connection. In this case, it can be sent without the aid of any 613 additional framing or demultiplexing. In other usages, or with other 614 extensions, it may be multiplexed with other data over a TCP 615 connection. In that case, STUN MUST be run ontop of some kind of 616 framing protocol, specified by the usage or extension, which allows 617 for the agent to extract complete STUN messages and complete 618 application layer messages. 620 For TLS-over-TCP, the TLS_RSA_WITH_AES_128_CBC_SHA ciphersuite MUST 621 be supported at a minimum. Implementations MAY also support any 622 other ciphersuite. When it receives the TLS Certificate message, the 623 client SHOULD verify the certificate and inspect the site identified 624 by the certificate. If the certificate is invalid, revoked, or if it 625 does not identify the appropriate party, the client MUST NOT send the 626 STUN message or otherwise proceed with the STUN transaction. The 627 client MUST verify the identity of the server. To do that, it 628 follows the identification procedures defined in Section 3.1 of RFC 629 2818 [RFC2818]. Those procedures assume the client is dereferencing 630 a URI. For purposes of usage with this specification, the client 631 treats the domain name or IP address used in Section 8.1 as the host 632 portion of the URI that has been dereferenced. If DNS was not used, 633 the client MUST be configured with a set of authorized domains whose 634 certificates will be accepted. 636 Reliability of STUN over TCP and TLS-over-TCP is handled by TCP 637 itself, and there are no retransmissions at the STUN protocol level. 638 However, for a request/response transaction, if the client has not 639 received a response after 7900ms, it considers the transaction to 640 have timed out. This value has been chosen to equalize the TCP and 641 UDP timeouts for the default initial RTO. 643 In addition, if the client is unable to establish the TCP connection, 644 or the TCP connection is reset or fails before a response is 645 received, any request/response transaction in progress is considered 646 to have failed 648 The client MAY send multiple transactions over a single TCP (or TLS- 649 over-TCP) connection, and it MAY send another request before 650 receiving a response to the previous. The client SHOULD keep the 651 connection open until it 653 o has no further STUN requests or indications to send over that 654 connection, and; 656 o has no plans to use any resources (such as a mapped address 657 (MAPPED-ADDRESS or XOR-MAPPED-ADDRESS) or relayed address 658 [I-D.ietf-behave-turn]) that were learned though STUN requests 659 sent over that connection, and; 661 o if multiplexing other application protocols over that port, has 662 finished using that other application, and; 664 o if using that learned port with a remote peer, has established 665 communications with that remote peer, as is required by some TCP 666 NAT traversal techniques (e.g., [I-D.ietf-mmusic-ice-tcp]). 668 At the server end, the server SHOULD keep the connection open, and 669 let the client close it. If a server becomes overloaded and needs to 670 close connections to free up resources, it SHOULD close an existing 671 connection rather than reject new connection requests. The server 672 SHOULD NOT close a connection if a request was received over that 673 connection for which a response was not sent. A server MUST NOT ever 674 open a connection back towards the client in order to send a 675 response. 677 7.3. Receiving a STUN Message 679 This section specifies the processing of a STUN message. The 680 processing specified here is for STUN messages as defined in this 681 specification; additional rules for backwards compatibility are 682 defined in in Section 12. Those additional procedures are optional, 683 and usages can elect to utilize them. First, a set of processing 684 operations are applied that are independent of the class. This is 685 followed by class-specific processing, described in the subsections 686 which follow. 688 When a STUN agent receives a STUN message, it first checks that the 689 message obeys the rules of Section 6. It checks that the first two 690 bits are 0, that the magic cookie field has the correct value, that 691 the message length is sensible, and that the method value is a 692 supported method. If the message-class is Success Response or Error 693 Response, the agent checks that the transaction ID matches a 694 transaction that is still in progress. If the FINGERPRINT extension 695 is being used, the agent checks that the FINGERPRINT attribute is 696 present and contains the correct value. If any errors are detected, 697 the message is silently discarded. In the case when STUN is being 698 multiplexed with another protocol, an error may indicate that this is 699 not really a STUN message; in this case, the agent should try to 700 parse the message as a different protocol. 702 The STUN agent then does any checks that are required by a 703 authentication mechanism that the usage has specified (see 704 Section 10. 706 Once the authentication checks are done, the STUN agent checks for 707 unknown attributes and known-but-unexpected attributes in the 708 message. Unknown comprehension-optional attributes MUST be ignored 709 by the agent. Known-but-unexpected attributes SHOULD be ignored by 710 the agent. Unknown comprehension-required attributes cause 711 processing that depends on the message-class and is described below. 713 At this point, further processing depends on the message class of the 714 request. 716 7.3.1. Processing a Request 718 If the request contains one or more unknown comprehension-required 719 attributes, the server replies with an error response with an error 720 code of 420 Unknown attributes, and includes an UNKNOWN-ATTRIBUTES 721 attribute in the response that lists the unknown comprehension- 722 required attributes. 724 The server then does any additional checking that the method or the 725 specific usage requires. If all the checks succeed, the server 726 formulates a success response as described below. 728 If the request uses UDP transport and is a retransmission of a 729 request for which the server has already generated a success response 730 within the last 10 seconds, the server MUST retransmit the same 731 success response. One way for a server to do this is to remember all 732 transaction IDs received over UDP and their corresponding responses 733 in the last 10 seconds. Another way is to reprocess the request and 734 recompute the response. The latter technique MUST only be applied to 735 requests which are idempotent and result in the same success response 736 for the same request. The Binding method is considered to idempotent 737 in this way (even though certain rare network events could cause the 738 reflexive transport address value to change). Extensions to STUN 739 SHOULD state whether their request types have this property or not. 741 7.3.1.1. Forming a Success or Error Response 743 When forming the response (success or error), the server follows the 744 rules of section 6. The method of the response is the same as that 745 of the request, and the message class is either "Success Response" or 746 "Error Response". 748 For an error response, the server MUST add an ERROR-CODE attribute 749 containing the error code specified in the processing above. The 750 reason phrase is not fixed, but SHOULD be something suitable for the 751 error code. For certain errors, additional attributes are added to 752 the message. These attributes are spelled out in the description 753 where the error code is specified. For example, for an error code of 754 420 Unknown Attribute, the server MUST include an UNKNOWN-ATTRIBUTES 755 attribute. Certain authentication errors also cause attributes to be 756 added (see Section 10). Extensions may define other errors and/or 757 additional attributes to add in error cases. 759 If the server authenticated the request using an authentication 760 mechanism, then the server SHOULD add the appropriate authentication 761 attributes to the response (see Section 10). 763 The server also adds any attributes required by the specific method 764 or usage. In addition, the server SHOULD add a SERVER attribute to 765 the message. 767 For the Binding method, no additional checking is required unless the 768 usage specifies otherwise. When forming the success response, the 769 server adds a XOR-MAPPED-ADDRESS attribute to the response, where the 770 contents of the attribute are the source transport address of the 771 request message. For UDP, this is the source IP address and source 772 UDP port of the request message. For TCP and TLS-over-TCP, this is 773 the source IP address and source TCP port of the TCP connection as 774 seen by the server. 