idnits 2.17.1 draft-ietf-behave-rfc3489bis-08.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 2073. -- Found old boilerplate from RFC 3979, Section 5, paragraph 1 on line 2084. -- Found old boilerplate from RFC 3979, Section 5, paragraph 2 on line 2091. -- Found old boilerplate from RFC 3979, Section 5, paragraph 3 on line 2097. 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 27, 2007) is 6118 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 28, 2008 R. Mahy 7 Plantronics 8 P. Matthews 9 Avaya 10 D. Wing 11 Cisco 12 July 27, 2007 14 Session Traversal Utilities for (NAT) (STUN) 15 draft-ietf-behave-rfc3489bis-08 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 28, 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 . . . . . . . . . . . . . . . . . . . . . . . . . 8 70 6. STUN Message Structure . . . . . . . . . . . . . . . . . . . . 9 71 7. Base Protocol Procedures . . . . . . . . . . . . . . . . . . . 12 72 7.1. Forming a Request or an Indication . . . . . . . . . . . 12 73 7.2. Sending the Request or Indication . . . . . . . . . . . . 12 74 7.2.1. Sending over UDP . . . . . . . . . . . . . . . . . . . 12 75 7.2.2. Sending over TCP or TLS-over-TCP . . . . . . . . . . . 13 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 . . . . . . . 16 79 7.3.1.2. Sending the Success or Error Response . . . . . . 17 80 7.3.2. Processing an Indication . . . . . . . . . . . . . . . 17 81 7.3.3. Processing a Success Response . . . . . . . . . . . . 17 82 7.3.4. Processing an Error Response . . . . . . . . . . . . . 18 83 8. FINGERPRINT Mechanism . . . . . . . . . . . . . . . . . . . . 18 84 9. DNS Discovery of a Server . . . . . . . . . . . . . . . . . . 19 85 10. Authentication and Message-Integrity Mechanisms . . . . . . . 20 86 10.1. Short-Term Credential Mechanism . . . . . . . . . . . . . 20 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 . . . . . . . . . . . . . 22 91 10.2.1. Forming a Request . . . . . . . . . . . . . . . . . . 23 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 . . . . . . . . . . . . . . . . . . 25 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 . . . . . . . . . . . . . . . . . . . . . . . . . 27 102 14. STUN Attributes . . . . . . . . . . . . . . . . . . . . . . . 28 103 14.1. MAPPED-ADDRESS . . . . . . . . . . . . . . . . . . . . . 29 104 14.2. XOR-MAPPED-ADDRESS . . . . . . . . . . . . . . . . . . . 30 105 14.3. USERNAME . . . . . . . . . . . . . . . . . . . . . . . . 31 106 14.4. MESSAGE-INTEGRITY . . . . . . . . . . . . . . . . . . . . 31 107 14.5. FINGERPRINT . . . . . . . . . . . . . . . . . . . . . . . 32 108 14.6. ERROR-CODE . . . . . . . . . . . . . . . . . . . . . . . 32 109 14.7. REALM . . . . . . . . . . . . . . . . . . . . . . . . . . 34 110 14.8. NONCE . . . . . . . . . . . . . . . . . . . . . . . . . . 34 111 14.9. UNKNOWN-ATTRIBUTES . . . . . . . . . . . . . . . . . . . 34 112 14.10. SERVER . . . . . . . . . . . . . . . . . . . . . . . . . 35 113 14.11. ALTERNATE-SERVER . . . . . . . . . . . . . . . . . . . . 35 114 15. Security Considerations . . . . . . . . . . . . . . . . . . . 35 115 15.1. Attacks against the Protocol . . . . . . . . . . . . . . 35 116 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 . . . . . . . . . . . . . . . 37 123 15.3. Hash Agility Plan . . . . . . . . . . . . . . . . . . . . 38 124 16. IAB Considerations . . . . . . . . . . . . . . . . . . . . . . 38 125 17. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 38 126 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 . . . . . . . . . . . . . . . . 42 134 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 tool for dealing with NATs. It 142 provides a means for an endpoint to determine the IP address and port 143 allocated by a NAT that corresponds to its private IP address and 144 port. It also provides a way for an endpoint to keep a NAT binding 145 alive. With some extensions, the protocol can be used to do 146 connectivity checks between two endpoints [I-D.ietf-mmusic-ice], or 147 to relay packets between two endpoints [I-D.ietf-behave-turn]. 149 In keeping with its tool nature, this specification defines an 150 extensible packet format, defines operation over several transport 151 protocols, and provides for two forms of authentication. 153 STUN is intended to be used in context of one or more NAT traversal 154 solutions. These solutions are known as STUN usages. Each usage 155 describes how STUN is utilized to achieve the NAT traversal solution. 156 Typically, a usage indicates when STUN messages get sent, which 157 optional attributes to include, what server is used, and what 158 authentication mechanism is to be used. Interactive Connectivity 159 Establishment (ICE) [I-D.ietf-mmusic-ice] is one usage of ICE. SIP 160 Outbound [I-D.ietf-sip-outbound] is another usage of ICE. In some 161 cases, a usage will require extensions to STUN. A STUN extension can 162 be in the form of new methods, attributes, or error response codes. 163 More information on STUN usages can be found in Section 13. 165 2. Evolution from RFC 3489 167 STUN was originally defined in RFC 3489 [RFC3489]. That 168 specification, sometimes referred to as "classic STUN", represented 169 itself as a complete solution to the NAT traversal problem. In that 170 solution, a client would discover whether it was behind a NAT, 171 determine its NAT type, discover its IP address and port on the 172 public side of the outermost NAT, and then utilize that IP address 173 and port within the body of protocols, such as the Session Initiation 174 Protocol (SIP) [RFC3261]. However, experience since the publication 175 of RFC 3489 has found that classic STUN simply does not work 176 sufficiently well to be a deployable solution. The address and port 177 learned through classic STUN are sometimes usable for communications 178 with a peer, and sometimes not. Classic STUN provided no way to 179 discover whether it would, in fact, work or not, and it provided no 180 remedy in cases where it did not. Furthermore, classic STUN's 181 algorithm for classification of NAT types was found to be faulty, as 182 many NATs did not fit cleanly into the types defined there. Classic 183 STUN also had security vulnerabilities which required an extremely 184 complicated mechanism to address, and despite the complexity of the 185 mechanism, were not fully remedied. 187 For these reasons, this specification obsoletes RFC 3489, and instead 188 describes STUN as a tool that is utilized as part of a complete NAT 189 traversal solution. ICE is a complete NAT traversal solution for 190 protocols based on the offer/answer [RFC3264] methodology, such as 191 SIP. SIP Outbound is a complete solution for traversal of SIP 192 signaling, and it uses STUN in a very different way. Though it is 193 possible that a protocol may be able to use STUN by itself (classic 194 STUN) as a traversal solution, such usage is not described here and 195 is strongly discouraged for the reasons described above. 197 The on-the-wire protocol described here is changed only slightly from 198 classic STUN. The protocol now runs over TCP in addition to UDP. 199 Extensibility was added to the protocol in a more structured way. A 200 magic-cookie mechanism for demultiplexing STUN with application 201 protocols was added by stealing 32 bits from the 128 bit transaction 202 ID defined in RFC 3489, allowing the change to be backwards 203 compatible. Mapped addresses are encoded using a new exclusive-or 204 format. There are other, more minor changes. See Section 18 for a 205 more complete listing. 207 Due to the change in scope, STUN has also been renamed from "Simple 208 Traversal of UDP Through NAT" to "Session Traversal Utilities for 209 NAT". The acronym remains STUN, which is all anyone ever remembers 210 anyway. 212 3. Overview of Operation 214 This section is descriptive only. 216 /--------\ 217 // STUN \\ 218 | Agent | 219 \\ (server) // 220 \--------/ 222 +----------------+ Public Internet 223 ................| NAT 2 |....................... 224 +----------------+ 226 +----------------+ Private NET 2 227 ................| NAT 1 |....................... 228 +----------------+ 230 /--------\ 231 // STUN \\ 232 | Agent | 233 \\ (client) // Private NET 1 234 \--------/ 236 Figure 1: One possible STUN Configuration 238 One possible STUN configuration is shown in Figure 1. In this 239 configuration, there are two entities (called STUN agents) that 240 implement the STUN protocol. The lower agent in the figure is 241 connected to private network 1. This network connects to private 242 network 2 through NAT 1. Private network 2 connects to the public 243 Internet through NAT 2. The upper agent in the figure resides on the 244 public Internet. 246 STUN is a client-server protocol. It supports two types of 247 transactions. One is a request/response transaction in which a 248 client sends a request to a server, and the server returns a 249 response. The second is an indication transaction in which a client 250 sends an indication to the server and the server does not respond. 251 Both types of transactions include a transaction ID, which is a 252 randomly selected 96-bit number. For request/response transactions, 253 this transaction ID allows the client to associate the response with 254 the request that generated it; for indications, this simply serves as 255 a debugging aid. 257 All STUN messages start with a fixed header that includes a method, a 258 class, and the transaction ID. The method indicates which of the 259 various requests or indications this is; this specification defines 260 just one method, Binding, but other methods are expected to be 261 defined in other documents. The class indicates whether this is a 262 request, a success response, an error response, or an indication. 263 Following the fixed header comes zero or more attributes, which are 264 type-length-value extensions that convey additional information for 265 the specific message. 267 This document defines a single method called Binding. The Binding 268 method can be used either in request/response transactions or in 269 indication transactions. When used in request/response transactions, 270 the Binding method can be used to determine the particular "binding" 271 a NAT has allocated to a STUN client. When used in either request/ 272 response or in indication transactions, the Binding method can also 273 be used to keep these "bindings" alive. 275 In the Binding request/response transaction, a Binding Request is 276 sent from a STUN client to a STUN server. When the Binding Request 277 arrives at the STUN server, it may have passed through one or more 278 NATs between the STUN client and the STUN server (in Figure 1, there 279 were two such NATs). As the Binding Request message passes through a 280 NAT, the NAT will modify the source transport address (that is, the 281 source IP address and the source port) of the packet. As a result, 282 the source transport address of the request received by the server 283 will be the public IP address and port created by the NAT closest to 284 the server. This is called a reflexive transport address. The STUN 285 server copies that source transport address into an XOR-MAPPED- 286 ADDRESS attribute in the STUN Binding Response and sends the Binding 287 Response back to the the STUN client. As this packet passes back 288 through a NAT, the NAT will modify the destination transport address 289 in the IP header, but the transport address in the XOR-MAPPED-ADDRESS 290 attribute within the body of the STUN response will remain untouched. 291 In this way, the client can learn its reflexive transport address 292 allocated by the outermost NAT with respect to the STUN server. 294 In some usages, STUN must be multiplexed with other protocols (e.g., 295 [I-D.ietf-mmusic-ice], [I-D.ietf-sip-outbound]). In these usages, 296 there must be a way to inspect a packet and determine if it is a STUN 297 packet or not. STUN provides three fields in the STUN header with 298 fixed values that can be used for this purpose. If this is not 299 sufficient, then STUN packets can also contain a FINGERPRINT value 300 which can further be used to distinguish the packets. 