idnits 2.17.1 draft-ietf-tram-turnbis-09.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- == There are 4 instances 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. -- The document has examples using IPv4 documentation addresses according to RFC6890, but does not use any IPv6 documentation addresses. Maybe there should be IPv6 examples, too? Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (October 30, 2016) is 2706 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Missing Reference: '0x4001' is mentioned on line 667, but not defined ** Obsolete normative reference: RFC 5389 (Obsoleted by RFC 8489) ** Obsolete normative reference: RFC 6145 (Obsoleted by RFC 7915) ** Obsolete normative reference: RFC 3697 (Obsoleted by RFC 6437) == Outdated reference: A later version (-21) exists of draft-ietf-tram-stunbis-08 -- Obsolete informational reference (is this intentional?): RFC 5245 (Obsoleted by RFC 8445, RFC 8839) -- Obsolete informational reference (is this intentional?): RFC 6156 (Obsoleted by RFC 8656) == Outdated reference: A later version (-20) exists of draft-ietf-tram-stun-pmtud-03 == Outdated reference: A later version (-12) exists of draft-ietf-tram-turn-server-discovery-10 -- Obsolete informational reference (is this intentional?): RFC 5766 (Obsoleted by RFC 8656) Summary: 3 errors (**), 0 flaws (~~), 6 warnings (==), 5 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 TRAM WG T. Reddy, Ed. 3 Internet-Draft Cisco Systems, Inc. 4 Intended status: Standards Track A. Johnston, Ed. 5 Expires: May 3, 2017 Unaffiliated 6 P. Matthews 7 Alcatel-Lucent 8 J. Rosenberg 9 jdrosen.net 10 October 30, 2016 12 Traversal Using Relays around NAT (TURN): Relay Extensions to Session 13 Traversal Utilities for NAT (STUN) 14 draft-ietf-tram-turnbis-09 16 Abstract 18 If a host is located behind a NAT, then in certain situations it can 19 be impossible for that host to communicate directly with other hosts 20 (peers). In these situations, it is necessary for the host to use 21 the services of an intermediate node that acts as a communication 22 relay. This specification defines a protocol, called TURN (Traversal 23 Using Relays around NAT), that allows the host to control the 24 operation of the relay and to exchange packets with its peers using 25 the relay. TURN differs from some other relay control protocols in 26 that it allows a client to communicate with multiple peers using a 27 single relay address. 29 The TURN protocol was designed to be used as part of the ICE 30 (Interactive Connectivity Establishment) approach to NAT traversal, 31 though it also can be used without ICE. 33 Status of This Memo 35 This Internet-Draft is submitted in full conformance with the 36 provisions of BCP 78 and BCP 79. 38 Internet-Drafts are working documents of the Internet Engineering 39 Task Force (IETF). Note that other groups may also distribute 40 working documents as Internet-Drafts. The list of current Internet- 41 Drafts is at http://datatracker.ietf.org/drafts/current/. 43 Internet-Drafts are draft documents valid for a maximum of six months 44 and may be updated, replaced, or obsoleted by other documents at any 45 time. It is inappropriate to use Internet-Drafts as reference 46 material or to cite them other than as "work in progress." 48 This Internet-Draft will expire on May 3, 2017. 50 Copyright Notice 52 Copyright (c) 2016 IETF Trust and the persons identified as the 53 document authors. All rights reserved. 55 This document is subject to BCP 78 and the IETF Trust's Legal 56 Provisions Relating to IETF Documents 57 (http://trustee.ietf.org/license-info) in effect on the date of 58 publication of this document. Please review these documents 59 carefully, as they describe your rights and restrictions with respect 60 to this document. Code Components extracted from this document must 61 include Simplified BSD License text as described in Section 4.e of 62 the Trust Legal Provisions and are provided without warranty as 63 described in the Simplified BSD License. 65 Table of Contents 67 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 4 68 2. Overview of Operation . . . . . . . . . . . . . . . . . . . . 6 69 2.1. Transports . . . . . . . . . . . . . . . . . . . . . . . 8 70 2.2. Allocations . . . . . . . . . . . . . . . . . . . . . . . 9 71 2.3. Permissions . . . . . . . . . . . . . . . . . . . . . . . 11 72 2.4. Send Mechanism . . . . . . . . . . . . . . . . . . . . . 12 73 2.5. Channels . . . . . . . . . . . . . . . . . . . . . . . . 14 74 2.6. Unprivileged TURN Servers . . . . . . . . . . . . . . . . 16 75 2.7. Avoiding IP Fragmentation . . . . . . . . . . . . . . . . 16 76 2.8. RTP Support . . . . . . . . . . . . . . . . . . . . . . . 18 77 2.9. Discovery of TURN server . . . . . . . . . . . . . . . . 18 78 2.9.1. TURN URI Scheme Semantics . . . . . . . . . . . . . . 18 79 3. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 18 80 4. General Behavior . . . . . . . . . . . . . . . . . . . . . . 20 81 5. Allocations . . . . . . . . . . . . . . . . . . . . . . . . . 22 82 6. Creating an Allocation . . . . . . . . . . . . . . . . . . . 24 83 6.1. Sending an Allocate Request . . . . . . . . . . . . . . . 24 84 6.2. Receiving an Allocate Request . . . . . . . . . . . . . . 25 85 6.3. Receiving an Allocate Success Response . . . . . . . . . 30 86 6.4. Receiving an Allocate Error Response . . . . . . . . . . 31 87 7. Refreshing an Allocation . . . . . . . . . . . . . . . . . . 33 88 7.1. Sending a Refresh Request . . . . . . . . . . . . . . . . 33 89 7.2. Receiving a Refresh Request . . . . . . . . . . . . . . . 34 90 7.3. Receiving a Refresh Response . . . . . . . . . . . . . . 34 91 8. Permissions . . . . . . . . . . . . . . . . . . . . . . . . . 35 92 9. CreatePermission . . . . . . . . . . . . . . . . . . . . . . 36 93 9.1. Forming a CreatePermission Request . . . . . . . . . . . 36 94 9.2. Receiving a CreatePermission Request . . . . . . . . . . 36 95 9.3. Receiving a CreatePermission Response . . . . . . . . . . 37 96 10. Send and Data Methods . . . . . . . . . . . . . . . . . . . . 37 97 10.1. Forming a Send Indication . . . . . . . . . . . . . . . 37 98 10.2. Receiving a Send Indication . . . . . . . . . . . . . . 38 99 10.3. Receiving a UDP Datagram . . . . . . . . . . . . . . . . 39 100 10.4. Receiving a Data Indication with DATA attribute . . . . 39 101 10.5. Receiving an ICMP Packet . . . . . . . . . . . . . . . . 40 102 10.6. Receiving a Data Indication with an ICMP attribute . . . 40 103 11. Channels . . . . . . . . . . . . . . . . . . . . . . . . . . 41 104 11.1. Sending a ChannelBind Request . . . . . . . . . . . . . 43 105 11.2. Receiving a ChannelBind Request . . . . . . . . . . . . 43 106 11.3. Receiving a ChannelBind Response . . . . . . . . . . . . 44 107 11.4. The ChannelData Message . . . . . . . . . . . . . . . . 45 108 11.5. Sending a ChannelData Message . . . . . . . . . . . . . 45 109 11.6. Receiving a ChannelData Message . . . . . . . . . . . . 46 110 11.7. Relaying Data from the Peer . . . . . . . . . . . . . . 47 111 12. Packet Translations . . . . . . . . . . . . . . . . . . . . . 47 112 12.1. IPv4-to-IPv6 Translations . . . . . . . . . . . . . . . 47 113 12.2. IPv6-to-IPv6 Translations . . . . . . . . . . . . . . . 48 114 12.3. IPv6-to-IPv4 Translations . . . . . . . . . . . . . . . 49 115 13. IP Header Fields . . . . . . . . . . . . . . . . . . . . . . 50 116 14. New STUN Methods . . . . . . . . . . . . . . . . . . . . . . 52 117 15. New STUN Attributes . . . . . . . . . . . . . . . . . . . . . 52 118 15.1. CHANNEL-NUMBER . . . . . . . . . . . . . . . . . . . . . 53 119 15.2. LIFETIME . . . . . . . . . . . . . . . . . . . . . . . . 53 120 15.3. XOR-PEER-ADDRESS . . . . . . . . . . . . . . . . . . . . 53 121 15.4. DATA . . . . . . . . . . . . . . . . . . . . . . . . . . 53 122 15.5. XOR-RELAYED-ADDRESS . . . . . . . . . . . . . . . . . . 53 123 15.6. REQUESTED-ADDRESS-FAMILY . . . . . . . . . . . . . . . . 54 124 15.7. EVEN-PORT . . . . . . . . . . . . . . . . . . . . . . . 54 125 15.8. REQUESTED-TRANSPORT . . . . . . . . . . . . . . . . . . 55 126 15.9. DONT-FRAGMENT . . . . . . . . . . . . . . . . . . . . . 55 127 15.10. RESERVATION-TOKEN . . . . . . . . . . . . . . . . . . . 55 128 15.11. ADDITIONAL-ADDRESS-FAMILY . . . . . . . . . . . . . . . 56 129 15.12. ADDRESS-ERROR-CODE Attribute . . . . . . . . . . . . . . 56 130 15.13. ICMP Attribute . . . . . . . . . . . . . . . . . . . . . 57 131 16. New STUN Error Response Codes . . . . . . . . . . . . . . . . 57 132 17. Detailed Example . . . . . . . . . . . . . . . . . . . . . . 58 133 18. Security Considerations . . . . . . . . . . . . . . . . . . . 65 134 18.1. Outsider Attacks . . . . . . . . . . . . . . . . . . . . 65 135 18.1.1. Obtaining Unauthorized Allocations . . . . . . . . . 65 136 18.1.2. Offline Dictionary Attacks . . . . . . . . . . . . . 66 137 18.1.3. Faked Refreshes and Permissions . . . . . . . . . . 66 138 18.1.4. Fake Data . . . . . . . . . . . . . . . . . . . . . 66 139 18.1.5. Impersonating a Server . . . . . . . . . . . . . . . 67 140 18.1.6. Eavesdropping Traffic . . . . . . . . . . . . . . . 67 141 18.1.7. TURN Loop Attack . . . . . . . . . . . . . . . . . . 68 142 18.2. Firewall Considerations . . . . . . . . . . . . . . . . 69 143 18.2.1. Faked Permissions . . . . . . . . . . . . . . . . . 69 144 18.2.2. Blacklisted IP Addresses . . . . . . . . . . . . . . 70 145 18.2.3. Running Servers on Well-Known Ports . . . . . . . . 70 147 18.3. Insider Attacks . . . . . . . . . . . . . . . . . . . . 70 148 18.3.1. DoS against TURN Server . . . . . . . . . . . . . . 70 149 18.3.2. Anonymous Relaying of Malicious Traffic . . . . . . 71 150 18.3.3. Manipulating Other Allocations . . . . . . . . . . . 71 151 18.4. Tunnel Amplification Attack . . . . . . . . . . . . . . 71 152 18.5. Other Considerations . . . . . . . . . . . . . . . . . . 72 153 19. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 72 154 20. IAB Considerations . . . . . . . . . . . . . . . . . . . . . 73 155 21. Changes since RFC 5766 . . . . . . . . . . . . . . . . . . . 75 156 22. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 75 157 23. References . . . . . . . . . . . . . . . . . . . . . . . . . 76 158 23.1. Normative References . . . . . . . . . . . . . . . . . . 76 159 23.2. Informative References . . . . . . . . . . . . . . . . . 77 160 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 80 162 1. Introduction 164 A host behind a NAT may wish to exchange packets with other hosts, 165 some of which may also be behind NATs. To do this, the hosts 166 involved can use "hole punching" techniques (see [RFC5128]) in an 167 attempt discover a direct communication path; that is, a 168 communication path that goes from one host to another through 169 intervening NATs and routers, but does not traverse any relays. 171 As described in [RFC5128] and [RFC4787], hole punching techniques 172 will fail if both hosts are behind NATs that are not well behaved. 173 For example, if both hosts are behind NATs that have a mapping 174 behavior of "address-dependent mapping" or "address- and port- 175 dependent mapping", then hole punching techniques generally fail. 177 When a direct communication path cannot be found, it is necessary to 178 use the services of an intermediate host that acts as a relay for the 179 packets. This relay typically sits in the public Internet and relays 180 packets between two hosts that both sit behind NATs. 182 This specification defines a protocol, called TURN, that allows a 183 host behind a NAT (called the TURN client) to request that another 184 host (called the TURN server) act as a relay. The client can arrange 185 for the server to relay packets to and from certain other hosts 186 (called peers) and can control aspects of how the relaying is done. 187 The client does this by obtaining an IP address and port on the 188 server, called the relayed transport address. When a peer sends a 189 packet to the relayed transport address, the server relays the packet 190 to the client. When the client sends a data packet to the server, 191 the server relays it to the appropriate peer using the relayed 192 transport address as the source. 194 A client using TURN must have some way to communicate the relayed 195 transport address to its peers, and to learn each peer's IP address 196 and port (more precisely, each peer's server-reflexive transport 197 address, see Section 2). How this is done is out of the scope of the 198 TURN protocol. One way this might be done is for the client and 199 peers to exchange email messages. Another way is for the client and 200 its peers to use a special-purpose "introduction" or "rendezvous" 201 protocol (see [RFC5128] for more details). 203 If TURN is used with ICE [RFC5245], then the relayed transport 204 address and the IP addresses and ports of the peers are included in 205 the ICE candidate information that the rendezvous protocol must 206 carry. For example, if TURN and ICE are used as part of a multimedia 207 solution using SIP [RFC3261], then SIP serves the role of the 208 rendezvous protocol, carrying the ICE candidate information inside 209 the body of SIP messages. If TURN and ICE are used with some other 210 rendezvous protocol, then [I-D.rosenberg-mmusic-ice-nonsip] provides 211 guidance on the services the rendezvous protocol must perform. 213 Though the use of a TURN server to enable communication between two 214 hosts behind NATs is very likely to work, it comes at a high cost to 215 the provider of the TURN server, since the server typically needs a 216 high-bandwidth connection to the Internet . As a consequence, it is 217 best to use a TURN server only when a direct communication path 218 cannot be found. When the client and a peer use ICE to determine the 219 communication path, ICE will use hole punching techniques to search 220 for a direct path first and only use a TURN server when a direct path 221 cannot be found. 223 TURN was originally invented to support multimedia sessions signaled 224 using SIP. Since SIP supports forking, TURN supports multiple peers 225 per relayed transport address; a feature not supported by other 226 approaches (e.g., SOCKS [RFC1928]). However, care has been taken to 227 make sure that TURN is suitable for other types of applications. 229 TURN was designed as one piece in the larger ICE approach to NAT 230 traversal. Implementors of TURN are urged to investigate ICE and 231 seriously consider using it for their application. However, it is 232 possible to use TURN without ICE. 234 TURN is an extension to the STUN (Session Traversal Utilities for 235 NAT) protocol [RFC5389]. Most, though not all, TURN messages are 236 STUN-formatted messages. A reader of this document should be 237 familiar with STUN. 239 2. Overview of Operation 241 This section gives an overview of the operation of TURN. It is non- 242 normative. 244 In a typical configuration, a TURN client is connected to a private 245 network [RFC1918] and through one or more NATs to the public 246 Internet. On the public Internet is a TURN server. Elsewhere in the 247 Internet are one or more peers with which the TURN client wishes to 248 communicate. These peers may or may not be behind one or more NATs. 249 The client uses the server as a relay to send packets to these peers 250 and to receive packets from these peers. 252 Peer A 253 Server-Reflexive +---------+ 254 Transport Address | | 255 192.0.2.150:32102 | | 256 | /| | 257 TURN | / ^| Peer A | 258 Client's Server | / || | 259 Host Transport Transport | // || | 260 Address Address | // |+---------+ 261 10.1.1.2:49721 192.0.2.15:3478 |+-+ // Peer A 262 | | ||N| / Host Transport 263 | +-+ | ||A|/ Address 264 | | | | v|T| 192.168.100.2:49582 265 | | | | /+-+ 266 +---------+| | | |+---------+ / +---------+ 267 | || |N| || | // | | 268 | TURN |v | | v| TURN |/ | | 269 | Client |----|A|----------| Server |------------------| Peer B | 270 | | | |^ | |^ ^| | 271 | | |T|| | || || | 272 +---------+ | || +---------+| |+---------+ 273 | || | | 274 | || | | 275 +-+| | | 276 | | | 277 | | | 278 Client's | Peer B 279 Server-Reflexive Relayed Transport 280 Transport Address Transport Address Address 281 192.0.2.1:7000 192.0.2.15:50000 192.0.2.210:49191 283 Figure 1 285 Figure 1 shows a typical deployment. In this figure, the TURN client 286 and the TURN server are separated by a NAT, with the client on the 287 private side and the server on the public side of the NAT. This NAT 288 is assumed to be a "bad" NAT; for example, it might have a mapping 289 property of "address-and-port-dependent mapping" (see [RFC4787]). 291 The client talks to the server from a (IP address, port) combination 292 called the client's HOST TRANSPORT ADDRESS. (The combination of an 293 IP address and port is called a TRANSPORT ADDRESS.) 295 The client sends TURN messages from its host transport address to a 296 transport address on the TURN server that is known as the TURN SERVER 297 TRANSPORT ADDRESS. The client learns the TURN server transport 298 address through some unspecified means (e.g., configuration), and 299 this address is typically used by many clients simultaneously. 301 Since the client is behind a NAT, the server sees packets from the 302 client as coming from a transport address on the NAT itself. This 303 address is known as the client's SERVER-REFLEXIVE transport address; 304 packets sent by the server to the client's server-reflexive transport 305 address will be forwarded by the NAT to the client's host transport 306 address. 308 The client uses TURN commands to create and manipulate an ALLOCATION 309 on the server. An allocation is a data structure on the server. 310 This data structure contains, amongst other things, the RELAYED 311 TRANSPORT ADDRESS for the allocation. The relayed transport address 312 is the transport address on the server that peers can use to have the 313 server relay data to the client. An allocation is uniquely 314 identified by its relayed transport address. 316 Once an allocation is created, the client can send application data 317 to the server along with an indication of to which peer the data is 318 to be sent, and the server will relay this data to the appropriate 319 peer. The client sends the application data to the server inside a 320 TURN message; at the server, the data is extracted from the TURN 321 message and sent to the peer in a UDP datagram. In the reverse 322 direction, a peer can send application data in a UDP datagram to the 323 relayed transport address for the allocation; the server will then 324 encapsulate this data inside a TURN message and send it to the client 325 along with an indication of which peer sent the data. Since the TURN 326 message always contains an indication of which peer the client is 327 communicating with, the client can use a single allocation to 328 communicate with multiple peers. 330 When the peer is behind a NAT, then the client must identify the peer 331 using its server-reflexive transport address rather than its host 332 transport address. For example, to send application data to Peer A 333 in the example above, the client must specify 192.0.2.150:32102 (Peer 334 A's server-reflexive transport address) rather than 335 192.168.100.2:49582 (Peer A's host transport address). 337 Each allocation on the server belongs to a single client and has 338 exactly one relayed transport address that is used only by that 339 allocation. Thus, when a packet arrives at a relayed transport 340 address on the server, the server knows for which client the data is 341 intended. 343 The client may have multiple allocations on a server at the same 344 time. 346 2.1. Transports 348 TURN, as defined in this specification, always uses UDP between the 349 server and the peer. However, this specification allows the use of 350 any one of UDP, TCP, Transport Layer Security (TLS) over TCP or 351 Datagram Transport Layer Security (DTLS) over UDP to carry the TURN 352 messages between the client and the server. 354 +----------------------------+---------------------+ 355 | TURN client to TURN server | TURN server to peer | 356 +----------------------------+---------------------+ 357 | UDP | UDP | 358 | TCP | UDP | 359 | TLS-over-TCP | UDP | 360 | DTLS-over-UDP | UDP | 361 +----------------------------+---------------------+ 363 If TCP or TLS-over-TCP is used between the client and the server, 364 then the server will convert between these transports and UDP 365 transport when relaying data to/from the peer. 367 Since this version of TURN only supports UDP between the server and 368 the peer, it is expected that most clients will prefer to use UDP 369 between the client and the server as well. That being the case, some 370 readers may wonder: Why also support TCP and TLS-over-TCP? 372 TURN supports TCP transport between the client and the server because 373 some firewalls are configured to block UDP entirely. These firewalls 374 block UDP but not TCP, in part because TCP has properties that make 375 the intention of the nodes being protected by the firewall more 376 obvious to the firewall. For example, TCP has a three-way handshake 377 that makes in clearer that the protected node really wishes to have 378 that particular connection established, while for UDP the best the 379 firewall can do is guess which flows are desired by using filtering 380 rules. Also, TCP has explicit connection teardown; while for UDP, 381 the firewall has to use timers to guess when the flow is finished. 383 TURN supports TLS-over-TCP transport and DTLS-over-UDP transport 384 between the client and the server because (D)TLS provides additional 385 security properties not provided by TURN's default digest 386 authentication; properties that some clients may wish to take 387 advantage of. In particular, (D)TLS provides a way for the client to 388 ascertain that it is talking to the correct server, and provides for 389 confidentiality of TURN control messages. TURN does not require 390 (D)TLS because the overhead of using (D)TLS is higher than that of 391 digest authentication; for example, using (D)TLS likely means that 392 most application data will be doubly encrypted (once by (D)TLS and 393 once to ensure it is still encrypted in the UDP datagram). 395 There is an extension to TURN for TCP transport between the server 396 and the peers [RFC6062]. For this reason, allocations that use UDP 397 between the server and the peers are known as UDP allocations, while 398 allocations that use TCP between the server and the peers are known 399 as TCP allocations. This specification describes only UDP 400 allocations. 