[眩晕系列 01] 眩晕协议原文-rfc5389 原文

原文地址：//tools.ietf.org/html/rfc5389  后续提供rfc5389协议中文版

中文翻译版：https://www.cnblogs.com/a4234613/p/15533464.html

内容如下：

---

 

Network Working Group J. Rosenberg
Request for Comments: 5389 Cisco
Obsoletes: 3489 R. Mahy
Category: Standards Track P. Matthew Unaffiliate D. Win Cisc October 2008

Session Traversal Utilities for NAT (STUN)

Status of This Memo

This document specifies an Internet standards track protocol for the Internet community, and requests discussion and suggestions for improvements. Please refer to the current edition of the "Internet Official Protocol Standards" (STD 1) for the standardization state and status of this protocol. Distribution of this memo is unlimited.

Abstract

Session Traversal Utilities for NAT (STUN) is a protocol that serves as a tool for other protocols in dealing with Network Address Translator (NAT) traversal. It can be used by an endpoint to determine the IP address and port allocated to it by a NAT. It can also be used to check connectivity between two endpoints, and as a keep-alive protocol to maintain NAT bindings. STUN works with many existing NATs, and does not require any special behavior from them.

STUN is not a NAT traversal solution by itself. Rather, it is a tool to be used in the context of a NAT traversal solution. This is an important change from the previous version of this specification (RFC 3489), which presented STUN as a complete solution.

This document obsoletes RFC 3489.

Table of Contents

1. Introduction
2. Evolution from RFC 3489
3. Overview of Operation
4. Terminology
5. Definitions
6. STUN Message Structure
7. Base Protocol Procedures
7.1. Forming a Request or an Indication
7.2. Sending the Request or Indication
7.2.1. Sending over UDP
7.2.2. Sending over TCP or TLS-over-TCP
7.3. Receiving a STUN Message
7.3.1. Processing a Request
7.3.1.1. Forming a Success or Error Response
7.3.1.2. Sending the Success or Error Response
7.3.2. Processing an Indication
7.3.3. Processing a Success Response
7.3.4. Processing an Error Response
8. FINGERPRINT Mechanism
9. DNS Discovery of a Server
10. Authentication and Message-Integrity Mechanisms
10.1. Short-Term Credential Mechanism
10.1.1. Forming a Request or Indication
10.1.2. Receiving a Request or Indication
10.1.3. Receiving a Response
10.2. Long-Term Credential Mechanism
10.2.1. Forming a Request
10.2.1.1. First Request
10.2.1.2. Subsequent Requests
10.2.2. Receiving a Request
10.2.3. Receiving a Response
11. ALTERNATE-SERVER Mechanism
12. Backwards Compatibility with RFC 3489
12.1. Changes to Client Processing
12.2. Changes to Server Processing
13. Basic Server Behavior
14. STUN Usages
15. STUN Attributes
15.1. MAPPED-ADDRESS
15.2. XOR-MAPPED-ADDRESS
15.3. USERNAME
15.4. MESSAGE-INTEGRITY
15.5. FINGERPRINT
15.6. ERROR-CODE
15.7. REALM
15.8. NONCE
15.9. UNKNOWN-ATTRIBUTES
15.10. SOFTWARE
15.11. ALTERNATE-SERVER
16. Security Considerations
16.1. Attacks against the Protocol
16.1.1. Outside Attacks
16.1.2. Inside Attacks
16.2. Attacks Affecting the Usage
16.2.1. Attack I: Distributed DoS (DDoS) against a
Target
16.2.2. Attack II: Silencing a Client
16.2.3. Attack III: Assuming the Identity of a Client
16.2.4. Attack IV: Eavesdropping
16.3. Hash Agility Plan
17. IAB Considerations
18. IANA Considerations
18.1. STUN Methods Registry
18.2. STUN Attribute Registry
18.3. STUN Error Code Registry
18.4. STUN UDP and TCP Port Numbers
19. Changes since RFC 3489
20. Contributors
21. Acknowledgements
22. References
22.1. Normative References
22.2. Informative References
Appendix A. C Snippet to Determine STUN Message Types

1. Introduction

The protocol defined in this specification, Session Traversal Utilities for NAT, provides a tool for dealing with NATs. It provides a means for an endpoint to determine the IP address and port allocated by a NAT that corresponds to its private IP address and port. It also provides a way for an endpoint to keep a NAT binding alive. With some extensions, the protocol can be used to do connectivity checks between two endpoints [MMUSIC-ICE], or to relay packets between two endpoints [BEHAVE-TURN].