302 STUN defines a set of optional procedures that a usage can decide to 303 use, called mechanisms. These mechanisms include DNS discovery, a 304 redirection technique to an alternate server, a fingerprint attribute 305 for demultiplexing, and two authentication and message integrity 306 exchanges. The authentication mechanisms revolve around the use of a 307 username, password, and message-integrity value. Two authentication 308 mechanisms, the long-term credential mechanism and the short-term 309 credential mechanism, are defined in this specification. Each usage 310 specifies the mechanisms allowed with that usage. 312 In the long-term credential mechanism, the client and server share a 313 pre-provisioned username and password and perform a digest challenge/ 314 response exchange inspired by (but differing in details) to the one 315 defined for HTTP [RFC2617]. In the short-term credential mechanism, 316 the client and the server exchange a username and password through 317 some out-of-band method prior to the STUN exchange. For example, in 318 the ICE usage [I-D.ietf-mmusic-ice] the two endpoints use out-of-band 319 signaling to exchange a username and password. These are used to 320 integrity protect and authenticate the request and response. There 321 is no challenge or nonce used. 323 4. Terminology 325 In this document, the key words "MUST", "MUST NOT", "REQUIRED", 326 "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", 327 and "OPTIONAL" are to be interpreted as described in BCP 14, RFC 2119 328 [RFC2119] and indicate requirement levels for compliant STUN 329 implementations. 331 5. Definitions 333 STUN Agent: An entity that implements the STUN protocol. Agents can 334 act as STUN clients for some transactions and as STUN servers for 335 other transactions. 337 STUN Client: A logical role in the STUN protocol. A STUN client 338 sends STUN requests or STUN indications, and receives STUN 339 responses. The term "STUN client" is also used colloquially to 340 refer to a STUN agent that only acts as a STUN client. 342 STUN Server: A logical role in the STUN protocol. A STUN server 343 receives STUN requests or STUN indications and sends STUN 344 responses. The term "STUN server" is also used colloquially to 345 refer to a STUN agent that only acts as a STUN server. 347 Transport Address: The combination of an IP address and port number 348 (such as a UDP or TCP port number). 350 Reflexive Transport Address: A transport address learned by a client 351 that identifies that client as seen by another host on an IP 352 network, typically a STUN server. When there is an intervening 353 NAT between the client and the other host, the reflexive transport 354 address represents the mapped address allocated to the client on 355 the public side of the NAT. Reflexive transport addresses are 356 learned from the mapped address attribute (MAPPED-ADDRESS or XOR- 357 MAPPED-ADDRESS) in STUN responses. 359 Mapped Address: Same meaning as Reflexive Address. This term is 360 retained only for for historic reasons and due to the naming of 361 the MAPPED-ADDRESS and XOR-MAPPED-ADDRESS attributes. 363 Long Term Credential: A username and associated password that 364 represent a shared secret between client and server. Long term 365 credentials are generally granted to the client when a subscriber 366 enrolls in a service and persist until the subscriber leaves the 367 service or explicitly changes the credential. 369 Long Term Password: The password from a long term credential. 371 Short Term Credential: A temporary username and associated password 372 which represent a shared secret between client and server. Short 373 term credentials are obtained through some kind of protocol 374 mechanism between the client server, preceding the STUN exchange. 375 A short term credential has an explicit temporal scope, which may 376 be based on a specific amount of time (such as 5 minutes) or on an 377 event (such as termination of a SIP dialog). The specific scope 378 of a short term credential is defined by the application usage. 380 Short Term Password: The password component of a short term 381 credential. 383 STUN Indication: A STUN message that does not receive a response 385 Attribute: The STUN term for a Type-Length-Value (TLV) object that 386 can be added to a STUN message. Attributes are divided into two 387 types: comprehension-required and comprehension-optional. STUN 388 agents can safely ignore comprehension-optional attributes they 389 don't understand, but cannot successfully process a message if it 390 contains comprehension-required attributes that are not 391 understood. 393 RTO: Retransmission TimeOut 395 6. STUN Message Structure 397 STUN messages are encoded in binary using network-oriented format 398 (most significant byte or octet first, also commonly known as big- 399 endian). The transmission order is described in detail in Appendix B 400 of RFC791 [RFC0791]. Unless otherwise noted, numeric constants are 401 in decimal (base 10). 403 All STUN messages MUST start with a 20-byte header followed by zero 404 or more Attributes. The STUN header contains a STUN message type, 405 magic cookie, transaction ID, and message length. 407 0 1 2 3 408 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 1 409 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 410 |0 0| STUN Message Type | Message Length | 411 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 412 | Magic Cookie | 413 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 414 | | 415 | Transaction ID (96 bits) | 416 | | 417 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 419 Figure 2: Format of STUN Message Header 421 The most significant two bits of every STUN message MUST be zeroes. 422 This can be used to differentiate STUN packets from other protocols 423 when STUN is multiplexed with other protocols on the same port. 425 The message type defines the message class (request, success 426 response, failure response, or indication) and the message method 427 (the primary function) of the STUN message. Although there are four 428 message classes, there are only two types of transactions in STUN: 429 request/response transactions (which consist of a request message and 430 a response message), and indication transactions (which consists a 431 single indication message). Response classes are split into error 432 and success responses to aid in quickly processing the STUN message. 434 The message type field is decomposed further into the following 435 structure: 437 +--+--+-+-+-+-+-+-+-+-+-+-+-+-+ 438 |M |M |M|M|M|C|M|M|M|C|M|M|M|M| 439 |11|10|9|8|7|1|6|5|4|0|3|2|1|0| 440 +--+--+-+-+-+-+-+-+-+-+-+-+-+-+ 442 Figure 3: Format of STUN Message Type Field 444 Here the bits in the message type field are shown as most-significant 445 (M11) through least-significant (M0). M11 through M0 represent a 12- 446 bit encoding of the method. C1 and C0 represent a 2 bit encoding of 447 the class. A class of 0b00 is a Request, a class of 0b01 is an 448 indication, a class of 0b10 is a success response, and a class of 449 0b11 is an error response. This specification defines a single 450 method, Binding. The method and class are orthogonal, so that four 451 each method, a request, success response, error response and 452 indication are defined for that method. 454 For example, a Binding Request has class=0b00 (request) and 455 method=0b000000000001 (Binding), and is encoded into the first 16 456 bits as 0x0001. A Binding response has class=0b10 (success response) 457 and method=0b000000000001, and is encoded into the first 16 bits as 458 0x0101. 460 Note: This unfortunate encoding is due to assignment of values in 461 [RFC3489] which did not consider encoding Indications, Success, 462 and Errors using bit fields. 464 The magic cookie field MUST contain the fixed value 0x2112A442 in 465 network byte order. In RFC 3489 [RFC3489], this field was part of 466 the transaction ID; placing the magic cookie in this location allows 467 a server to detect if the client will understand certain attributes 468 that were added in this revised specification. In addition, it aids 469 in distinguishing STUN packets from packets of other protocols when 470 STUN is multiplexed with those other protocols on the same port. 472 The transaction ID is a 96 bit identifier, used to uniquely identify 473 STUN transactions. The transaction ID is chosen by the STUN client. 474 It primarily serves to correlate requests with responses, though it 475 also plays a small role in helping to prevent certain types of 476 attacks. As such, the transaction ID MUST be uniformly and randomly 477 chosen from the interval 0 .. 2**96-1. Resends of the same request 478 reuse the same transaction ID, but the client MUST choose a new 479 transaction ID for new transactions unless the new request is bit- 480 wise identical to the previous request and sent from the same 481 transport address to the same IP address. Success and error 482 responses MUST carry the same transaction ID as their corresponding 483 request. When an agent is acting as a STUN server and STUN client on 484 the same port, the transaction IDs in requests sent by the agent have 485 no relationship to the transaction IDs in requests received by the 486 agent. 488 The message length MUST contain the size, in bytes, of the message 489 not including the 20 byte STUN header. Since all STUN attributes are 490 padded to a multiple of four bytes, the last two bits of this field 491 are always zero. This provides another way to distinguish STUN 492 packets from packets of other protocols. 494 Following the STUN fixed portion of the header are zero or more 495 attributes. Each attribute is TLV (type-length-value) encoded. The 496 details of the encoding, and of the attributes themselves is given in 497 Section 14. 499 7. Base Protocol Procedures 501 This section defines the base procedures of the STUN protocol. It 502 describes how messages are formed, how they are sent, and how they 503 are processed when they are received. It also defines the detailed 504 processing of the Binding method. Other sections in this document 505 describe optional procedures that a usage may elect to use in certain 506 situations. Other documents may define other extensions to STUN, by 507 adding new methods, new attributes, or new error response codes. 509 7.1. Forming a Request or an Indication 511 When formulating a request or indication message, the client MUST 512 follow the rules in Section 6 when creating the header. In addition, 513 the message class MUST be either "Request" or "Indication" (as 514 appropriate), and the method must be either Binding or some method 515 defined in another document. 517 The client then adds any attributes specified by the method or the 518 usage. For example, some usages may specify that the client use an 519 authentication method (Section 10) or the FINGERPRINT attribute 520 (Section 8). 522 For the Binding method with no authentication, no attributes are 523 required unless the usage specifies otherwise. 525 7.2. Sending the Request or Indication 527 The client then sends the request to the server. This document 528 specifies how to send STUN messages over UDP, TCP, or TLS-over-TCP; 529 other transport protocols may be added in the future. The STUN usage 530 must specify which transport protocol is used, and how the client 531 determines the IP address and port of the server. Section 9 532 describes a DNS-based method of determining the IP address and port 533 of a server which a usage may elect to use. STUN may be used with 534 anycast addresses, but only with UDP and in usages where 535 authentication is not used. 537 At any time, a client MAY have multiple outstanding STUN requests 538 with the same STUN server (that is, multiple transactions in 539 progress, with different transaction ids). 