402 In some applications for TURN, the client may send and receive 403 packets other than TURN packets on the host transport address it uses 404 to communicate with the server. This can happen, for example, when 405 using TURN with ICE. In these cases, the client can distinguish TURN 406 packets from other packets by examining the source address of the 407 arriving packet: those arriving from the TURN server will be TURN 408 packets. 410 2.2. Allocations 412 To create an allocation on the server, the client uses an Allocate 413 transaction. The client sends an Allocate request to the server, and 414 the server replies with an Allocate success response containing the 415 allocated relayed transport address. The client can include 416 attributes in the Allocate request that describe the type of 417 allocation it desires (e.g., the lifetime of the allocation). Since 418 relaying data has security implications, the server requires that the 419 client authenticate itself, typically using STUN's long-term 420 credential mechanism, to show that it is authorized to use the 421 server. 423 Once a relayed transport address is allocated, a client must keep the 424 allocation alive. To do this, the client periodically sends a 425 Refresh request to the server. TURN deliberately uses a different 426 method (Refresh rather than Allocate) for refreshes to ensure that 427 the client is informed if the allocation vanishes for some reason. 429 The frequency of the Refresh transaction is determined by the 430 lifetime of the allocation. The default lifetime of an allocation is 431 10 minutes -- this value was chosen to be long enough so that 432 refreshing is not typically a burden on the client, while expiring 433 allocations where the client has unexpectedly quit in a timely 434 manner. However, the client can request a longer lifetime in the 435 Allocate request and may modify its request in a Refresh request, and 436 the server always indicates the actual lifetime in the response. The 437 client must issue a new Refresh transaction within "lifetime" seconds 438 of the previous Allocate or Refresh transaction. Once a client no 439 longer wishes to use an allocation, it should delete the allocation 440 using a Refresh request with a requested lifetime of 0. 442 Both the server and client keep track of a value known as the 443 5-TUPLE. At the client, the 5-tuple consists of the client's host 444 transport address, the server transport address, and the transport 445 protocol used by the client to communicate with the server. At the 446 server, the 5-tuple value is the same except that the client's host 447 transport address is replaced by the client's server-reflexive 448 address, since that is the client's address as seen by the server. 450 Both the client and the server remember the 5-tuple used in the 451 Allocate request. Subsequent messages between the client and the 452 server use the same 5-tuple. In this way, the client and server know 453 which allocation is being referred to. If the client wishes to 454 allocate a second relayed transport address, it must create a second 455 allocation using a different 5-tuple (e.g., by using a different 456 client host address or port). 458 NOTE: While the terminology used in this document refers to 459 5-tuples, the TURN server can store whatever identifier it likes 460 that yields identical results. Specifically, an implementation 461 may use a file-descriptor in place of a 5-tuple to represent a TCP 462 connection. 464 TURN TURN Peer Peer 465 client server A B 466 |-- Allocate request --------------->| | | 467 | | | | 468 |<--------------- Allocate failure --| | | 469 | (401 Unauthorized) | | | 470 | | | | 471 |-- Allocate request --------------->| | | 472 | | | | 473 |<---------- Allocate success resp --| | | 474 | (192.0.2.15:50000) | | | 475 // // // // 476 | | | | 477 |-- Refresh request ---------------->| | | 478 | | | | 479 |<----------- Refresh success resp --| | | 480 | | | | 482 Figure 2 484 In Figure 2, the client sends an Allocate request to the server 485 without credentials. Since the server requires that all requests be 486 authenticated using STUN's long-term credential mechanism, the server 487 rejects the request with a 401 (Unauthorized) error code. The client 488 then tries again, this time including credentials (not shown). This 489 time, the server accepts the Allocate request and returns an Allocate 490 success response containing (amongst other things) the relayed 491 transport address assigned to the allocation. Sometime later, the 492 client decides to refresh the allocation and thus sends a Refresh 493 request to the server. The refresh is accepted and the server 494 replies with a Refresh success response. 496 2.3. Permissions 498 To ease concerns amongst enterprise IT administrators that TURN could 499 be used to bypass corporate firewall security, TURN includes the 500 notion of permissions. TURN permissions mimic the address-restricted 501 filtering mechanism of NATs that comply with [RFC4787]. 503 An allocation can have zero or more permissions. Each permission 504 consists of an IP address and a lifetime. When the server receives a 505 UDP datagram on the allocation's relayed transport address, it first 506 checks the list of permissions. If the source IP address of the 507 datagram matches a permission, the application data is relayed to the 508 client, otherwise the UDP datagram is silently discarded. 510 A permission expires after 5 minutes if it is not refreshed, and 511 there is no way to explicitly delete a permission. This behavior was 512 selected to match the behavior of a NAT that complies with [RFC4787]. 514 The client can install or refresh a permission using either a 515 CreatePermission request or a ChannelBind request. Using the 516 CreatePermission request, multiple permissions can be installed or 517 refreshed with a single request -- this is important for applications 518 that use ICE. For security reasons, permissions can only be 519 installed or refreshed by transactions that can be authenticated; 520 thus, Send indications and ChannelData messages (which are used to 521 send data to peers) do not install or refresh any permissions. 523 Note that permissions are within the context of an allocation, so 524 adding or expiring a permission in one allocation does not affect 525 other allocations. 527 2.4. Send Mechanism 529 There are two mechanisms for the client and peers to exchange 530 application data using the TURN server. The first mechanism uses the 531 Send and Data methods, the second way uses channels. Common to both 532 ways is the ability of the client to communicate with multiple peers 533 using a single allocated relayed transport address; thus, both ways 534 include a means for the client to indicate to the server which peer 535 should receive the data, and for the server to indicate to the client 536 which peer sent the data. 538 The Send mechanism uses Send and Data indications. Send indications 539 are used to send application data from the client to the server, 540 while Data indications are used to send application data from the 541 server to the client. 543 When using the Send mechanism, the client sends a Send indication to 544 the TURN server containing (a) an XOR-PEER-ADDRESS attribute 545 specifying the (server-reflexive) transport address of the peer and 546 (b) a DATA attribute holding the application data. When the TURN 547 server receives the Send indication, it extracts the application data 548 from the DATA attribute and sends it in a UDP datagram to the peer, 549 using the allocated relay address as the source address. Note that 550 there is no need to specify the relayed transport address, since it 551 is implied by the 5-tuple used for the Send indication. 553 In the reverse direction, UDP datagrams arriving at the relayed 554 transport address on the TURN server are converted into Data 555 indications and sent to the client, with the server-reflexive 556 transport address of the peer included in an XOR-PEER-ADDRESS 557 attribute and the data itself in a DATA attribute. Since the relayed 558 transport address uniquely identified the allocation, the server 559 knows which client should receive the data. 561 Some ICMP (Internet Control Message Protocol) packets arriving at the 562 relayed transport address on the TURN server may be converted into 563 Data indications and sent to the client, with the transport address 564 of the peer included in an XOR-PEER-ADDRESS attribute and the ICMP 565 type and code in a ICMP attribute. Data indications containing the 566 XOR-PEER-ADDRESS and ICMP attribute are also sent when using the 567 channel mechanism. 569 Send and Data indications cannot be authenticated, since the long- 570 term credential mechanism of STUN does not support authenticating 571 indications. This is not as big an issue as it might first appear, 572 since the client-to-server leg is only half of the total path to the 573 peer. Applications that want proper security should encrypt the data 574 sent between the client and a peer. 576 Because Send indications are not authenticated, it is possible for an 577 attacker to send bogus Send indications to the server, which will 578 then relay these to a peer. To partly mitigate this attack, TURN 579 requires that the client install a permission towards a peer before 580 sending data to it using a Send indication. 582 TURN TURN Peer Peer 583 client server A B 584 | | | | 585 |-- CreatePermission req (Peer A) -->| | | 586 |<-- CreatePermission success resp --| | | 587 | | | | 588 |--- Send ind (Peer A)-------------->| | | 589 | |=== data ===>| | 590 | | | | 591 | |<== data ====| | 592 |<-------------- Data ind (Peer A) --| | | 593 | | | | 594 | | | | 595 |--- Send ind (Peer B)-------------->| | | 596 | | dropped | | 597 | | | | 598 | |<== data ==================| 599 | dropped | | | 600 | | | | 602 Figure 3 604 In Figure 3, the client has already created an allocation and now 605 wishes to send data to its peers. The client first creates a 606 permission by sending the server a CreatePermission request 607 specifying Peer A's (server-reflexive) IP address in the XOR-PEER- 608 ADDRESS attribute; if this was not done, the server would not relay 609 data between the client and the server. The client then sends data 610 to Peer A using a Send indication; at the server, the application 611 data is extracted and forwarded in a UDP datagram to Peer A, using 612 the relayed transport address as the source transport address. When 613 a UDP datagram from Peer A is received at the relayed transport 614 address, the contents are placed into a Data indication and forwarded 615 to the client. Later, the client attempts to exchange data with Peer 616 B; however, no permission has been installed for Peer B, so the Send 617 indication from the client and the UDP datagram from the peer are 618 both dropped by the server. 620 2.5. Channels 622 For some applications (e.g., Voice over IP), the 36 bytes of overhead 623 that a Send indication or Data indication adds to the application 624 data can substantially increase the bandwidth required between the 625 client and the server. To remedy this, TURN offers a second way for 626 the client and server to associate data with a specific peer. 628 This second way uses an alternate packet format known as the 629 ChannelData message. The ChannelData message does not use the STUN 630 header used by other TURN messages, but instead has a 4-byte header 631 that includes a number known as a channel number. Each channel 632 number in use is bound to a specific peer and thus serves as a 633 shorthand for the peer's host transport address. 635 To bind a channel to a peer, the client sends a ChannelBind request 636 to the server, and includes an unbound channel number and the 637 transport address of the peer. Once the channel is bound, the client 638 can use a ChannelData message to send the server data destined for 639 the peer. Similarly, the server can relay data from that peer 640 towards the client using a ChannelData message. 642 Channel bindings last for 10 minutes unless refreshed -- this 643 lifetime was chosen to be longer than the permission lifetime. 644 Channel bindings are refreshed by sending another ChannelBind request 645 rebinding the channel to the peer. Like permissions (but unlike 646 allocations), there is no way to explicitly delete a channel binding; 647 the client must simply wait for it to time out. 649 TURN TURN Peer Peer 650 client server A B 651 | | | | 652 |-- ChannelBind req ---------------->| | | 653 | (Peer A to 0x4001) | | | 654 | | | | 655 |<---------- ChannelBind succ resp --| | | 656 | | | | 657 |-- [0x4001] data ------------------>| | | 658 | |=== data ===>| | 659 | | | | 660 | |<== data ====| | 661 |<------------------ [0x4001] data --| | | 662 | | | | 663 |--- Send ind (Peer A)-------------->| | | 664 | |=== data ===>| | 665 | | | | 666 | |<== data ====| | 667 |<------------------ [0x4001] data --| | | 668 | | | | 670 Figure 4 672 Figure 4 shows the channel mechanism in use. The client has already 673 created an allocation and now wishes to bind a channel to Peer A. To 674 do this, the client sends a ChannelBind request to the server, 675 specifying the transport address of Peer A and a channel number 676 (0x4001). After that, the client can send application data 677 encapsulated inside ChannelData messages to Peer A: this is shown as 678 "[0x4001] data" where 0x4001 is the channel number. When the 679 ChannelData message arrives at the server, the server transfers the 680 data to a UDP datagram and sends it to Peer A (which is the peer 681 bound to channel number 0x4001). 683 In the reverse direction, when Peer A sends a UDP datagram to the 684 relayed transport address, this UDP datagram arrives at the server on 685 the relayed transport address assigned to the allocation. Since the 686 UDP datagram was received from Peer A, which has a channel number 687 assigned to it, the server encapsulates the data into a ChannelData 688 message when sending the data to the client. 690 Once a channel has been bound, the client is free to intermix 691 ChannelData messages and Send indications. In the figure, the client 692 later decides to use a Send indication rather than a ChannelData 693 message to send additional data to Peer A. The client might decide 694 to do this, for example, so it can use the DONT-FRAGMENT attribute 695 (see the next section). However, once a channel is bound, the server 696 will always use a ChannelData message, as shown in the call flow. 698 Note that ChannelData messages can only be used for peers to which 699 the client has bound a channel. In the example above, Peer A has 700 been bound to a channel, but Peer B has not, so application data to 701 and from Peer B would use the Send mechanism. 703 2.6. Unprivileged TURN Servers 705 This version of TURN is designed so that the server can be 706 implemented as an application that runs in user space under commonly 707 available operating systems without requiring special privileges. 708 This design decision was made to make it easy to deploy a TURN 709 server: for example, to allow a TURN server to be integrated into a 710 peer-to-peer application so that one peer can offer NAT traversal 711 services to another peer. 713 This design decision has the following implications for data relayed 714 by a TURN server: 716 o The value of the Diffserv field may not be preserved across the 717 server; 719 o The Time to Live (TTL) field may be reset, rather than 720 decremented, across the server; 722 o The Explicit Congestion Notification (ECN) field may be reset by 723 the server; 725 o There is no end-to-end fragmentation, since the packet is re- 726 assembled at the server. 728 Future work may specify alternate TURN semantics that address these 729 limitations. 731 2.7. Avoiding IP Fragmentation 733 For reasons described in [Frag-Harmful], applications, especially 734 those sending large volumes of data, should try hard to avoid having 735 their packets fragmented. Applications using TCP can more or less 736 ignore this issue because fragmentation avoidance is now a standard 737 part of TCP, but applications using UDP (and thus any application 738 using this version of TURN) must handle fragmentation avoidance 739 themselves. 741 The application running on the client and the peer can take one of 742 two approaches to avoid IP fragmentation. 744 The first approach is to avoid sending large amounts of application 745 data in the TURN messages/UDP datagrams exchanged between the client 746 and the peer. This is the approach taken by most VoIP (Voice-over- 747 IP) applications. In this approach, the application exploits the 748 fact that the IP specification [RFC0791] specifies that IP packets up 749 to 576 bytes should never need to be fragmented. 751 The exact amount of application data that can be included while 752 avoiding fragmentation depends on the details of the TURN session 753 between the client and the server: whether UDP, TCP, or (D)TLS 754 transport is used, whether ChannelData messages or Send/Data 755 indications are used, and whether any additional attributes (such as 756 the DONT-FRAGMENT attribute) are included. Another factor, which is 757 hard to determine, is whether the MTU is reduced somewhere along the 758 path for other reasons, such as the use of IP-in-IP tunneling. 760 As a guideline, sending a maximum of 500 bytes of application data in 761 a single TURN message (by the client on the client-to-server leg) or 762 a UDP datagram (by the peer on the peer-to-server leg) will generally 763 avoid IP fragmentation. To further reduce the chance of 764 fragmentation, it is recommended that the client use ChannelData 765 messages when transferring significant volumes of data, since the 766 overhead of the ChannelData message is less than Send and Data 767 indications. 769 The second approach the client and peer can take to avoid 770 fragmentation is to use a path MTU discovery algorithm to determine 771 the maximum amount of application data that can be sent without 772 fragmentation. The classic path MTU discovery algorithm defined in 773 [RFC1191] may not be able to discover the MTU of the transmission 774 path between the client and the peer since: 776 - a probe packet with DF bit set to test a path for a larger MTU 777 can be dropped by routers, or 779 - ICMP error messages can be dropped by middle boxes. 781 As a result, the client and server need to use a path MTU discovery 782 algorithm that does not require ICMP messages. The Packetized Path 783 MTU Discovery algorithm defined in [RFC4821] is one such algorithm. 785 [I-D.ietf-tram-stun-pmtud] is an implementation of [RFC4821] that is 786 using STUN to discover the PMTUD, and so may be a suitable approach 787 to be used in conjunction with a TURN server, together with the DONT- 788 FRAGMENT attribute. When the client includes the DONT-FRAGMENT 789 attribute in a Send indication, this tells the server to set the DF 790 bit in the resulting UDP datagram that it sends to the peer. Since 791 some servers may be unable to set the DF bit, the client should also 792 include this attribute in the Allocate request -- any server that 793 does not support the DONT-FRAGMENT attribute will indicate this by 794 rejecting the Allocate request. 796 2.8. RTP Support 798 One of the envisioned uses of TURN is as a relay for clients and 799 peers wishing to exchange real-time data (e.g., voice or video) using 800 RTP. To facilitate the use of TURN for this purpose, TURN includes 801 some special support for older versions of RTP. 803 Old versions of RTP [RFC3550] required that the RTP stream be on an 804 even port number and the associated RTP Control Protocol (RTCP) 805 stream, if present, be on the next highest port. To allow clients to 806 work with peers that still require this, TURN allows the client to 807 request that the server allocate a relayed transport address with an 808 even port number, and to optionally request the server reserve the 809 next-highest port number for a subsequent allocation. 811 2.9. Discovery of TURN server 813 Methods of TURN server discovery, including using anycast, are 814 described in [I-D.ietf-tram-turn-server-discovery]. The syntax of 815 the "turn" and "turn" URIs are defined in Section 3.1 of [RFC7065]. 817 2.9.1. TURN URI Scheme Semantics 819 The "turn" and "turns" URI schemes are used to designate a TURN 820 server (also known as a relay) on Internet hosts accessible using the 821 TURN protocol. The TURN protocol supports sending messages over UDP, 822 TCP, TLS-over-TCP or DTLS-over-UDP. The "turns" URI scheme MUST be 823 used when TURN is run over TLS-over-TCP or in DTLS-over-UDP, and the 824 "turn" scheme MUST be used otherwise. The required part of 825 the "turn" URI denotes the TURN server host. The part, if 826 present, denotes the port on which the TURN server is awaiting 827 connection requests. If it is absent, the default port is 3478 for 828 both UDP and TCP. The default port for TURN over TLS and TURN over 829 DTLS is 5349. 831 3. Terminology 833 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 834 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 835 document are to be interpreted as described in RFC 2119 [RFC2119]. 837 Readers are expected to be familiar with [RFC5389] and the terms 838 defined there. 840 The following terms are used in this document: 842 TURN: The protocol spoken between a TURN client and a TURN server. 843 It is an extension to the STUN protocol [RFC5389]. The protocol 844 allows a client to allocate and use a relayed transport address. 846 TURN client: A STUN client that implements this specification. 848 TURN server: A STUN server that implements this specification. It 849 relays data between a TURN client and its peer(s). 851 Peer: A host with which the TURN client wishes to communicate. The 852 TURN server relays traffic between the TURN client and its 853 peer(s). The peer does not interact with the TURN server using 854 the protocol defined in this document; rather, the peer receives 855 data sent by the TURN server and the peer sends data towards the 856 TURN server. 858 Transport Address: The combination of an IP address and a port. 860 Host Transport Address: A transport address on a client or a peer. 862 Server-Reflexive Transport Address: A transport address on the 863 "public side" of a NAT. This address is allocated by the NAT to 864 correspond to a specific host transport address. 