In keeping with its tool nature, this specification defines an extensible packet format, defines operation over several transport protocols, and provides for two forms of authentication.

STUN is intended to be used in context of one or more NAT traversal solutions. These solutions are known as STUN usages. Each usage describes how STUN is utilized to achieve the NAT traversal solution. Typically, a usage indicates when STUN messages get sent, which optional attributes to include, what server is used, and what authentication mechanism is to be used. Interactive Connectivity Establishment (ICE) [MMUSIC-ICE] is one usage of STUN. SIP Outbound [SIP-OUTBOUND] is another usage of STUN. In some cases, a usage will require extensions to STUN. A STUN extension can be in the form of new methods, attributes, or error response codes. More information on STUN usages can be found in Section 14.

2. Evolution from RFC 3489

STUN was originally defined in RFC 3489 [RFC3489]. That specification, sometimes referred to as "classic STUN", represented itself as a complete solution to the NAT traversal problem. In that solution, a client would discover whether it was behind a NAT, determine its NAT type, discover its IP address and port on the public side of the outermost NAT, and then utilize that IP address and port within the body of protocols, such as the Session Initiation Protocol (SIP) [RFC3261]. However, experience since the publication of RFC 3489 has found that classic STUN simply does not work sufficiently well to be a deployable solution. The address and port learned through classic STUN are sometimes usable for communications with a peer, and sometimes not. Classic STUN provided no way to discover whether it would, in fact, work or not, and it provided no remedy in cases where it did not. Furthermore, classic STUN's algorithm for classification of NAT types was found to be faulty, as many NATs did not fit cleanly into the types defined there.

Classic STUN also had a security vulnerability -- attackers could provide the client with incorrect mapped addresses under certain topologies and constraints, and this was fundamentally not solvable through any cryptographic means. Though this problem remains with this specification, those attacks are now mitigated through the use of more complete solutions that make use of STUN.

For these reasons, this specification obsoletes RFC 3489, and instead describes STUN as a tool that is utilized as part of a complete NAT traversal solution. ICE [MMUSIC-ICE] is a complete NAT traversal solution for protocols based on the offer/answer [RFC3264] methodology, such as SIP. SIP Outbound [SIP-OUTBOUND] is a complete solution for traversal of SIP signaling, and it uses STUN in a very different way. Though it is possible that a protocol may be able to use STUN by itself (classic STUN) as a traversal solution, such usage is not described here and is strongly discouraged for the reasons described above.

The on-the-wire protocol described here is changed only slightly from classic STUN. The protocol now runs over TCP in addition to UDP. Extensibility was added to the protocol in a more structured way. A magic cookie mechanism for demultiplexing STUN with application protocols was added by stealing 32 bits from the 128-bit transaction ID defined in RFC 3489, allowing the change to be backwards compatible. Mapped addresses are encoded using a new exclusive-or format. There are other, more minor changes. See Section 19 for a more complete listing.

Due to the change in scope, STUN has also been renamed from "Simple Traversal of UDP through NAT" to "Session Traversal Utilities for NAT". The acronym remains STUN, which is all anyone ever remembers anyway.

3. Overview of Operation

This section is descriptive only.

 

 

 

                        Figure 1: One Possible STUN Configuration

One possible STUN configuration is shown in Figure 1. In this configuration, there are two entities (called STUN agents) that implement the STUN protocol. The lower agent in the figure is the client, and is connected to private network 1. This network connects to private network 2 through NAT 1. Private network 2 connects to the public Internet through NAT 2. The upper agent in the figure is the server, and resides on the public Internet.

STUN is a client-server protocol. It supports two types of transactions. One is a request/response transaction in which a client sends a request to a server, and the server returns a response. The second is an indication transaction in which either agent -- client or server -- sends an indication that generates no response. Both types of transactions include a transaction ID, which is a randomly selected 96-bit number. For request/response transactions, this transaction ID allows the client to associate the response with the request that generated it; for indications, the transaction ID serves as a debugging aid.