541 7.2.1. Sending over UDP 543 When running STUN over UDP it is possible that the STUN message might 544 be dropped by the network. Reliability of STUN request/response 545 transactions is accomplished through retransmissions of the request 546 message by the client application itself. STUN indications are not 547 retransmitted; thus indication transactions over UDP are not 548 reliable. 550 A client SHOULD retransmit a STUN request message starting with an 551 interval of RTO ("Retransmission TimeOut"), doubling after each 552 retransmission. The RTO is an estimate of the round-trip-time, and 553 is computed as described in RFC 2988 [RFC2988], with two exceptions. 554 First, the initial value for RTO SHOULD be configurable (rather than 555 the 3s recommended in RFC 2988) and SHOULD be greater than 100ms. In 556 fixed- line access links, a value of 100ms is RECOMMENDED. Secondly, 557 the value of RTO MUST NOT be rounded up to the nearest second. 558 Rather, a 1ms accuracy MUST be maintained. As with TCP, the usage of 559 Karn's algorithm is RECOMMENDED. When applied to STUN, it means that 560 RTT estimates SHOULD NOT be computed from STUN transactions which 561 result in the retransmission of a request. 563 The value for RTO SHOULD be cached by an client after the completion 564 of the transaction, and used as the starting value for RTO for the 565 next transaction to the same server (based on equality of IP 566 address). The value SHOULD be considered stale and discarded after 567 10 minutes. 569 Retransmissions continue until a response is received, or until a 570 total of 7 requests have been sent. If, after the last request, a 571 duration equal to 16 times the RTO has passed without a response, the 572 client SHOULD consider the transaction to have failed. A STUN 573 transaction over UDP is also considered failed if there has been a 574 transport failure of some sort, such as a fatal ICMP error. For 575 example, assuming an RTO of 100ms, requests would be sent at times 576 0ms, 100ms, 300ms, 700ms, 1500ms, 3100ms, and 6300ms. If the client 577 has not received a response after 7900ms, the client will consider 578 the transaction to have timed out. 580 7.2.2. Sending over TCP or TLS-over-TCP 582 For TCP and TLS-over-TCP, the client opens a TCP connection to the 583 server. 585 In some usage of STUN, STUN is sent as the only protocol over the TCP 586 connection. In this case, it can be sent without the aid of any 587 additional framing or demultiplexing. In other usages, or with other 588 extensions, it may be multiplexed with other data over a TCP 589 connection. In that case, STUN MUST be run on top of some kind of 590 framing protocol, specified by the usage or extension, which allows 591 for the agent to extract complete STUN messages and complete 592 application layer messages. 594 For TLS-over-TCP, the TLS_RSA_WITH_AES_128_CBC_SHA ciphersuite MUST 595 be supported at a minimum. Implementations MAY also support any 596 other ciphersuite. When it receives the TLS Certificate message, the 597 client SHOULD verify the certificate and inspect the site identified 598 by the certificate. If the certificate is invalid, revoked, or if it 599 does not identify the appropriate party, the client MUST NOT send the 600 STUN message or otherwise proceed with the STUN transaction. The 601 client MUST verify the identity of the server. To do that, it 602 follows the identification procedures defined in Section 3.1 of RFC 603 2818 [RFC2818]. Those procedures assume the client is dereferencing 604 a URI. For purposes of usage with this specification, the client 605 treats the domain name or IP address used in Section 8.1 as the host 606 portion of the URI that has been dereferenced. If DNS was not used, 607 the client MUST be configured with a set of authorized domains whose 608 certificates will be accepted. 610 Reliability of STUN over TCP and TLS-over-TCP is handled by TCP 611 itself, and there are no retransmissions at the STUN protocol level. 612 However, for a request/response transaction, if the client has not 613 received a response 7900ms after it sent the SYN to establish the 614 connection, it considers the transaction to have timed out. This 615 value has been chosen to equalize the TCP and UDP timeouts for the 616 default initial RTO. 618 In addition, if the client is unable to establish the TCP connection, 619 or the TCP connection is reset or fails before a response is 620 received, any request/response transaction in progress is considered 621 to have failed 623 The client MAY send multiple transactions over a single TCP (or TLS- 624 over-TCP) connection, and it MAY send another request before 625 receiving a response to the previous. The client SHOULD keep the 626 connection open until it 628 o has no further STUN requests or indications to send over that 629 connection, and; 631 o has no plans to use any resources (such as a mapped address 632 (MAPPED-ADDRESS or XOR-MAPPED-ADDRESS) or relayed address 633 [I-D.ietf-behave-turn]) that were learned though STUN requests 634 sent over that connection, and; 636 o if multiplexing other application protocols over that port, has 637 finished using that other application, and; 639 o if using that learned port with a remote peer, has established 640 communications with that remote peer, as is required by some TCP 641 NAT traversal techniques (e.g., [I-D.ietf-mmusic-ice-tcp]). 643 At the server end, the server SHOULD keep the connection open, and 644 let the client close it. If a server becomes overloaded and needs to 645 close connections to free up resources, it SHOULD close an existing 646 connection rather than reject new connection requests. The server 647 SHOULD NOT close a connection if a request was received over that 648 connection for which a response was not sent. A server MUST NOT ever 649 open a connection back towards the client in order to send a 650 response. 652 7.3. Receiving a STUN Message 654 This section specifies the processing of a STUN message. The 655 processing specified here is for STUN messages as defined in this 656 specification; additional rules for backwards compatibility are 657 defined in in Section 12. Those additional procedures are optional, 658 and usages can elect to utilize them. First, a set of processing 659 operations are applied that are independent of the class. This is 660 followed by class-specific processing, described in the subsections 661 which follow. 663 When a STUN agent receives a STUN message, it first checks that the 664 message obeys the rules of Section 6. It checks that the first two 665 bits are 0, that the magic cookie field has the correct value, that 666 the message length is sensible, and that the method value is a 667 supported method. If the message-class is Success Response or Error 668 Response, the agent checks that the transaction ID matches a 669 transaction that is still in progress. If the FINGERPRINT extension 670 is being used, the agent checks that the FINGERPRINT attribute is 671 present and contains the correct value. If any errors are detected, 672 the message is silently discarded. In the case when STUN is being 673 multiplexed with another protocol, an error may indicate that this is 674 not really a STUN message; in this case, the agent should try to 675 parse the message as a different protocol. 677 The STUN agent then does any checks that are required by a 678 authentication mechanism that the usage has specified (see 679 Section 10. 681 Once the authentication checks are done, the STUN agent checks for 682 unknown attributes and known-but-unexpected attributes in the 683 message. Unknown comprehension-optional attributes MUST be ignored 684 by the agent. Known-but-unexpected attributes SHOULD be ignored by 685 the agent. Unknown comprehension-required attributes cause 686 processing that depends on the message-class and is described below. 688 At this point, further processing depends on the message class of the 689 request. 691 7.3.1. Processing a Request 693 If the request contains one or more unknown comprehension-required 694 attributes, the server replies with an error response with an error 695 code of 420 (Unknown Attribute), and includes an UNKNOWN-ATTRIBUTES 696 attribute in the response that lists the unknown comprehension- 697 required attributes. 699 The server then does any additional checking that the method or the 700 specific usage requires. If all the checks succeed, the server 701 formulates a success response as described below. 703 If the request uses UDP transport and is a retransmission of a 704 request for which the server has already generated a success response 705 within the last 10 seconds, the server MUST retransmit the same 706 success response. One way for a server to do this is to remember all 707 transaction IDs received over UDP and their corresponding responses 708 in the last 10 seconds. Another way is to reprocess the request and 709 recompute the response. The latter technique MUST only be applied to 710 requests which are idempotent and result in the same success response 711 for the same request. The Binding method is considered to idempotent 712 in this way (even though certain rare network events could cause the 713 reflexive transport address value to change). Extensions to STUN 714 SHOULD state whether their request types have this property or not. 716 7.3.1.1. Forming a Success or Error Response 718 When forming the response (success or error), the server follows the 719 rules of section 6. The method of the response is the same as that 720 of the request, and the message class is either "Success Response" or 721 "Error Response". 723 For an error response, the server MUST add an ERROR-CODE attribute 724 containing the error code specified in the processing above. The 725 reason phrase is not fixed, but SHOULD be something suitable for the 726 error code. For certain errors, additional attributes are added to 727 the message. These attributes are spelled out in the description 728 where the error code is specified. For example, for an error code of 729 420 (Unknown Attribute), the server MUST include an UNKNOWN- 730 ATTRIBUTES attribute. Certain authentication errors also cause 731 attributes to be added (see Section 10). Extensions may define other 732 errors and/or additional attributes to add in error cases. 734 If the server authenticated the request using an authentication 735 mechanism, then the server SHOULD add the appropriate authentication 736 attributes to the response (see Section 10). 738 The server also adds any attributes required by the specific method 739 or usage. In addition, the server SHOULD add a SERVER attribute to 740 the message. 742 For the Binding method, no additional checking is required unless the 743 usage specifies otherwise. When forming the success response, the 744 server adds a XOR-MAPPED-ADDRESS attribute to the response, where the 745 contents of the attribute are the source transport address of the 746 request message. For UDP, this is the source IP address and source 747 UDP port of the request message. For TCP and TLS-over-TCP, this is 748 the source IP address and source TCP port of the TCP connection as 749 seen by the server. 751 7.3.1.2. Sending the Success or Error Response 753 The response (success or error) is sent over the same transport as 754 the request was received on. If the request was received over UDP, 755 the destination IP address and port of the response is the source IP 756 address and port of the received request message, and the source IP 757 address and port of the response is equal to the destination IP 758 address and port of the received request message. If the request was 759 received over TCP or TLS-over-TCP, the response is sent back on the 760 same TCP connection as the request was received on. 762 7.3.2. Processing an Indication 764 If the indication contains unknown comprehension-required attributes, 765 the indication is discarded and processing ceases. 767 The server then does any additional checking that the method or the 768 specific usage requires. If all the checks succeed, the server then 769 processes the indication. No response is generated for an 770 indication. 