776 7.3.1.2. Sending the Success or Error Response 778 The response (success or error) is sent over the same transport as 779 the request was received on. If the request was received over UDP, 780 the destination IP address and port of the response is the source IP 781 address and port of the received request message, and the source IP 782 address and port of the response is equal to the destination IP 783 address and port of the received request message. If the request was 784 received over TCP or TLS-over-TCP, the response is sent back on the 785 same TCP connection as the request was received on. 787 7.3.2. Processing an Indication 789 If the indication contains unknown comprehension-required attributes, 790 the indication is discarded and processing ceases. 792 The server then does any additional checking that the method or the 793 specific usage requires. If all the checks succeed, the server then 794 processes the indication. No response is generated for an 795 indication. 797 For the Binding method, no additional checking or processing is 798 required, unless the usage specifies otherwise. The mere receipt of 799 the message by the server has refreshed the "bindings" in the 800 intervening NATs. 802 Since indications are not re-transmitted over UDP (unlike requests), 803 there is no need to handle re-transmissions of indications at the 804 server. 806 7.3.3. Processing a Success Response 808 If the success response contains unknown comprehension-required 809 attributes, the response is discarded and the transaction is 810 considered to have failed. 812 The client then does any additional checking that the method or the 813 specific usage requires. If all the checks succeed, the client then 814 processes the success response. 816 For the Binding method, the client checks that the XOR-MAPPED-ADDRESS 817 attribute is present in the response. The client checks the address 818 family specified. If it is an unsupported address family, the 819 attribute SHOULD be ignored. If it is an unexpected but supported 820 address family (for example, the Binding transaction was sent over 821 IPv4, but the address family specified is IPv6), then the client MAY 822 accept and use the value. 824 7.3.4. Processing an Error Response 826 If the error response contains unknown comprehension-required 827 attributes, or if the error response does not contain an ERROR-CODE 828 attribute, then the transaction is simply considered to have failed. 830 The client then does any processing specified by the authentication 831 mechanism (see Section 10). This may result in a new transaction 832 attempt. 834 The processing at this point depends on the error-code, the method, 835 and the usage; the following are the default rules: 837 o If the error code is 300 through 399, the client SHOULD consider 838 the transaction as failed unless the ALTERNATE-SERVER extension is 839 being used. See Section 11. 841 o If the error code is 400 through 499, the client declares the 842 transaction failed; in the case of 420, the response should 843 contain a UNKNOWN-ATTRIBUTES attribute that gives additional 844 information. 846 o If the error code is 500 through 599, the client MAY resend the 847 request; clients that do so MUST limit the number of times they do 848 this. Any other error code causes the client to consider the 849 transaction failed. 851 8. FINGERPRINT Mechanism 853 This section describes an optional mechanism for STUN that aids in 854 distinguishing STUN messages from packets of other protocols when the 855 two are multiplexed on the same transport address. This mechanism is 856 optional, and a STUN usage must describe if and when it is used. 858 In some usages, STUN messages are multiplexed on the same transport 859 address as other protocols, such as RTP. In order to apply the 860 processing described in Section 7, STUN messages must first be 861 separated from the application packets. Section 6 describes three 862 fixed fields in the STUN header that can be used for this purpose. 863 However, in some cases, these three fixed fields may not be 864 sufficient. 866 When the FINGERPRINT extension is used, an agent includes the 867 FINGERPRINT attribute in messages it sends to another agent. 868 Section 14.5 describes the placement and value of this attribute. 869 When the agent receives what it believes is a STUN message, then, in 870 addition to other basic checks, the agent also checks that the 871 message contains a FINGERPRINT attribute and that the attribute 872 contains the correct value (see Section 7.3. This additional check 873 helps the agent detect messages of other protocols that might 874 otherwise seem to be STUN messages. 876 9. DNS Discovery of a Server 878 This section describes an optional procedure for STUN that allows a 879 client to use DNS to determine the IP address and port of a server. 880 A STUN usage must describe if and when this extension is used. To 881 use this procedure, the client must have a domain name and a service 882 name; the usage must also describe how the client obtains these. 884 When a client wishes to locate a STUN server in the public Internet 885 that accepts Binding Request/Response transactions, the SRV service 886 name is "stun". STUN usages MAY define additional DNS SRV service 887 names. 889 The domain name is resolved to a transport address using the SRV 890 procedures specified in [RFC2782]. The DNS SRV service name is the 891 service name provided as input to this procedure. The protocol in 892 the SRV lookup is the transport protocol the client will run STUN 893 over: "udp" for UDP, "tcp" for TCP, and "tls" for TLS-over-TCP. If, 894 in the future, additional SRV records are defined for TLS over other 895 transport protocols, those will need to utilize an SRV transport 896 token of the form "tls-foo" for transport protocol "foo". 898 The procedures of RFC 2782 are followed to determine the server to 899 contact. RFC 2782 spells out the details of how a set of SRV records 900 are sorted and then tried. However, RFC2782 only states that the 901 client should "try to connect to the (protocol, address, service)" 902 without giving any details on what happens in the event of failure. 903 When following these procedures, if the STUN transaction times out 904 without receipt of a response, the client SHOULD retry the request to 905 the next server in the list of servers from the DNS SRV response. 906 Such a retry is only possible for request/response transmissions, 907 since indication transactions generate no response or timeout. 909 The default port for STUN requests is 3478, for both TCP and UDP. 910 Administrators SHOULD use this port in their SRV records for UDP and 911 TCP, but MAY use others. There is no default port for STUN over TLS, 912 however a STUN server SHOULD use a port number for TLS different from 913 3478 so that the server can determine whether the first message it 914 will receive after the TCP connection is set up, is a STUN message or 915 a TLS message. 917 If no SRV records were found, the client performs an A or AAAA record 918 lookup of the domain name. The result will be a list of IP 919 addresses, each of which can be contacted at the default port using 920 UDP or TCP, independent of the STUN usage. For usages that require 921 TLS, lack of SRV records is equivalent to a failure of the 922 transaction, since the request or indication MUST NOT be sent unless 923 SRV records provided a transport address specifically for TLS. 925 10. Authentication and Message-Integrity Mechanisms 927 This section defines two mechanisms for STUN that a client and server 928 can use to provide authentication and message-integrity; these two 929 mechanisms are known as the short-term credential mechanism and the 930 long-term credential mechanism. These two mechanisms are optional, 931 and each usage must specify if and when these mechanisms are used. 932 An overview of these two mechanisms is given in . 934 Each mechanism specifies the additional processing required to use 935 that mechanism, extending the processing specified in Section 7. The 936 additional processing occurs in three different places: when forming 937 a message; when receiving a message immediately after the the basic 938 checks have been performed; and when doing the detailed processing of 939 error responses. 941 10.1. Short-Term Credential Mechanism 943 The short-term credential mechanism assumes that, prior to the STUN 944 transaction, the client and server have used some other protocol to 945 exchange a credential in the form of a username and password. This 946 credential is time-limited. The time-limit is defined by the usage. 