866 Relayed Transport Address: A transport address on the TURN server 867 that is used for relaying packets between the client and a peer. 868 A peer sends to this address on the TURN server, and the packet is 869 then relayed to the client. 871 TURN Server Transport Address: A transport address on the TURN 872 server that is used for sending TURN messages to the server. This 873 is the transport address that the client uses to communicate with 874 the server. 876 Peer Transport Address: The transport address of the peer as seen by 877 the server. When the peer is behind a NAT, this is the peer's 878 server-reflexive transport address. 880 Allocation: The relayed transport address granted to a client 881 through an Allocate request, along with related state, such as 882 permissions and expiration timers. 884 5-tuple: The combination (client IP address and port, server IP 885 address and port, and transport protocol (currently one of UDP, 886 TCP, or (D)TLS)) used to communicate between the client and the 887 server. The 5-tuple uniquely identifies this communication 888 stream. The 5-tuple also uniquely identifies the Allocation on 889 the server. 891 Channel: A channel number and associated peer transport address. 892 Once a channel number is bound to a peer's transport address, the 893 client and server can use the more bandwidth-efficient ChannelData 894 message to exchange data. 896 Permission: The IP address and transport protocol (but not the port) 897 of a peer that is permitted to send traffic to the TURN server and 898 have that traffic relayed to the TURN client. The TURN server 899 will only forward traffic to its client from peers that match an 900 existing permission. 902 Realm: A string used to describe the server or a context within the 903 server. The realm tells the client which username and password 904 combination to use to authenticate requests. 906 Nonce: A string chosen at random by the server and included in the 907 message-digest. To prevent reply attacks, the server should 908 change the nonce regularly. 910 4. General Behavior 912 This section contains general TURN processing rules that apply to all 913 TURN messages. 915 TURN is an extension to STUN. All TURN messages, with the exception 916 of the ChannelData message, are STUN-formatted messages. All the 917 base processing rules described in [RFC5389] apply to STUN-formatted 918 messages. This means that all the message-forming and message- 919 processing descriptions in this document are implicitly prefixed with 920 the rules of [RFC5389]. 922 [RFC5389] specifies an authentication mechanism called the long-term 923 credential mechanism. TURN servers and clients MUST implement this 924 mechanism. The server MUST demand that all requests from the client 925 be authenticated using this mechanism, or that a equally strong or 926 stronger mechanism for client authentication is used. 928 Note that the long-term credential mechanism applies only to requests 929 and cannot be used to authenticate indications; thus, indications in 930 TURN are never authenticated. If the server requires requests to be 931 authenticated, then the server's administrator MUST choose a realm 932 value that will uniquely identify the username and password 933 combination that the client must use, even if the client uses 934 multiple servers under different administrations. The server's 935 administrator MAY choose to allocate a unique username to each 936 client, or MAY choose to allocate the same username to more than one 937 client (for example, to all clients from the same department or 938 company). For each allocation, the server SHOULD generate a new 939 random nonce when the allocation is first attempted following the 940 randomness recommendations in [RFC4086] and SHOULD expire the nonce 941 at least once every hour during the lifetime of the allocation. 943 All requests after the initial Allocate must use the same username as 944 that used to create the allocation, to prevent attackers from 945 hijacking the client's allocation. Specifically, if the server 946 requires the use of the long-term credential mechanism, and if a non- 947 Allocate request passes authentication under this mechanism, and if 948 the 5-tuple identifies an existing allocation, but the request does 949 not use the same username as used to create the allocation, then the 950 request MUST be rejected with a 441 (Wrong Credentials) error. 952 When a TURN message arrives at the server from the client, the server 953 uses the 5-tuple in the message to identify the associated 954 allocation. For all TURN messages (including ChannelData) EXCEPT an 955 Allocate request, if the 5-tuple does not identify an existing 956 allocation, then the message MUST either be rejected with a 437 957 Allocation Mismatch error (if it is a request) or silently ignored 958 (if it is an indication or a ChannelData message). A client 959 receiving a 437 error response to a request other than Allocate MUST 960 assume the allocation no longer exists. 962 [RFC5389] defines a number of attributes, including the SOFTWARE and 963 FINGERPRINT attributes. The client SHOULD include the SOFTWARE 964 attribute in all Allocate and Refresh requests and MAY include it in 965 any other requests or indications. The server SHOULD include the 966 SOFTWARE attribute in all Allocate and Refresh responses (either 967 success or failure) and MAY include it in other responses or 968 indications. The client and the server MAY include the FINGERPRINT 969 attribute in any STUN-formatted messages defined in this document. 971 TURN does not use the backwards-compatibility mechanism described in 972 [RFC5389]. 974 TURN, as defined in this specification, supports both IPv4 and IPv6. 975 IPv6 support in TURN includes IPv4-to-IPv6, IPv6-to-IPv6, and IPv6- 976 to-IPv4 relaying. The REQUESTED-ADDRESS-FAMILY attribute allows a 977 client to explicitly request the address type the TURN server will 978 allocate (e.g., an IPv4-only node may request the TURN server to 979 allocate an IPv6 address). The ADDITIONAL-ADDRESS-FAMILY attribute 980 allows a client to request the server to allocate one IPv4 and one 981 IPv6 relay address in a single Allocate request. This saves local 982 ports on the client and reduces the number of messages sent between 983 the client and the TURN server. 985 By default, TURN runs on the same ports as STUN: 3478 for TURN over 986 UDP and TCP, and 5349 for TURN over (D)TLS. However, TURN has its 987 own set of Service Record (SRV) names: "turn" for UDP and TCP, and 988 "turns" for (D)TLS. Either the SRV procedures or the ALTERNATE- 989 SERVER procedures, both described in Section 6, can be used to run 990 TURN on a different port. 992 To ensure interoperability, a TURN server MUST support the use of UDP 993 transport between the client and the server, and SHOULD support the 994 use of TCP and (D)TLS transport. 996 When UDP transport is used between the client and the server, the 997 client will retransmit a request if it does not receive a response 998 within a certain timeout period. Because of this, the server may 999 receive two (or more) requests with the same 5-tuple and same 1000 transaction id. STUN requires that the server recognize this case 1001 and treat the request as idempotent (see [RFC5389]). Some 1002 implementations may choose to meet this requirement by remembering 1003 all received requests and the corresponding responses for 40 seconds. 1004 Other implementations may choose to reprocess the request and arrange 1005 that such reprocessing returns essentially the same response. To aid 1006 implementors who choose the latter approach (the so-called "stateless 1007 stack approach"), this specification includes some implementation 1008 notes on how this might be done. Implementations are free to choose 1009 either approach or choose some other approach that gives the same 1010 results. 1012 When TCP transport is used between the client and the server, it is 1013 possible that a bit error will cause a length field in a TURN packet 1014 to become corrupted, causing the receiver to lose synchronization 1015 with the incoming stream of TURN messages. A client or server that 1016 detects a long sequence of invalid TURN messages over TCP transport 1017 SHOULD close the corresponding TCP connection to help the other end 1018 detect this situation more rapidly. 1020 To mitigate either intentional or unintentional denial-of-service 1021 attacks against the server by clients with valid usernames and 1022 passwords, it is RECOMMENDED that the server impose limits on both 1023 the number of allocations active at one time for a given username and 1024 on the amount of bandwidth those allocations can use. The server 1025 should reject new allocations that would exceed the limit on the 1026 allowed number of allocations active at one time with a 486 1027 (Allocation Quota Exceeded) (see Section 6.2), and should discard 1028 application data traffic that exceeds the bandwidth quota. 1030 5. Allocations 1032 All TURN operations revolve around allocations, and all TURN messages 1033 are associated with an allocation. An allocation conceptually 1034 consists of the following state data: 1036 o the relayed transport address; 1038 o the 5-tuple: (client's IP address, client's port, server IP 1039 address, server port, transport protocol); 1041 o the authentication information; 1043 o the time-to-expiry; 1045 o a list of permissions; 1047 o a list of channel to peer bindings. 1049 The relayed transport address is the transport address allocated by 1050 the server for communicating with peers, while the 5-tuple describes 1051 the communication path between the client and the server. On the 1052 client, the 5-tuple uses the client's host transport address; on the 1053 server, the 5-tuple uses the client's server-reflexive transport 1054 address. 1056 Both the relayed transport address and the 5-tuple MUST be unique 1057 across all allocations, so either one can be used to uniquely 1058 identify the allocation. 1060 The authentication information (e.g., username, password, realm, and 1061 nonce) is used to both verify subsequent requests and to compute the 1062 message integrity of responses. The username, realm, and nonce 1063 values are initially those used in the authenticated Allocate request 1064 that creates the allocation, though the server can change the nonce 1065 value during the lifetime of the allocation using a 438 (Stale Nonce) 1066 reply. Note that, rather than storing the password explicitly, for 1067 security reasons, it may be desirable for the server to store the key 1068 value, which is an secure hash over the username, realm, and password 1069 (see [I-D.ietf-tram-stunbis]). 1071 The time-to-expiry is the time in seconds left until the allocation 1072 expires. Each Allocate or Refresh transaction sets this timer, which 1073 then ticks down towards 0. By default, each Allocate or Refresh 1074 transaction resets this timer to the default lifetime value of 600 1075 seconds (10 minutes), but the client can request a different value in 1076 the Allocate and Refresh request. Allocations can only be refreshed 1077 using the Refresh request; sending data to a peer does not refresh an 1078 allocation. When an allocation expires, the state data associated 1079 with the allocation can be freed. 1081 The list of permissions is described in Section 8 and the list of 1082 channels is described in Section 11. 1084 6. Creating an Allocation 1086 An allocation on the server is created using an Allocate transaction. 1088 6.1. Sending an Allocate Request 1090 The client forms an Allocate request as follows. 1092 The client first picks a host transport address. It is RECOMMENDED 1093 that the client pick a currently unused transport address, typically 1094 by allowing the underlying OS to pick a currently unused port for a 1095 new socket. 1097 The client then picks a transport protocol to use between the client 1098 and the server. The transport protocol MUST be one of UDP, TCP, TLS- 1099 over-TCP or DTLS-over-UDP. Since this specification only allows UDP 1100 between the server and the peers, it is RECOMMENDED that the client 1101 pick UDP unless it has a reason to use a different transport. One 1102 reason to pick a different transport would be that the client 1103 believes, either through configuration or by experiment, that it is 1104 unable to contact any TURN server using UDP. See Section 2.1 for 1105 more discussion. 1107 The client also picks a server transport address, which SHOULD be 1108 done as follows. The client uses the procedures described in 1109 [I-D.ietf-tram-turn-server-discovery] to discover a TURN server and 1110 TURN server resolution mechanism defined in [RFC5928] to get a list 1111 of server transport addresses that can be tried to create a TURN 1112 allocation. 1114 The client MUST include a REQUESTED-TRANSPORT attribute in the 1115 request. This attribute specifies the transport protocol between the 1116 server and the peers (note that this is NOT the transport protocol 1117 that appears in the 5-tuple). In this specification, the REQUESTED- 1118 TRANSPORT type is always UDP. This attribute is included to allow 1119 future extensions to specify other protocols. 1121 If the client wishes to obtain a relayed transport address of a 1122 specific address type then it includes a REQUESTED-ADDRESS-FAMILY 1123 attribute in the request. This attribute indicates the specific 1124 address type the client wishes the TURN server to allocate. Clients 1125 MUST NOT include more than one REQUESTED-ADDRESS-FAMILY attribute in 1126 an Allocate request. Clients MUST NOT include a REQUESTED-ADDRESS- 1127 FAMILY attribute in an Allocate request that contains a RESERVATION- 1128 TOKEN attribute, for the reasons outlined in [RFC6156]. 1130 If the client wishes to obtain one IPv6 and one IPv4 relayed 1131 transport addresses then it includes an ADDITIONAL-ADDRESS-FAMILY 1132 attribute in the request. This attribute specifies that the server 1133 must allocate both address types. The attribute value in the 1134 ADDITIONAL-ADDRESS-FAMILY MUST be set to 0x02 (IPv6 address family). 1135 Clients MUST NOT include REQUESTED-ADDRESS-FAMILY and ADDITIONAL- 1136 ADDRESS-FAMILY attributes in the same request. Clients MUST NOT 1137 include ADDITIONAL-ADDRESS-FAMILY attribute in a Allocate request 1138 that contains a RESERVATION-TOKEN attribute. Clients MUST NOT 1139 include ADDITIONAL-ADDRESS-FAMILY attribute in a Allocate request 1140 that contains a EVEN-PORT attribute with the R bit set to 1. 1142 If the client wishes the server to initialize the time-to-expiry 1143 field of the allocation to some value other than the default 1144 lifetime, then it MAY include a LIFETIME attribute specifying its 1145 desired value. This is just a hint, and the server may elect to use 1146 a different value. Note that the server will ignore requests to 1147 initialize the field to less than the default value. 1149 If the client wishes to later use the DONT-FRAGMENT attribute in one 1150 or more Send indications on this allocation, then the client SHOULD 1151 include the DONT-FRAGMENT attribute in the Allocate request. This 1152 allows the client to test whether this attribute is supported by the 1153 server. 1155 If the client requires the port number of the relayed transport 1156 address be even, the client includes the EVEN-PORT attribute. If 1157 this attribute is not included, then the port can be even or odd. By 1158 setting the R bit in the EVEN-PORT attribute to 1, the client can 1159 request that the server reserve the next highest port number (on the 1160 same IP address) for a subsequent allocation. If the R bit is 0, no 1161 such request is made. 1163 The client MAY also include a RESERVATION-TOKEN attribute in the 1164 request to ask the server to use a previously reserved port for the 1165 allocation. If the RESERVATION-TOKEN attribute is included, then the 1166 client MUST omit the EVEN-PORT attribute. 1168 Once constructed, the client sends the Allocate request on the 1169 5-tuple. 1171 6.2. Receiving an Allocate Request 1173 When the server receives an Allocate request, it performs the 1174 following checks: 1176 1. The server MUST require that the request be authenticated. This 1177 authentication MUST be done using the long-term credential 1178 mechanism of [RFC5389] unless the client and server agree to use 1179 another mechanism through some procedure outside the scope of 1180 this document. 1182 2. The server checks if the 5-tuple is currently in use by an 1183 existing allocation. If yes, the server rejects the request 1184 with a 437 (Allocation Mismatch) error. 1186 3. The server checks if the request contains a REQUESTED-TRANSPORT 1187 attribute. If the REQUESTED-TRANSPORT attribute is not included 1188 or is malformed, the server rejects the request with a 400 (Bad 1189 Request) error. Otherwise, if the attribute is included but 1190 specifies a protocol other that UDP, the server rejects the 1191 request with a 442 (Unsupported Transport Protocol) error. 1193 4. The request may contain a DONT-FRAGMENT attribute. If it does, 1194 but the server does not support sending UDP datagrams with the 1195 DF bit set to 1 (see Section 13), then the server treats the 1196 DONT-FRAGMENT attribute in the Allocate request as an unknown 1197 comprehension-required attribute. 1199 5. The server checks if the request contains a RESERVATION-TOKEN 1200 attribute. If yes, and the request also contains an EVEN-PORT 1201 or REQUESTED-ADDRESS-FAMILY or ADDITIONAL-ADDRESS-FAMILY 1202 attribute, the server rejects the request with a 400 (Bad 1203 Request) error. Otherwise, it checks to see if the token is 1204 valid (i.e., the token is in range and has not expired and the 1205 corresponding relayed transport address is still available). If 1206 the token is not valid for some reason, the server rejects the 1207 request with a 508 (Insufficient Capacity) error. 1209 6. The server checks if the request contains both REQUESTED- 1210 ADDRESS-FAMILY and ADDITIONAL-ADDRESS-FAMILY attributes, then 1211 the server rejects the request with a 400 (Bad Request) error. 1213 7. If the server does not support the address family requested by 1214 the client in REQUESTED-ADDRESS-FAMILY or is disabled by local 1215 policy, it MUST generate an Allocate error response, and it MUST 1216 include an ERROR-CODE attribute with the 440 (Address Family not 1217 Supported) response code. If the REQUESTED-ADDRESS-FAMILY 1218 attribute is absent, the server MUST allocate an IPv4 relayed 1219 transport address for the TURN client. 1221 8. The server checks if the request contains an EVEN-PORT attribute 1222 with the R bit set to 1. If yes, and the request also contains 1223 an ADDITIONAL- ADDRESS-FAMILY attribute, the server rejects the 1224 request with a 400 (Bad Request) error. Otherwise, the server 1225 checks if it can satisfy the request (i.e., can allocate a 1226 relayed transport address as described below). If the server 1227 cannot satisfy the request, then the server rejects the request 1228 with a 508 (Insufficient Capacity) error. 1230 9. The server checks if the request contains an ADDITIONAL-ADDRESS- 1231 FAMILY attribute. If yes, and the attribute value is 0x01 (IPv4 1232 address family), then the server rejects the request with a 400 1233 (Bad Request) error. Otherwise, and the server checks if it can 1234 allocate relayed transport addresses of both address types. If 1235 the server cannot satisfy the request, then the server rejects 1236 the request with a 508 (Insufficient Capacity) error. If the 1237 server can partially meet the request, i.e. if it can only 1238 allocate one relayed transport address of a specific address 1239 type, then it includes ADDRESS-ERROR-CODE attribute in the 1240 response to inform the client the reason for partial failure of 1241 the request. The error code value signaled in the ADDRESS- 1242 ERROR-CODE attribute could be 440 (Address Family not Supported) 1243 or 508 (Insufficient Capacity). 1245 10. At any point, the server MAY choose to reject the request with a 1246 486 (Allocation Quota Reached) error if it feels the client is 1247 trying to exceed some locally defined allocation quota. The 1248 server is free to define this allocation quota any way it 1249 wishes, but SHOULD define it based on the username used to 1250 authenticate the request, and not on the client's transport 1251 address. 1253 11. Also at any point, the server MAY choose to reject the request 1254 with a 300 (Try Alternate) error if it wishes to redirect the 1255 client to a different server. The use of this error code and 1256 attribute follow the specification in [RFC5389]. 1258 If all the checks pass, the server creates the allocation. The 1259 5-tuple is set to the 5-tuple from the Allocate request, while the 1260 list of permissions and the list of channels are initially empty. 1262 The server chooses a relayed transport address for the allocation as 1263 follows: 1265 o If the request contains a RESERVATION-TOKEN attribute, the server 1266 uses the previously reserved transport address corresponding to 1267 the included token (if it is still available). Note that the 1268 reservation is a server-wide reservation and is not specific to a 1269 particular allocation, since the Allocate request containing the 1270 RESERVATION-TOKEN uses a different 5-tuple than the Allocate 1271 request that made the reservation. The 5-tuple for the Allocate 1272 request containing the RESERVATION-TOKEN attribute can be any 1273 allowed 5-tuple; it can use a different client IP address and 1274 port, a different transport protocol, and even different server IP 1275 address and port (provided, of course, that the server IP address 1276 and port are ones on which the server is listening for TURN 1277 requests). 1279 o If the request contains an EVEN-PORT attribute with the R bit set 1280 to 0, then the server allocates a relayed transport address with 1281 an even port number. 1283 o If the request contains an EVEN-PORT attribute with the R bit set 1284 to 1, then the server looks for a pair of port numbers N and N+1 1285 on the same IP address, where N is even. Port N is used in the 1286 current allocation, while the relayed transport address with port 1287 N+1 is assigned a token and reserved for a future allocation. The 1288 server MUST hold this reservation for at least 30 seconds, and MAY 1289 choose to hold longer (e.g., until the allocation with port N 1290 expires). The server then includes the token in a RESERVATION- 1291 TOKEN attribute in the success response. 