772 For the Binding method, no additional checking or processing is 773 required, unless the usage specifies otherwise. The mere receipt of 774 the message by the server has refreshed the "bindings" in the 775 intervening NATs. 777 Since indications are not re-transmitted over UDP (unlike requests), 778 there is no need to handle re-transmissions of indications at the 779 server. 781 7.3.3. Processing a Success Response 783 If the success response contains unknown comprehension-required 784 attributes, the response is discarded and the transaction is 785 considered to have failed. 787 The client then does any additional checking that the method or the 788 specific usage requires. If all the checks succeed, the client then 789 processes the success response. 791 For the Binding method, the client checks that the XOR-MAPPED-ADDRESS 792 attribute is present in the response. The client checks the address 793 family specified. If it is an unsupported address family, the 794 attribute SHOULD be ignored. If it is an unexpected but supported 795 address family (for example, the Binding transaction was sent over 796 IPv4, but the address family specified is IPv6), then the client MAY 797 accept and use the value. 799 7.3.4. Processing an Error Response 801 If the error response contains unknown comprehension-required 802 attributes, or if the error response does not contain an ERROR-CODE 803 attribute, then the transaction is simply considered to have failed. 805 The client then does any processing specified by the authentication 806 mechanism (see Section 10). This may result in a new transaction 807 attempt. 809 The processing at this point depends on the error-code, the method, 810 and the usage; the following are the default rules: 812 o If the error code is 300 through 399, the client SHOULD consider 813 the transaction as failed unless the ALTERNATE-SERVER extension is 814 being used. See Section 11. 816 o If the error code is 400 through 499, the client declares the 817 transaction failed; in the case of 420 (Unknown Attribute), the 818 response should contain a UNKNOWN-ATTRIBUTES attribute that gives 819 additional information. 821 o If the error code is 500 through 599, the client MAY resend the 822 request; clients that do so MUST limit the number of times they do 823 this. 825 Any other error code causes the client to consider the transaction 826 failed. 828 8. FINGERPRINT Mechanism 830 This section describes an optional mechanism for STUN that aids in 831 distinguishing STUN messages from packets of other protocols when the 832 two are multiplexed on the same transport address. This mechanism is 833 optional, and a STUN usage must describe if and when it is used. 835 In some usages, STUN messages are multiplexed on the same transport 836 address as other protocols, such as RTP. In order to apply the 837 processing described in Section 7, STUN messages must first be 838 separated from the application packets. Section 6 describes three 839 fixed fields in the STUN header that can be used for this purpose. 840 However, in some cases, these three fixed fields may not be 841 sufficient. 843 When the FINGERPRINT extension is used, an agent includes the 844 FINGERPRINT attribute in messages it sends to another agent. 845 Section 14.5 describes the placement and value of this attribute. 846 When the agent receives what it believes is a STUN message, then, in 847 addition to other basic checks, the agent also checks that the 848 message contains a FINGERPRINT attribute and that the attribute 849 contains the correct value (see Section 7.3. This additional check 850 helps the agent detect messages of other protocols that might 851 otherwise seem to be STUN messages. 853 9. DNS Discovery of a Server 855 This section describes an optional procedure for STUN that allows a 856 client to use DNS to determine the IP address and port of a server. 857 A STUN usage must describe if and when this extension is used. To 858 use this procedure, the client must have a domain name and a service 859 name; the usage must also describe how the client obtains these. 861 When a client wishes to locate a STUN server in the public Internet 862 that accepts Binding Request/Response transactions, the SRV service 863 name is "stun". STUN usages MAY define additional DNS SRV service 864 names. 866 The domain name is resolved to a transport address using the SRV 867 procedures specified in [RFC2782]. The DNS SRV service name is the 868 service name provided as input to this procedure. The protocol in 869 the SRV lookup is the transport protocol the client will run STUN 870 over: "udp" for UDP, "tcp" for TCP, and "tls" for TLS-over-TCP. If, 871 in the future, additional SRV records are defined for TLS over other 872 transport protocols, those will need to utilize an SRV transport 873 token of the form "tls-foo" for transport protocol "foo". 875 The procedures of RFC 2782 are followed to determine the server to 876 contact. RFC 2782 spells out the details of how a set of SRV records 877 are sorted and then tried. However, RFC2782 only states that the 878 client should "try to connect to the (protocol, address, service)" 879 without giving any details on what happens in the event of failure. 880 When following these procedures, if the STUN transaction times out 881 without receipt of a response, the client SHOULD retry the request to 882 the next server in the list of servers from the DNS SRV response. 883 Such a retry is only possible for request/response transmissions, 884 since indication transactions generate no response or timeout. 886 The default port for STUN requests is 3478, for both TCP and UDP. 887 Administrators SHOULD use this port in their SRV records for UDP and 888 TCP, but MAY use others. There is no default port for STUN over TLS, 889 however a STUN server SHOULD use a port number for TLS different from 890 3478 so that the server can determine whether the first message it 891 will receive after the TCP connection is set up, is a STUN message or 892 a TLS message. 894 If no SRV records were found, the client performs an A or AAAA record 895 lookup of the domain name. The result will be a list of IP 896 addresses, each of which can be contacted at the default port using 897 UDP or TCP, independent of the STUN usage. For usages that require 898 TLS, lack of SRV records is equivalent to a failure of the 899 transaction, since the request or indication MUST NOT be sent unless 900 SRV records provided a transport address specifically for TLS. 902 10. Authentication and Message-Integrity Mechanisms 904 This section defines two mechanisms for STUN that a client and server 905 can use to provide authentication and message-integrity; these two 906 mechanisms are known as the short-term credential mechanism and the 907 long-term credential mechanism. These two mechanisms are optional, 908 and each usage must specify if and when these mechanisms are used. 909 Consequently, both clients and servers will know which mechanism (if 910 any) to follow based on knowledge of which usage applies. For 911 example, a STUN server on the public Internet supporting ICE would 912 have no authentication, whereas the STUN server functionality in an 913 agent supporting connectivity checks would utilize short term 914 credentials. An overview of these two mechanisms is given in 915 Section 3. 917 Each mechanism specifies the additional processing required to use 918 that mechanism, extending the processing specified in Section 7. The 919 additional processing occurs in three different places: when forming 920 a message; when receiving a message immediately after the the basic 921 checks have been performed; and when doing the detailed processing of 922 error responses. 924 10.1. Short-Term Credential Mechanism 926 The short-term credential mechanism assumes that, prior to the STUN 927 transaction, the client and server have used some other protocol to 928 exchange a credential in the form of a username and password. This 929 credential is time-limited. The time-limit is defined by the usage. 930 As an example, in the ICE usage [I-D.ietf-mmusic-ice], the two 931 endpoints use out-of-band signaling to agree on a username and 932 password, and this username and password is applicable for the 933 duration of the media session. 935 This credential is used to form a message integrity check in each 936 request and in many responses. There is no challenge and response as 937 in the long term mechanism; consequently, replay is prevented by 938 virtue of the time-limited nature of the credential. 940 10.1.1. Forming a Request or Indication 942 For a request or indication message, the agent MUST include the 943 USERNAME and MESSAGE-INTEGRITY attributes in the message. The HMAC 944 for the MESSAGE-INTEGRITY attribute is computed as described in 945 Section 14.4. The key for the HMAC is the password. Note that the 946 password is never included in the request or indication. 948 10.1.2. Receiving a Request or Indication 950 After the agent has done the basic processing of a message, the agent 951 performs the checks listed below in order specified: 953 o If the message does not contain both a MESSAGE-INTEGRITY and a 954 USERNAME attribute: 956 * If the message is a request, the server MUST reject the request 957 with an error response. This response MUST use an error code 958 of 400 (Bad Request). 960 * If the message is an indication, the server MUST silently 961 discard the indication. 963 o If the USERNAME does not contain a username value currently valid 964 within the server: 966 * If the message is a request, the server MUST reject the request 967 with an error response. This response MUST use an error code 968 of 401 (Unauthorized). 970 * If the message is an indication, the server MUST silently 971 discard the indication. 973 o Using the password associated with the username, compute the value 974 for the message-integrity as described in Section 14.4. If the 975 resulting value does not match the contents of the MESSAGE- 976 INTEGRITY attribute: 978 * If the message is a request, the server MUST reject the request 979 with an error response. This response MUST use an error code 980 of 401 (Unauthorized). 982 * If the message is an indication, the server MUST silently 983 discard the indication. 985 If these checks pass, the server continues to process the request or 986 indication. Any response generated by the server MUST include the 987 MESSAGE-INTEGRITY attribute, computed using the password utilized to 988 authenticate the request. The response MUST NOT contain the USERNAME 989 attribute. 991 If any of the checks fail, the server MUST NOT include a MESSAGE- 992 INTEGRITY or USERNAME attribute in the error response. This is 993 because, in these failure cases, the server cannot determine the 994 shared secret necessary to compute MESSAGE-INTEGRITY. 996 10.1.3. Receiving a Response 998 The client looks for the MESSAGE-INTEGRITY attribute in the response. 999 If present, the client computes the message integrity over the 1000 response as defined in Section 14.4, using the same password it 1001 utilized for the request. If the resulting value matches the 1002 contents of the MESSAGE-INTEGRITY attribute, the response is 1003 considered authenticated. If the value does not match, or if 1004 MESSAGE-INTEGRITY was absent, the response MUST be discarded, as if 1005 it was never received. This means that retransmits, if applicable, 1006 will continue. 1008 10.2. Long-term Credential Mechanism 1010 The long-term credential mechanism relies on a long term credential, 1011 in the form of a username and password, that are shared between 1012 client and server. The credential is considered long-term since it 1013 is assumed that it is provisioned for a user, and remains in effect 1014 until the user is no longer a subscriber of the system, or is 1015 changed. This is basically a traditional "log-in" username and 1016 password given to users. 1018 Because these usernames and passwords are expected to be valid for 1019 extended periods of time, replay prevention is provided in the form 1020 of a digest challenge. In this mechanism, the client initially sends 1021 a request, without offering any credentials or any integrity checks. 