947 As an example, in the ICE usage [I-D.ietf-mmusic-ice], the two 948 endpoints use out-of-band signaling to agree on a username and 949 password, and this username and password is applicable for the 950 duration of the media session. 952 This credential is used to form a message integrity check in each 953 request and in many responses. There is no challenge and response as 954 in the long term mechanism; consequently, replay is prevented by 955 virtue of the time-limited nature of the credential. 957 10.1.1. Forming a Request or Indication 959 For a request or indication message, the agent MUST include the 960 USERNAME and MESSAGE-INTEGRITY attributes in the message. The HMAC 961 for the MESSAGE-INTEGRITY attribute is computed as described in 962 Section 14.4. The key for the HMAC is the password. Note that the 963 password is never included in the request or indication. 965 10.1.2. Receiving a Request or Indication 967 After the agent has done the basic processing of a message, the agent 968 performs the checks listed below in order specified: 970 o If the message does not contain both a MESSAGE-INTEGRITY and a 971 USERNAME attribute: 973 * If the message is a request, the server MUST reject the request 974 with an error response. This response MUST use an error code 975 of 400. 977 * If the message is an indication, the server MUST silently 978 discard the indication. 980 o If the USERNAME does not contain a username value currently valid 981 within the server: 983 * If the message is a request, the server MUST reject the request 984 with an error response. This response MUST use an error code 985 of 401. 987 * If the message is an indication, the server MUST silently 988 discard the indication. 990 o Using the password associated with the username, compute the value 991 for the message-integrity as described in Section 14.4. If the 992 resulting value does not match the contents of the MESSAGE- 993 INTEGRITY attribute: 995 * If the message is a request, the server MUST reject the request 996 with an error response. This response MUST use an error code 997 of 431. 999 * If the message is an indication, the server MUST silently 1000 discard the indication. 1002 If these checks pass, the server continues to process the request or 1003 indication. Any response generated by the server MUST include the 1004 MESSAGE-INTEGRITY attribute, computed using the username and password 1005 utilized to authenticate the request. 1007 If any of the checks fail, the server MUST NOT include a MESSAGE- 1008 INTEGRITY or USERNAME attribute in the error response. 1010 10.1.3. Receiving a Response 1012 The processing here takes place prior to the processing in 1013 Section 7.3.3 or Section 7.3.4. 1015 The client looks for the MESSAGE-INTEGRITY attribute in the response. 1016 If present, the client computes the message integrity over the 1017 response as defined in Section 14.4, using the same password it 1018 utilized for the request. If the resulting value matches the 1019 contents of the MESSAGE-INTEGRITY attribute, the response is 1020 considered authenticated. If the value does not match, or if 1021 MESSAGE-INTEGRITY was absent, the response MUST be discarded, as if 1022 it was never received. This means that retransmits, if applicable, 1023 will continue. 1025 10.2. Long-term Credential Mechanism 1027 The long-term credential mechanism relies on a long term credential, 1028 in the form of a username and password, that are shared between 1029 client and server. The credential is considered long-term since it 1030 is assumed that it is provisioned for a user, and remains in effect 1031 until the user is no longer a subscriber of the system, or is 1032 changed. This is basically a traditional "log-in" username and 1033 password given to users. 1035 Because these usernames and passwords are expected to be valid for 1036 extended periods of time, replay prevention is provided in the form 1037 of a digest challenge. In this mechanism, the client initially sends 1038 a request, without offering any credentials or any integrity checks. 1039 The server rejects this request, providing the user a realm (used to 1040 guide the user or agent in selection of a username and password) and 1041 a nonce. The nonce provides the replay protection. It is a cookie, 1042 selected by the server, and encoded in such a way as to indicate a 1043 duration of validity or client identity from which it is valid. The 1044 client retries the request, this time including its username, the 1045 realm, and echoing the nonce provided by the server. The client also 1046 includes a message-integrity, which provides an HMAC over the entire 1047 request, including the nonce. The server validates the nonce, and 1048 checks the message-integrity. If they match, the request is 1049 authenticated. If the nonce is no longer valid, it is considered 1050 "stale", and the server rejects the request, providing a new nonce. 1052 In subsequent requests to the same server, the client reuses the 1053 nonce, username, realm and password it used previously. In this way, 1054 subsequent requests are not rejected until the nonce becomes invalid 1055 by the server, in which case the rejection provides a new nonce to 1056 the client. 1058 Note that the long-term credential mechanism cannot be used to 1059 protect indications, since indications cannot be challenged. Usages 1060 utilizing indications must either use a short-term credential, or 1061 omit authentication and message integrity for them. 1063 Since the long-term credential mechanism is susceptible to offline 1064 dictionary attacks, deployments SHOULD utilize strong passwords. 1066 For STUN servers used in conjunction with SIP servers, it is 1067 desirable to use the same credentials for authentication to the SIP 1068 server and STUN server. Typically, SIP systems utilizing SIP's 1069 digest authentication mechanism do not actually store the password in 1070 the database. Rather, they store a value called H(A1), which is 1071 computed as: 1073 H(A1) = MD5(username ":" realm ":" password) 1075 If a system wishes to utilize this credential, the STUN password 1076 would be computed by taking the user-entered username and password, 1077 and using H(A1) as the STUN password. It is RECOMMENDED that clients 1078 utilize this construction for the STUN password. 1080 10.2.1. Forming a Request 1082 There are two cases when forming a request. In the first case, this 1083 is the first request from the client to the server (as identified by 1084 its IP address and port). In the second case, the client is 1085 submitting a subsequent request once a previous request/response 1086 transaction has completed successfully. 1088 10.2.1.1. First Request 1090 If the client has not completed a successful request/response 1091 transaction with the server, it SHOULD omit the USERNAME, MESSAGE- 1092 INTEGRITY, REALM, and NONCE attributes. In other words, the very 1093 first request is sent as if there were no authentication or message 1094 integrity applied. 1096 10.2.1.2. Subsequent Requests 1098 Once a request/response transaction has completed successfully, the 1099 client will have been been presented a realm and nonce by the server, 1100 and selected a username and password with which it authenticated. 1101 The client SHOULD cache the username, password, realm, and nonce for 1102 subsequent communications with the server. When the client sends a 1103 subsequent request, it SHOULD include the USERNAME, REALM, and NONCE 1104 attributes with these cached values. It SHOULD include a MESSAGE- 1105 INTEGRITY attributed, computed as described in Section 14.4 using the 1106 cached password as the key. 1108 10.2.2. Receiving a Request 1110 After the server has done the basic processing of a request, it 1111 performs the checks listed below in the order specified: 1113 o If the message: 1115 * does not contain a MESSAGE-INTEGRITY attribute, 1117 * OR, it contains a USERNAME whose value is not a valid username, 1119 the server MUST generate an error response with an error code of 1120 401. This response MUST include a REALM value. It is RECOMENDED 1121 that the REALM value by the domain name of the provider of the 1122 STUN server. The response MUST include a NONCE, selected by the 1123 server. 1125 o If the message contains a MESSAGE-INTEGRITY attribute, but is 1126 missing the USERNAME, REALM or NONCE attributes, the server MUST 1127 generate an error response with an error code of 400. 