All STUN messages start with a fixed header that includes a method, a class, and the transaction ID. The method indicates which of the various requests or indications this is; this specification defines just one method, Binding, but other methods are expected to be defined in other documents. The class indicates whether this is a request, a success response, an error response, or an indication. Following the fixed header comes zero or more attributes, which are Type-Length-Value extensions that convey additional information for the specific message.

This document defines a single method called Binding. The Binding method can be used either in request/response transactions or in indication transactions. When used in request/response transactions, the Binding method can be used to determine the particular "binding" a NAT has allocated to a STUN client. When used in either request/ response or in indication transactions, the Binding method can also be used to keep these "bindings" alive.

In the Binding request/response transaction, a Binding request is sent from a STUN client to a STUN server. When the Binding request arrives at the STUN server, it may have passed through one or more NATs between the STUN client and the STUN server (in Figure 1, there were two such NATs). As the Binding request message passes through a NAT, the NAT will modify the source transport address (that is, the source IP address and the source port) of the packet. As a result, the source transport address of the request received by the server will be the public IP address and port created by the NAT closest to the server. This is called a reflexive transport address. The STUN server copies that source transport address into an XOR-MAPPED- ADDRESS attribute in the STUN Binding response and sends the Binding response back to the STUN client. As this packet passes back through a NAT, the NAT will modify the destination transport address in the IP header, but the transport address in the XOR-MAPPED-ADDRESS attribute within the body of the STUN response will remain untouched. In this way, the client can learn its reflexive transport address allocated by the outermost NAT with respect to the STUN server.

In some usages, STUN must be multiplexed with other protocols (e.g., [MMUSIC-ICE], [SIP-OUTBOUND]). In these usages, there must be a way to inspect a packet and determine if it is a STUN packet or not. STUN provides three fields in the STUN header with fixed values that can be used for this purpose. If this is not sufficient, then STUN packets can also contain a FINGERPRINT value, which can further be used to distinguish the packets.

STUN defines a set of optional procedures that a usage can decide to use, called mechanisms. These mechanisms include DNS discovery, a redirection technique to an alternate server, a fingerprint attribute for demultiplexing, and two authentication and message-integrity exchanges. The authentication mechanisms revolve around the use of a username, password, and message-integrity value. Two authentication mechanisms, the long-term credential mechanism and the short-term credential mechanism, are defined in this specification. Each usage specifies the mechanisms allowed with that usage.

In the long-term credential mechanism, the client and server share a pre-provisioned username and password and perform a digest challenge/ response exchange inspired by (but differing in details) to the one defined for HTTP [RFC2617]. In the short-term credential mechanism, the client and the server exchange a username and password through some out-of-band method prior to the STUN exchange. For example, in the ICE usage [MMUSIC-ICE] the two endpoints use out-of-band signaling to exchange a username and password. These are used to integrity protect and authenticate the request and response. There is no challenge or nonce used.

4. Terminology

In this document, the key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" are to be interpreted as described in BCP 14, RFC 2119 [RFC2119] and indicate requirement levels for compliant STUN implementations.

5. Definitions

STUN Agent: A STUN agent is an entity that implements the STUN protocol. The entity can be either a STUN client or a STUN server.

STUN Client: A STUN client is an entity that sends STUN requests and receives STUN responses. A STUN client can also send indications. In this specification, the terms STUN client and client are synonymous.

STUN Server: A STUN server is an entity that receives STUN requests and sends STUN responses. A STUN server can also send indications. In this specification, the terms STUN server and server are synonymous.

Transport Address: The combination of an IP address and port number (such as a UDP or TCP port number).

Reflexive Transport Address: A transport address learned by a client that identifies that client as seen by another host on an IP network, typically a STUN server. When there is an intervening NAT between the client and the other host, the reflexive transport address represents the mapped address allocated to the client on the public side of the NAT. Reflexive transport addresses are learned from the mapped address attribute (MAPPED-ADDRESS or XOR- MAPPED-ADDRESS) in STUN responses.