1022 The server rejects this request, providing the user a realm (used to 1023 guide the user or agent in selection of a username and password) and 1024 a nonce. The nonce provides the replay protection. It is a cookie, 1025 selected by the server, and encoded in such a way as to indicate a 1026 duration of validity or client identity from which it is valid. The 1027 client retries the request, this time including its username, the 1028 realm, and echoing the nonce provided by the server. The client also 1029 includes a message-integrity, which provides an HMAC over the entire 1030 request, including the nonce. The server validates the nonce, and 1031 checks the message-integrity. If they match, the request is 1032 authenticated. If the nonce is no longer valid, it is considered 1033 "stale", and the server rejects the request, providing a new nonce. 1035 In subsequent requests to the same server, the client reuses the 1036 nonce, username, realm and password it used previously. In this way, 1037 subsequent requests are not rejected until the nonce becomes invalid 1038 by the server, in which case the rejection provides a new nonce to 1039 the client. 1041 Note that the long-term credential mechanism cannot be used to 1042 protect indications, since indications cannot be challenged. Usages 1043 utilizing indications must either use a short-term credential, or 1044 omit authentication and message integrity for them. 1046 Since the long-term credential mechanism is susceptible to offline 1047 dictionary attacks, deployments SHOULD utilize strong passwords. 1049 For STUN servers used in conjunction with SIP servers, it is 1050 desirable to use the same credentials for authentication to the SIP 1051 server and STUN server. Typically, SIP systems utilizing SIP's 1052 digest authentication mechanism do not actually store the password in 1053 the database. Rather, they store a value called H(A1), which is 1054 computed as: 1056 H(A1) = MD5(username ":" realm ":" password) 1058 If a system wishes to utilize this credential, the STUN password 1059 would be computed by taking the user-entered username and password, 1060 and using H(A1) as the STUN password. It is RECOMMENDED that clients 1061 utilize this construction for the STUN password. 1063 10.2.1. Forming a Request 1065 There are two cases when forming a request. In the first case, this 1066 is the first request from the client to the server (as identified by 1067 its IP address and port). In the second case, the client is 1068 submitting a subsequent request once a previous request/response 1069 transaction has completed successfully. 1071 10.2.1.1. First Request 1073 If the client has not completed a successful request/response 1074 transaction with the server, it SHOULD omit the USERNAME, MESSAGE- 1075 INTEGRITY, REALM, and NONCE attributes. In other words, the very 1076 first request is sent as if there were no authentication or message 1077 integrity applied. 1079 10.2.1.2. Subsequent Requests 1081 Once a request/response transaction has completed successfully, the 1082 client will have been been presented a realm and nonce by the server, 1083 and selected a username and password with which it authenticated. 1084 The client SHOULD cache the username, password, realm, and nonce for 1085 subsequent communications with the server. When the client sends a 1086 subsequent request, it SHOULD include the USERNAME, REALM, and NONCE 1087 attributes with these cached values. It SHOULD include a MESSAGE- 1088 INTEGRITY attributed, computed as described in Section 14.4 using the 1089 cached password as the key. 1091 10.2.2. Receiving a Request 1093 After the server has done the basic processing of a request, it 1094 performs the checks listed below in the order specified: 1096 o If the message: 1098 * does not contain a MESSAGE-INTEGRITY attribute, 1100 * OR, it contains a USERNAME whose value is not a valid username, 1102 the server MUST generate an error response with an error code of 1103 401 (Unauthorized). This response MUST include a REALM value. It 1104 is RECOMMENDED that the REALM value be the domain name of the 1105 provider of the STUN server. The response MUST include a NONCE, 1106 selected by the server. 1108 o If the message contains a MESSAGE-INTEGRITY attribute, but is 1109 missing the USERNAME, REALM or NONCE attributes, the server MUST 1110 generate an error response with an error code of 400 (Bad 1111 Request). 1113 o If the NONCE is no longer valid, the server MUST generate an error 1114 response with an error code of 438 (Stale Nonce). This response 1115 MUST include a NONCE and REALM attribute. 1117 o Using the password associated with the username in the USERNAME 1118 attribute, compute the value for the message-integrity as 1119 described in Section 14.4. If the resulting value does not match 1120 the contents of the MESSAGE-INTEGRITY attribute, the server MUST 1121 reject the request with an error response. This response MUST use 1122 an error code of 401 (Unauthorized). It MUST include a REALM and 1123 NONCE attribute. 1125 If these checks pass, the server continues to process the request or 1126 indication. Any response generated by the server MUST include the 1127 MESSAGE-INTEGRITY attribute, computed using the username and password 1128 utilized to authenticate the request. The REALM, NONCE, and USERNAME 1129 attributes SHOULD NOT be included. 1131 10.2.3. Receiving a Response 1133 If the response is an error response, with an error code of 401 1134 (Unauthorized), the client SHOULD retry the request with a new 1135 transaction. This request MUST contain a USERNAME, determined by the 1136 client as the appropriate username for the REALM from the error 1137 response. The request MUST contain the REALM, copied from the error 1138 response. The request MUST contain the NONCE, copied from the error 1139 response. The request MUST contain the MESSAGE-INTEGRITY attribute, 1140 computed using the password associated with the username in the 1141 USERNAME attribute. The client MUST NOT perform this retry if it is 1142 not changing the USERNAME or REALM or its associated password, from 1143 the previous attempt. 1145 If the response is an error response with an error code of 438 (Stale 1146 Nonce), the client MUST retry the request, using the new NONCE 1147 supplied in the 438 (Stale Nonce) response. This retry MUST also 1148 include the USERNAME, REALM and MESSAGE-INTEGRITY. 1150 The client looks for the MESSAGE-INTEGRITY attribute in the response 1151 (either success or failure). If present, the client computes the 1152 message integrity over the response as defined in Section 14.4, using 1153 the same password it utilized for the request. If the resulting 1154 value matches the contents of the MESSAGE-INTEGRITY attribute, the 1155 response is considered authenticated. If the value does not match, 1156 or if MESSAGE-INTEGRITY was absent, the response MUST be discarded, 1157 as if it was never received. This means that retransmits, if 1158 applicable, will continue. 1160 11. ALTERNATE-SERVER Mechanism 1162 This section describes a mechanism in STUN that allows a server to 1163 redirect a client to another server. This extension is optional, and 1164 a usage must define if and when this extension is used. To prevent 1165 denial-of-service attacks, this extension MUST only be used in 1166 situations where the client and server are using an authentication 1167 and message-integrity mechanism. 1169 A server using this extension redirects a client to another server by 1170 replying to a request message with an error response message with an 1171 error code of 300 (Try Alternate). The server MUST include a 1172 ALTERNATE-SERVER attribute in the error response. The error response 1173 message MUST be authenticated, which in practice means the request 1174 message must have passed the authentication checks. 1176 A client using this extension handles a 300 (Try Alternate) error 1177 code as follows. If the error response has passed the authentication 1178 checks, then the client looks for a ALTERNATE-SERVER attribute in the 1179 error response. If one is found, then the client considers the 1180 current transaction as failed, and re-attempts the request with the 1181 server specified in the attribute. The client SHOULD reuse any 1182 authentication credentials from the old request in the new 1183 transaction. 1185 12. Backwards Compatibility with RFC 3489 1187 This section define procedures that allow a degree of backwards 1188 compatible with the original protocol defined in RFC 3489 [RFC3489]. 1189 This mechanism is optional, meant to be utilized only in cases where 1190 a new client can connect to an old server, or vice-a-versa. A usage 1191 must define if and when this procedure is used. 1193 Section 18 lists all the changes between this specification and RFC 1194 3489 [RFC3489]. However, not all of these differences are important, 1195 because "classic STUN" was only used in a few specific ways. For the 1196 purposes of this extension, the important changes are the following. 1197 In RFC 3489: 1199 o UDP was the only supported transport; 1201 o The field that is now the Magic Cookie field was a part of the 1202 transaction id field, and transaction ids were 128 bits long; 1204 o The XOR-MAPPED-ADDRESS attribute did not exist, and the Binding 1205 method used the MAPPED-ADDRESS attribute instead; 1207 o There were two comprehension-required attributes, RESPONSE-ADDRESS 1208 and CHANGE-REQUEST, that have been removed from this 1209 specification; 1211 * These attributes are now part of the NAT Behavior Discovery 1212 usage. 1214 12.1. Changes to Client Processing 1216 A client that wants to interoperate with a [RFC3489] server SHOULD 1217 send a request message that uses the Binding method, contains no 1218 attributes, and uses UDP as the transport protocol to the server. If 1219 successful, the success response received from the server will 1220 contain a MAPPED-ADDRESS attribute rather than an XOR-MAPPED-ADDRESS 1221 attribute; other than this change, the processing of the response is 1222 identical to the procedures described above. 1224 12.2. Changes to Server Processing 1226 A STUN server can detect when a given Binding Request message was 1227 sent from an RFC 3489 [RFC3489] client by the absence of the correct 1228 value in the Magic Cookie field. When the server detects an RFC 3489 1229 client, it SHOULD copy the value seen in the Magic Cookie field in 1230 the Binding Request to the Magic Cookie field in the Binding Response 1231 message, and insert a MAPPED-ADDRESS attribute instead of an XOR- 1232 MAPPED-ADDRESS attribute. 1234 The client might, in rare situations, include either the RESPONSE- 1235 ADDRESS or CHANGE-REQUEST attributes. In these situations, the 1236 server will view these as unknown comprehension-required attributes 1237 and reply with an error response. Since the mechanisms utilizing 1238 those attributes are no longer supported, this behavior is 1239 acceptable. 1241 13. STUN Usages 1243 STUN by itself is not a solution to the NAT traversal problem. 1244 Rather, STUN defines a tool that can be used inside a larger 1245 solution. The term "STUN Usage" is used for any solution that uses 1246 STUN as a component. 1248 At the time of writing, three STUN usages are defined: Interactive 1249 Connectivity Establishment (ICE) [I-D.ietf-mmusic-ice], Client- 1250 initiated connections for SIP [I-D.ietf-sip-outbound], and NAT 1251 Behavior Discovery [I-D.ietf-behave-nat-behavior-discovery]. Other 1252 STUN usages may be defined in the future. 1254 A STUN usage defines how STUN is actually utilized - when to send 1255 requests, what to do with the responses, and which optional 1256 procedures defined here (or in an extension to STUN) are to be used. 1257 A usage would also define: 1259 o Which STUN methods are used; 1260 o What authentication and message integrity mechanisms are used; 1262 o What mechanisms are used to distinguish STUN messages from other 1263 messages. When STUN is run over TCP, a framing mechanism may be 1264 required; 1266 o How a STUN client determines the IP address and port of the STUN 1267 server; 1269 o Whether backwards compatibility to RFC 3489 is required; 1271 o What optional attributes defined here (such as FINGERPRINT and 1272 ALTERNATE-SERVER) or in other extensions are required. 