1293 o Otherwise, the server allocates any available relayed transport 1294 address. 1296 In all cases, the server SHOULD only allocate ports from the range 1297 49152 - 65535 (the Dynamic and/or Private Port range [Port-Numbers]), 1298 unless the TURN server application knows, through some means not 1299 specified here, that other applications running on the same host as 1300 the TURN server application will not be impacted by allocating ports 1301 outside this range. This condition can often be satisfied by running 1302 the TURN server application on a dedicated machine and/or by 1303 arranging that any other applications on the machine allocate ports 1304 before the TURN server application starts. In any case, the TURN 1305 server SHOULD NOT allocate ports in the range 0 - 1023 (the Well- 1306 Known Port range) to discourage clients from using TURN to run 1307 standard services. 1309 NOTE: The use of randomized port assignments to avoid certain 1310 types of attacks is described in [RFC6056]. It is RECOMMENDED 1311 that a TURN server implement a randomized port assignment 1312 algorithm from [RFC6056]. This is especially applicable to 1313 servers that choose to pre-allocate a number of ports from the 1314 underlying OS and then later assign them to allocations; for 1315 example, a server may choose this technique to implement the EVEN- 1316 PORT attribute. 1318 The server determines the initial value of the time-to-expiry field 1319 as follows. If the request contains a LIFETIME attribute, then the 1320 server computes the minimum of the client's proposed lifetime and the 1321 server's maximum allowed lifetime. If this computed value is greater 1322 than the default lifetime, then the server uses the computed lifetime 1323 as the initial value of the time-to-expiry field. Otherwise, the 1324 server uses the default lifetime. It is RECOMMENDED that the server 1325 use a maximum allowed lifetime value of no more than 3600 seconds (1 1326 hour). Servers that implement allocation quotas or charge users for 1327 allocations in some way may wish to use a smaller maximum allowed 1328 lifetime (perhaps as small as the default lifetime) to more quickly 1329 remove orphaned allocations (that is, allocations where the 1330 corresponding client has crashed or terminated or the client 1331 connection has been lost for some reason). Also, note that the time- 1332 to-expiry is recomputed with each successful Refresh request, and 1333 thus the value computed here applies only until the first refresh. 1335 Once the allocation is created, the server replies with a success 1336 response. The success response contains: 1338 o An XOR-RELAYED-ADDRESS attribute containing the relayed transport 1339 address. 1341 o A LIFETIME attribute containing the current value of the time-to- 1342 expiry timer. 1344 o A RESERVATION-TOKEN attribute (if a second relayed transport 1345 address was reserved). 1347 o An XOR-MAPPED-ADDRESS attribute containing the client's IP address 1348 and port (from the 5-tuple). 1350 NOTE: The XOR-MAPPED-ADDRESS attribute is included in the response 1351 as a convenience to the client. TURN itself does not make use of 1352 this value, but clients running ICE can often need this value and 1353 can thus avoid having to do an extra Binding transaction with some 1354 STUN server to learn it. 1356 The response (either success or error) is sent back to the client on 1357 the 5-tuple. 1359 NOTE: When the Allocate request is sent over UDP, section 7.3.1 of 1360 [RFC5389] requires that the server handle the possible 1361 retransmissions of the request so that retransmissions do not 1362 cause multiple allocations to be created. Implementations may 1363 achieve this using the so-called "stateless stack approach" as 1364 follows. To detect retransmissions when the original request was 1365 successful in creating an allocation, the server can store the 1366 transaction id that created the request with the allocation data 1367 and compare it with incoming Allocate requests on the same 1368 5-tuple. Once such a request is detected, the server can stop 1369 parsing the request and immediately generate a success response. 1370 When building this response, the value of the LIFETIME attribute 1371 can be taken from the time-to-expiry field in the allocate state 1372 data, even though this value may differ slightly from the LIFETIME 1373 value originally returned. In addition, the server may need to 1374 store an indication of any reservation token returned in the 1375 original response, so that this may be returned in any 1376 retransmitted responses. 1378 For the case where the original request was unsuccessful in 1379 creating an allocation, the server may choose to do nothing 1380 special. Note, however, that there is a rare case where the 1381 server rejects the original request but accepts the retransmitted 1382 request (because conditions have changed in the brief intervening 1383 time period). If the client receives the first failure response, 1384 it will ignore the second (success) response and believe that an 1385 allocation was not created. An allocation created in this matter 1386 will eventually timeout, since the client will not refresh it. 1387 Furthermore, if the client later retries with the same 5-tuple but 1388 different transaction id, it will receive a 437 (Allocation 1389 Mismatch), which will cause it to retry with a different 5-tuple. 1390 The server may use a smaller maximum lifetime value to minimize 1391 the lifetime of allocations "orphaned" in this manner. 1393 6.3. Receiving an Allocate Success Response 1395 If the client receives an Allocate success response, then it MUST 1396 check that the mapped address and the relayed transport address are 1397 part of an address family that the client understands and is prepared 1398 to handle. If these two addresses are not part of an address family 1399 which the client is prepared to handle, then the client MUST delete 1400 the allocation (Section 7) and MUST NOT attempt to create another 1401 allocation on that server until it believes the mismatch has been 1402 fixed. 1404 Otherwise, the client creates its own copy of the allocation data 1405 structure to track what is happening on the server. In particular, 1406 the client needs to remember the actual lifetime received back from 1407 the server, rather than the value sent to the server in the request. 1408 The client must also remember the 5-tuple used for the request and 1409 the username and password it used to authenticate the request to 1410 ensure that it reuses them for subsequent messages. The client also 1411 needs to track the channels and permissions it establishes on the 1412 server. 1414 The client will probably wish to send the relayed transport address 1415 to peers (using some method not specified here) so the peers can 1416 communicate with it. The client may also wish to use the server- 1417 reflexive address it receives in the XOR-MAPPED-ADDRESS attribute in 1418 its ICE processing. 1420 6.4. Receiving an Allocate Error Response 1422 If the client receives an Allocate error response, then the 1423 processing depends on the actual error code returned: 1425 o (Request timed out): There is either a problem with the server, or 1426 a problem reaching the server with the chosen transport. The 1427 client considers the current transaction as having failed but MAY 1428 choose to retry the Allocate request using a different transport 1429 (e.g., TCP instead of UDP). 1431 o 300 (Try Alternate): The server would like the client to use the 1432 server specified in the ALTERNATE-SERVER attribute instead. The 1433 client considers the current transaction as having failed, but 1434 SHOULD try the Allocate request with the alternate server before 1435 trying any other servers (e.g., other servers discovered using the 1436 SRV procedures). When trying the Allocate request with the 1437 alternate server, the client follows the ALTERNATE-SERVER 1438 procedures specified in [RFC5389]. 1440 o 400 (Bad Request): The server believes the client's request is 1441 malformed for some reason. The client considers the current 1442 transaction as having failed. The client MAY notify the user or 1443 operator and SHOULD NOT retry the request with this server until 1444 it believes the problem has been fixed. 1446 o 401 (Unauthorized): If the client has followed the procedures of 1447 the long-term credential mechanism and still gets this error, then 1448 the server is not accepting the client's credentials. In this 1449 case, the client considers the current transaction as having 1450 failed and SHOULD notify the user or operator. The client SHOULD 1451 NOT send any further requests to this server until it believes the 1452 problem has been fixed. 1454 o 403 (Forbidden): The request is valid, but the server is refusing 1455 to perform it, likely due to administrative restrictions. The 1456 client considers the current transaction as having failed. The 1457 client MAY notify the user or operator and SHOULD NOT retry the 1458 same request with this server until it believes the problem has 1459 been fixed. 1461 o 420 (Unknown Attribute): If the client included a DONT-FRAGMENT 1462 attribute in the request and the server rejected the request with 1463 a 420 error code and listed the DONT-FRAGMENT attribute in the 1464 UNKNOWN-ATTRIBUTES attribute in the error response, then the 1465 client now knows that the server does not support the DONT- 1466 FRAGMENT attribute. The client considers the current transaction 1467 as having failed but MAY choose to retry the Allocate request 1468 without the DONT-FRAGMENT attribute. 1470 o 437 (Allocation Mismatch): This indicates that the client has 1471 picked a 5-tuple that the server sees as already in use. One way 1472 this could happen is if an intervening NAT assigned a mapped 1473 transport address that was used by another client that recently 1474 crashed. The client considers the current transaction as having 1475 failed. The client SHOULD pick another client transport address 1476 and retry the Allocate request (using a different transaction id). 1477 The client SHOULD try three different client transport addresses 1478 before giving up on this server. Once the client gives up on the 1479 server, it SHOULD NOT try to create another allocation on the 1480 server for 2 minutes. 1482 o 438 (Stale Nonce): See the procedures for the long-term credential 1483 mechanism [RFC5389]. 1485 o 440 (Address Family not Supported): The server does not support 1486 the address family requested by the client. If the client 1487 receives an Allocate error response with the 440 (Unsupported 1488 Address Family) error code, the client MUST NOT retry the request. 1490 o 441 (Wrong Credentials): The client should not receive this error 1491 in response to a Allocate request. The client MAY notify the user 1492 or operator and SHOULD NOT retry the same request with this server 1493 until it believes the problem has been fixed. 1495 o 442 (Unsupported Transport Address): The client should not receive 1496 this error in response to a request for a UDP allocation. The 1497 client MAY notify the user or operator and SHOULD NOT reattempt 1498 the request with this server until it believes the problem has 1499 been fixed. 1501 o 486 (Allocation Quota Reached): The server is currently unable to 1502 create any more allocations with this username. The client 1503 considers the current transaction as having failed. The client 1504 SHOULD wait at least 1 minute before trying to create any more 1505 allocations on the server. 1507 o 508 (Insufficient Capacity): The server has no more relayed 1508 transport addresses available, or has none with the requested 1509 properties, or the one that was reserved is no longer available. 1510 The client considers the current operation as having failed. If 1511 the client is using either the EVEN-PORT or the RESERVATION-TOKEN 1512 attribute, then the client MAY choose to remove or modify this 1513 attribute and try again immediately. Otherwise, the client SHOULD 1514 wait at least 1 minute before trying to create any more 1515 allocations on this server. 1517 An unknown error response MUST be handled as described in [RFC5389]. 1519 7. Refreshing an Allocation 1521 A Refresh transaction can be used to either (a) refresh an existing 1522 allocation and update its time-to-expiry or (b) delete an existing 1523 allocation. 1525 If a client wishes to continue using an allocation, then the client 1526 MUST refresh it before it expires. It is suggested that the client 1527 refresh the allocation roughly 1 minute before it expires. If a 1528 client no longer wishes to use an allocation, then it SHOULD 1529 explicitly delete the allocation. A client MAY refresh an allocation 1530 at any time for other reasons. 1532 7.1. Sending a Refresh Request 1534 If the client wishes to immediately delete an existing allocation, it 1535 includes a LIFETIME attribute with a value of 0. All other forms of 1536 the request refresh the allocation. 1538 When refreshing a dual allocation, the client includes REQUESTED- 1539 ADDRESS-FAMILY attribute indicating the address family type that 1540 should be refreshed. If no REQUESTED-ADDRESS-FAMILY is included then 1541 the request should be treated as applying to all current allocations. 1542 The client MUST only include family types it previously allocated and 1543 has not yet deleted. This process can also be used to delete an 1544 allocation of a specific address type, by setting the lifetime of 1545 that refresh request to 0. Deleting a single allocation destroys any 1546 permissions or channels associated with that particular allocation; 1547 it MUST NOT affect any permissions or channels associated with 1548 allocations for the other address family. 1550 The Refresh transaction updates the time-to-expiry timer of an 1551 allocation. If the client wishes the server to set the time-to- 1552 expiry timer to something other than the default lifetime, it 1553 includes a LIFETIME attribute with the requested value. The server 1554 then computes a new time-to-expiry value in the same way as it does 1555 for an Allocate transaction, with the exception that a requested 1556 lifetime of 0 causes the server to immediately delete the allocation. 1558 7.2. Receiving a Refresh Request 1560 When the server receives a Refresh request, it processes it as per 1561 Section 4 plus the specific rules mentioned here. 1563 If the server receives a Refresh Request with an REQUESTED-ADDRESS- 1564 FAMILY attribute and the attribute value does not match the address 1565 family of the allocation, the server MUST reply with a 443 (Peer 1566 Address Family Mismatch) Refresh error response. 1568 The server computes a value called the "desired lifetime" as follows: 1569 if the request contains a LIFETIME attribute and the attribute value 1570 is 0, then the "desired lifetime" is 0. Otherwise, if the request 1571 contains a LIFETIME attribute, then the server computes the minimum 1572 of the client's requested lifetime and the server's maximum allowed 1573 lifetime. If this computed value is greater than the default 1574 lifetime, then the "desired lifetime" is the computed value. 1575 Otherwise, the "desired lifetime" is the default lifetime. 1577 Subsequent processing depends on the "desired lifetime" value: 1579 o If the "desired lifetime" is 0, then the request succeeds and the 1580 allocation is deleted. 1582 o If the "desired lifetime" is non-zero, then the request succeeds 1583 and the allocation's time-to-expiry is set to the "desired 1584 lifetime". 1586 If the request succeeds, then the server sends a success response 1587 containing: 1589 o A LIFETIME attribute containing the current value of the time-to- 1590 expiry timer. 1592 NOTE: A server need not do anything special to implement 1593 idempotency of Refresh requests over UDP using the "stateless 1594 stack approach". Retransmitted Refresh requests with a non-zero 1595 "desired lifetime" will simply refresh the allocation. A 1596 retransmitted Refresh request with a zero "desired lifetime" will 1597 cause a 437 (Allocation Mismatch) response if the allocation has 1598 already been deleted, but the client will treat this as equivalent 1599 to a success response (see below). 1601 7.3. Receiving a Refresh Response 1603 If the client receives a success response to its Refresh request with 1604 a non-zero lifetime, it updates its copy of the allocation data 1605 structure with the time-to-expiry value contained in the response. 1607 If the client receives a 437 (Allocation Mismatch) error response to 1608 a request to delete the allocation, then the allocation no longer 1609 exists and it should consider its request as having effectively 1610 succeeded. 1612 8. Permissions 1614 For each allocation, the server keeps a list of zero or more 1615 permissions. Each permission consists of an IP address and an 1616 associated time-to-expiry. While a permission exists, all peers 1617 using the IP address in the permission are allowed to send data to 1618 the client. The time-to-expiry is the number of seconds until the 1619 permission expires. Within the context of an allocation, a 1620 permission is uniquely identified by its associated IP address. 1622 By sending either CreatePermission requests or ChannelBind requests, 1623 the client can cause the server to install or refresh a permission 1624 for a given IP address. This causes one of two things to happen: 1626 o If no permission for that IP address exists, then a permission is 1627 created with the given IP address and a time-to-expiry equal to 1628 Permission Lifetime. 1630 o If a permission for that IP address already exists, then the time- 1631 to-expiry for that permission is reset to Permission Lifetime. 1633 The Permission Lifetime MUST be 300 seconds (= 5 minutes). 1635 Each permission's time-to-expiry decreases down once per second until 1636 it reaches 0; at which point, the permission expires and is deleted. 1638 CreatePermission and ChannelBind requests may be freely intermixed on 1639 a permission. A given permission may be initially installed and/or 1640 refreshed with a CreatePermission request, and then later refreshed 1641 with a ChannelBind request, or vice versa. 1643 When a UDP datagram arrives at the relayed transport address for the 1644 allocation, the server extracts the source IP address from the IP 1645 header. The server then compares this address with the IP address 1646 associated with each permission in the list of permissions for the 1647 allocation. If no match is found, relaying is not permitted, and the 1648 server silently discards the UDP datagram. If an exact match is 1649 found, then the permission check is considered to have succeeded and 1650 the server continues to process the UDP datagram as specified 1651 elsewhere (Section 10.3). Note that only addresses are compared and 1652 port numbers are not considered. 1654 The permissions for one allocation are totally unrelated to the 1655 permissions for a different allocation. If an allocation expires, 1656 all its permissions expire with it. 1658 NOTE: Though TURN permissions expire after 5 minutes, many NATs 1659 deployed at the time of publication expire their UDP bindings 1660 considerably faster. Thus, an application using TURN will 1661 probably wish to send some sort of keep-alive traffic at a much 1662 faster rate. Applications using ICE should follow the keep-alive 1663 guidelines of ICE [RFC5245], and applications not using ICE are 1664 advised to do something similar. 1666 9. CreatePermission 1668 TURN supports two ways for the client to install or refresh 1669 permissions on the server. This section describes one way: the 1670 CreatePermission request. 1672 A CreatePermission request may be used in conjunction with either the 1673 Send mechanism in Section 10 or the Channel mechanism in Section 11. 1675 9.1. Forming a CreatePermission Request 1677 The client who wishes to install or refresh one or more permissions 1678 can send a CreatePermission request to the server. 1680 When forming a CreatePermission request, the client MUST include at 1681 least one XOR-PEER-ADDRESS attribute, and MAY include more than one 1682 such attribute. The IP address portion of each XOR-PEER-ADDRESS 1683 attribute contains the IP address for which a permission should be 1684 installed or refreshed. The port portion of each XOR-PEER-ADDRESS 1685 attribute will be ignored and can be any arbitrary value. The 1686 various XOR-PEER-ADDRESS attributes can appear in any order. The 1687 client MUST only include XOR-PEER-ADDRESS attributes with addresses 1688 of the same address family as that of the relayed transport address 1689 for the allocation. For dual allocations obtained using the 1690 ADDITIONAL-FAMILY-ADDRESS attribute, the client can include XOR-PEER- 1691 ADDRESS attributes with addresses of IPv4 and IPv6 address families. 1693 9.2. Receiving a CreatePermission Request 1695 When the server receives the CreatePermission request, it processes 1696 as per Section 4 plus the specific rules mentioned here. 1698 The message is checked for validity. The CreatePermission request 1699 MUST contain at least one XOR-PEER-ADDRESS attribute and MAY contain 1700 multiple such attributes. If no such attribute exists, or if any of 1701 these attributes are invalid, then a 400 (Bad Request) error is 1702 returned. If the request is valid, but the server is unable to 1703 satisfy the request due to some capacity limit or similar, then a 508 1704 (Insufficient Capacity) error is returned. 1706 If an XOR-PEER-ADDRESS attribute contains an address of an address 1707 family that is not the same as that of the relayed transport address 1708 for the allocation, the server MUST generate an error response with 1709 the 443 (Peer Address Family Mismatch) response code. 1711 The server MAY impose restrictions on the IP address allowed in the 1712 XOR-PEER-ADDRESS attribute -- if a value is not allowed, the server 1713 rejects the request with a 403 (Forbidden) error. 1715 If the message is valid and the server is capable of carrying out the 1716 request, then the server installs or refreshes a permission for the 1717 IP address contained in each XOR-PEER-ADDRESS attribute as described 1718 in Section 8. The port portion of each attribute is ignored and may 1719 be any arbitrary value. 1721 The server then responds with a CreatePermission success response. 1722 There are no mandatory attributes in the success response. 