1129 o If the NONCE is no longer valid, the server MUST generate an error 1130 response with an error code of 438 (Stale Nonce). This response 1131 MUST include a NONCE and REALM attribute. 1133 o Using the password associated with the username in the USERNAME 1134 attribute, compute the value for the message-integrity as 1135 described in Section 14.4. If the resulting value does not match 1136 the contents of the MESSAGE-INTEGRITY attribute, the server MUST 1137 reject the request with an error response. This response MUST use 1138 an error code of 401. It MUST include a REALM and NONCE 1139 attribute. 1141 If these checks pass, the server continues to process the request or 1142 indication. Any response generated by the server MUST include the 1143 MESSAGE-INTEGRITY attribute, computed using the username and password 1144 utilized to authenticate the request. The REALM, NONCE, and USERNAME 1145 attributes SHOULD NOT be included. 1147 10.2.3. Receiving a Response 1149 The processing here takes place prior to the processing in 1150 Section 7.3.3 or Section 7.3.4. 1152 If the response is an error response, with an error code of 401 1153 (Unauthorized), the client SHOULD retry the request with a new 1154 transaction. This request MUST contain a USERNAME, determined by the 1155 client as the appropriate username for the REALM from the error 1156 response. The request MUST contain the REALM, copied from the error 1157 response. The request MUST contain the NONCE, copied from the error 1158 response. The request MUST contain the MESSAGE-INTEGRITY attribute, 1159 computed using the password associated with the username in the 1160 USERNAME attribute. The client MUST NOT perform this retry if it is 1161 not changing the USERNAME or REALM or its associated password, from 1162 the previous attempt. 1164 If the response is an error response with an error code of 438, the 1165 client MUST retry the request, using the new NONCE supplied in the 1166 438 response. This retry MUST also include the USERNAME, REALM and 1167 MESSAGE-INTEGRITY. 1169 The client looks for the MESSAGE-INTEGRITY attribute in the response 1170 (either success or failure). If present, the client computes the 1171 message integrity over the response as defined in Section 14.4, using 1172 the same password it utilized for the request. If the resulting 1173 value matches the contents of the MESSAGE-INTEGRITY attribute, the 1174 response is considered authenticated. If the value does not match, 1175 or if MESSAGE-INTEGRITY was absent, the response MUST be discarded, 1176 as if it was never received. This means that retransmits, if 1177 applicable, will continue. 1179 11. ALTERNATE-SERVER Mechanism 1181 This section describes a mechanism in STUN that allows a server to 1182 redirect a client to another server. This extension is optional, and 1183 a usage must define if and when this extension is used. To prevent 1184 denial-of-service attacks, this extension MUST only be used in 1185 situations where the client and server are using an authentication 1186 and message-integrity mechanism. 1188 A server using this extension redirects a client to another server by 1189 replying to a request message with an error response message with an 1190 error code of 300 (Try Alternate). The server MUST include a 1191 ALTERNATE-SERVER attribute in the error response. The error response 1192 message MUST be authenticated, which in practice means the request 1193 message must have passed the authentication checks. 1195 A client using this extension handles a 300 (Try Alternate) error 1196 code as follows. If the error response has passed the authentication 1197 checks, then the client looks for a ALTERNATE-SERVER attribute in the 1198 error response. If one is found, then the client considers the 1199 current transaction as failed, and re-attempts the request with the 1200 server specified in the attribute. The client SHOULD reuse any 1201 authentication credentials from the old request in the new 1202 transaction. 1204 12. Backwards Compatibility with RFC 3489 1206 This section define procedures that allow a degree of backwards 1207 compatible with the original protocol defined in RFC 3489 [RFC3489]. 1208 This mechanism is optional, meant to be utilized only in cases where 1209 a new client can connect to an old server, or vice-a-versa. A usage 1210 must define if and when this procedure is used. 1212 Section 18 lists all the changes between this specification and RFC 1213 3489 [RFC3489]. However, not all of these differences are important, 1214 because "classic STUN" was only used in a few specific ways. For the 1215 purposes of this extension, the important changes are the following. 1216 In RFC 3489: 1218 o UDP was the only supported transport; 1220 o The field that is now the Magic Cookie field was a part of the 1221 transaction id field, and transaction ids were 128 bits long; 1223 o The XOR-MAPPED-ADDRESS attribute did not exist, and the Binding 1224 method used the MAPPED-ADDRESS attribute instead 1226 o There were two comprehension-required attributes, RESPONSE-ADDRESS 1227 and CHANGE-REQUEST, that have been removed from this 1228 specification. 1230 * These attributes are now part of the NAT Behavior Discovery 1231 usage. 1233 12.1. Changes to Client Processing 1235 A client that wants to interoperate with a [RFC3489] server SHOULD 1236 send a request message that uses the Binding method, contains no 1237 attributes, and uses UDP as the transport protocol to the server. If 1238 successful, the success response received from the server will 1239 contain a MAPPED-ADDRESS attribute rather than an XOR-MAPPED-ADDRESS 1240 attribute; other than this change, the processing of the response is 1241 identical to the procedures described above. 1243 12.2. Changes to Server Processing 1245 A STUN server can detect when a given Binding Request message was 1246 sent from an RFC 3489 [RFC3489] client by the absence of the correct 1247 value in the Magic Cookie field. When the server detects an RFC 3489 1248 client, it SHOULD copy the value seen in the Magic Cookie field in 1249 the Binding Request to the Magic Cookie field in the Binding Response 1250 message, and insert a MAPPED-ADDRESS attribute instead of an XOR- 1251 MAPPED-ADDRESS attribute. 1253 The client might, in rare situations, include either the RESPONSE- 1254 ADDRESS or CHANGE-REQUEST attributes. In these situations, the 1255 server will view these as unknown comprehension-required attributes 1256 and reply with an error response. Since the mechanisms utilizing 1257 those attributes are no longer supported, this behavior is 1258 acceptable. 1260 13. STUN Usages 1262 STUN by itself is not a solution to the NAT traversal problem. 1263 Rather, STUN defines a toolkit of functions that can be used inside a 1264 larger solution. The term "STUN Usage" is used for any solution that 1265 uses STUN as a component. 1267 At the time of writing, three STUN usages are defined: Interactive 1268 Connectivity Establishment (ICE) [I-D.ietf-mmusic-ice], Client- 1269 initiated connections for SIP [I-D.ietf-sip-outbound], and NAT 1270 Behavior Discovery [I-D.ietf-behave-nat-behavior-discovery]. Other 1271 STUN usages may be defined in the future. 1273 A STUN usage defines how STUN is actually utilized - when to send 1274 requests, what to do with the responses, and which optional 1275 procedures defined here (or in an extension to STUN) are to be used. 1276 A usage would also define: 1278 o Which STUN methods are used; 1280 o What authentication and message integrity mechanisms are used; 1282 o What mechanisms are used to distinguish STUN messages from other 1283 messages. When STUN is run over TCP, a framing mechanism may be 1284 required; 1286 o How a STUN client determines the IP address and port of the STUN 1287 server; 1289 o Whether backwards compatibility to RFC 3489 is required; 1291 o What optional attributes defined here (such as FINGERPRINT and 1292 ALTERNATE-SERVER) or in other extensions are required. 1294 In addition, any STUN usage must consider the security implications 1295 of using STUN in that usage. A number of attacks against STUN are 1296 known (see the Security Considerations section in this document) and 1297 any usage must consider how these attacks can be thwarted or 1298 mitigated. 1300 Finally, a usage must consider whether its usage of STUN is an 1301 example of the Unilateral Self-Address Fixing approach to NAT 1302 traversal, and if so, address the questions raised in RFC 3424. 1304 14. STUN Attributes 1306 After the STUN header are zero or more attributes. Each attribute 1307 MUST be TLV encoded, with a 16 bit type, 16 bit length, and value. 