1274 In addition, any STUN usage must consider the security implications 1275 of using STUN in that usage. A number of attacks against STUN are 1276 known (see the Security Considerations section in this document) and 1277 any usage must consider how these attacks can be thwarted or 1278 mitigated. 1280 Finally, a usage must consider whether its usage of STUN is an 1281 example of the Unilateral Self-Address Fixing approach to NAT 1282 traversal, and if so, address the questions raised in RFC 3424. 1284 14. STUN Attributes 1286 After the STUN header are zero or more attributes. Each attribute 1287 MUST be TLV encoded, with a 16 bit type, 16 bit length, and value. 1288 Each STUN attribute MUST end on a 32 bit boundary. As mentioned 1289 above, all fields in an attribute are transmitted most significant 1290 bit first. 1292 0 1 2 3 1293 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 1294 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1295 | Type | Length | 1296 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1297 | Value (variable) .... 1298 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1300 Figure 5: Format of STUN Attributes 1302 The value in the Length field MUST contain the length of the Value 1303 part of the attribute, prior to padding, measured in bytes. Since 1304 STUN aligns attributes on 32 bit boundaries, attributes whose content 1305 is not a multiple of 4 bytes are padded with 1, 2 or 3 bytes of 1306 padding so that its value contains a multiple of 4 bytes. The 1307 padding bits are ignored, and may be any value. 1309 Any attribute type MAY appear more than once in a STUN message. 1310 Unless specified otherwise, the order of appearance is significant: 1311 only the first occurance needs to be processed by a receiver, and any 1312 duplicates MAY be ignored by a receiver. 1314 To allow future revisions of this specification to add new attributes 1315 if needed, the attribute space is divided into two ranges. 1316 Attributes with type values between 0x0000 and 0x7FFF are 1317 comprehension-required attributes, which means that the STUN agent 1318 cannot successfully process the message unless it understands the 1319 attribute. Attributes with type values between 0x8000 and 0xFFFF are 1320 comprehension-optional attributes, which means that those attributes 1321 can be ignored by the STUN agent if it does not understand them. 1323 The STUN Attribute types defined by this specification are: 1325 Comprehension-required range (0x0000-0x7FFF): 1326 0x0000: (Reserved) 1327 0x0001: MAPPED-ADDRESS 1328 0x0006: USERNAME 1329 0x0007: (Reserved; was PASSWORD) 1330 0x0008: MESSAGE-INTEGRITY 1331 0x0009: ERROR-CODE 1332 0x000A: UNKNOWN-ATTRIBUTES 1333 0x0014: REALM 1334 0x0015: NONCE 1335 0x0020: XOR-MAPPED-ADDRESS 1337 Comprehension-optional range (0x8000-0xFFFF) 1338 0x8022: SERVER 1339 0x8023: ALTERNATE-SERVER 1340 0x8028: FINGERPRINT 1342 The rest of this section describes the format of the various 1343 attributes defined in this specification. 1345 14.1. MAPPED-ADDRESS 1347 The MAPPED-ADDRESS attribute indicates a reflexive transport address 1348 of the client. It consists of an eight bit address family, and a 1349 sixteen bit port, followed by a fixed length value representing the 1350 IP address. If the address family is IPv4, the address MUST be 32 1351 bits. If the address family is IPv6, the address MUST be 128 bits. 1352 All fields must be in network byte order. 1354 The format of the MAPPED-ADDRESS attribute is: 1356 0 1 2 3 1357 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 1358 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1359 |0 0 0 0 0 0 0 0| Family | Port | 1360 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1361 | | 1362 | Address (32 bits or 128 bits) | 1363 | | 1364 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1366 Figure 7: Format of MAPPED-ADDRESS attribute 1368 The address family can take on the following values: 1370 0x01:IPv4 1371 0x02:IPv6 1373 The first 8 bits of the MAPPED-ADDRESS MUST be set to 0 and MUST be 1374 ignored by receivers. These bits are present for aligning parameters 1375 on natural 32 bit boundaries. 1377 This attribute is used only by servers for achieving backwards 1378 compatibility with RFC 3489 [RFC3489] clients. 1380 14.2. XOR-MAPPED-ADDRESS 1382 The XOR-MAPPED-ADDRESS attribute is identical to the MAPPED-ADDRESS 1383 attribute, except that the reflexive transport address is obfuscated 1384 through the XOR function. 1386 The format of the XOR-MAPPED-ADDRESS is: 1388 0 1 2 3 1389 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 1390 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1391 |x x x x x x x x| Family | X-Port | 1392 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1393 | X-Address (Variable) 1394 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1396 Figure 9: Format of XOR-MAPPED-ADDRESS Attribute 1398 The Family represents the IP address family, and is encoded 1399 identically to the Family in MAPPED-ADDRESS. 1401 X-Port is the mapped port, exclusive or'd with most significant 16 1402 bits of the magic cookie. If the IP address family is IPv4, 1403 X-Address is the mapped IP address exclusive or'd with the magic 1404 cookie. If the IP address family is IPv6, the X-Address is the 1405 mapped IP address exclusively or'ed with the magic cookie and the 96- 1406 bit transaction ID. 1408 For example, using the "^" character to indicate exclusive or, if the 1409 IP address is 192.168.1.1 (0xc0a80101) and the port is 5555 (0x15B3), 1410 the X-Port would be 0x15B3 ^ 0x2112 = 0x34A1, and the X-Address would 1411 be 0xc0a80101 ^ 0x2112A442 = 0xe1baa543. 1413 The rules for encoding and processing the first 8 bits of the 1414 attribute's value, the rules for handling multiple occurrences of the 1415 attribute, and the rules for processing addresses families are the 1416 same as for MAPPED-ADDRESS. 1418 NOTE: XOR-MAPPED-ADDRESS and MAPPED-ADDRESS differ only in their 1419 encoding of the transport address. The former encodes the transport 1420 address by exclusive-or'ing it with the magic cookie. The latter 1421 encodes it directly in binary. RFC 3489 originally specified only 1422 MAPPED-ADDRESS. However, deployment experience found that some NATs 1423 rewrite the 32-bit binary payloads containing the NAT's public IP 1424 address, such as STUN's MAPPED-ADDRESS attribute, in the well-meaning 1425 but misguided attempt at providing a generic ALG function. Such 1426 behavior interferes with the operation of STUN and also causes 1427 failure of STUN's message integrity checking. 1429 14.3. USERNAME 1431 The USERNAME attribute is used for message integrity. It identifies 1432 the username and password combination used in the message integrity 1433 check. 1435 The value of USERNAME is a variable length value. It MUST contain a 1436 UTF-8 encoded sequence of less than 128 characters (which can be as 1437 long as 763 bytes). 1439 14.4. MESSAGE-INTEGRITY 1441 The MESSAGE-INTEGRITY attribute contains an HMAC-SHA1 [RFC2104] of 1442 the STUN message. The MESSAGE-INTEGRITY attribute can be present in 1443 any STUN message type. Since it uses the SHA1 hash, the HMAC will be 1444 20 bytes. The text used as input to HMAC is the STUN message, 1445 including the header, up to and including the attribute preceding the 1446 MESSAGE-INTEGRITY attribute. With the exception of the FINGERPRINT 1447 attribute, which appears after MESSAGE-INTEGRITY, agents MUST ignore 1448 all other attributes that follow MESSAGE-INTEGRITY. 1450 The key used as input to HMAC is the password. 1452 Based on the rules above, the hash includes the length field from the 1453 STUN message header. This length indicates the length of the entire 1454 message, including the MESSAGE-INTEGRITY attribute itself. 1455 Consequently, the MESSAGE-INTEGRITY attribute MUST be inserted into 1456 the message (with dummy content) prior to the computation of the 1457 integrity check. Once the computation is performed, the value of the 1458 attribute can be filled in. This ensures the length has the correct 1459 value when the hash is performed. Similarly, when validating the 1460 MESSAGE-INTEGRITY, the length field should be adjusted to point to 1461 the end of the MESSAGE-INTEGRITY attribute prior to calculating the 1462 HMAC. Such adjustment is necessary when attributes, such as 1463 FINTERPRINT, appear after MESSAGE-INTEGRITY. 1465 14.5. FINGERPRINT 1467 The FINGERPRINT attribute may be present in all STUN messages. The 1468 value of the attribute is computed as the CRC-32 of the STUN message 1469 up to (but excluding) the FINGERPRINT attribute itself, xor-d with 1470 the 32 bit value 0x5354554e (the XOR helps in cases where an 1471 application packet is also using CRC-32 in it). The 32 bit CRC is 1472 the one defined in ITU V.42 [ITU.V42.1994], which has a generator 1473 polynomial of x32+x26+x23+x22+x16+x12+x11+x10+x8+x7+x5+x4+x2+x+1. 1474 When present, the FINGERPRINT attribute MUST be the last attribute in 1475 the message, and thus will appear after MESSAGE-INTEGRITY. 1477 The FINGERPRINT attribute can aid in distinguishing STUN packets from 1478 packets of other protocols. See Section 8. 1480 When using the FINGERPRINT attribute in a message, the attribute is 1481 first placed into the message with a dummy value, then the CRC is 1482 computed, and then the value of the attribute is updated. If the 1483 MESSAGE-INTEGRITY attribute is also present, then it must be present 1484 with the correct message-integrity value before the CRC is computed, 1485 since the CRC is done over the value of the MESSAGE-INTEGRITY 1486 attribute as well. 1488 14.6. ERROR-CODE 1490 The ERROR-CODE attribute is used in Error Response messages. It 1491 contains a numeric error code value in the range of 300 to 699 plus a 1492 textual reason phrase encoded in UTF-8, and is consistent in its code 1493 assignments and semantics with SIP [RFC3261] and HTTP [RFC2616]. The 1494 reason phrase is meant for user consumption, and can be anything 1495 appropriate for the error code. Recommended reason phrases for the 1496 defined error codes are presented below. The reason phrase MUST be a 1497 UTF-8 encoded sequence of less than 128 characters (which can be as 1498 long as 763 bytes). 1500 To facilitate processing, the class of the error code (the hundreds 1501 digit) is encoded separately from the rest of the code. 1503 0 1 2 3 1504 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 1505 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1506 | Reserved, should be 0 |Class| Number | 1507 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1508 | Reason Phrase (variable) .. 1509 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1511 The Reserved bits SHOULD be 0, and are for alignment on 32-bit 1512 boundaries. Receivers MUST ignore these bits. The Class represents 1513 the hundreds digit of the error code. The value MUST be between 3 1514 and 6. The number represents the error code modulo 100, and its 1515 value MUST be between 0 and 99. 1517 The following error codes, along with their recommended reason 1518 phrases (in brackets) are defined: 1520 300 Try Alternate: The client should contact an alternate server for 1521 this request. This error response MUST only be sent if the 1522 request included a USERNAME attribute and a valid MESSAGE- 1523 INTEGRITY attribute; otherwise it MUST NOT be sent and error 1524 code 400 (Bad Request) is suggested. This error response MUST 1525 be protected with the MESSAGE-INTEGRITY attribute, and receivers 1526 MUST validate the MESSAGE-INTEGRITY of this response before 1527 redirecting themselves to an alternate server. 