Mapped Address: Same meaning as reflexive address. This term is retained only for historic reasons and due to the naming of the MAPPED-ADDRESS and XOR-MAPPED-ADDRESS attributes.

Long-Term Credential: A username and associated password that represent a shared secret between client and server. Long-term credentials are generally granted to the client when a subscriber enrolls in a service and persist until the subscriber leaves the service or explicitly changes the credential.

Long-Term Password: The password from a long-term credential.

Short-Term Credential: A temporary username and associated password that represent a shared secret between client and server. Short- term credentials are obtained through some kind of protocol mechanism between the client and server, preceding the STUN exchange. A short-term credential has an explicit temporal scope, which may be based on a specific amount of time (such as 5 minutes) or on an event (such as termination of a SIP dialog). The specific scope of a short-term credential is defined by the application usage.

Short-Term Password: The password component of a short-term credential.

STUN Indication: A STUN message that does not receive a response.

Attribute: The STUN term for a Type-Length-Value (TLV) object that can be added to a STUN message. Attributes are divided into two types: comprehension-required and comprehension-optional. STUN agents can safely ignore comprehension-optional attributes they don't understand, but cannot successfully process a message if it contains comprehension-required attributes that are not understood.

RTO: Retransmission TimeOut, which defines the initial period of time between transmission of a request and the first retransmit of that request.

6. STUN Message Structure

STUN messages are encoded in binary using network-oriented format (most significant byte or octet first, also commonly known as big- endian). The transmission order is described in detail in Appendix B of RFC 791 [RFC0791]. Unless otherwise noted, numeric constants are in decimal (base 10).

All STUN messages MUST start with a 20-byte header followed by zero or more Attributes. The STUN header contains a STUN message type, magic cookie, transaction ID, and message length.

 

                          Figure 2: Format of STUN Message Header

The most significant 2 bits of every STUN message MUST be zeroes. This can be used to differentiate STUN packets from other protocols when STUN is multiplexed with other protocols on the same port.

The message type defines the message class (request, success response, failure response, or indication) and the message method (the primary function) of the STUN message. Although there are four message classes, there are only two types of transactions in STUN: request/response transactions (which consist of a request message and a response message) and indication transactions (which consist of a single indication message). Response classes are split into error and success responses to aid in quickly processing the STUN message.

The message type field is decomposed further into the followin structure:

 

 Figure 3: Format of STUN Message Type Field

Here the bits in the message type field are shown as most significant (M11) through least significant (M0). M11 through M0 represent a 12- bit encoding of the method. C1 and C0 represent a 2-bit encoding of the class. A class of 0b00 is a request, a class of 0b01 is an indication, a class of 0b10 is a success response, and a class of 0b11 is an error response. This specification defines a single method, Binding. The method and class are orthogonal, so that for each method, a request, success response, error response, and indication are possible for that method. Extensions defining new methods MUST indicate which classes are permitted for that method.

For example, a Binding request has class="0b00" (request) and method=0b000000000001 (Binding) and is encoded into the first 16 bits as 0x0001. A Binding response has class="0b10" (success response) and method=0b000000000001, and is encoded into the first 16 bits as 0x0101.

Note: This unfortunate encoding is due to assignment of values in [RFC3489] that did not consider encoding Indications, Success, and Errors using bit fields.

The magic cookie field MUST contain the fixed value 0x2112A442 in network byte order. In RFC 3489 [RFC3489], this field was part of the transaction ID; placing the magic cookie in this location allows a server to detect if the client will understand certain attributes that were added in this revised specification. In addition, it aids in distinguishing STUN packets from packets of other protocols when STUN is multiplexed with those other protocols on the same port.

The transaction ID is a 96-bit identifier, used to uniquely identify STUN transactions. For request/response transactions, the transaction ID is chosen by the STUN client for the request and echoed by the server in the response. For indications, it is chosen by the agent sending the indication. It primarily serves to correlate requests with responses, though it also plays a small role

in helping to prevent certain types of attacks. The server also uses the transaction ID as a key to identify each transaction uniquely across all clients. As such, the transaction ID MUST be uniformly and randomly chosen from the interval 0 .. 2**96-1, and SHOULD be cryptographically random. Resends of the same request reuse the same transaction ID, but the client MUST choose a new transaction ID for new transactions unless the new request is bit-wise identical to the previous request and sent from the same transport address to the same IP address. Success and error responses MUST carry the same transaction ID as their corresponding request. When an agent is acting as a STUN server and STUN client on the same port, the transaction IDs in requests sent by the agent have no relationship to the transaction IDs in requests received by the agent.