1308 Each STUN attribute MUST end on a 32 bit boundary. As mentioned 1309 above, all fields in an attribute are transmitted most significant 1310 bit first. 1312 0 1 2 3 1313 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 1314 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1315 | Type | Length | 1316 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1317 | Value (variable) .... 1318 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1320 Figure 5: Format of STUN Attributes 1322 The value in the Length field MUST contain the length of the Value 1323 part of the attribute, prior to padding, measured in bytes. Since 1324 STUN aligns attributes on 32 bit boundaries, attributes whose content 1325 is not a multiple of 4 bytes are padded with 1, 2 or 3 bytes of 1326 padding so that its value contains a multiple of 4 bytes. The 1327 padding bits are ignored, and may be any value. 1329 Any attribute type MAY appear more than once in a STUN message. 1330 Unless specified otherwise, the order of appearance is significant: 1331 only the first occurance needs to be processed by a receiver, and any 1332 duplicates MAY be ignored by a receiver. 1334 To allow future revisions of this specification to add new attributes 1335 if needed, the attribute space is divided into two ranges. 1336 Attributes with type values between 0x0000 and 0x7FFF are 1337 comprehension-required attributes, which means that the STUN agent 1338 cannot successfully process the message unless it understands the 1339 attribute. Attributes with type values between 0x8000 and 0xFFFF are 1340 comprehension-optional attributes, which means that those attributes 1341 can be ignored by the STUN agent if it does not understand them. 1343 The STUN Attribute types defined by this specification are: 1345 Comprehension-required range (0x0000-0x7FFF): 1346 0x0000: (Reserved) 1347 0x0001: MAPPED-ADDRESS 1348 0x0006: USERNAME 1349 0x0007: (Reserved; was PASSWORD) 1350 0x0008: MESSAGE-INTEGRITY 1351 0x0009: ERROR-CODE 1352 0x000A: UNKNOWN-ATTRIBUTES 1353 0x0014: REALM 1354 0x0015: NONCE 1355 0x0020: XOR-MAPPED-ADDRESS 1357 Comprehension-optional range (0x8000-0xFFFF) 1358 0x8022: SERVER 1359 0x8023: ALTERNATE-SERVER 1360 0x8028: FINGERPRINT 1362 The rest of this section describes the format of the various 1363 attributes defined in this specification. 1365 14.1. MAPPED-ADDRESS 1367 The MAPPED-ADDRESS attribute indicates a reflexive transport address 1368 of the client. It consists of an eight bit address family, and a 1369 sixteen bit port, followed by a fixed length value representing the 1370 IP address. If the address family is IPv4, the address MUST be 32 1371 bits. If the address family is IPv6, the address MUST be 128 bits. 1372 All fields must be in network byte order. 1374 The format of the MAPPED-ADDRESS attribute is: 1376 0 1 2 3 1377 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 1378 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1379 |0 0 0 0 0 0 0 0| Family | Port | 1380 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1381 | | 1382 | Address (32 bits or 128 bits) | 1383 | | 1384 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1386 Figure 7: Format of MAPPED-ADDRESS attribute 1388 The address family can take on the following values: 1390 0x01:IPv4 1391 0x02:IPv6 1393 The first 8 bits of the MAPPED-ADDRESS MUST be set to 0 and MUST be 1394 ignored by receivers. These bits are present for aligning parameters 1395 on natural 32 bit boundaries. 1397 This attribute is used only by servers for achieving backwards 1398 compatibility with RFC 3489 [RFC3489] clients. 1400 14.2. XOR-MAPPED-ADDRESS 1402 The XOR-MAPPED-ADDRESS attribute is identical to the MAPPED-ADDRESS 1403 attribute, except that the reflexive transport address is obfuscated 1404 through the XOR function. 1406 The format of the XOR-MAPPED-ADDRESS is: 1408 0 1 2 3 1409 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 1410 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1411 |x x x x x x x x| Family | X-Port | 1412 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1413 | X-Address (Variable) 1414 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1416 Figure 9: Format of XOR-MAPPED-ADDRESS Attribute 1418 The Family represents the IP address family, and is encoded 1419 identically to the Family in MAPPED-ADDRESS. 1421 X-Port is the mapped port, exclusive or'd with most significant 16 1422 bits of the magic cookie. If the IP address family is IPv4, 1423 X-Address is the mapped IP address exclusive or'd with the magic 1424 cookie. If the IP address family is IPv6, the X-Address is the 1425 mapped IP address exclusively or'ed with the magic cookie and the 96- 1426 bit transaction ID. 1428 For example, using the "^" character to indicate exclusive or, if the 1429 IP address is 192.168.1.1 (0xc0a80101) and the port is 5555 (0x15B3), 1430 the X-Port would be 0x15B3 ^ 0x2112 = 0x34A1, and the X-Address would 1431 be 0xc0a80101 ^ 0x2112A442 = 0xe1baa543. 1433 The rules for encoding and processing the first 8 bits of the 1434 attribute's value, the rules for handling multiple occurrences of the 1435 attribute, and the rules for processing addresses families are the 1436 same as for MAPPED-ADDRESS. 1438 NOTE: XOR-MAPPED-ADDRESS and MAPPED-ADDRESS differ only in their 1439 encoding of the transport address. The former encodes the transport 1440 address by exclusive-or'ing it with the magic cookie. The latter 1441 encodes it directly in binary. RFC 3489 originally specified only 1442 MAPPED-ADDRESS. However, deployment experience found that some NATs 1443 rewrite the 32-bit binary payloads containing the NAT's public IP 1444 address, such as STUN's MAPPED-ADDRESS attribute, in the well-meaning 1445 but misguided attempt at providing a generic ALG function. Such 1446 behavior interferes with the operation of STUN and also causes 1447 failure of STUN's message integrity checking. 1449 14.3. USERNAME 1451 The USERNAME attribute is used for message integrity. It identifies 1452 the username and password combination used in the message integrity 1453 check. 1455 The value of USERNAME is a variable length value. It MUST contain a 1456 UTF-8 encoded sequence of less than 128 characters. 1458 14.4. MESSAGE-INTEGRITY 1460 The MESSAGE-INTEGRITY attribute contains an HMAC-SHA1 [RFC2104] of 1461 the STUN message. The MESSAGE-INTEGRITY attribute can be present in 1462 any STUN message type. Since it uses the SHA1 hash, the HMAC will be 1463 20 bytes. The text used as input to HMAC is the STUN message, 1464 including the header, up to and including the attribute preceding the 1465 MESSAGE-INTEGRITY attribute. With the exception of the FINGERPRINT 1466 attribute, which appears after MESSAGE-INTEGRITY, agents MUST ignore 1467 all other attributes that follow MESSAGE-INTEGRITY. 1469 The key used as input to HMAC is the password. 1471 Since the hash is computed over the entire STUN message, it includes 1472 the length field from the STUN message header. This length indicates 1473 the length of the entire message, including the MESSAGE-INTEGRITY 1474 attribute itself. Consequently, the MESSAGE-INTEGRITY attribute MUST 1475 be inserted into the message (with dummy content) prior to the 1476 computation of the integrity check. Once the computation is 1477 performed, the value of the attribute can be filled in. This ensures 1478 the length has the correct value when the hash is performed. 1479 Similarly, when validating the MESSAGE-INTEGRITY, the length field 1480 should be adjusted to point to the end of the MESSAGE-INTEGRITY 1481 attribute prior to calculating the HMAC. Such adjustment is 1482 necessary when attributes, such as FINTERPRINT, appear after MESSAGE- 1483 INTEGRITY. 1485 14.5. FINGERPRINT 1487 The FINGERPRINT attribute may be present in all STUN messages. The 1488 value of the attribute is computed as the CRC-32 of the STUN message 1489 up to (but excluding) the FINGERPRINT attribute itself, xor-d with 1490 the 32 bit value 0x5354554e (the XOR helps in cases where an 1491 application packet is also using CRC-32 in it). The 32 bit CRC is 1492 the one defined in ITU V.42 [ITU.V42.1994], which has a generator 1493 polynomial of x32+x26+x23+x22+x16+x12+x11+x10+x8+x7+x5+x4+x2+x+1. 1494 When present, the FINGERPRINT attribute MUST be the last attribute in 1495 the message, and thus will appear after MESSAGE-INTEGRITY. 1497 The FINGERPRINT attribute can aid in distinguishing STUN packets from 1498 packets of other protocols. See Section 8. 