1529 Note: failure to generate and validate message-integrity 1530 for a 300 response allows an on-path attacker to falsify a 1531 300 response thus causing subsequent STUN messages to be 1532 sent to a victim. 1534 400 Bad Request: The request was malformed. The client SHOULD NOT 1535 retry the request without modification from the previous 1536 attempt. The server may not be able to generate a valid 1537 MESSAGE-INTEGRITY for this error, so the client MUST NOT expect 1538 a valid MESSAGE-INTEGRITY attribute on this response. 1540 401 Unauthorized: The request did not contain the expected MESSAGE- 1541 INTEGRITY attribute. The server MAY include the MESSAGE- 1542 INTEGRITY attribute in its error response. 1544 420 Unknown Attribute: The server received STUN packet containing a 1545 comprehension-required attribute which it did not understand. 1546 The server MUST put this unknown attribute in the UNKNOWN- 1547 ATTRIBUTE attribute of its error response. 1549 438 Stale Nonce: The NONCE used by the client was no longer valid. 1550 The client should retry, using the NONCE provided in the 1551 response. 1553 500 Server Error: The server has suffered a temporary error. The 1554 client should try again. 1556 14.7. REALM 1558 The REALM attribute may be present in requests and responses. It 1559 contains text which meets the grammar for "realm-value" as described 1560 in RFC 3261 [RFC3261] but without the double quotes and their 1561 surrounding whitespace. That is, it is an unquoted realm-value. It 1562 MUST be a UTF-8 encoded sequence of less than 128 characters (which 1563 can be as long as 763 bytes). 1565 Presence of the REALM attribute in a request indicates that long-term 1566 credentials are being used for authentication. Presence in certain 1567 error responses indicates that the server wishes the client to use a 1568 long-term credential for authentication. 1570 14.8. NONCE 1572 The NONCE attribute may be present in requests and responses. It 1573 contains a sequence of qdtext or quoted-pair, which are defined in 1574 RFC 3261 [RFC3261]. See RFC 2617 [RFC2617], Section 4.3, for 1575 guidance on selection of nonce values in a server. It MUST be less 1576 than 128 characters (which can be as long as 763 bytes). 1578 14.9. UNKNOWN-ATTRIBUTES 1580 The UNKNOWN-ATTRIBUTES attribute is present only in an error response 1581 when the response code in the ERROR-CODE attribute is 420. 1583 The attribute contains a list of 16 bit values, each of which 1584 represents an attribute type that was not understood by the server. 1586 0 1 2 3 1587 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 1588 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1589 | Attribute 1 Type | Attribute 2 Type | 1590 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1591 | Attribute 3 Type | Attribute 4 Type ... 1592 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1594 Figure 11: Format of UNKNOWN-ATTRIBUTES attribute 1596 Note: In [RFC3489], this field was padded to 32 by duplicating the 1597 last attribute. In this version of the specification, the normal 1598 padding rules for attributes are used instead. 1600 14.10. SERVER 1602 The server attribute contains a textual description of the software 1603 being used by the server, including manufacturer and version number. 1604 The attribute has no impact on operation of the protocol, and serves 1605 only as a tool for diagnostic and debugging purposes. The value of 1606 SERVER is variable length. It MUST be a UTF-8 encoded sequence of 1607 less than 128 characters (which can be as long as 763 bytes). 1609 14.11. ALTERNATE-SERVER 1611 The alternate server represents an alternate transport address 1612 identifying a different STUN server which the STUN client should try. 1614 It is encoded in the same way as MAPPED-ADDRESS, and thus refers to a 1615 single server by IP address. The IP address family MUST be identical 1616 to that of the source IP address of the request. 1618 This attribute MUST only appear in an error response that contains a 1619 MESSAGE-INTEGRITY attribute. This prevents it from being used in 1620 denial-of-service attacks. 1622 15. Security Considerations 1624 15.1. Attacks against the Protocol 1625 15.1.1. Outside Attacks 1627 An attacker can try to modify STUN messages in transit, in order to 1628 cause a failure in STUN operation. These attacks are detected for 1629 both requests and responses through the message integrity mechanism, 1630 using either a short term or long term credential. Of course, once 1631 detected, the manipulated packets will be dropped, causing the STUN 1632 transaction to effectively fail. This attack is possible only by an 1633 on-path attacker. 1635 An attacker that can observe, but not modify STUN messages in-transit 1636 (for example, an attacker present on a shared access medium, such as 1637 Wi-Fi), can see a STUN request, and then immediately send a STUN 1638 response, typically an error response, in order to disrupt STUN 1639 processing. This attack is also prevented for messages that utilize 1640 MESSAGE-INTEGRITY. However, some error responses, those related to 1641 authentication in particular, cannot be protected by MESSAGE- 1642 INTEGRITY. When STUN itself is run over a secure transport protocol 1643 (e.g., TLS), these attacks are completely mitigated. 1645 15.1.2. Inside Attacks 1647 A rogue client may try to launch a DoS attack against a server by 1648 sending it a large number of STUN requests. Fortunately, STUN 1649 requests can be processed statelessly by a server, making such 1650 attacks hard to launch. 1652 A rogue client may use a STUN server as a reflector, sending it 1653 requests with a falsified source IP address and port. In such a 1654 case, the response would be delivered to that source IP and port. 1655 There is no amplification with this attack, and it is mitigated by 1656 ingress source address filtering. 1658 15.2. Attacks Affecting the Usage 1660 This section lists attacks that might be launched against a usage of 1661 STUN. Each STUN usage must consider whether these attacks are 1662 applicable to it, and if so, discuss counter-measures. 1664 Most of the attacks in this section revolve around an attacker 1665 modifying the reflexive address learned by a STUN client through a 1666 Binding Request/Binding Response transaction. Since the usage of the 1667 reflexive address is a function of the usage, the applicability and 1668 remediation of these attacks is usage-specific. In common 1669 situations, modification of the reflexive address by an on-path 1670 attacker is easy to do. Consider, for example, the common situation 1671 where STUN is run directly over UDP. In this case, an on-path 1672 attacker can modify the source IP address of the Binding Request 1673 before it arrives at the STUN server. The STUN server will then 1674 return this IP address in the XOR-MAPPED-ADDRESS attribute to the 1675 client. Protecting against this attack by using a message-integrity 1676 check is impossible, since a message-integrity value cannot cover the 1677 source IP address, since the intervening NAT must be able to modify 1678 this value. Instead, one solution to preventing the attacks listed 1679 below is for the client to verify the reflexive address learned, as 1680 is done in ICE [I-D.ietf-mmusic-ice]. Other usages may use other 1681 means to prevent these attacks. 1683 15.2.1. Attack I: DDoS Against a Target 1685 In this attack, the attacker provides one or more clients with the 1686 same faked reflexive address that points to the intended target. 1687 This will trick the STUN clients into thinking that their reflexive 1688 addresses are equal to that of the target. If the clients hand out 1689 that reflexive address in order to receive traffic on it (for 1690 example, in SIP messages), the traffic will instead be sent to the 1691 target. This attack can provide substantial amplification, 1692 especially when used with clients that are using STUN to enable 1693 multimedia applications. 1695 15.2.2. Attack II: Silencing a Client 1697 In this attack, the attacker provides a STUN client with a faked 1698 reflexive address. The reflexive address it provides is a transport 1699 address that routes to nowhere. As a result, the client won't 1700 receive any of the packets it expects to receive when it hands out 1701 the reflexive address. This exploitation is not very interesting for 1702 the attacker. It impacts a single client, which is frequently not 1703 the desired target. Moreover, any attacker that can mount the attack 1704 could also deny service to the client by other means, such as 1705 preventing the client from receiving any response from the STUN 1706 server, or even a DHCP server. 1708 15.2.3. Attack III: Assuming the Identity of a Client 1710 This attack is similar to attack II. However, the faked reflexive 1711 address points to the attacker itself. This allows the attacker to 1712 receive traffic which was destined for the client. 1714 15.2.4. Attack IV: Eavesdropping 1716 In this attack, the attacker forces the client to use a reflexive 1717 address that routes to itself. It then forwards any packets it 1718 receives to the client. This attack would allow the attacker to 1719 observe all packets sent to the client. However, in order to launch 1720 the attack, the attacker must have already been able to observe 1721 packets from the client to the STUN server. In most cases (such as 1722 when the attack is launched from an access network), this means that 1723 the attacker could already observe packets sent to the client. This 1724 attack is, as a result, only useful for observing traffic by 1725 attackers on the path from the client to the STUN server, but not 1726 generally on the path of packets being routed towards the client. 1728 15.3. Hash Agility Plan 1730 This specification uses SHA-1 for computation of the message 1731 integrity. If, at a later time, SHA-1 is found to be compromised, 1732 the following is the remedy that will be applied. 1734 We will define a STUN extension which introduces a new message 1735 integrity attribute, computed using a new hash. Clients would be 1736 required to include both the new and old message integrity attributes 1737 in their requests or indications. A new server will utilize the new 1738 message integrity attribute, and an old one, the old. After a 1739 transition period where mixed implementations are in deployment, the 1740 old message-integrity attribute will be deprecated by another 1741 specification, and clients will cease including it in requests. 1743 16. IAB Considerations 1745 The IAB has studied the problem of "Unilateral Self Address Fixing" 1746 (UNSAF), which is the general process by which a client attempts to 1747 determine its address in another realm on the other side of a NAT 1748 through a collaborative protocol reflection mechanism (RFC3424 1749 [RFC3424]). STUN can be used to perform this function using a 1750 Binding Request/Response transaction if one agent is behind a NAT and 1751 the other is on the public side of the NAT. 1753 The IAB has mandated that protocols developed for this purpose 1754 document a specific set of considerations. Because some STUN usages 1755 provide UNSAF functions (such as ICE [I-D.ietf-mmusic-ice] ), and 1756 others do not (such as SIP Outbound [I-D.ietf-sip-outbound]), answers 1757 to these considerations need to be addressed by the usages 1758 themselves. 1760 17. IANA Considerations 1762 IANA is hereby requested to create three new registries: a STUN 1763 methods registry, a STUN Attributes registry, and a STUN Error Codes 1764 registry. 1766 17.1. STUN Methods Registry 1768 A STUN method is a hex number in the range 0x000 - 0x3FF. The 1769 encoding of STUN method into a STUN message is described in 1770 Section 6. 1772 The initial STUN methods are: 1774 0x000: (Reserved) 1775 0x001: Binding 1776 0x002: (Reserved; was SharedSecret) 1778 STUN methods in the range 0x000 - 0x1FF are assigned by IETF 1779 Consensus [RFC2434]. STUN methods in the range 0x200 - 0x3FF are 1780 assigned on a First Come First Served basis [RFC2434] 1782 17.2. STUN Attribute Registry 1784 A STUN Attribute type is a hex number in the range 0x0000 - 0xFFFF. 