The message length MUST contain the size, in bytes, of the message not including the 20-byte STUN header. Since all STUN attributes are padded to a multiple of 4 bytes, the last 2 bits of this field are always zero. This provides another way to distinguish STUN packets from packets of other protocols.

Following the STUN fixed portion of the header are zero or mor attributes. Each attribute is TLV (Type-Length-Value) encoded. Th details of the encoding, and of the attributes themselves are give in Section 15.

7. Base Protocol Procedures

This section defines the base procedures of the STUN protocol. I describes how messages are formed, how they are sent, and how the are processed when they are received. It also defines the detaile processing of the Binding method. Other sections in this documen describe optional procedures that a usage may elect to use in certai situations. Other documents may define other extensions to STUN, b adding new methods, new attributes, or new error response codes.

7.1. Forming a Request or an Indication

When formulating a request or indication message, the agent MUST follow the rules in Section 6 when creating the header. In addition,the message class MUST be either "Request" or "Indication" (a appropriate), and the method must be either Binding or some metho defined in another document.

The agent then adds any attributes specified by the method or th usage. For example, some usages may specify that the agent use a authentication method (Section 10) or the FINGERPRINT attribut (Section 8).

If the agent is sending a request, it SHOULD add a SOFTWARE attribut to the request. Agents MAY include a SOFTWARE attribute i indications, depending on the method. Extensions to STUN shoul discuss whether SOFTWARE is useful in new indications.

For the Binding method with no authentication, no attributes ar required unless the usage specifies otherwise.

All STUN messages sent over UDP SHOULD be less than the path MTU, i known. If the path MTU is unknown, messages SHOULD be the smaller o 576 bytes and the first-hop MTU for IPv4 [RFC1122] and 1280 bytes fo IPv6 [RFC2460]. This value corresponds to the overall size of the I packet. Consequently, for IPv4, the actual STUN message would nee to be less than 548 bytes (576 minus 20-byte IP header, minus 8-byt UDP header, assuming no IP options are used). STUN provides n ability to handle the case where the request is under the MTU but th response would be larger than the MTU. It is not envisioned tha this limitation will be an issue for STUN. The MTU limitation is SHOULD, and not a MUST, to account for cases where STUN itself i being used to probe for MTU characteristics [BEHAVE-NAT]. Outside o this or similar applications, the MTU constraint MUST be followed.

7.2. Sending the Request or Indication

The agent then sends the request or indication. This documen specifies how to send STUN messages over UDP, TCP, or TLS-over-TCP;
other transport protocols may be added in the future. The STUN usag must specify which transport protocol is used, and how the agen determines the IP address and port of the recipient. Section describes a DNS-based method of determining the IP address and por of a server that a usage may elect to use. STUN may be used wit anycast addresses, but only with UDP and in usages wher authentication is not used.

At any time, a client MAY have multiple outstanding STUN request with the same STUN server (that is, multiple transactions i progress, with different transaction IDs). Absent other limits t the rate of new transactions (such as those specified by ICE fo connectivity checks or when STUN is run over TCP), a client SHOUL space new transactions to a server by RTO and SHOULD limit itself t ten outstanding transactions to the same server.

7.2.1. Sending over UDP

When running STUN over UDP, it is possible that the STUN messag might be dropped by the network. Reliability of STUN request/
response transactions is accomplished through retransmissions of the

request message by the client application itself. STUN indication are not retransmitted; thus, indication transactions over UDP are no reliable.