1500 When using the FINGERPRINT attribute in a message, the attribute is 1501 first placed into the message with a dummy value, then the CRC is 1502 computed, and then the value of the attribute is updated. If the 1503 MESSAGE-INTEGRITY attribute is also present, then it must be present 1504 with the correct message-integrity value before the CRC is computed, 1505 since the CRC is done over the value of the MESSAGE-INTEGRITY 1506 attribute as well. 1508 14.6. ERROR-CODE 1510 The ERROR-CODE attribute is used in Error Response messages. It 1511 contains a numeric error code value in the range of 300 to 699 plus a 1512 textual reason phrase encoded in UTF-8, and is consistent in its code 1513 assignments and semantics with SIP [RFC3261] and HTTP [RFC2616]. The 1514 reason phrase is meant for user consumption, and can be anything 1515 appropriate for the error code. Recommended reason phrases for the 1516 defined error codes are presented below. The reason phrase MUST be a 1517 UTF-8 encoded sequence of less than 128 characters. 1519 To facilitate processing, the class of the error code (the hundreds 1520 digit) is encoded separately from the rest of the code. 1522 0 1 2 3 1523 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 1524 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1525 | Reserved, should be 0 |Class| Number | 1526 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1527 | Reason Phrase (variable) .. 1528 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1530 The Reserved bits SHOULD be 0, and are for alignment on 32-bit 1531 boundaries. Receivers MUST ignore these bits. The Class represents 1532 the hundreds digit of the error code. The value MUST be between 3 1533 and 6. The number represents the error code modulo 100, and its 1534 value MUST be between 0 and 99. 1536 The following error codes, along with their recommended reason 1537 phrases (in brackets) are defined: 1539 300 Try Alternate: The client should contact an alternate server for 1540 this request. This error response MUST only be sent if the 1541 request included a USERNAME attribute and a valid MESSAGE- 1542 INTEGRITY attribute; otherwise it MUST NOT be sent and error 1543 code 400 is suggested. This error response MUST be protected 1544 with the MESSAGE-INTEGRITY attribute, and receivers MUST 1545 validate the MESSAGE-INTEGRITY of this response before 1546 redirecting themselves to an alternate server. 1548 Note: failure to generate and validate message-integrity 1549 for a 300 response allows an on-path attacker to falsify a 1550 300 response thus causing subsequent STUN messages to be 1551 sent to a victim. 1553 400 Bad Request: The request was malformed. The client SHOULD NOT 1554 retry the request without modification from the previous 1555 attempt. The server may not be able to generate a valid 1556 MESSAGE-INTEGRITY for this error, so the client MUST NOT expect 1557 a valid MESSAGE-INTEGRITY attribute on this response. 1559 401 Unauthorized: The request did not contain the expected MESSAGE- 1560 INTEGRITY attribute. The server MAY include the MESSAGE- 1561 INTEGRITY attribute in its error response. 1563 420 Unknown Attribute: The server received STUN packet containing a 1564 comprehension-required attribute which it did not understand. 1565 The server MUST put this unknown attribute in the UNKNOWN- 1566 ATTRIBUTE attribute of its error response. 1568 438 Stale Nonce: The NONCE used by the client was no longer valid. 1569 The client should retry, using the NONCE provided in the 1570 response. 1572 500 Server Error: The server has suffered a temporary error. The 1573 client should try again. 1575 14.7. REALM 1577 The REALM attribute may be present in requests and responses. It 1578 contains text which meets the grammar for "realm-value" as described 1579 in RFC 3261 [RFC3261] but without the double quotes and their 1580 surrounding whitespace. That is, it is an unquoted realm-value. It 1581 MUST be a UTF-8 encoded sequence of less than 128 characters. 1583 Presence of the REALM attribute in a request indicates that long-term 1584 credentials are being used for authentication. Presence in certain 1585 error responses indicates that the server wishes the client to use a 1586 long-term credential for authentication. 1588 14.8. NONCE 1590 The NONCE attribute may be present in requests and responses. It 1591 contains a sequence of qdtext or quoted-pair, which are defined in 1592 RFC 3261 [RFC3261]. See RFC 2617 [RFC2617], Section 4.3, for 1593 guidance on selection of nonce values in a server. It MUST be less 1594 than 128 characters. 1596 14.9. UNKNOWN-ATTRIBUTES 1598 The UNKNOWN-ATTRIBUTES attribute is present only in an error response 1599 when the response code in the ERROR-CODE attribute is 420. 1601 The attribute contains a list of 16 bit values, each of which 1602 represents an attribute type that was not understood by the server. 1604 0 1 2 3 1605 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 1606 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1607 | Attribute 1 Type | Attribute 2 Type | 1608 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1609 | Attribute 3 Type | Attribute 4 Type ... 1610 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1612 Figure 11: Format of UNKNOWN-ATTRIBUTES attribute 1614 Note: In [RFC3489], this field was padded to 32 by duplicating the 1615 last attribute. In this version of the specification, the normal 1616 padding rules for attributes are used instead. 1618 14.10. SERVER 1620 The server attribute contains a textual description of the software 1621 being used by the server, including manufacturer and version number. 1622 The attribute has no impact on operation of the protocol, and serves 1623 only as a tool for diagnostic and debugging purposes. The value of 1624 SERVER is variable length. It MUST be a UTF-8 encoded sequence of 1625 less than 128 characters 1627 14.11. ALTERNATE-SERVER 1629 The alternate server represents an alternate transport address 1630 identifying a different STUN server which the STUN client should try. 1632 It is encoded in the same way as MAPPED-ADDRESS, and thus refers to a 1633 single server by IP address. The IP address family MUST be identical 1634 to that of the source IP address of the request. 1636 This attribute MUST only appear in an error response that contains a 1637 MESSAGE-INTEGRITY attribute. This prevents it from being used in 1638 denial-of-service attacks. 1640 15. Security Considerations 1642 15.1. Attacks against the Protocol 1644 15.1.1. Outside Attacks 1646 An attacker can try to modify STUN messages in transit, in order to 1647 cause a failure in STUN operation. These attacks are prevented for 1648 both requests and responses through the message integrity mechanism, 1649 using either a short term or long term credential. 1651 An attacker that can observe, but not modify STUN messages in-transit 1652 (for example, an attacker present on a shared access medium, such as 1653 Wi-Fi), can see a STUN request, and then immediately send a STUN 1654 response, typically an error response, in order to disrupt STUN 1655 processing. This attack is also prevented for messages that utilize 1656 MESSAGE-INTEGRITY. However, some error responses, those related to 1657 authentication in particular, cannot be protected by MESSAGE- 1658 INTEGRITY. When STUN itself is run over a secure transport protocol 1659 (e.g., TLS), these attacks are completely mitigated. 1661 15.1.2. Inside Attacks 1663 A rogue client may try to launch a DoS attack against a server by 1664 sending it a large number of STUN requests. Fortunately, STUN 1665 requests can be processed statelessly by a server, making such 1666 attacks hard to launch. 1668 15.2. Attacks Affecting the Usage 1670 This section lists attacks that might be launched against a usage of 1671 STUN. Each STUN usage must consider whether these attacks are 1672 applicable to it, and if so, discuss counter-measures. 1674 Most of the attacks in this section revolve around an attacker 1675 modifying the reflexive address learned by a STUN client through a 1676 Binding Request/Binding Response transaction. Since the usage of the 1677 reflexive address is a function of the usage, the applicability and 1678 remediation of these attacks is usage-specific. In common 1679 situations, modification of the reflexive address by an on-path 1680 attacker is easy to do. Consider, for example, the common situation 1681 where STUN is run directly over UDP. In this case, an on-path 1682 attacker can modify the source IP address of the Binding Request 1683 before it arrives at the STUN server. The STUN server will then 1684 return this IP address in the XOR-MAPPED-ADDRESS attribute to the 1685 client. Protecting against this attack by using a message-integrity 1686 check is impossible, since a message-integrity value cannot cover the 1687 source IP address, since the intervening NAT must be able to modify 1688 this value. Instead, one solution to preventing the attacks listed 1689 below is for the client to verify the reflexive address learned, as 1690 is done in ICE [I-D.ietf-mmusic-ice]. Other usages may use other 1691 means to prevent these attacks. 