1785 STUN attribute types in the range 0x0000 - 0x7FFF are considered 1786 comprehension-required; STUN attribute types in the range 0x8000 - 1787 0xFFFF are considered comprehension-optional. A STUN agent handles 1788 unknown comprehension-required and comprehension-optional attributes 1789 differently. 1791 The initial STUN Attributes types are: 1793 Comprehension-required range (0x0000-0x7FFF): 1794 0x0000: (Reserved) 1795 0x0001: MAPPED-ADDRESS 1796 0x0006: USERNAME 1797 0x0007: (Reserved; was PASSWORD) 1798 0x0008: MESSAGE-INTEGRITY 1799 0x0009: ERROR-CODE 1800 0x000A: UNKNOWN-ATTRIBUTES 1801 0x0014: REALM 1802 0x0015: NONCE 1803 0x0020: XOR-MAPPED-ADDRESS 1805 Comprehension-optional range (0x8000-0xFFFF) 1806 0x8022: SERVER 1807 0x8023: ALTERNATE-SERVER 1808 0x8028: FINGERPRINT 1810 STUN Attribute types in the first half of the comprehension-required 1811 range (0x0000 - 0x3FFF) and in the first half of the comprehension- 1812 optional range (0x8000 - 0xBFFF) are assigned by IETF Consensus 1813 [RFC2434]. STUN Attribute types in the second half of the 1814 comprehension-required range (0x4000 - 0x7FFF) and in the second half 1815 of the comprehension-optional range (0xC000 - 0xFFFF) are assigned on 1816 a First Come First Served basis [RFC2434]. 1818 17.3. STUN Error Code Registry 1820 A STUN Error code is a number in the range 0 - 699. STUN error codes 1821 are accompanied by a textual reason phrase in UTF-8 which is intended 1822 only for human consumption and can be anything appropriate; this 1823 document proposes only suggested values. 1825 STUN error codes are consistent in codepoint assignments and 1826 semantics with SIP [RFC3261] and HTTP [RFC2616]. 1828 The initial values in this registry are given in Section 14.6. 1830 New STUN error codes are assigned on a Specification-Required basis 1831 [RFC2434]. The specification must carefully consider how clients 1832 that do not understand this error code will process it before 1833 granting the request. See the rules in Section 7.3.4. 1835 18. Changes Since RFC 3489 1837 This specification obsoletes RFC3489 [RFC3489]. This specification 1838 differs from RFC3489 in the following ways: 1840 o Removed the notion that STUN is a complete NAT traversal solution. 1841 STUN is now a tool that can be used to produce a NAT traversal 1842 solution. As a consequence, changed the name of the protocol to 1843 Session Traversal Utilities for NAT. 1845 o Introduced the concept of STUN usages, and described what a usage 1846 of STUN must document. 1848 o Removed the usage of STUN for NAT type detection and binding 1849 lifetime discovery. These techniques have proven overly brittle 1850 due to wider variations in the types of NAT devices than described 1851 in this document. Removed the RESPONSE-ADDRESS, CHANGED-ADDRESS, 1852 CHANGE-REQUEST, SOURCE-ADDRESS, and REFLECTED-FROM attributes. 1854 o Added a fixed 32-bit magic cookie and reduced length of 1855 transaction ID by 32 bits. The magic cookie begins at the same 1856 offset as the original transaction ID. 1858 o Added the XOR-MAPPED-ADDRESS attribute, which is included in 1859 Binding Responses if the magic cookie is present in the request. 1860 Otherwise the RFC3489 behavior is retained (that is, Binding 1861 Response includes MAPPED-ADDRESS). See discussion in XOR-MAPPED- 1862 ADDRESS regarding this change. 1864 o Introduced formal structure into the Message Type header field, 1865 with an explicit pair of bits for indication of request, response, 1866 error response or indication. Consequently, the message type 1867 field is split into the class (one of the previous four) and 1868 method. 1870 o Explicitly point out that the most significant two bits of STUN 1871 are 0b00, allowing easy differentiation with RTP packets when used 1872 with ICE. 1874 o Added the FINGERPRINT attribute to provide a method of definitely 1875 detecting the difference between STUN and another protocol when 1876 the two protocols are multiplexed together. 1878 o Added support for IPv6. Made it clear that an IPv4 client could 1879 get a v6 mapped address, and vice-a-versa. 1881 o Added long-term credential-based authentication. 1883 o Added the SERVER, REALM, NONCE, and ALTERNATE-SERVER attributes. 1885 o Removed the SharedSecret method, and thus the PASSWORD attribute. 1886 This method was almost never implemented and is not needed with 1887 current usages. 1889 o Removed recommendation to continue listening for STUN Responses 1890 for 10 seconds in an attempt to recognize an attack. 1892 o Changed transaction timers to be more TCP friendly. 1894 o Removed the STUN example that centered around the separation of 1895 the control and media planes. Instead, provided more information 1896 on using STUN with protocols. 1898 o Defined a generic padding mechanism that changes the 1899 interpretation of the length attribute. This would, in theory, 1900 break backwards compatibility. However, the mechanism in RFC 3489 1901 never worked for the few attributes that weren't aligned naturally 1902 on 32 bit boundaries. 1904 o REALM, USERNAME, SERVER, reason phrases and NONCE limited to 127 1905 characters. 1907 19. Acknowledgements 1909 The authors would like to thank Cedric Aoun, Pete Cordell, Cullen 1910 Jennings, Bob Penfield, Xavier Marjou, Bruce Lowekamp and Chris 1911 Sullivan for their comments, and Baruch Sterman and Alan Hawrylyshen 1912 for initial implementations. Thanks for Leslie Daigle, Allison 1913 Mankin, Eric Rescorla, and Henning Schulzrinne for IESG and IAB input 1914 on this work. 1916 20. References 1918 20.1. Normative References 1920 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1921 Requirement Levels", BCP 14, RFC 2119, March 1997. 1923 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, 1924 September 1981. 1926 [RFC2782] Gulbrandsen, A., Vixie, P., and L. Esibov, "A DNS RR for 1927 specifying the location of services (DNS SRV)", RFC 2782, 1928 February 2000. 1930 [RFC2818] Rescorla, E., "HTTP Over TLS", RFC 2818, May 2000. 1932 [RFC2617] Franks, J., Hallam-Baker, P., Hostetler, J., Lawrence, S., 1933 Leach, P., Luotonen, A., and L. Stewart, "HTTP 1934 Authentication: Basic and Digest Access Authentication", 1935 RFC 2617, June 1999. 1937 [RFC2988] Paxson, V. and M. Allman, "Computing TCP's Retransmission 1938 Timer", RFC 2988, November 2000. 1940 [ITU.V42.1994] 1941 International Telecommunications Union, "Error-correcting 1942 Procedures for DCEs Using Asynchronous-to-Synchronous 1943 Conversion", ITU-T Recommendation V.42, 1994. 1945 20.2. Informational References 1947 [RFC2104] Krawczyk, H., Bellare, M., and R. Canetti, "HMAC: Keyed- 1948 Hashing for Message Authentication", RFC 2104, 1949 February 1997. 1951 [RFC3261] Rosenberg, J., Schulzrinne, H., Camarillo, G., Johnston, 1952 A., Peterson, J., Sparks, R., Handley, M., and E. 1953 Schooler, "SIP: Session Initiation Protocol", RFC 3261, 1954 June 2002. 1956 [RFC2616] Fielding, R., Gettys, J., Mogul, J., Frystyk, H., 1957 Masinter, L., Leach, P., and T. Berners-Lee, "Hypertext 1958 Transfer Protocol -- HTTP/1.1", RFC 2616, June 1999. 1960 [I-D.ietf-mmusic-ice] 1961 Rosenberg, J., "Interactive Connectivity Establishment 1962 (ICE): A Protocol for Network Address Translator (NAT) 1963 Traversal for Offer/Answer Protocols", 1964 draft-ietf-mmusic-ice-16 (work in progress), June 2007. 1966 [RFC3489] Rosenberg, J., Weinberger, J., Huitema, C., and R. Mahy, 1967 "STUN - Simple Traversal of User Datagram Protocol (UDP) 1968 Through Network Address Translators (NATs)", RFC 3489, 1969 March 2003. 1971 [I-D.ietf-behave-turn] 1972 Rosenberg, J., "Obtaining Relay Addresses from Simple 1973 Traversal Underneath NAT (STUN)", 1974 draft-ietf-behave-turn-03 (work in progress), March 2007. 1976 [I-D.ietf-sip-outbound] 1977 Jennings, C. and R. Mahy, "Managing Client Initiated 1978 Connections in the Session Initiation Protocol (SIP)", 1979 draft-ietf-sip-outbound-09 (work in progress), June 2007. 1981 [I-D.ietf-behave-nat-behavior-discovery] 1982 MacDonald, D. and B. Lowekamp, "NAT Behavior Discovery 1983 Using STUN", draft-ietf-behave-nat-behavior-discovery-00 1984 (work in progress), February 2007. 1986 [I-D.ietf-mmusic-ice-tcp] 1987 Rosenberg, J., "TCP Candidates with Interactive 1988 Connectivity Establishment (ICE", 1989 draft-ietf-mmusic-ice-tcp-03 (work in progress), 1990 March 2007. 1992 [RFC3264] Rosenberg, J. and H. Schulzrinne, "An Offer/Answer Model 1993 with Session Description Protocol (SDP)", RFC 3264, 1994 June 2002. 1996 [RFC3424] Daigle, L. and IAB, "IAB Considerations for UNilateral 1997 Self-Address Fixing (UNSAF) Across Network Address 1998 Translation", RFC 3424, November 2002. 2000 [RFC2434] Narten, T. and H. Alvestrand, "Guidelines for Writing an 2001 IANA Considerations Section in RFCs", BCP 26, RFC 2434, 2002 October 1998. 2004 Appendix A. C Snippet to Determine STUN Message Types 2006 Given an 16-bit STUN message type value in host byte order in 2007 msg_type parameter, below are C macros to determine the STUN message 2008 types: 2010 #define IS_REQUEST(msg_type) (((msg_type) & 0x0110) == 0x0000) 2011 #define IS_INDICATION(msg_type) (((msg_type) & 0x0110) == 0x0010) 2012 #define IS_SUCCESS_RESP(msg_type) (((msg_type) & 0x0110) == 0x0100) 2013 #define IS_ERR_RESP(msg_type) (((msg_type) & 0x0110) == 0x0110) 2015 Authors' Addresses 2017 Jonathan Rosenberg 2018 Cisco 2019 Edison, NJ 2020 US 2022 Email: jdrosen@cisco.com 2023 URI: http://www.jdrosen.net 2025 Christian Huitema 2026 Microsoft 2027 One Microsoft Way 2028 Redmond, WA 98052 2029 US 2031 Email: huitema@microsoft.com 2033 Rohan Mahy 2034 Plantronics 2035 345 Encinal Street 2036 Santa Cruz, CA 95060 2037 US 2039 Email: rohan@ekabal.com 2040 Philip Matthews 2041 Avaya 2042 1135 Innovation Drive 2043 Ottawa, Ontario K2K 3G7 2044 Canada 2046 Phone: +1 613 592 4343 x224 2047 Fax: 2048 Email: philip_matthews@magma.ca 2049 URI: 2051 Dan Wing 2052 Cisco 2053 771 Alder Drive 2054 San Jose, CA 95035 2055 US 2057 Email: dwing@cisco.com 2059 Full Copyright Statement 2061 Copyright (C) The IETF Trust (2007). 2063 This document is subject to the rights, licenses and restrictions 2064 contained in BCP 78, and except as set forth therein, the authors 2065 retain all their rights. 2067 This document and the information contained herein are provided on an 2068 "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS 2069 OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND 2070 THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS 2071 OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF 2072 THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED 2073 WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. 2075 Intellectual Property 2077 The IETF takes no position regarding the validity or scope of any 2078 Intellectual Property Rights or other rights that might be claimed to 2079 pertain to the implementation or use of the technology described in 2080 this document or the extent to which any license under such rights 2081 might or might not be available; nor does it represent that it has 2082 made any independent effort to identify any such rights. Information 2083 on the procedures with respect to rights in RFC documents can be 2084 found in BCP 78 and BCP 79. 2086 Copies of IPR disclosures made to the IETF Secretariat and any 2087 assurances of licenses to be made available, or the result of an 2088 attempt made to obtain a general license or permission for the use of 2089 such proprietary rights by implementers or users of this 2090 specification can be obtained from the IETF on-line IPR repository at 2091 http://www.ietf.org/ipr. 2093 The IETF invites any interested party to bring to its attention any 2094 copyrights, patents or patent applications, or other proprietary 2095 rights that may cover technology that may be required to implement 2096 this standard. Please address the information to the IETF at 2097 ietf-ipr@ietf.org. 2099 Acknowledgment 2101 Funding for the RFC Editor function is provided by the IETF 2102 Administrative Support Activity (IASA).