A client SHOULD retransmit a STUN request message starting with a interval of RTO ("Retransmission TimeOut"), doubling after eac retransmission. The RTO is an estimate of the round-trip time (RTT),and is computed as described in RFC 2988 [RFC2988], with tw exceptions. First, the initial value for RTO SHOULD be configurabl (rather than the 3 s recommended in RFC 2988) and SHOULD be greate than 500 ms. The exception cases for this "SHOULD" are when othe mechanisms are used to derive congestion thresholds (such as the one defined in ICE for fixed rate streams), or when STUN is used in non-Internet environments with known network capacities. In fixed-lin access links, a value of 500 ms is RECOMMENDED. Second, the value o RTO SHOULD NOT be rounded up to the nearest second. Rather, a 1 m accuracy SHOULD be maintained. As with TCP, the usage of Karn' algorithm is RECOMMENDED [KARN87]. When applied to STUN, it mean that RTT estimates SHOULD NOT be computed from STUN transactions tha result in the retransmission of a request.

The value for RTO SHOULD be cached by a client after the completio of the transaction, and used as the starting value for RTO for th next transaction to the same server (based on equality of I address). The value SHOULD be considered stale and discarded afte 10 minutes.

Retransmissions continue until a response is received, or until total of Rc requests have been sent. Rc SHOULD be configurable an SHOULD have a default of 7. If, after the last request, a duratio equal to Rm times the RTO has passed without a response (providin ample time to get a response if only this final request actuall succeeds), the client SHOULD consider the transaction to have failed.Rm SHOULD be configurable and SHOULD have a default of 16. A STU transaction over UDP is also considered failed if there has been hard ICMP error [RFC1122]. For example, assuming an RTO of 500 ms,requests would be sent at times 0 ms, 500 ms, 1500 ms, 3500 ms, 750 ms, 15500 ms, and 31500 ms. If the client has not received response after 39500 ms, the client will consider the transaction t have timed out.

7.2.2. Sending over TCP or TLS-over-TCP

For TCP and TLS-over-TCP, the client opens a TCP connection to th server.

In some usages of STUN, STUN is sent as the only protocol over th TCP connection. In this case, it can be sent without the aid of an additional framing or demultiplexing. In other usages, or with othe extensions, it may be multiplexed with other data over a TC connection. In that case, STUN MUST be run on top of some kind o framing protocol, specified by the usage or extension, which allow for the agent to extract complete STUN messages and complet application layer messages. The STUN service running on the well-known port or ports discovered through the DNS procedures i Section 9 is for STUN alone, and not for STUN multiplexed with othe data. Consequently, no framing protocols are used in connections t those servers. When additional framing is utilized, the usage wil specify how the client knows to apply it and what port to connect to.For example, in the case of ICE connectivity checks, this informatio is learned through out-of-band negotiation between client and server.

When STUN is run by itself over TLS-over-TCP, th TLS_RSA_WITH_AES_128_CBC_SHA ciphersuite MUST be implemented at minimum. Implementations MAY also support any other ciphersuite.When it receives the TLS Certificate message, the client SHOUL verify the certificate and inspect the site identified by th certificate. If the certificate is invalid or revoked, or if it doe not identify the appropriate party, the client MUST NOT send the STU message or otherwise proceed with the STUN transaction. The clien MUST verify the identity of the server. To do that, it follows th identification procedures defined in Section 3.1 of RFC 281 [RFC2818]. Those procedures assume the client is dereferencing URI. For purposes of usage with this specification, the clien treats the domain name or IP address used in Section 8.1 as the hos portion of the URI that has been dereferenced. Alternatively, client MAY be configured with a set of domains or IP addresses tha are trusted; if a certificate is received that identifies one o those domains or IP addresses, the client considers the identity o the server to be verified.

When STUN is run multiplexed with other protocols over a TLS-over-TC connection, the mandatory ciphersuites and TLS handling procedure operate as defined by those protocols.

Reliability of STUN over TCP and TLS-over-TCP is handled by TC itself, and there are no retransmissions at the STUN protocol level.However, for a request/response transaction, if the client has no received a response by Ti seconds after it sent the SYN to establis the connection, it considers the transaction to have timed out. T SHOULD be configurable and SHOULD have a default of 39.5s. Thi value has been chosen to equalize the TCP and UDP timeouts for th default initial RTO.

In addition, if the client is unable to establish the TCP connection,or the TCP connection is reset or fails before a response i received, any request/response transaction in progress is considere to have failed.