1724 NOTE: A server need not do anything special to implement 1725 idempotency of CreatePermission requests over UDP using the 1726 "stateless stack approach". Retransmitted CreatePermission 1727 requests will simply refresh the permissions. 1729 9.3. Receiving a CreatePermission Response 1731 If the client receives a valid CreatePermission success response, 1732 then the client updates its data structures to indicate that the 1733 permissions have been installed or refreshed. 1735 10. Send and Data Methods 1737 TURN supports two mechanisms for sending and receiving data from 1738 peers. This section describes the use of the Send and Data 1739 mechanisms, while Section 11 describes the use of the Channel 1740 mechanism. 1742 10.1. Forming a Send Indication 1744 The client can use a Send indication to pass data to the server for 1745 relaying to a peer. A client may use a Send indication even if a 1746 channel is bound to that peer. However, the client MUST ensure that 1747 there is a permission installed for the IP address of the peer to 1748 which the Send indication is being sent; this prevents a third party 1749 from using a TURN server to send data to arbitrary destinations. 1751 When forming a Send indication, the client MUST include an XOR-PEER- 1752 ADDRESS attribute and a DATA attribute. The XOR-PEER-ADDRESS 1753 attribute contains the transport address of the peer to which the 1754 data is to be sent, and the DATA attribute contains the actual 1755 application data to be sent to the peer. 1757 The client MAY include a DONT-FRAGMENT attribute in the Send 1758 indication if it wishes the server to set the DF bit on the UDP 1759 datagram sent to the peer. 1761 10.2. Receiving a Send Indication 1763 When the server receives a Send indication, it processes as per 1764 Section 4 plus the specific rules mentioned here. 1766 The message is first checked for validity. The Send indication MUST 1767 contain both an XOR-PEER-ADDRESS attribute and a DATA attribute. If 1768 one of these attributes is missing or invalid, then the message is 1769 discarded. Note that the DATA attribute is allowed to contain zero 1770 bytes of data. 1772 The Send indication may also contain the DONT-FRAGMENT attribute. If 1773 the server is unable to set the DF bit on outgoing UDP datagrams when 1774 this attribute is present, then the server acts as if the DONT- 1775 FRAGMENT attribute is an unknown comprehension-required attribute 1776 (and thus the Send indication is discarded). 1778 The server also checks that there is a permission installed for the 1779 IP address contained in the XOR-PEER-ADDRESS attribute. If no such 1780 permission exists, the message is discarded. Note that a Send 1781 indication never causes the server to refresh the permission. 1783 The server MAY impose restrictions on the IP address and port values 1784 allowed in the XOR-PEER-ADDRESS attribute -- if a value is not 1785 allowed, the server silently discards the Send indication. 1787 If everything is OK, then the server forms a UDP datagram as follows: 1789 o the source transport address is the relayed transport address of 1790 the allocation, where the allocation is determined by the 5-tuple 1791 on which the Send indication arrived; 1793 o the destination transport address is taken from the XOR-PEER- 1794 ADDRESS attribute; 1796 o the data following the UDP header is the contents of the value 1797 field of the DATA attribute. 1799 The handling of the DONT-FRAGMENT attribute (if present), is 1800 described in Section 13. 1802 The resulting UDP datagram is then sent to the peer. 1804 10.3. Receiving a UDP Datagram 1806 When the server receives a UDP datagram at a currently allocated 1807 relayed transport address, the server looks up the allocation 1808 associated with the relayed transport address. The server then 1809 checks to see whether the set of permissions for the allocation allow 1810 the relaying of the UDP datagram as described in Section 8. 1812 If relaying is permitted, then the server checks if there is a 1813 channel bound to the peer that sent the UDP datagram (see 1814 Section 11). If a channel is bound, then processing proceeds as 1815 described in Section 11.7. 1817 If relaying is permitted but no channel is bound to the peer, then 1818 the server forms and sends a Data indication. The Data indication 1819 MUST contain both an XOR-PEER-ADDRESS and a DATA attribute. The DATA 1820 attribute is set to the value of the 'data octets' field from the 1821 datagram, and the XOR-PEER-ADDRESS attribute is set to the source 1822 transport address of the received UDP datagram. The Data indication 1823 is then sent on the 5-tuple associated with the allocation. 1825 10.4. Receiving a Data Indication with DATA attribute 1827 When the client receives a Data indication with DATA attribute, it 1828 checks that the Data indication contains an XOR-PEER-ADDRESS 1829 attribute, and discards the indication if it does not. The client 1830 SHOULD also check that the XOR-PEER-ADDRESS attribute value contains 1831 an IP address with which the client believes there is an active 1832 permission, and discard the Data indication otherwise. Note that the 1833 DATA attribute is allowed to contain zero bytes of data. 1835 NOTE: The latter check protects the client against an attacker who 1836 somehow manages to trick the server into installing permissions 1837 not desired by the client. 1839 If the Data indication passes the above checks, the client delivers 1840 the data octets inside the DATA attribute to the application, along 1841 with an indication that they were received from the peer whose 1842 transport address is given by the XOR-PEER-ADDRESS attribute. 1844 10.5. Receiving an ICMP Packet 1846 When the server receives an ICMP packet, the server verifies that the 1847 type is either 3, 11 or 12 for an ICMPv4 [RFC0792] packet or either 1848 1, 2, or 3 for an ICMPv6 [RFC4443] packet. It also verifies that the 1849 IP packet in the ICMP packet payload contains a UDP header. If 1850 either of these conditions fail, then the ICMP packet is silently 1851 dropped. 1853 The server looks up the allocation whose relayed transport address 1854 corresponds to the encapsulated packet's source IP address and UDP 1855 port. If no such allocation exists, the packet is silently dropped. 1856 The server then checks to see whether the set of permissions for the 1857 allocation allows the relaying of the ICMP packet. For ICMP packets, 1858 the source IP address MUST NOT be checked against the permissions 1859 list as it would be for UDP packets. Instead, the server extracts 1860 the destination IP address from the encapsulated IP header. The 1861 server then compares this address with the IP address associated with 1862 each permission in the list of permissions for the allocation. If no 1863 match is found, relaying is not permitted, and the server silently 1864 discards the ICMP packet. Note that only addresses are compared and 1865 port numbers are not considered. 1867 If relaying is permitted then the server forms and sends a Data 1868 indication. The Data indication MUST contain both an XOR-PEER- 1869 ADDRESS and an ICMP attribute. The ICMP attribute is set to the 1870 value of the type and code fields from the ICMP packet. The IP 1871 address portion of XOR-PEER-ADDRESS attribute is set to the 1872 destination IP address in the encapsulated IP header. At the time of 1873 writing of this specification, Socket APIs on some operating systems 1874 do not deliver the destination port in the encapsulated UDP header to 1875 applications without superuser privileges. If destination port in 1876 the encapsulated UDP header is available to the server then the port 1877 portion of XOR-PEER-ADDRESS attribute is set to the destination port 1878 otherwise the port portion is set to 0. The Data indication is then 1879 sent on the 5-tuple associated with the allocation. 1881 10.6. Receiving a Data Indication with an ICMP attribute 1883 When the client receives a Data indication with an ICMP attribute, it 1884 checks that the Data indication contains an XOR-PEER-ADDRESS 1885 attribute, and discards the indication if it does not. The client 1886 SHOULD also check that the XOR-PEER-ADDRESS attribute value contains 1887 an IP address with an active permission, and discard the Data 1888 indication otherwise. 1890 If the Data indication passes the above checks, the client signals 1891 the application of the error condition, along with an indication that 1892 it was received from the peer whose transport address is given by the 1893 XOR-PEER-ADDRESS attribute. The application can make sense of the 1894 meaning of the type and code values in the ICMP attribute by using 1895 the family field in the XOR-PEER-ADDRESS attribute. 1897 11. Channels 1899 Channels provide a way for the client and server to send application 1900 data using ChannelData messages, which have less overhead than Send 1901 and Data indications. 1903 The ChannelData message (see Section 11.4) starts with a two-byte 1904 field that carries the channel number. The values of this field are 1905 allocated as follows: 1907 0x0000 through 0x3FFF: These values can never be used for channel 1908 numbers. 1910 0x4000 through 0x7FFF: These values are the allowed channel 1911 numbers (16,384 possible values). 1913 0x8000 through 0xFFFF: These values are reserved for future use. 1915 Because of this division, ChannelData messages can be distinguished 1916 from STUN-formatted messages (e.g., Allocate request, Send 1917 indication, etc.) by examining the first two bits of the message: 1919 0b00: STUN-formatted message (since the first two bits of a STUN- 1920 formatted message are always zero). 1922 0b01: ChannelData message (since the channel number is the first 1923 field in the ChannelData message and channel numbers fall in the 1924 range 0x4000 - 0x7FFF). 1926 0b10: Reserved 1928 0b11: Reserved 1930 The reserved values may be used in the future to extend the range of 1931 channel numbers. Thus, an implementation MUST NOT assume that a TURN 1932 message always starts with a 0 bit. 1934 Channel bindings are always initiated by the client. The client can 1935 bind a channel to a peer at any time during the lifetime of the 1936 allocation. The client may bind a channel to a peer before 1937 exchanging data with it, or after exchanging data with it (using Send 1938 and Data indications) for some time, or may choose never to bind a 1939 channel to it. The client can also bind channels to some peers while 1940 not binding channels to other peers. 1942 Channel bindings are specific to an allocation, so that the use of a 1943 channel number or peer transport address in a channel binding in one 1944 allocation has no impact on their use in a different allocation. If 1945 an allocation expires, all its channel bindings expire with it. 1947 A channel binding consists of: 1949 o a channel number; 1951 o a transport address (of the peer); and 1953 o A time-to-expiry timer. 1955 Within the context of an allocation, a channel binding is uniquely 1956 identified either by the channel number or by the peer's transport 1957 address. Thus, the same channel cannot be bound to two different 1958 transport addresses, nor can the same transport address be bound to 1959 two different channels. 1961 A channel binding lasts for 10 minutes unless refreshed. Refreshing 1962 the binding (by the server receiving a ChannelBind request rebinding 1963 the channel to the same peer) resets the time-to-expiry timer back to 1964 10 minutes. 1966 When the channel binding expires, the channel becomes unbound. Once 1967 unbound, the channel number can be bound to a different transport 1968 address, and the transport address can be bound to a different 1969 channel number. To prevent race conditions, the client MUST wait 5 1970 minutes after the channel binding expires before attempting to bind 1971 the channel number to a different transport address or the transport 1972 address to a different channel number. 1974 When binding a channel to a peer, the client SHOULD be prepared to 1975 receive ChannelData messages on the channel from the server as soon 1976 as it has sent the ChannelBind request. Over UDP, it is possible for 1977 the client to receive ChannelData messages from the server before it 1978 receives a ChannelBind success response. 1980 In the other direction, the client MAY elect to send ChannelData 1981 messages before receiving the ChannelBind success response. Doing 1982 so, however, runs the risk of having the ChannelData messages dropped 1983 by the server if the ChannelBind request does not succeed for some 1984 reason (e.g., packet lost if the request is sent over UDP, or the 1985 server being unable to fulfill the request). A client that wishes to 1986 be safe should either queue the data or use Send indications until 1987 the channel binding is confirmed. 1989 11.1. Sending a ChannelBind Request 1991 A channel binding is created or refreshed using a ChannelBind 1992 transaction. A ChannelBind transaction also creates or refreshes a 1993 permission towards the peer (see Section 8). 1995 To initiate the ChannelBind transaction, the client forms a 1996 ChannelBind request. The channel to be bound is specified in a 1997 CHANNEL-NUMBER attribute, and the peer's transport address is 1998 specified in an XOR-PEER-ADDRESS attribute. Section 11.2 describes 1999 the restrictions on these attributes. The client MUST only include 2000 an XOR-PEER-ADDRESS attribute with an address of the same address 2001 family as that of the relayed transport address for the allocation. 2003 Rebinding a channel to the same transport address that it is already 2004 bound to provides a way to refresh a channel binding and the 2005 corresponding permission without sending data to the peer. Note 2006 however, that permissions need to be refreshed more frequently than 2007 channels. 2009 11.2. Receiving a ChannelBind Request 2011 When the server receives a ChannelBind request, it processes as per 2012 Section 4 plus the specific rules mentioned here. 2014 The server checks the following: 2016 o The request contains both a CHANNEL-NUMBER and an XOR-PEER-ADDRESS 2017 attribute; 2019 o The channel number is in the range 0x4000 through 0x7FFE 2020 (inclusive); 2022 o The channel number is not currently bound to a different transport 2023 address (same transport address is OK); 2025 o The transport address is not currently bound to a different 2026 channel number. 2028 o If the XOR-PEER-ADDRESS attribute contains an address of an 2029 address family that is not the same as that of the relayed 2030 transport address for the allocation, the server MUST generate an 2031 error response with the 443 (Peer Address Family Mismatch) 2032 response code. 2034 If any of these tests fail, the server replies with a 400 (Bad 2035 Request) error. 2037 The server MAY impose restrictions on the IP address and port values 2038 allowed in the XOR-PEER-ADDRESS attribute -- if a value is not 2039 allowed, the server rejects the request with a 403 (Forbidden) error. 2041 If the request is valid, but the server is unable to fulfill the 2042 request due to some capacity limit or similar, the server replies 2043 with a 508 (Insufficient Capacity) error. 2045 Otherwise, the server replies with a ChannelBind success response. 2046 There are no required attributes in a successful ChannelBind 2047 response. 2049 If the server can satisfy the request, then the server creates or 2050 refreshes the channel binding using the channel number in the 2051 CHANNEL-NUMBER attribute and the transport address in the XOR-PEER- 2052 ADDRESS attribute. The server also installs or refreshes a 2053 permission for the IP address in the XOR-PEER-ADDRESS attribute as 2054 described in Section 8. 2056 NOTE: A server need not do anything special to implement 2057 idempotency of ChannelBind requests over UDP using the "stateless 2058 stack approach". Retransmitted ChannelBind requests will simply 2059 refresh the channel binding and the corresponding permission. 2060 Furthermore, the client must wait 5 minutes before binding a 2061 previously bound channel number or peer address to a different 2062 channel, eliminating the possibility that the transaction would 2063 initially fail but succeed on a retransmission. 2065 11.3. Receiving a ChannelBind Response 2067 When the client receives a ChannelBind success response, it updates 2068 its data structures to record that the channel binding is now active. 2069 It also updates its data structures to record that the corresponding 2070 permission has been installed or refreshed. 2072 If the client receives a ChannelBind failure response that indicates 2073 that the channel information is out-of-sync between the client and 2074 the server (e.g., an unexpected 400 "Bad Request" response), then it 2075 is RECOMMENDED that the client immediately delete the allocation and 2076 start afresh with a new allocation. 2078 11.4. The ChannelData Message 2080 The ChannelData message is used to carry application data between the 2081 client and the server. It has the following format: 2083 0 1 2 3 2084 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 2085 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2086 | Channel Number | Length | 2087 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2088 | | 2089 / Application Data / 2090 / / 2091 | | 2092 | +-------------------------------+ 2093 | | 2094 +-------------------------------+ 2096 The Channel Number field specifies the number of the channel on which 2097 the data is traveling, and thus the address of the peer that is 2098 sending or is to receive the data. 2100 The Length field specifies the length in bytes of the application 2101 data field (i.e., it does not include the size of the ChannelData 2102 header). Note that 0 is a valid length. 2104 The Application Data field carries the data the client is trying to 2105 send to the peer, or that the peer is sending to the client. 2107 11.5. Sending a ChannelData Message 2109 Once a client has bound a channel to a peer, then when the client has 2110 data to send to that peer it may use either a ChannelData message or 2111 a Send indication; that is, the client is not obligated to use the 2112 channel when it exists and may freely intermix the two message types 2113 when sending data to the peer. The server, on the other hand, MUST 2114 use the ChannelData message if a channel has been bound to the peer. 2115 The server uses a Data indication to signal the XOR-PEER-ADDRESS and 2116 ICMP attributes to the client even if a channel has been bound to the 2117 peer. 2119 The fields of the ChannelData message are filled in as described in 2120 Section 11.4. 2122 Over TCP and TLS-over-TCP, the ChannelData message MUST be padded to 2123 a multiple of four bytes in order to ensure the alignment of 2124 subsequent messages. The padding is not reflected in the length 2125 field of the ChannelData message, so the actual size of a ChannelData 2126 message (including padding) is (4 + Length) rounded up to the nearest 2127 multiple of 4. Over UDP, the padding is not required but MAY be 2128 included. 2130 The ChannelData message is then sent on the 5-tuple associated with 2131 the allocation. 2133 11.6. Receiving a ChannelData Message 2135 The receiver of the ChannelData message uses the first two bits to 2136 distinguish it from STUN-formatted messages, as described above. If 2137 the message uses a value in the reserved range (0x8000 through 2138 0xFFFF), then the message is silently discarded. 2140 If the ChannelData message is received in a UDP datagram, and if the 2141 UDP datagram is too short to contain the claimed length of the 2142 ChannelData message (i.e., the UDP header length field value is less 2143 than the ChannelData header length field value + 4 + 8), then the 2144 message is silently discarded. 2146 If the ChannelData message is received over TCP or over TLS-over-TCP, 2147 then the actual length of the ChannelData message is as described in 2148 Section 11.5. 2150 If the ChannelData message is received on a channel that is not bound 2151 to any peer, then the message is silently discarded. 2153 On the client, it is RECOMMENDED that the client discard the 2154 ChannelData message if the client believes there is no active 2155 permission towards the peer. On the server, the receipt of a 2156 ChannelData message MUST NOT refresh either the channel binding or 2157 the permission towards the peer. 2159 On the server, if no errors are detected, the server relays the 2160 application data to the peer by forming a UDP datagram as follows: 2162 o the source transport address is the relayed transport address of 2163 the allocation, where the allocation is determined by the 5-tuple 2164 on which the ChannelData message arrived; 2166 o the destination transport address is the transport address to 2167 which the channel is bound; 2169 o the data following the UDP header is the contents of the data 2170 field of the ChannelData message. 2172 The resulting UDP datagram is then sent to the peer. Note that if 2173 the Length field in the ChannelData message is 0, then there will be 2174 no data in the UDP datagram, but the UDP datagram is still formed and 2175 sent. 2177 11.7. Relaying Data from the Peer 2179 When the server receives a UDP datagram on the relayed transport 2180 address associated with an allocation, the server processes it as 2181 described in Section 10.3. If that section indicates that a 2182 ChannelData message should be sent (because there is a channel bound 2183 to the peer that sent to the UDP datagram), then the server forms and 2184 sends a ChannelData message as described in Section 11.5. 2186 When the server receives an ICMP packet, the server processes it as 2187 described in Section 10.5. A Data indication MUST be sent regardless 2188 if there is a channel bound to the peer that was the destination of 2189 the UDP datagram that triggered the reception of the ICMP packet. 2191 12. Packet Translations 2193 As discussed in Section 2.6, translations in TURN are designed so 2194 that a TURN server can be implemented as an application that runs in 2195 userland under commonly available operating systems and that does not 2196 require special privileges. The translations specified in the 2197 following sections follow this principle. 2199 The descriptions below have two parts: a preferred behavior and an 2200 alternate behavior. The server SHOULD implement the preferred 2201 behavior. Otherwise, the server MUST implement the alternate 2202 behavior and MUST NOT do anything else for the reasons detailed in 2203 [RFC6145]. 