1693 15.2.1. Attack I: DDoS Against a Target 1695 In this attack, the attacker provides one or more clients with the 1696 same faked reflexive address that points to the intended target. 1697 This will trick the STUN clients into thinking that their reflexive 1698 addresses are equal to that of the target. If the clients hand out 1699 that reflexive address in order to receive traffic on it (for 1700 example, in SIP messages), the traffic will instead be sent to the 1701 target. This attack can provide substantial amplification, 1702 especially when used with clients that are using STUN to enable 1703 multimedia applications. 1705 15.2.2. Attack II: Silencing a Client 1707 In this attack, the attacker provides a STUN client with a faked 1708 reflexive address. The reflexive address it provides is a transport 1709 address that routes to nowhere. As a result, the client won't 1710 receive any of the packets it expects to receive when it hands out 1711 the reflexive address. This exploitation is not very interesting for 1712 the attacker. It impacts a single client, which is frequently not 1713 the desired target. Moreover, any attacker that can mount the attack 1714 could also deny service to the client by other means, such as 1715 preventing the client from receiving any response from the STUN 1716 server, or even a DHCP server. 1718 15.2.3. Attack III: Assuming the Identity of a Client 1720 This attack is similar to attack III. However, the faked reflexive 1721 address points to the attacker itself. This allows the attacker to 1722 receive traffic which was destined for the client. 1724 15.2.4. Attack IV: Eavesdropping 1726 In this attack, the attacker forces the client to use a reflexive 1727 address that routes to itself. It then forwards any packets it 1728 receives to the client. This attack would allow the attacker to 1729 observe all packets sent to the client. However, in order to launch 1730 the attack, the attacker must have already been able to observe 1731 packets from the client to the STUN server. In most cases (such as 1732 when the attack is launched from an access network), this means that 1733 the attacker could already observe packets sent to the client. This 1734 attack is, as a result, only useful for observing traffic by 1735 attackers on the path from the client to the STUN server, but not 1736 generally on the path of packets being routed towards the client. 1738 15.3. Hash Agility Plan 1740 This specification uses SHA-1 for computation of the message 1741 integrity. If, at a later time, SHA-1 is found to be compromised, 1742 the following is the remedy that will be applied. 1744 We will define a STUN extension which introduces a new message 1745 integrity attribute, computed using a new hash. Clients would be 1746 required to include both the new and old message integrity attributes 1747 in their requests or indications. A new server will utilize the new 1748 message integrity attribute, and an old one, the old. After a 1749 transition period where mixed implementations are in deployment, the 1750 old message-integrity attribute will be deprecated by another 1751 specification, and clients will cease including it in requests. 1753 16. IAB Considerations 1755 The IAB has studied the problem of "Unilateral Self Address Fixing" 1756 (UNSAF), which is the general process by which a client attempts to 1757 determine its address in another realm on the other side of a NAT 1758 through a collaborative protocol reflection mechanism (RFC3424 1759 [RFC3424]). STUN can be used to perform this function using a 1760 BindingRequest/BindingResponse transaction if one agent is behind a 1761 NAT and the other is on the public side of the NAT. 1763 The IAB has mandated that protocols developed for this purpose 1764 document a specific set of considerations. Because some STUN usages 1765 provide UNSAF functions (such as ICE [I-D.ietf-mmusic-ice] ), and 1766 others do not (such as SIP Outbound [I-D.ietf-sip-outbound]), answers 1767 to these considerations need to be addressed by the usages 1768 themselves. 1770 17. IANA Considerations 1772 IANA is hereby requested to create three new registries: a STUN 1773 methods registry, a STUN Attributes registry, and a STUN Error Codes 1774 registry. 1776 17.1. STUN Methods Registry 1778 A STUN method is a hex number in the range 0x000 - 0x3FF. The 1779 encoding of STUN method into a STUN message is described in 1780 Section 6. 1782 The initial STUN methods are: 1784 0x000: (Reserved) 1785 0x001: Binding 1786 0x002: (Reserved; was SharedSecret) 1788 STUN methods in the range 0x000 - 0x1FF are assigned by IETF 1789 Consensus [RFC2434]. STUN methods in the range 0x200 - 0x3FF are 1790 assigned on a First Come First Served basis [RFC2434] 1792 17.2. STUN Attribute Registry 1794 A STUN Attribute type is a hex number in the range 0x0000 - 0xFFFF. 1795 STUN attribute types in the range 0x0000 - 0x7FFF are considered 1796 comprehension-required; STUN attribute types in the range 0x8000 - 1797 0xFFFF are considered comprehension-optional. A STUN agent handles 1798 unknown comprehension-required and comprehension-optional attributes 1799 differently. 1801 The initial STUN Attributes types are: 1803 Comprehension-required range (0x0000-0x7FFF): 1804 0x0000: (Reserved) 1805 0x0001: MAPPED-ADDRESS 1806 0x0006: USERNAME 1807 0x0007: (Reserved; was PASSWORD) 1808 0x0008: MESSAGE-INTEGRITY 1809 0x0009: ERROR-CODE 1810 0x000A: UNKNOWN-ATTRIBUTES 1811 0x0014: REALM 1812 0x0015: NONCE 1813 0x0020: XOR-MAPPED-ADDRESS 1815 Comprehension-optional range (0x8000-0xFFFF) 1816 0x8022: SERVER 1817 0x8023: ALTERNATE-SERVER 1818 0x8028: FINGERPRINT 1820 STUN Attribute types in the first half of the comprehension-required 1821 range (0x0000 - 0x3FFF) and in the first half of the comprehension- 1822 optional range (0x8000 - 0xBFFF) are assigned by IETF Consensus 1823 [RFC2434]. STUN Attribute types in the second half of the 1824 comprehension-required range (0x4000 - 0x7FFF) and in the second half 1825 of the comprehension-optional range (0xC000 - 0xFFFF) are assigned on 1826 a First Come First Served basis [RFC2434]. 1828 17.3. STUN Error Code Registry 1830 A STUN Error code is a number in the range 0 - 699. STUN error codes 1831 are accompanied by a textual reason phrase in UTF-8 which is intended 1832 only for human consumption and can be anything appropriate; this 1833 document proposes only suggested values. 1835 STUN error codes are consistent in codepoint assignments and 1836 semantics with SIP [RFC3261] and HTTP [RFC2616]. 1838 The initial values in this registry are given in Section 14.6. 1840 New STUN error codes are assigned on a Specification-Required basis 1841 [RFC2434]. The specification must carefully consider how clients 1842 that do not understand this error code will process it before 1843 granting the request. See the rules in Section 7.3.4. 1845 18. Changes Since RFC 3489 1847 This specification obsoletes RFC3489 [RFC3489]. This specification 1848 differs from RFC3489 in the following ways: 1850 o Removed the notion that STUN is a complete NAT traversal solution. 1851 STUN is now a toolkit that can be used to produce a NAT traversal 1852 solution. As a consequence, changed the name of the protocol to 1853 Session Traversal Utilities for NAT. 1855 o Introduced the concept of STUN usages, and described what a usage 1856 of STUN must document. 1858 o Removed the usage of STUN for NAT type detection and binding 1859 lifetime discovery. These techniques have proven overly brittle 1860 due to wider variations in the types of NAT devices than described 1861 in this document. Removed the RESPONSE-ADDRESS, CHANGED-ADDRESS, 1862 CHANGE-REQUEST, SOURCE-ADDRESS, and REFLECTED-FROM attributes. 1864 o Added a fixed 32-bit magic cookie and reduced length of 1865 transaction ID by 32 bits. The magic cookie begins at the same 1866 offset as the original transaction ID. 1868 o Added the XOR-MAPPED-ADDRESS attribute, which is included in 1869 Binding Responses if the magic cookie is present in the request. 1870 Otherwise the RFC3489 behavior is retained (that is, Binding 1871 Response includes MAPPED-ADDRESS). See discussion in XOR-MAPPED- 1872 ADDRESS regarding this change. 