The client MAY send multiple transactions over a single TCP (or TLS-over-TCP) connection, and it MAY send another request befor receiving a response to the previous. The client SHOULD keep th connection open until it:

o has no further STUN requests or indications to send over tha connection, and

o has no plans to use any resources (such as a mapped addres (MAPPED-ADDRESS or XOR-MAPPED-ADDRESS) or relayed addres [BEHAVE-TURN]) that were learned though STUN requests sent ove that connection, and

o if multiplexing other application protocols over that port, ha finished using that other application, and

o if using that learned port with a remote peer, has establishe communications with that remote peer, as is required by some TC NAT traversal techniques (e.g., [MMUSIC-ICE-TCP]).

At the server end, the server SHOULD keep the connection open, an let the client close it, unless the server has determined that th connection has timed out (for example, due to the clien disconnecting from the network). Bindings learned by the client wil remain valid in intervening NATs only while the connection remain open. Only the client knows how long it needs the binding. Th server SHOULD NOT close a connection if a request was received ove that connection for which a response was not sent. A server MUST NO ever open a connection back towards the client in order to send response. Servers SHOULD follow best practices regarding connectio management in cases of overload.

7.3. Receiving a STUN Message

This section specifies the processing of a STUN message. Th processing specified here is for STUN messages as defined in thi specification; additional rules for backwards compatibility ar defined in Section 12. Those additional procedures are optional, an usages can elect to utilize them. First, a set of processin operations is applied that is independent of the class. This i followed by class-specific processing, described in the subsection that follow.

When a STUN agent receives a STUN message, it first checks that th message obeys the rules of Section 6. It checks that the first tw bits are 0, that the magic cookie field has the correct value, tha the message length is sensible, and that the method value is supported method. It checks that the message class is allowed fo the particular method. If the message class is "Success Response" o "Error Response", the agent checks that the transaction ID matches transaction that is still in progress. If the FINGERPRINT extensio is being used, the agent checks that the FINGERPRINT attribute i present and contains the correct value. If any errors are detected,the message is silently discarded. In the case when STUN is bein multiplexed with another protocol, an error may indicate that this i not really a STUN message; in this case, the agent should try t parse the message as a different protocol.

The STUN agent then does any checks that are required by authentication mechanism that the usage has specified (se Section 10).

Once the authentication checks are done, the STUN agent checks fo unknown attributes and known-but-unexpected attributes in th message. Unknown comprehension-optional attributes MUST be ignore by the agent. Known-but-unexpected attributes SHOULD be ignored b the agent. Unknown comprehension-required attributes caus processing that depends on the message class and is described below.

At this point, further processing depends on the message class of th request.

7.3.1. Processing a Request

If the request contains one or more unknown comprehension-require attributes, the server replies with an error response with an erro code of 420 (Unknown Attribute), and includes an UNKNOWN-ATTRIBUTE attribute in the response that lists the unknown comprehension-required attributes.

The server then does any additional checking that the method or th specific usage requires. If all the checks succeed, the serve formulates a success response as described below.

When run over UDP, a request received by the server could be th first request of a transaction, or a retransmission. The server MUS respond to retransmissions such that the following property i preserved: if the client receives the response to the retransmissio and not the response that was sent to the original request, th overall state on the client and server is identical to the case wher only the response to the original retransmission is received, or

where both responses are received (in which case the client will us the first). The easiest way to meet this requirement is for th server to remember all transaction IDs received over UDP and thei corresponding responses in the last 40 seconds. However, thi requires the server to hold state, and will be inappropriate for an requests which are not authenticated. Another way is to reproces the request and recompute the response. The latter technique MUS only be applied to requests that are idempotent (a request i considered idempotent when the same request can be safely repeate without impacting the overall state of the system) and result in th same success response for the same request. The Binding method i considered to be idempotent. Note that there are certain rar network events that could cause the reflexive transport address valu to change, resulting in a different mapped address in differen success responses. Extensions to STUN MUST discuss the implication of request retransmissions on servers that do not store transactio state.

7.3.1.1. Forming a Success or Error Response

When forming the response (success or error), the server follows th rules of Section 6. The method of the response is the same as tha of the request, and the message class is either "Success Response" o "Error Response".