2205 12.1. IPv4-to-IPv6 Translations 2207 Traffic Class 2209 Preferred behavior: As specified in Section 4 of [RFC6145]. 2211 Alternate behavior: The relay sets the Traffic Class to the 2212 default value for outgoing packets. 2214 Flow Label 2216 Preferred behavior: The relay sets the Flow label to 0. The relay 2217 can choose to set the Flow label to a different value if it 2218 supports the IPv6 Flow Label field[RFC3697]. 2220 Alternate behavior: the relay sets the Flow label to the default 2221 value for outgoing packets. 2223 Hop Limit 2225 Preferred behavior: As specified in Section 4 of [RFC6145]. 2227 Alternate behavior: The relay sets the Hop Limit to the default 2228 value for outgoing packets. 2230 Fragmentation 2232 Preferred behavior: As specified in Section 4 of [RFC6145]. 2234 Alternate behavior: The relay assembles incoming fragments. The 2235 relay follows its default behavior to send outgoing packets. 2237 For both preferred and alternate behavior, the DONT-FRAGMENT 2238 attribute MUST be ignored by the server. 2240 Extension Headers 2242 Preferred behavior: The relay sends outgoing packet without any 2243 IPv6 extension headers, with the exception of the Fragmentation 2244 header as described above. 2246 Alternate behavior: Same as preferred. 2248 12.2. IPv6-to-IPv6 Translations 2250 Flow Label 2252 The relay should consider that it is handling two different IPv6 2253 flows. Therefore, the Flow label [RFC3697] SHOULD NOT be copied as 2254 part of the translation. 2256 Preferred behavior: The relay sets the Flow label to 0. The relay 2257 can choose to set the Flow label to a different value if it 2258 supports the IPv6 Flow Label field[RFC3697]. 2260 Alternate behavior: The relay sets the Flow label to the default 2261 value for outgoing packets. 2263 Hop Limit 2265 Preferred behavior: The relay acts as a regular router with 2266 respect to decrementing the Hop Limit and generating an ICMPv6 2267 error if it reaches zero. 2269 Alternate behavior: The relay sets the Hop Limit to the default 2270 value for outgoing packets. 2272 Fragmentation 2274 Preferred behavior: If the incoming packet did not include a 2275 Fragment header and the outgoing packet size does not exceed the 2276 outgoing link's MTU, the relay sends the outgoing packet without a 2277 Fragment header. 2279 If the incoming packet did not include a Fragment header and the 2280 outgoing packet size exceeds the outgoing link's MTU, the relay 2281 drops the outgoing packet and send an ICMP message of type 2 code 2282 0 ("Packet too big") to the sender of the incoming packet. If 2283 the packet is being sent to the peer, the relay reduces the MTU 2284 reported in the ICMP message by 48 bytes to allow room for the 2285 overhead of a Data indication. 2287 If the incoming packet included a Fragment header and the outgoing 2288 packet size (with a Fragment header included) does not exceed the 2289 outgoing link's MTU, the relay sends the outgoing packet with a 2290 Fragment header. The relay sets the fields of the Fragment header 2291 as appropriate for a packet originating from the server. 2293 If the incoming packet included a Fragment header and the outgoing 2294 packet size exceeds the outgoing link's MTU, the relay MUST 2295 fragment the outgoing packet into fragments of no more than 1280 2296 bytes. The relay sets the fields of the Fragment header as 2297 appropriate for a packet originating from the server. 2299 Alternate behavior: The relay assembles incoming fragments. The 2300 relay follows its default behavior to send outgoing packets. 2302 For both preferred and alternate behavior, the DONT-FRAGMENT 2303 attribute MUST be ignored by the server. 2305 Extension Headers 2307 Preferred behavior: The relay sends outgoing packet without any 2308 IPv6 extension headers, with the exception of the Fragmentation 2309 header as described above. 2311 Alternate behavior: Same as preferred. 2313 12.3. IPv6-to-IPv4 Translations 2315 Type of Service and Precedence 2317 Preferred behavior: As specified in Section 5 of [RFC6145]. 2319 Alternate behavior: The relay sets the Type of Service and 2320 Precedence to the default value for outgoing packets. 2322 Time to Live 2324 Preferred behavior: As specified in Section 5 of [RFC6145]. 2326 Alternate behavior: The relay sets the Time to Live to the default 2327 value for outgoing packets. 2329 Fragmentation 2331 Preferred behavior: As specified in Section 5 of [RFC6145]. 2332 Additionally, when the outgoing packet's size exceeds the 2333 outgoing link's MTU, the relay needs to generate an ICMP error 2334 (ICMPv6 Packet Too Big) reporting the MTU size. If the packet is 2335 being sent to the peer, the relay SHOULD reduce the MTU reported 2336 in the ICMP message by 48 bytes to allow room for the overhead of 2337 a Data indication. 2339 Alternate behavior: The relay assembles incoming fragments. The 2340 relay follows its default behavior to send outgoing packets. 2342 For both preferred and alternate behavior, the DONT-FRAGMENT 2343 attribute MUST be ignored by the server. 2345 13. IP Header Fields 2347 This section describes how the server sets various fields in the IP 2348 header when relaying between the client and the peer or vice versa. 2349 The descriptions in this section apply: (a) when the server sends a 2350 UDP datagram to the peer, or (b) when the server sends a Data 2351 indication or ChannelData message to the client over UDP transport. 2352 The descriptions in this section do not apply to TURN messages sent 2353 over TCP or TLS transport from the server to the client. 2355 The descriptions below have two parts: a preferred behavior and an 2356 alternate behavior. The server SHOULD implement the preferred 2357 behavior, but if that is not possible for a particular field, then it 2358 SHOULD implement the alternative behavior. 2360 Time to Live (TTL) field 2362 Preferred Behavior: If the incoming value is 0, then the drop the 2363 incoming packet. Otherwise, set the outgoing Time to Live/Hop 2364 Count to one less than the incoming value. 2366 Alternate Behavior: Set the outgoing value to the default for 2367 outgoing packets. 2369 Differentiated Services Code Point (DSCP) field [RFC2474] 2371 Preferred Behavior: Set the outgoing value to the incoming value, 2372 unless the server includes a differentiated services classifier 2373 and marker [RFC2474]. 2375 Alternate Behavior: Set the outgoing value to a fixed value, which 2376 by default is Best Effort unless configured otherwise. 2378 In both cases, if the server is immediately adjacent to a 2379 differentiated services classifier and marker, then DSCP MAY be 2380 set to any arbitrary value in the direction towards the 2381 classifier. 2383 Explicit Congestion Notification (ECN) field [RFC3168] 2385 Preferred Behavior: Set the outgoing value to the incoming value, 2386 UNLESS the server is doing Active Queue Management, the incoming 2387 ECN field is ECT(1) (=0b01) or ECT(0) (=0b10), and the server 2388 wishes to indicate that congestion has been experienced, in which 2389 case set the outgoing value to CE (=0b11). 2391 Alternate Behavior: Set the outgoing value to Not-ECT (=0b00). 2393 IPv4 Fragmentation fields 2395 Preferred Behavior: When the server sends a packet to a peer in 2396 response to a Send indication containing the DONT-FRAGMENT 2397 attribute, then set the DF bit in the outgoing IP header to 1. In 2398 all other cases when sending an outgoing packet containing 2399 application data (e.g., Data indication, ChannelData message, or 2400 DONT-FRAGMENT attribute not included in the Send indication), copy 2401 the DF bit from the DF bit of the incoming packet that contained 2402 the application data. 2404 Set the other fragmentation fields (Identification, More 2405 Fragments, Fragment Offset) as appropriate for a packet 2406 originating from the server. 2408 Alternate Behavior: As described in the Preferred Behavior, except 2409 always assume the incoming DF bit is 0. 2411 In both the Preferred and Alternate Behaviors, the resulting 2412 packet may be too large for the outgoing link. If this is the 2413 case, then the normal fragmentation rules apply [RFC1122]. 2415 IPv4 Options 2417 Preferred Behavior: The outgoing packet is sent without any IPv4 2418 options. 2420 Alternate Behavior: Same as preferred. 2422 14. New STUN Methods 2424 This section lists the codepoints for the new STUN methods defined in 2425 this specification. See elsewhere in this document for the semantics 2426 of these new methods. 2428 0x003 : Allocate (only request/response semantics defined) 2429 0x004 : Refresh (only request/response semantics defined) 2430 0x006 : Send (only indication semantics defined) 2431 0x007 : Data (only indication semantics defined) 2432 0x008 : CreatePermission (only request/response semantics defined 2433 0x009 : ChannelBind (only request/response semantics defined) 2435 15. New STUN Attributes 2437 This STUN extension defines the following new attributes: 2439 0x000C: CHANNEL-NUMBER 2440 0x000D: LIFETIME 2441 0x0010: Reserved (was BANDWIDTH) 2442 0x0012: XOR-PEER-ADDRESS 2443 0x0013: DATA 2444 0x0016: XOR-RELAYED-ADDRESS 2445 0x0017: REQUESTED-ADDRESS-FAMILY 2446 0x0018: EVEN-PORT 2447 0x0019: REQUESTED-TRANSPORT 2448 0x001A: DONT-FRAGMENT 2449 0x0021: Reserved (was TIMER-VAL) 2450 0x0022: RESERVATION-TOKEN 2451 TBD-CA: ADDITIONAL-ADDRESS-FAMILY 2452 TBD-CA: ADDRESS-ERROR-CODE 2453 TBD-CA: ICMP 2455 Some of these attributes have lengths that are not multiples of 4. 2456 By the rules of STUN, any attribute whose length is not a multiple of 2457 4 bytes MUST be immediately followed by 1 to 3 padding bytes to 2458 ensure the next attribute (if any) would start on a 4-byte boundary 2459 (see [RFC5389]). 2461 15.1. CHANNEL-NUMBER 2463 The CHANNEL-NUMBER attribute contains the number of the channel. The 2464 value portion of this attribute is 4 bytes long and consists of a 2465 16-bit unsigned integer, followed by a two-octet RFFU (Reserved For 2466 Future Use) field, which MUST be set to 0 on transmission and MUST be 2467 ignored on reception. 2469 0 1 2 3 2470 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 2471 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2472 | Channel Number | RFFU = 0 | 2473 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2475 15.2. LIFETIME 2477 The LIFETIME attribute represents the duration for which the server 2478 will maintain an allocation in the absence of a refresh. The value 2479 portion of this attribute is 4-bytes long and consists of a 32-bit 2480 unsigned integral value representing the number of seconds remaining 2481 until expiration. 2483 15.3. XOR-PEER-ADDRESS 2485 The XOR-PEER-ADDRESS specifies the address and port of the peer as 2486 seen from the TURN server. (For example, the peer's server-reflexive 2487 transport address if the peer is behind a NAT.) It is encoded in the 2488 same way as XOR-MAPPED-ADDRESS [RFC5389]. 2490 15.4. DATA 2492 The DATA attribute is present in all Send and Data indications. The 2493 value portion of this attribute is variable length and consists of 2494 the application data (that is, the data that would immediately follow 2495 the UDP header if the data was been sent directly between the client 2496 and the peer). If the length of this attribute is not a multiple of 2497 4, then padding must be added after this attribute. 2499 15.5. XOR-RELAYED-ADDRESS 2501 The XOR-RELAYED-ADDRESS is present in Allocate responses. It 2502 specifies the address and port that the server allocated to the 2503 client. It is encoded in the same way as XOR-MAPPED-ADDRESS 2504 [RFC5389]. 2506 15.6. REQUESTED-ADDRESS-FAMILY 2508 This attribute is used by clients to request the allocation of a 2509 specific address type from a server. The following is the format of 2510 the REQUESTED-ADDRESS-FAMILY attribute. Note that TURN attributes 2511 are TLV (Type-Length-Value) encoded, with a 16-bit type, a 16-bit 2512 length, and a variable-length value. 2514 0 1 2 3 2515 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 2516 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2517 | Type | Length | 2518 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2519 | Family | Reserved | 2520 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2522 Type: the type of the REQUESTED-ADDRESS-FAMILY attribute is 0x0017. 2523 As specified in [RFC5389], attributes with values between 0x0000 2524 and 0x7FFF are comprehension-required, which means that the client 2525 or server cannot successfully process the message unless it 2526 understands the attribute. 2528 Length: this 16-bit field contains the length of the attribute in 2529 bytes. The length of this attribute is 4 bytes. 2531 Family: there are two values defined for this field and specified in 2532 [RFC5389], Section 15.1: 0x01 for IPv4 addresses and 0x02 for IPv6 2533 addresses. 2535 Reserved: at this point, the 24 bits in the Reserved field MUST be 2536 set to zero by the client and MUST be ignored by the server. 2538 The REQUEST-ADDRESS-TYPE attribute MAY only be present in Allocate 2539 requests. 2541 15.7. EVEN-PORT 2543 This attribute allows the client to request that the port in the 2544 relayed transport address be even, and (optionally) that the server 2545 reserve the next-higher port number. The value portion of this 2546 attribute is 1 byte long. Its format is: 2548 0 2549 0 1 2 3 4 5 6 7 2550 +-+-+-+-+-+-+-+-+ 2551 |R| RFFU | 2552 +-+-+-+-+-+-+-+-+ 2554 The value contains a single 1-bit flag: 2556 R: If 1, the server is requested to reserve the next-higher port 2557 number (on the same IP address) for a subsequent allocation. If 2558 0, no such reservation is requested. 2560 The other 7 bits of the attribute's value must be set to zero on 2561 transmission and ignored on reception. 2563 Since the length of this attribute is not a multiple of 4, padding 2564 must immediately follow this attribute. 2566 15.8. REQUESTED-TRANSPORT 2568 This attribute is used by the client to request a specific transport 2569 protocol for the allocated transport address. The value of this 2570 attribute is 4 bytes with the following format: 2572 0 1 2 3 2573 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 2574 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2575 | Protocol | RFFU | 2576 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2578 The Protocol field specifies the desired protocol. The codepoints 2579 used in this field are taken from those allowed in the Protocol field 2580 in the IPv4 header and the NextHeader field in the IPv6 header 2581 [Protocol-Numbers]. This specification only allows the use of 2582 codepoint 17 (User Datagram Protocol). 2584 The RFFU field MUST be set to zero on transmission and MUST be 2585 ignored on reception. It is reserved for future uses. 2587 15.9. DONT-FRAGMENT 2589 This attribute is used by the client to request that the server set 2590 the DF (Don't Fragment) bit in the IP header when relaying the 2591 application data onward to the peer. This attribute has no value 2592 part and thus the attribute length field is 0. 2594 15.10. RESERVATION-TOKEN 2596 The RESERVATION-TOKEN attribute contains a token that uniquely 2597 identifies a relayed transport address being held in reserve by the 2598 server. The server includes this attribute in a success response to 2599 tell the client about the token, and the client includes this 2600 attribute in a subsequent Allocate request to request the server use 2601 that relayed transport address for the allocation. 2603 The attribute value is 8 bytes and contains the token value. 2605 15.11. ADDITIONAL-ADDRESS-FAMILY 2607 This attribute is used by clients to request the allocation of a IPv4 2608 and IPv6 address type from a server. It is encoded in the same way 2609 as REQUESTED-ADDRESS-FAMILY Section 15.6. The ADDITIONAL-ADDRESS- 2610 FAMILY attribute MAY be present in Allocate request. The attribute 2611 value of 0x02 (IPv6 address) is the only valid value in Allocate 2612 request. 2614 15.12. ADDRESS-ERROR-CODE Attribute 2616 This attribute is used by servers to signal the reason for not 2617 allocating the requested address family. The following is the format 2618 of the ADDRESS-ERROR-CODE attribute. 2620 0 1 2 3 2621 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 2622 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2623 | Type | Length | 2624 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2625 | Family | Rsvd |Class| Number | 2626 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2627 | Reason Phrase (variable) .. 2628 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2630 Type: the type of the ADDRESS-ERROR-CODE attribute is TBD-CA. As 2631 specified in [RFC5389], attributes with values between 0x8000 and 2632 0xFFFF are comprehension-optional, which means that the client or 2633 server can safely ignore the attribute if they don't understand 2634 it. 2636 Length: this 16-bit field contains the length of the attribute in 2637 bytes. 2639 Family: there are two values defined for this field and specified in 2640 [RFC5389], Section 15.1: 0x01 for IPv4 addresses and 0x02 for IPv6 2641 addresses. 2643 Reserved: at this point, the 13 bits in the Reserved field MUST be 2644 set to zero by the client and MUST be ignored by the server. 2646 Class: The Class represents the hundreds digit of the error code and 2647 is defined in section 15.6 of [RFC5389]. 2649 Number: this 8-bit field contains the reason server cannot allocate 2650 one of the requested address types. The error code values could 2651 be either 440 (unsupported address family) or 508 (insufficient 2652 capacity). The number representation is defined in section 15.6 2653 of [RFC5389]. 2655 Reason Phrase: The recommended reason phrases for error codes 440 2656 and 508 are explained in Section 16. 2658 The ADDRESS-ERROR-CODE attribute MAY only be present in Allocate 2659 responses. 2661 15.13. ICMP Attribute 2663 This attribute is used by servers to signal the reason an UDP packet 2664 was dropped. The following is the format of the ICMP attribute. 2666 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 2667 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2668 | Reserved | Type | Code | 2669 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2671 Reserved: This field MUST be set to 0 when sent, and MUST be ignored 2672 when received. 2674 Type: The field contains the value in the ICMP type. Its 2675 interpretation depends whether the ICMP was received over IPv4 or 2676 IPv6. 2678 Code: The field contains the value in the ICMP code. Its 2679 interpretation depends whether the ICMP was received over IPv4 or 2680 IPv6. 2682 16. New STUN Error Response Codes 2684 This document defines the following new error response codes: 2686 403 (Forbidden): The request was valid but cannot be performed due 2687 to administrative or similar restrictions. 2689 437 (Allocation Mismatch): A request was received by the server that 2690 requires an allocation to be in place, but no allocation exists, 2691 or a request was received that requires no allocation, but an 2692 allocation exists. 2694 440 (Address Family not Supported): The server does not support the 2695 address family requested by the client. 2697 441 (Wrong Credentials): The credentials in the (non-Allocate) 2698 request do not match those used to create the allocation. 2700 442 (Unsupported Transport Protocol): The Allocate request asked the 2701 server to use a transport protocol between the server and the peer 2702 that the server does not support. NOTE: This does NOT refer to 2703 the transport protocol used in the 5-tuple. 2705 443 (Peer Address Family Mismatch). A peer address is part of a 2706 different address family than that of the relayed transport 2707 address of the allocation. 2709 486 (Allocation Quota Reached): No more allocations using this 2710 username can be created at the present time. 2712 508 (Insufficient Capacity): The server is unable to carry out the 2713 request due to some capacity limit being reached. In an Allocate 2714 response, this could be due to the server having no more relayed 2715 transport addresses available at that time, having none with the 2716 requested properties, or the one that corresponds to the specified 2717 reservation token is not available. 2719 17. Detailed Example 2721 This section gives an example of the use of TURN, showing in detail 2722 the contents of the messages exchanged. The example uses the network 2723 diagram shown in the Overview (Figure 1). 2725 For each message, the attributes included in the message and their 2726 values are shown. For convenience, values are shown in a human- 2727 readable format rather than showing the actual octets; for example, 2728 "XOR-RELAYED-ADDRESS=192.0.2.15:9000" shows that the XOR-RELAYED- 2729 ADDRESS attribute is included with an address of 192.0.2.15 and a 2730 port of 9000, here the address and port are shown before the xor-ing 2731 is done. For attributes with string-like values (e.g., 2732 SOFTWARE="Example client, version 1.03" and 2733 NONCE="adl7W7PeDU4hKE72jdaQvbAMcr6h39sm"), the value of the attribute 2734 is shown in quotes for readability, but these quotes do not appear in 2735 the actual value. 