1874 o Introduced formal structure into the Message Type header field, 1875 with an explicit pair of bits for indication of request, response, 1876 error response or indication. Consequently, the message type 1877 field is split into the class (one of the previous four) and 1878 method. 1880 o Explicitly point out that the most significant two bits of STUN 1881 are 0b00, allowing easy differentiation with RTP packets when used 1882 with ICE. 1884 o Added the FINGERPRINT attribute to provide a method of definitely 1885 detecting the difference between STUN and another protocol when 1886 the two protocols are multiplexed together. 1888 o Added support for IPv6. Made it clear that an IPv4 client could 1889 get a v6 mapped address, and vice-a-versa. 1891 o Added long-term credential-based authentication. 1893 o Added the SERVER, REALM, NONCE, and ALTERNATE-SERVER attributes. 1895 o Removed the SharedSecret method, and thus the PASSWORD attribute. 1896 This method was almost never implemented and is not needed with 1897 current usages. 1899 o Removed recommendation to continue listening for STUN Responses 1900 for 10 seconds in an attempt to recognize an attack. 1902 o Changed transaction timers to be more TCP friendly. 1904 o Removed the STUN example that centered around the separation of 1905 the control and media planes. Instead, provided more information 1906 on using STUN with protocols. 1908 o Defined a generic padding mechanism that changes the 1909 interpretation of the length attribute. This would, in theory, 1910 break backwards compatibility. However, the mechanism in RFC 3489 1911 never worked for the few attributes that weren't aligned naturally 1912 on 32 bit boundaries. 1914 o REALM, USERNAME, SERVER, reason phrases and NONCE limited to 127 1915 characters. 1917 19. Acknowledgements 1919 The authors would like to thank Cedric Aoun, Pete Cordell, Cullen 1920 Jennings, Bob Penfield, Xavier Marjou, Bruce Lowekamp and Chris 1921 Sullivan for their comments, and Baruch Sterman and Alan Hawrylyshen 1922 for initial implementations. Thanks for Leslie Daigle, Allison 1923 Mankin, Eric Rescorla, and Henning Schulzrinne for IESG and IAB input 1924 on this work. 1926 20. References 1928 20.1. Normative References 1930 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1931 Requirement Levels", BCP 14, RFC 2119, March 1997. 1933 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, 1934 September 1981. 1936 [RFC2782] Gulbrandsen, A., Vixie, P., and L. Esibov, "A DNS RR for 1937 specifying the location of services (DNS SRV)", RFC 2782, 1938 February 2000. 1940 [RFC2818] Rescorla, E., "HTTP Over TLS", RFC 2818, May 2000. 1942 [RFC2617] Franks, J., Hallam-Baker, P., Hostetler, J., Lawrence, S., 1943 Leach, P., Luotonen, A., and L. Stewart, "HTTP 1944 Authentication: Basic and Digest Access Authentication", 1945 RFC 2617, June 1999. 1947 [RFC2988] Paxson, V. and M. Allman, "Computing TCP's Retransmission 1948 Timer", RFC 2988, November 2000. 1950 [ITU.V42.1994] 1951 International Telecommunications Union, "Error-correcting 1952 Procedures for DCEs Using Asynchronous-to-Synchronous 1953 Conversion", ITU-T Recommendation V.42, 1994. 1955 20.2. Informational References 1957 [RFC2104] Krawczyk, H., Bellare, M., and R. Canetti, "HMAC: Keyed- 1958 Hashing for Message Authentication", RFC 2104, 1959 February 1997. 1961 [RFC3261] Rosenberg, J., Schulzrinne, H., Camarillo, G., Johnston, 1962 A., Peterson, J., Sparks, R., Handley, M., and E. 1963 Schooler, "SIP: Session Initiation Protocol", RFC 3261, 1964 June 2002. 1966 [RFC2616] Fielding, R., Gettys, J., Mogul, J., Frystyk, H., 1967 Masinter, L., Leach, P., and T. Berners-Lee, "Hypertext 1968 Transfer Protocol -- HTTP/1.1", RFC 2616, June 1999. 1970 [I-D.ietf-mmusic-ice] 1971 Rosenberg, J., "Interactive Connectivity Establishment 1972 (ICE): A Protocol for Network Address Translator (NAT) 1973 Traversal for Offer/Answer Protocols", 1974 draft-ietf-mmusic-ice-16 (work in progress), June 2007. 1976 [RFC3489] Rosenberg, J., Weinberger, J., Huitema, C., and R. Mahy, 1977 "STUN - Simple Traversal of User Datagram Protocol (UDP) 1978 Through Network Address Translators (NATs)", RFC 3489, 1979 March 2003. 1981 [I-D.ietf-behave-turn] 1982 Rosenberg, J., "Obtaining Relay Addresses from Simple 1983 Traversal Underneath NAT (STUN)", 1984 draft-ietf-behave-turn-03 (work in progress), March 2007. 1986 [I-D.ietf-sip-outbound] 1987 Jennings, C. and R. Mahy, "Managing Client Initiated 1988 Connections in the Session Initiation Protocol (SIP)", 1989 draft-ietf-sip-outbound-09 (work in progress), June 2007. 1991 [I-D.ietf-behave-nat-behavior-discovery] 1992 MacDonald, D. and B. Lowekamp, "NAT Behavior Discovery 1993 Using STUN", draft-ietf-behave-nat-behavior-discovery-00 1994 (work in progress), February 2007. 1996 [I-D.ietf-mmusic-ice-tcp] 1997 Rosenberg, J., "TCP Candidates with Interactive 1998 Connectivity Establishment (ICE", 1999 draft-ietf-mmusic-ice-tcp-03 (work in progress), 2000 March 2007. 2002 [RFC3264] Rosenberg, J. and H. Schulzrinne, "An Offer/Answer Model 2003 with Session Description Protocol (SDP)", RFC 3264, 2004 June 2002. 2006 [RFC3424] Daigle, L. and IAB, "IAB Considerations for UNilateral 2007 Self-Address Fixing (UNSAF) Across Network Address 2008 Translation", RFC 3424, November 2002. 2010 [RFC2434] Narten, T. and H. Alvestrand, "Guidelines for Writing an 2011 IANA Considerations Section in RFCs", BCP 26, RFC 2434, 2012 October 1998. 2014 Appendix A. C Snippet to Determine STUN Message Types 2016 Given an 16-bit STUN message type value in host byte order in 2017 msg_type parameter, below are C macros to determine the STUN message 2018 types: 2020 #define IS_REQUEST(msg_type) (((msg_type) & 0x0110) == 0x0000) 2021 #define IS_INDICATION(msg_type) (((msg_type) & 0x0110) == 0x0010) 2022 #define IS_SUCCESS_RESP(msg_type) (((msg_type) & 0x0110) == 0x0100) 2023 #define IS_ERR_RESP(msg_type) (((msg_type) & 0x0110) == 0x0110) 2025 Authors' Addresses 2027 Jonathan Rosenberg 2028 Cisco 2029 Edison, NJ 2030 US 2032 Email: jdrosen@cisco.com 2033 URI: http://www.jdrosen.net 2034 Christian Huitema 2035 Microsoft 2036 One Microsoft Way 2037 Redmond, WA 98052 2038 US 2040 Email: huitema@microsoft.com 2042 Rohan Mahy 2043 Plantronics 2044 345 Encinal Street 2045 Santa Cruz, CA 95060 2046 US 2048 Email: rohan@ekabal.com 2050 Philip Matthews 2051 Avaya 2052 1135 Innovation Drive 2053 Ottawa, Ontario K2K 3G7 2054 Canada 2056 Phone: +1 613 592 4343 x224 2057 Fax: 2058 Email: philip_matthews@magma.ca 2059 URI: 2061 Dan Wing 2062 Cisco 2063 771 Alder Drive 2064 San Jose, CA 95035 2065 US 2067 Email: dwing@cisco.com 2069 Full Copyright Statement 2071 Copyright (C) The IETF Trust (2007). 2073 This document is subject to the rights, licenses and restrictions 2074 contained in BCP 78, and except as set forth therein, the authors 2075 retain all their rights. 2077 This document and the information contained herein are provided on an 2078 "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS 2079 OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND 2080 THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS 2081 OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF 2082 THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED 2083 WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. 2085 Intellectual Property 2087 The IETF takes no position regarding the validity or scope of any 2088 Intellectual Property Rights or other rights that might be claimed to 2089 pertain to the implementation or use of the technology described in 2090 this document or the extent to which any license under such rights 2091 might or might not be available; nor does it represent that it has 2092 made any independent effort to identify any such rights. Information 2093 on the procedures with respect to rights in RFC documents can be 2094 found in BCP 78 and BCP 79. 2096 Copies of IPR disclosures made to the IETF Secretariat and any 2097 assurances of licenses to be made available, or the result of an 2098 attempt made to obtain a general license or permission for the use of 2099 such proprietary rights by implementers or users of this 2100 specification can be obtained from the IETF on-line IPR repository at 2101 http://www.ietf.org/ipr. 2103 The IETF invites any interested party to bring to its attention any 2104 copyrights, patents or patent applications, or other proprietary 2105 rights that may cover technology that may be required to implement 2106 this standard. Please address the information to the IETF at 2107 ietf-ipr@ietf.org. 2109 Acknowledgment 2111 Funding for the RFC Editor function is provided by the IETF 2112 Administrative Support Activity (IASA).