2737 TURN TURN Peer Peer 2738 client server A B 2739 | | | | 2740 |--- Allocate request -------------->| | | 2741 | Transaction-Id=0xA56250D3F17ABE679422DE85 | | 2742 | SOFTWARE="Example client, version 1.03" | | 2743 | LIFETIME=3600 (1 hour) | | | 2744 | REQUESTED-TRANSPORT=17 (UDP) | | | 2745 | DONT-FRAGMENT | | | 2746 | | | | 2747 |<-- Allocate error response --------| | | 2748 | Transaction-Id=0xA56250D3F17ABE679422DE85 | | 2749 | SOFTWARE="Example server, version 1.17" | | 2750 | ERROR-CODE=401 (Unauthorized) | | | 2751 | REALM="example.com" | | | 2752 | NONCE="adl7W7PeDU4hKE72jdaQvbAMcr6h39sm" | | 2753 | | | | 2754 |--- Allocate request -------------->| | | 2755 | Transaction-Id=0xC271E932AD7446A32C234492 | | 2756 | SOFTWARE="Example client 1.03" | | | 2757 | LIFETIME=3600 (1 hour) | | | 2758 | REQUESTED-TRANSPORT=17 (UDP) | | | 2759 | DONT-FRAGMENT | | | 2760 | USERNAME="George" | | | 2761 | REALM="example.com" | | | 2762 | NONCE="adl7W7PeDU4hKE72jdaQvbAMcr6h39sm" | | 2763 | MESSAGE-INTEGRITY=... | | | 2764 | | | | 2765 |<-- Allocate success response ------| | | 2766 | Transaction-Id=0xC271E932AD7446A32C234492 | | 2767 | SOFTWARE="Example server, version 1.17" | | 2768 | LIFETIME=1200 (20 minutes) | | | 2769 | XOR-RELAYED-ADDRESS=192.0.2.15:50000 | | 2770 | XOR-MAPPED-ADDRESS=192.0.2.1:7000 | | 2771 | MESSAGE-INTEGRITY=... | | | 2773 The client begins by selecting a host transport address to use for 2774 the TURN session; in this example, the client has selected 2775 10.1.1.2:49721 as shown in Figure 1. The client then sends an 2776 Allocate request to the server at the server transport address. The 2777 client randomly selects a 96-bit transaction id of 2778 0xA56250D3F17ABE679422DE85 for this transaction; this is encoded in 2779 the transaction id field in the fixed header. The client includes a 2780 SOFTWARE attribute that gives information about the client's 2781 software; here the value is "Example client, version 1.03" to 2782 indicate that this is version 1.03 of something called the Example 2783 client. The client includes the LIFETIME attribute because it wishes 2784 the allocation to have a longer lifetime than the default of 10 2785 minutes; the value of this attribute is 3600 seconds, which 2786 corresponds to 1 hour. The client must always include a REQUESTED- 2787 TRANSPORT attribute in an Allocate request and the only value allowed 2788 by this specification is 17, which indicates UDP transport between 2789 the server and the peers. The client also includes the DONT-FRAGMENT 2790 attribute because it wishes to use the DONT-FRAGMENT attribute later 2791 in Send indications; this attribute consists of only an attribute 2792 header, there is no value part. We assume the client has not 2793 recently interacted with the server, thus the client does not include 2794 USERNAME, REALM, NONCE, or MESSAGE-INTEGRITY attribute. Finally, 2795 note that the order of attributes in a message is arbitrary (except 2796 for the MESSAGE-INTEGRITY and FINGERPRINT attributes) and the client 2797 could have used a different order. 2799 Servers require any request to be authenticated. Thus, when the 2800 server receives the initial Allocate request, it rejects the request 2801 because the request does not contain the authentication attributes. 2802 Following the procedures of the long-term credential mechanism of 2803 STUN [RFC5389], the server includes an ERROR-CODE attribute with a 2804 value of 401 (Unauthorized), a REALM attribute that specifies the 2805 authentication realm used by the server (in this case, the server's 2806 domain "example.com"), and a nonce value in a NONCE attribute. The 2807 server also includes a SOFTWARE attribute that gives information 2808 about the server's software. 2810 The client, upon receipt of the 401 error, re-attempts the Allocate 2811 request, this time including the authentication attributes. The 2812 client selects a new transaction id, and then populates the new 2813 Allocate request with the same attributes as before. The client 2814 includes a USERNAME attribute and uses the realm value received from 2815 the server to help it determine which value to use; here the client 2816 is configured to use the username "George" for the realm 2817 "example.com". The client also includes the REALM and NONCE 2818 attributes, which are just copied from the 401 error response. 2819 Finally, the client includes a MESSAGE-INTEGRITY attribute as the 2820 last attribute in the message, whose value is a Hashed Message 2821 Authentication Code - Secure Hash Algorithm 1 (HMAC-SHA1) hash over 2822 the contents of the message (shown as just "..." above); this HMAC- 2823 SHA1 computation includes a password value. Thus, an attacker cannot 2824 compute the message integrity value without somehow knowing the 2825 secret password. 2827 The server, upon receipt of the authenticated Allocate request, 2828 checks that everything is OK, then creates an allocation. The server 2829 replies with an Allocate success response. The server includes a 2830 LIFETIME attribute giving the lifetime of the allocation; here, the 2831 server has reduced the client's requested 1-hour lifetime to just 20 2832 minutes, because this particular server doesn't allow lifetimes 2833 longer than 20 minutes. The server includes an XOR-RELAYED-ADDRESS 2834 attribute whose value is the relayed transport address of the 2835 allocation. The server includes an XOR-MAPPED-ADDRESS attribute 2836 whose value is the server-reflexive address of the client; this value 2837 is not used otherwise in TURN but is returned as a convenience to the 2838 client. The server includes a MESSAGE-INTEGRITY attribute to 2839 authenticate the response and to ensure its integrity; note that the 2840 response does not contain the USERNAME, REALM, and NONCE attributes. 2841 The server also includes a SOFTWARE attribute. 2843 TURN TURN Peer Peer 2844 client server A B 2845 |--- CreatePermission request ------>| | | 2846 | Transaction-Id=0xE5913A8F460956CA277D3319 | | 2847 | XOR-PEER-ADDRESS=192.0.2.150:0 | | | 2848 | USERNAME="George" | | | 2849 | REALM="example.com" | | | 2850 | NONCE="adl7W7PeDU4hKE72jdaQvbAMcr6h39sm" | | 2851 | MESSAGE-INTEGRITY=... | | | 2852 | | | | 2853 |<-- CreatePermission success resp.--| | | 2854 | Transaction-Id=0xE5913A8F460956CA277D3319 | | 2855 | MESSAGE-INTEGRITY=... | | | 2857 The client then creates a permission towards Peer A in preparation 2858 for sending it some application data. This is done through a 2859 CreatePermission request. The XOR-PEER-ADDRESS attribute contains 2860 the IP address for which a permission is established (the IP address 2861 of peer A); note that the port number in the attribute is ignored 2862 when used in a CreatePermission request, and here it has been set to 2863 0; also, note how the client uses Peer A's server-reflexive IP 2864 address and not its (private) host address. The client uses the same 2865 username, realm, and nonce values as in the previous request on the 2866 allocation. Though it is allowed to do so, the client has chosen not 2867 to include a SOFTWARE attribute in this request. 2869 The server receives the CreatePermission request, creates the 2870 corresponding permission, and then replies with a CreatePermission 2871 success response. Like the client, the server chooses not to include 2872 the SOFTWARE attribute in its reply. Again, note how success 2873 responses contain a MESSAGE-INTEGRITY attribute (assuming the server 2874 uses the long-term credential mechanism), but no USERNAME, REALM, and 2875 NONCE attributes. 2877 TURN TURN Peer Peer 2878 client server A B 2879 |--- Send indication --------------->| | | 2880 | Transaction-Id=0x1278E9ACA2711637EF7D3328 | | 2881 | XOR-PEER-ADDRESS=192.0.2.150:32102 | | 2882 | DONT-FRAGMENT | | | 2883 | DATA=... | | | 2884 | |-- UDP dgm ->| | 2885 | | data=... | | 2886 | | | | 2887 | |<- UDP dgm --| | 2888 | | data=... | | 2889 |<-- Data indication ----------------| | | 2890 | Transaction-Id=0x8231AE8F9242DA9FF287FEFF | | 2891 | XOR-PEER-ADDRESS=192.0.2.150:32102 | | 2892 | DATA=... | | | 2894 The client now sends application data to Peer A using a Send 2895 indication. Peer A's server-reflexive transport address is specified 2896 in the XOR-PEER-ADDRESS attribute, and the application data (shown 2897 here as just "...") is specified in the DATA attribute. The client 2898 is doing a form of path MTU discovery at the application layer and 2899 thus specifies (by including the DONT-FRAGMENT attribute) that the 2900 server should set the DF bit in the UDP datagram to send to the peer. 2901 Indications cannot be authenticated using the long-term credential 2902 mechanism of STUN, so no MESSAGE-INTEGRITY attribute is included in 2903 the message. An application wishing to ensure that its data is not 2904 altered or forged must integrity-protect its data at the application 2905 level. 2907 Upon receipt of the Send indication, the server extracts the 2908 application data and sends it in a UDP datagram to Peer A, with the 2909 relayed transport address as the source transport address of the 2910 datagram, and with the DF bit set as requested. Note that, had the 2911 client not previously established a permission for Peer A's server- 2912 reflexive IP address, then the server would have silently discarded 2913 the Send indication instead. 2915 Peer A then replies with its own UDP datagram containing application 2916 data. The datagram is sent to the relayed transport address on the 2917 server. When this arrives, the server creates a Data indication 2918 containing the source of the UDP datagram in the XOR-PEER-ADDRESS 2919 attribute, and the data from the UDP datagram in the DATA attribute. 2920 The resulting Data indication is then sent to the client. 2922 TURN TURN Peer Peer 2923 client server A B 2924 |--- ChannelBind request ----------->| | | 2925 | Transaction-Id=0x6490D3BC175AFF3D84513212 | | 2926 | CHANNEL-NUMBER=0x4000 | | | 2927 | XOR-PEER-ADDRESS=192.0.2.210:49191 | | 2928 | USERNAME="George" | | | 2929 | REALM="example.com" | | | 2930 | NONCE="adl7W7PeDU4hKE72jdaQvbAMcr6h39sm" | | 2931 | MESSAGE-INTEGRITY=... | | | 2932 | | | | 2933 |<-- ChannelBind success response ---| | | 2934 | Transaction-Id=0x6490D3BC175AFF3D84513212 | | 2935 | MESSAGE-INTEGRITY=... | | | 2937 The client now binds a channel to Peer B, specifying a free channel 2938 number (0x4000) in the CHANNEL-NUMBER attribute, and Peer B's 2939 transport address in the XOR-PEER-ADDRESS attribute. As before, the 2940 client re-uses the username, realm, and nonce from its last request 2941 in the message. 2943 Upon receipt of the request, the server binds the channel number to 2944 the peer, installs a permission for Peer B's IP address, and then 2945 replies with ChannelBind success response. 2947 TURN TURN Peer Peer 2948 client server A B 2949 |--- ChannelData ------------------->| | | 2950 | Channel-number=0x4000 |--- UDP datagram --------->| 2951 | Data=... | Data=... | 2952 | | | | 2953 | |<-- UDP datagram ----------| 2954 | | Data=... | | 2955 |<-- ChannelData --------------------| | | 2956 | Channel-number=0x4000 | | | 2957 | Data=... | | | 2959 The client now sends a ChannelData message to the server with data 2960 destined for Peer B. The ChannelData message is not a STUN message, 2961 and thus has no transaction id. Instead, it has only three fields: a 2962 channel number, data, and data length; here the channel number field 2963 is 0x4000 (the channel the client just bound to Peer B). When the 2964 server receives the ChannelData message, it checks that the channel 2965 is currently bound (which it is) and then sends the data onward to 2966 Peer B in a UDP datagram, using the relayed transport address as the 2967 source transport address and 192.0.2.210:49191 (the value of the XOR- 2968 PEER-ADDRESS attribute in the ChannelBind request) as the destination 2969 transport address. 2971 Later, Peer B sends a UDP datagram back to the relayed transport 2972 address. This causes the server to send a ChannelData message to the 2973 client containing the data from the UDP datagram. The server knows 2974 to which client to send the ChannelData message because of the 2975 relayed transport address at which the UDP datagram arrived, and 2976 knows to use channel 0x4000 because this is the channel bound to 2977 192.0.2.210:49191. Note that if there had not been any channel 2978 number bound to that address, the server would have used a Data 2979 indication instead. 2981 TURN TURN Peer Peer 2982 client server A B 2983 |--- Refresh request --------------->| | | 2984 | Transaction-Id=0x0864B3C27ADE9354B4312414 | | 2985 | SOFTWARE="Example client 1.03" | | | 2986 | USERNAME="George" | | | 2987 | REALM="example.com" | | | 2988 | NONCE="adl7W7PeDU4hKE72jdaQvbAMcr6h39sm" | | 2989 | MESSAGE-INTEGRITY=... | | | 2990 | | | | 2991 |<-- Refresh error response ---------| | | 2992 | Transaction-Id=0x0864B3C27ADE9354B4312414 | | 2993 | SOFTWARE="Example server, version 1.17" | | 2994 | ERROR-CODE=438 (Stale Nonce) | | | 2995 | REALM="example.com" | | | 2996 | NONCE="npSw1Xw239bBwGYhjNWgz2yH47sxB2j" | | 2997 | | | | 2998 |--- Refresh request --------------->| | | 2999 | Transaction-Id=0x427BD3E625A85FC731DC4191 | | 3000 | SOFTWARE="Example client 1.03" | | | 3001 | USERNAME="George" | | | 3002 | REALM="example.com" | | | 3003 | NONCE="npSw1Xw239bBwGYhjNWgz2yH47sxB2j" | | 3004 | MESSAGE-INTEGRITY=... | | | 3005 | | | | 3006 |<-- Refresh success response -------| | | 3007 | Transaction-Id=0x427BD3E625A85FC731DC4191 | | 3008 | SOFTWARE="Example server, version 1.17" | | 3009 | LIFETIME=600 (10 minutes) | | | 3011 Sometime before the 20 minute lifetime is up, the client refreshes 3012 the allocation. This is done using a Refresh request. As before, 3013 the client includes the latest username, realm, and nonce values in 3014 the request. The client also includes the SOFTWARE attribute, 3015 following the recommended practice of always including this attribute 3016 in Allocate and Refresh messages. When the server receives the 3017 Refresh request, it notices that the nonce value has expired, and so 3018 replies with 438 (Stale Nonce) error given a new nonce value. The 3019 client then reattempts the request, this time with the new nonce 3020 value. This second attempt is accepted, and the server replies with 3021 a success response. Note that the client did not include a LIFETIME 3022 attribute in the request, so the server refreshes the allocation for 3023 the default lifetime of 10 minutes (as can be seen by the LIFETIME 3024 attribute in the success response). 3026 18. Security Considerations 3028 This section considers attacks that are possible in a TURN 3029 deployment, and discusses how they are mitigated by mechanisms in the 3030 protocol or recommended practices in the implementation. 3032 Most of the attacks on TURN are mitigated by the server requiring 3033 requests be authenticated. Thus, this specification requires the use 3034 of authentication. The mandatory-to-implement mechanism is the long- 3035 term credential mechanism of STUN. Other authentication mechanisms 3036 of equal or stronger security properties may be used. However, it is 3037 important to ensure that they can be invoked in an inter-operable 3038 way. 3040 18.1. Outsider Attacks 3042 Outsider attacks are ones where the attacker has no credentials in 3043 the system, and is attempting to disrupt the service seen by the 3044 client or the server. 3046 18.1.1. Obtaining Unauthorized Allocations 3048 An attacker might wish to obtain allocations on a TURN server for any 3049 number of nefarious purposes. A TURN server provides a mechanism for 3050 sending and receiving packets while cloaking the actual IP address of 3051 the client. This makes TURN servers an attractive target for 3052 attackers who wish to use it to mask their true identity. 3054 An attacker might also wish to simply utilize the services of a TURN 3055 server without paying for them. Since TURN services require 3056 resources from the provider, it is anticipated that their usage will 3057 come with a cost. 3059 These attacks are prevented using the long-term credential mechanism, 3060 which allows the TURN server to determine the identity of the 3061 requestor and whether the requestor is allowed to obtain the 3062 allocation. 3064 18.1.2. Offline Dictionary Attacks 3066 The long-term credential mechanism used by TURN is subject to offline 3067 dictionary attacks. An attacker that is capable of eavesdropping on 3068 a message exchange between a client and server can determine the 3069 password by trying a number of candidate passwords and seeing if one 3070 of them is correct. This attack works when the passwords are low 3071 entropy, such as a word from the dictionary. This attack can be 3072 mitigated by using strong passwords with large entropy. In 3073 situations where even stronger mitigation is required, (D)TLS 3074 transport between the client and the server can be used. 3076 18.1.3. Faked Refreshes and Permissions 3078 An attacker might wish to attack an active allocation by sending it a 3079 Refresh request with an immediate expiration, in order to delete it 3080 and disrupt service to the client. This is prevented by 3081 authentication of refreshes. Similarly, an attacker wishing to send 3082 CreatePermission requests to create permissions to undesirable 3083 destinations is prevented from doing so through authentication. The 3084 motivations for such an attack are described in Section 18.2. 3086 18.1.4. Fake Data 3088 An attacker might wish to send data to the client or the peer, as if 3089 they came from the peer or client, respectively. To do that, the 3090 attacker can send the client a faked Data Indication or ChannelData 3091 message, or send the TURN server a faked Send Indication or 3092 ChannelData message. 3094 Since indications and ChannelData messages are not authenticated, 3095 this attack is not prevented by TURN. However, this attack is 3096 generally present in IP-based communications and is not substantially 3097 worsened by TURN. Consider a normal, non-TURN IP session between 3098 hosts A and B. An attacker can send packets to B as if they came 3099 from A by sending packets towards A with a spoofed IP address of B. 3100 This attack requires the attacker to know the IP addresses of A and 3101 B. With TURN, an attacker wishing to send packets towards a client 3102 using a Data indication needs to know its IP address (and port), the 3103 IP address and port of the TURN server, and the IP address and port 3104 of the peer (for inclusion in the XOR-PEER-ADDRESS attribute). To 3105 send a fake ChannelData message to a client, an attacker needs to 3106 know the IP address and port of the client, the IP address and port 3107 of the TURN server, and the channel number. This particular 3108 combination is mildly more guessable than in the non-TURN case. 3110 These attacks are more properly mitigated by application-layer 3111 authentication techniques. In the case of real-time traffic, usage 3112 of SRTP [RFC3711] prevents these attacks. 3114 In some situations, the TURN server may be situated in the network 3115 such that it is able to send to hosts to which the client cannot 3116 directly send. This can happen, for example, if the server is 3117 located behind a firewall that allows packets from outside the 3118 firewall to be delivered to the server, but not to other hosts behind 3119 the firewall. In these situations, an attacker could send the server 3120 a Send indication with an XOR-PEER-ADDRESS attribute containing the 3121 transport address of one of the other hosts behind the firewall. If 3122 the server was to allow relaying of traffic to arbitrary peers, then 3123 this would provide a way for the attacker to attack arbitrary hosts 3124 behind the firewall. 3126 To mitigate this attack, TURN requires that the client establish a 3127 permission to a host before sending it data. Thus, an attacker can 3128 only attack hosts with which the client is already communicating, 3129 unless the attacker is able to create authenticated requests. 3130 Furthermore, the server administrator may configure the server to 3131 restrict the range of IP addresses and ports to which it will relay 3132 data. To provide even greater security, the server administrator can 3133 require that the client use (D)TLS for all communication between the 3134 client and the server. 3136 18.1.5. Impersonating a Server 3138 When a client learns a relayed address from a TURN server, it uses 3139 that relayed address in application protocols to receive traffic. 3140 Therefore, an attacker wishing to intercept or redirect that traffic 3141 might try to impersonate a TURN server and provide the client with a 3142 faked relayed address. 3144 This attack is prevented through the long-term credential mechanism, 3145 which provides message integrity for responses in addition to 3146 verifying that they came from the server. Furthermore, an attacker 3147 cannot replay old server responses as the transaction id in the STUN 3148 header prevents this. Replay attacks are further thwarted through 3149 frequent changes to the nonce value. 3151 18.1.6. Eavesdropping Traffic 3153 TURN concerns itself primarily with authentication and message 3154 integrity. Confidentiality is only a secondary concern, as TURN 3155 control messages do not include information that is particularly 3156 sensitive. The primary protocol content of the messages is the IP 3157 address of the peer. If it is important to prevent an eavesdropper 3158 on a TURN connection from learning this, TURN can be run over (D)TLS. 3160 Confidentiality for the application data relayed by TURN is best 3161 provided by the application protocol itself, since running TURN over 3162 (D)TLS does not protect application data between the server and the 3163 peer. If confidentiality of application data is important, then the 3164 application should encrypt or otherwise protect its data. For 3165 example, for real-time media, confidentiality can be provided by 3166 using SRTP. 3168 18.1.7. TURN Loop Attack 3170 An attacker might attempt to cause data packets to loop indefinitely 3171 between two TURN servers. The attack goes as follows. First, the 3172 attacker sends an Allocate request to server A, using the source 3173 address of server B. Server A will send its response to server B, 3174 and for the attack to succeed, the attacker must have the ability to 3175 either view or guess the contents of this response, so that the 3176 attacker can learn the allocated relayed transport address. The 3177 attacker then sends an Allocate request to server B, using the source 3178 address of server A. Again, the attacker must be able to view or 3179 guess the contents of the response, so it can send learn the 3180 allocated relayed transport address. Using the same spoofed source 3181 address technique, the attacker then binds a channel number on server 3182 A to the relayed transport address on server B, and similarly binds 3183 the same channel number on server B to the relayed transport address 3184 on server A. Finally, the attacker sends a ChannelData message to 3185 server A. 3187 The result is a data packet that loops from the relayed transport 3188 address on server A to the relayed transport address on server B, 3189 then from server B's transport address to server A's transport 3190 address, and then around the loop again. 3192 This attack is mitigated as follows. By requiring all requests to be 3193 authenticated and/or by randomizing the port number allocated for the 3194 relayed transport address, the server forces the attacker to either 3195 intercept or view responses sent to a third party (in this case, the 3196 other server) so that the attacker can authenticate the requests and 3197 learn the relayed transport address. Without one of these two 3198 measures, an attacker can guess the contents of the responses without 3199 needing to see them, which makes the attack much easier to perform. 3200 Furthermore, by requiring authenticated requests, the server forces 3201 the attacker to have credentials acceptable to the server, which 3202 turns this from an outsider attack into an insider attack and allows 3203 the attack to be traced back to the client initiating it. 3205 The attack can be further mitigated by imposing a per-username limit 3206 on the bandwidth used to relay data by allocations owned by that 3207 username, to limit the impact of this attack on other allocations. 3208 More mitigation can be achieved by decrementing the TTL when relaying 3209 data packets (if the underlying OS allows this). 3211 18.2. Firewall Considerations 3213 A key security consideration of TURN is that TURN should not weaken 3214 the protections afforded by firewalls deployed between a client and a 3215 TURN server. It is anticipated that TURN servers will often be 3216 present on the public Internet, and clients may often be inside 3217 enterprise networks with corporate firewalls. If TURN servers 3218 provide a 'backdoor' for reaching into the enterprise, TURN will be 3219 blocked by these firewalls. 3221 TURN servers therefore emulate the behavior of NAT devices that 3222 implement address-dependent filtering [RFC4787], a property common in 3223 many firewalls as well. When a NAT or firewall implements this 3224 behavior, packets from an outside IP address are only allowed to be 3225 sent to an internal IP address and port if the internal IP address 3226 and port had recently sent a packet to that outside IP address. TURN 3227 servers introduce the concept of permissions, which provide exactly 3228 this same behavior on the TURN server. An attacker cannot send a 3229 packet to a TURN server and expect it to be relayed towards the 3230 client, unless the client has tried to contact the attacker first. 3232 It is important to note that some firewalls have policies that are 3233 even more restrictive than address-dependent filtering. Firewalls 3234 can also be configured with address- and port-dependent filtering, or 3235 can be configured to disallow inbound traffic entirely. In these 3236 cases, if a client is allowed to connect the TURN server, 3237 communications to the client will be less restrictive than what the 3238 firewall would normally allow. 3240 18.2.1. Faked Permissions 3242 In firewalls and NAT devices, permissions are granted implicitly 3243 through the traversal of a packet from the inside of the network 3244 towards the outside peer. Thus, a permission cannot, by definition, 3245 be created by any entity except one inside the firewall or NAT. With 3246 TURN, this restriction no longer holds. Since the TURN server sits 3247 outside the firewall, at attacker outside the firewall can now send a 3248 message to the TURN server and try to create a permission for itself. 3250 This attack is prevented because all messages that create permissions 3251 (i.e., ChannelBind and CreatePermission) are authenticated. 3253 18.2.2. Blacklisted IP Addresses 3255 Many firewalls can be configured with blacklists that prevent a 3256 client behind the firewall from sending packets to, or receiving 3257 packets from, ranges of blacklisted IP addresses. This is 3258 accomplished by inspecting the source and destination addresses of 3259 packets entering and exiting the firewall, respectively. 3261 This feature is also present in TURN, since TURN servers are allowed 3262 to arbitrarily restrict the range of addresses of peers that they 3263 will relay to. 3265 18.2.3. Running Servers on Well-Known Ports 3267 A malicious client behind a firewall might try to connect to a TURN 3268 server and obtain an allocation which it then uses to run a server. 3269 For example, a client might try to run a DNS server or FTP server. 3271 This is not possible in TURN. A TURN server will never accept 3272 traffic from a peer for which the client has not installed a 3273 permission. Thus, peers cannot just connect to the allocated port in 3274 order to obtain the service. 3276 18.3. Insider Attacks 3278 In insider attacks, a client has legitimate credentials but defies 3279 the trust relationship that goes with those credentials. These 3280 attacks cannot be prevented by cryptographic means but need to be 3281 considered in the design of the protocol. 3283 18.3.1. DoS against TURN Server 3285 A client wishing to disrupt service to other clients might obtain an 3286 allocation and then flood it with traffic, in an attempt to swamp the 3287 server and prevent it from servicing other legitimate clients. This 3288 is mitigated by the recommendation that the server limit the amount 3289 of bandwidth it will relay for a given username. This won't prevent 3290 a client from sending a large amount of traffic, but it allows the 3291 server to immediately discard traffic in excess. 3293 Since each allocation uses a port number on the IP address of the 3294 TURN server, the number of allocations on a server is finite. An 3295 attacker might attempt to consume all of them by requesting a large 3296 number of allocations. This is prevented by the recommendation that 3297 the server impose a limit of the number of allocations active at a 3298 time for a given username. 3300 18.3.2. Anonymous Relaying of Malicious Traffic 3302 TURN servers provide a degree of anonymization. A client can send 3303 data to peers without revealing its own IP address. TURN servers may 3304 therefore become attractive vehicles for attackers to launch attacks 3305 against targets without fear of detection. Indeed, it is possible 3306 for a client to chain together multiple TURN servers, such that any 3307 number of relays can be used before a target receives a packet. 3309 Administrators who are worried about this attack can maintain logs 3310 that capture the actual source IP and port of the client, and perhaps 3311 even every permission that client installs. This will allow for 3312 forensic tracing to determine the original source, should it be 3313 discovered that an attack is being relayed through a TURN server. 3315 18.3.3. Manipulating Other Allocations 3317 An attacker might attempt to disrupt service to other users of the 3318 TURN server by sending Refresh requests or CreatePermission requests 3319 that (through source address spoofing) appear to be coming from 3320 another user of the TURN server. TURN prevents this by requiring 3321 that the credentials used in CreatePermission, Refresh, and 3322 ChannelBind messages match those used to create the initial 3323 allocation. Thus, the fake requests from the attacker will be 3324 rejected. 3326 18.4. Tunnel Amplification Attack 3328 An attacker might attempt to cause data packets to loop numerous 3329 times between a TURN server and a tunnel between IPv4 and IPv6. The 3330 attack goes as follows. 3332 Suppose an attacker knows that a tunnel endpoint will forward 3333 encapsulated packets from a given IPv6 address (this doesn't 3334 necessarily need to be the tunnel endpoint's address). Suppose he 3335 then spoofs two packets from this address: 3337 1. An Allocate request asking for a v4 address, and 3339 2. A ChannelBind request establishing a channel to the IPv4 address 3340 of the tunnel endpoint 3342 Then he has set up an amplification attack: 3344 o The TURN relay will re-encapsulate IPv6 UDP data in v4 and send it 3345 to the tunnel endpoint 3347 o The tunnel endpoint will de-encapsulate packets from the v4 3348 interface and send them to v6 3350 So if the attacker sends a packet of the following form... 3352 IPv6: src=2001:DB9::1 dst=2001:DB8::2 3353 UDP: 3354 TURN: 3355 IPv6: src=2001:DB9::1 dst=2001:DB8::2 3356 UDP: 3357 TURN: 3358 IPv6: src=2001:DB9::1 dst=2001:DB8::2 3359 UDP: 3360 TURN: 3361 ... 3363 Then the TURN relay and the tunnel endpoint will send it back and 3364 forth until the last TURN header is consumed, at which point the TURN 3365 relay will send an empty packet, which the tunnel endpoint will drop. 3367 The amplification potential here is limited by the MTU, so it's not 3368 huge: IPv6+UDP+TURN takes 334 bytes, so a four-to-one amplification 3369 out of a 1500-byte packet is possible. But the attacker could still 3370 increase traffic volume by sending multiple packets or by 3371 establishing multiple channels spoofed from different addresses 3372 behind the same tunnel endpoint. 3374 The attack is mitigated as follows. It is RECOMMENDED that TURN 3375 relays not accept allocation or channel binding requests from 3376 addresses known to be tunneled, and that they not forward data to 3377 such addresses. In particular, a TURN relay MUST NOT accept Teredo 3378 or 6to4 addresses in these requests. 3380 18.5. Other Considerations 3382 Any relay addresses learned through an Allocate request will not 3383 operate properly with IPsec Authentication Header (AH) [RFC4302] in 3384 transport or tunnel mode. However, tunnel-mode IPsec Encapsulating 3385 Security Payload (ESP) [RFC4303] should still operate. 3387 19. IANA Considerations 3389 Since TURN is an extension to STUN [RFC5389], the methods, 3390 attributes, and error codes defined in this specification are new 3391 methods, attributes, and error codes for STUN. IANA has added these 3392 new protocol elements to the IANA registry of STUN protocol elements. 3394 The codepoints for the new STUN methods defined in this specification 3395 are listed in Section 14. 3397 The codepoints for the new STUN attributes defined in this 3398 specification are listed in Section 15. 3400 The codepoints for the new STUN error codes defined in this 3401 specification are listed in Section 16. 3403 IANA has allocated the SRV service name of "turn" for TURN over UDP 3404 or TCP, and the service name of "turns" for TURN over (D)TLS. 3406 IANA has created a registry for TURN channel numbers, initially 3407 populated as follows: 3409 o 0x0000 through 0x3FFF: Reserved and not available for use, since 3410 they conflict with the STUN header. 3412 o 0x4000 through 0x7FFF: A TURN implementation is free to use 3413 channel numbers in this range. 3415 o 0x8000 through 0xFFFF: Unassigned. 3417 Any change to this registry must be made through an IETF Standards 3418 Action. 3420 [Paragraphs in braces should be removed by the RFC Editor upon 3421 publication] 3423 [The ADDITIONAL-ADDRESS-FAMILY, ADDRESS-ERROR-CODE and ICMP 3424 attributes requires that IANA allocate a value in the "STUN 3425 attributes Registry" from the comprehension- optional range 3426 (0x8000-0xFFFF), to be replaced for TBD-CA throughout this document] 3428 [The SendErr method requires that IANA allocate a value in the "STUN 3429 Methods Registry" from the range (0x000-0x7FF), to be replaced for 3430 TBD-DA throughout this document] 3432 20. IAB Considerations 3434 The IAB has studied the problem of "Unilateral Self Address Fixing" 3435 (UNSAF), which is the general process by which a client attempts to 3436 determine its address in another realm on the other side of a NAT 3437 through a collaborative protocol-reflection mechanism [RFC3424]. The 3438 TURN extension is an example of a protocol that performs this type of 3439 function. The IAB has mandated that any protocols developed for this 3440 purpose document a specific set of considerations. These 3441 considerations and the responses for TURN are documented in this 3442 section. 3444 Consideration 1: Precise definition of a specific, limited-scope 3445 problem that is to be solved with the UNSAF proposal. A short-term 3446 fix should not be generalized to solve other problems. Such 3447 generalizations lead to the prolonged dependence on and usage of the 3448 supposed short-term fix -- meaning that it is no longer accurate to 3449 call it "short-term". 3451 Response: TURN is a protocol for communication between a relay (= 3452 TURN server) and its client. The protocol allows a client that is 3453 behind a NAT to obtain and use a public IP address on the relay. As 3454 a convenience to the client, TURN also allows the client to determine 3455 its server-reflexive transport address. 3457 Consideration 2: Description of an exit strategy/transition plan. 3458 The better short-term fixes are the ones that will naturally see less 3459 and less use as the appropriate technology is deployed. 3461 Response: TURN will no longer be needed once there are no longer any 3462 NATs. Unfortunately, as of the date of publication of this document, 3463 it no longer seems very likely that NATs will go away any time soon. 3464 However, the need for TURN will also decrease as the number of NATs 3465 with the mapping property of Endpoint-Independent Mapping [RFC4787] 3466 increases. 3468 Consideration 3: Discussion of specific issues that may render 3469 systems more "brittle". For example, approaches that involve using 3470 data at multiple network layers create more dependencies, increase 3471 debugging challenges, and make it harder to transition. 3473 Response: TURN is "brittle" in that it requires the NAT bindings 3474 between the client and the server to be maintained unchanged for the 3475 lifetime of the allocation. This is typically done using keep- 3476 alives. If this is not done, then the client will lose its 3477 allocation and can no longer exchange data with its peers. 3479 Consideration 4: Identify requirements for longer-term, sound 3480 technical solutions; contribute to the process of finding the right 3481 longer-term solution. 3483 Response: The need for TURN will be reduced once NATs implement the 3484 recommendations for NAT UDP behavior documented in [RFC4787]. 3485 Applications are also strongly urged to use ICE [RFC5245] to 3486 communicate with peers; though ICE uses TURN, it does so only as a 3487 last resort, and uses it in a controlled manner. 3489 Consideration 5: Discussion of the impact of the noted practical 3490 issues with existing deployed NATs and experience reports. 3492 Response: Some NATs deployed today exhibit a mapping behavior other 3493 than Endpoint-Independent mapping. These NATs are difficult to work 3494 with, as they make it difficult or impossible for protocols like ICE 3495 to use server-reflexive transport addresses on those NATs. A client 3496 behind such a NAT is often forced to use a relay protocol like TURN 3497 because "UDP hole punching" techniques [RFC5128] do not work. 3499 21. Changes since RFC 5766 3501 This section lists the major changes in the TURN protocol from the 3502 original [RFC5766] specification. 3504 o IPv6 support. 3506 o REQUESTED-ADDRESS-FAMILY, ADDITIONAL-ADDRESS-FAMILY, AND ADDRESS- 3507 ERRR-CODE attributes. 3509 o 440 (Address Family not Supported) and 443 (Peer Address Family 3510 Mismatch) responses. 3512 o Description of the tunnel amplification attack. 3514 o DTLS support. 3516 o More details on packet translations. 3518 o Add support for receiving ICMP packets. 3520 o Updates PMTUD. 3522 22. Acknowledgements 3524 Most of the text in this note comes from the original TURN 3525 specification, [RFC5766]. The authors would like to thank Rohan Mahy 3526 co-author of orginal TURN specification and everyone who had 3527 contributed to that document. 3529 Thanks to Justin Uberti, Pal Martinsen, Oleg Moskalenko, Aijun Wang 3530 and Simon Perreault for their help on SSODA mechanism. Authors would 3531 like to thank Gonzalo Salgueiro, Simon Perreault, Jonathan Lennox and 3532 Oleg Moskalenko for comments and review. The authors would like to 3533 thank Marc for his contributions to the text. 3535 23. References 3537 23.1. Normative References 3539 [RFC5389] Rosenberg, J., Mahy, R., Matthews, P., and D. Wing, 3540 "Session Traversal Utilities for NAT (STUN)", RFC 5389, 3541 DOI 10.17487/RFC5389, October 2008, 3542 . 3544 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 3545 Requirement Levels", BCP 14, RFC 2119, 3546 DOI 10.17487/RFC2119, March 1997, 3547 . 3549 [RFC2474] Nichols, K., Blake, S., Baker, F., and D. Black, 3550 "Definition of the Differentiated Services Field (DS 3551 Field) in the IPv4 and IPv6 Headers", RFC 2474, 3552 DOI 10.17487/RFC2474, December 1998, 3553 . 3555 [RFC3168] Ramakrishnan, K., Floyd, S., and D. Black, "The Addition 3556 of Explicit Congestion Notification (ECN) to IP", 3557 RFC 3168, DOI 10.17487/RFC3168, September 2001, 3558 . 3560 [RFC1122] Braden, R., Ed., "Requirements for Internet Hosts - 3561 Communication Layers", STD 3, RFC 1122, 3562 DOI 10.17487/RFC1122, October 1989, 3563 . 3565 [RFC6145] Li, X., Bao, C., and F. Baker, "IP/ICMP Translation 3566 Algorithm", RFC 6145, DOI 10.17487/RFC6145, April 2011, 3567 . 3569 [RFC3697] Rajahalme, J., Conta, A., Carpenter, B., and S. Deering, 3570 "IPv6 Flow Label Specification", RFC 3697, 3571 DOI 10.17487/RFC3697, March 2004, 3572 . 3574 [RFC7065] Petit-Huguenin, M., Nandakumar, S., Salgueiro, G., and P. 3575 Jones, "Traversal Using Relays around NAT (TURN) Uniform 3576 Resource Identifiers", RFC 7065, DOI 10.17487/RFC7065, 3577 November 2013, . 3579 [RFC0792] Postel, J., "Internet Control Message Protocol", STD 5, 3580 RFC 792, DOI 10.17487/RFC0792, September 1981, 3581 . 3583 [RFC4443] Conta, A., Deering, S., and M. Gupta, Ed., "Internet 3584 Control Message Protocol (ICMPv6) for the Internet 3585 Protocol Version 6 (IPv6) Specification", RFC 4443, 3586 DOI 10.17487/RFC4443, March 2006, 3587 . 3589 [I-D.ietf-tram-stunbis] 3590 Petit-Huguenin, M., Salgueiro, G., Rosenberg, J., Wing, 3591 D., Mahy, R., and P. Matthews, "Session Traversal 3592 Utilities for NAT (STUN)", draft-ietf-tram-stunbis-08 3593 (work in progress), June 2016. 3595 23.2. Informative References 3597 [RFC1191] Mogul, J. and S. Deering, "Path MTU discovery", RFC 1191, 3598 DOI 10.17487/RFC1191, November 1990, 3599 . 3601 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, 3602 DOI 10.17487/RFC0791, September 1981, 3603 . 3605 [RFC1918] Rekhter, Y., Moskowitz, B., Karrenberg, D., de Groot, G., 3606 and E. Lear, "Address Allocation for Private Internets", 3607 BCP 5, RFC 1918, DOI 10.17487/RFC1918, February 1996, 3608 . 3610 [RFC3424] Daigle, L., Ed. and IAB, "IAB Considerations for 3611 UNilateral Self-Address Fixing (UNSAF) Across Network 3612 Address Translation", RFC 3424, DOI 10.17487/RFC3424, 3613 November 2002, . 3615 [RFC4787] Audet, F., Ed. and C. Jennings, "Network Address 3616 Translation (NAT) Behavioral Requirements for Unicast 3617 UDP", BCP 127, RFC 4787, DOI 10.17487/RFC4787, January 3618 2007, . 3620 [RFC5245] Rosenberg, J., "Interactive Connectivity Establishment 3621 (ICE): A Protocol for Network Address Translator (NAT) 3622 Traversal for Offer/Answer Protocols", RFC 5245, 3623 DOI 10.17487/RFC5245, April 2010, 3624 . 3626 [RFC6062] Perreault, S., Ed. and J. Rosenberg, "Traversal Using 3627 Relays around NAT (TURN) Extensions for TCP Allocations", 3628 RFC 6062, DOI 10.17487/RFC6062, November 2010, 3629 . 3631 [RFC6156] Camarillo, G., Novo, O., and S. Perreault, Ed., "Traversal 3632 Using Relays around NAT (TURN) Extension for IPv6", 3633 RFC 6156, DOI 10.17487/RFC6156, April 2011, 3634 . 3636 [RFC6056] Larsen, M. and F. Gont, "Recommendations for Transport- 3637 Protocol Port Randomization", BCP 156, RFC 6056, 3638 DOI 10.17487/RFC6056, January 2011, 3639 . 3641 [RFC5128] Srisuresh, P., Ford, B., and D. Kegel, "State of Peer-to- 3642 Peer (P2P) Communication across Network Address 3643 Translators (NATs)", RFC 5128, DOI 10.17487/RFC5128, March 3644 2008, . 3646 [RFC1928] Leech, M., Ganis, M., Lee, Y., Kuris, R., Koblas, D., and 3647 L. Jones, "SOCKS Protocol Version 5", RFC 1928, 3648 DOI 10.17487/RFC1928, March 1996, 3649 . 3651 [RFC3550] Schulzrinne, H., Casner, S., Frederick, R., and V. 3652 Jacobson, "RTP: A Transport Protocol for Real-Time 3653 Applications", STD 64, RFC 3550, DOI 10.17487/RFC3550, 3654 July 2003, . 3656 [RFC3711] Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K. 3657 Norrman, "The Secure Real-time Transport Protocol (SRTP)", 3658 RFC 3711, DOI 10.17487/RFC3711, March 2004, 3659 . 3661 [RFC4302] Kent, S., "IP Authentication Header", RFC 4302, 3662 DOI 10.17487/RFC4302, December 2005, 3663 . 3665 [RFC4303] Kent, S., "IP Encapsulating Security Payload (ESP)", 3666 RFC 4303, DOI 10.17487/RFC4303, December 2005, 3667 . 3669 [RFC4821] Mathis, M. and J. Heffner, "Packetization Layer Path MTU 3670 Discovery", RFC 4821, DOI 10.17487/RFC4821, March 2007, 3671 . 3673 [RFC3261] Rosenberg, J., Schulzrinne, H., Camarillo, G., Johnston, 3674 A., Peterson, J., Sparks, R., Handley, M., and E. 3675 Schooler, "SIP: Session Initiation Protocol", RFC 3261, 3676 DOI 10.17487/RFC3261, June 2002, 3677 . 3679 [I-D.rosenberg-mmusic-ice-nonsip] 3680 Rosenberg, J., "Guidelines for Usage of Interactive 3681 Connectivity Establishment (ICE) by non Session Initiation 3682 Protocol (SIP) Protocols", draft-rosenberg-mmusic-ice- 3683 nonsip-01 (work in progress), July 2008. 3685 [I-D.ietf-tram-stun-pmtud] 3686 Petit-Huguenin, M. and G. Salgueiro, "Path MTU Discovery 3687 Using Session Traversal Utilities for NAT (STUN)", draft- 3688 ietf-tram-stun-pmtud-03 (work in progress), October 2016. 3690 [I-D.ietf-tram-turn-server-discovery] 3691 Patil, P., Reddy, T., and D. Wing, "TURN Server Auto 3692 Discovery", draft-ietf-tram-turn-server-discovery-10 (work 3693 in progress), October 2016. 3695 [RFC4086] Eastlake 3rd, D., Schiller, J., and S. Crocker, 3696 "Randomness Requirements for Security", BCP 106, RFC 4086, 3697 DOI 10.17487/RFC4086, June 2005, 3698 . 3700 [RFC5766] Mahy, R., Matthews, P., and J. Rosenberg, "Traversal Using 3701 Relays around NAT (TURN): Relay Extensions to Session 3702 Traversal Utilities for NAT (STUN)", RFC 5766, 3703 DOI 10.17487/RFC5766, April 2010, 3704 . 3706 [RFC5928] Petit-Huguenin, M., "Traversal Using Relays around NAT 3707 (TURN) Resolution Mechanism", RFC 5928, 3708 DOI 10.17487/RFC5928, August 2010, 3709 . 3711 [Port-Numbers] 3712 "IANA Port Numbers Registry", 2005, 3713 . 3715 [Frag-Harmful] 3716 "Fragmentation Considered Harmful", . 3719 [Protocol-Numbers] 3720 "IANA Protocol Numbers Registry", 2005, 3721 . 3723 Authors' Addresses 3725 Tirumaleswar Reddy (editor) 3726 Cisco Systems, Inc. 3727 Cessna Business Park, Varthur Hobl 3728 Sarjapur Marathalli Outer Ring Road 3729 Bangalore, Karnataka 560103 3730 India 3732 Email: tireddy@cisco.com 3734 Alan Johnston (editor) 3735 Unaffiliated 3736 Bellevue, WA 3737 USA 3739 Email: alan.b.johnston@gmail.com 3741 Philip Matthews 3742 Alcatel-Lucent 3743 600 March Road 3744 Ottawa, Ontario 3745 Canada 3747 Email: philip_matthews@magma.ca 3749 Jonathan Rosenberg 3750 jdrosen.net 3751 Edison, NJ 3752 USA 3754 Email: jdrosen@jdrosen.net 3755 URI: http://www.jdrosen.net