For an error response, the server MUST add an ERROR-CODE attribut containing the error code specified in the processing above. Th reason phrase is not fixed, but SHOULD be something suitable for th error code. For certain errors, additional attributes are added t the message. These attributes are spelled out in the descriptio where the error code is specified. For example, for an error code o 420 (Unknown Attribute), the server MUST include an UNKNOWN-ATTRIBUTES attribute. Certain authentication errors also caus attributes to be added (see Section 10). Extensions may define othe errors and/or additional attributes to add in error cases.

If the server authenticated the request using an authenticatio mechanism, then the server SHOULD add the appropriate authenticatio attributes to the response (see Section 10).

The server also adds any attributes required by the specific metho or usage. In addition, the server SHOULD add a SOFTWARE attribute t the message.

For the Binding method, no additional checking is required unless th usage specifies otherwise. When forming the success response, th server adds a XOR-MAPPED-ADDRESS attribute to the response, where th contents of the attribute are the source transport address of the

request message. For UDP, this is the source IP address and sourc UDP port of the request message. For TCP and TLS-over-TCP, this i the source IP address and source TCP port of the TCP connection a seen by the server.

7.3.1.2. Sending the Success or Error Response

The response (success or error) is sent over the same transport a the request was received on. If the request was received over UDP,the destination IP address and port of the response are the source I address and port of the received request message, and the source I address and port of the response are equal to the destination I address and port of the received request message. If the request wa received over TCP or TLS-over-TCP, the response is sent back on th same TCP connection as the request was received on.

7.3.2. Processing an Indication

If the indication contains unknown comprehension-required attributes,the indication is discarded and processing ceases.

The agent then does any additional checking that the method or th specific usage requires. If all the checks succeed, the agent the processes the indication. No response is generated for a indication.

For the Binding method, no additional checking or processing i required, unless the usage specifies otherwise. The mere receipt o the message by the agent has refreshed the "bindings" in th intervening NATs.

Since indications are not re-transmitted over UDP (unlike requests),there is no need to handle re-transmissions of indications at th sending agent.

7.3.3. Processing a Success Response

If the success response contains unknown comprehension-require attributes, the response is discarded and the transaction i considered to have failed.

The client then does any additional checking that the method or th specific usage requires. If all the checks succeed, the client the processes the success response.

For the Binding method, the client checks that the XOR-MAPPED-ADDRES attribute is present in the response. The client checks the addres family specified. If it is an unsupported address family, the

attribute SHOULD be ignored. If it is an unexpected but supporte address family (for example, the Binding transaction was sent ove IPv4, but the address family specified is IPv6), then the client MA accept and use the value.

7.3.4. Processing an Error Response

If the error response contains unknown comprehension-require attributes, or if the error response does not contain an ERROR-COD attribute, then the transaction is simply considered to have failed.

The client then does any processing specified by the authenticatio mechanism (see Section 10). This may result in a new transactio attempt.

The processing at this point depends on the error code, the method,and the usage; the following are the default rules:

o If the error code is 300 through 399, the client SHOULD conside the transaction as failed unless the ALTERNATE-SERVER extension i being used. See Section 11.

o If the error code is 400 through 499, the client declares th transaction failed; in the case of 420 (Unknown Attribute), th response should contain a UNKNOWN-ATTRIBUTES attribute that give additional information.

o If the error code is 500 through 599, the client MAY resend th request; clients that do so MUST limit the number of times they d this.

Any other error code causes the client to consider the transactio failed.

8. FINGERPRINT Mechanism

This section describes an optional mechanism for STUN that aids i distinguishing STUN messages from packets of other protocols when th two are multiplexed on the same transport address. This mechanism i optional, and a STUN usage must describe if and when it is used. Th FINGERPRINT mechanism is not backwards compatible with RFC 3489, an cannot be used in environments where such compatibility is required.

In some usages, STUN messages are multiplexed on the same transpor address as other protocols, such as the Real Time Transport Protoco (RTP). In order to apply the processing described in Section 7, STU messages must first be separated from the application packets.

Section 6 describes three fixed fields in the STUN header that can b used for this purpose. However, in some cases, these three fixe fields may not be sufficient.

When the FINGERPRINT extension is used, an agent includes th FINGERPRINT attribute in messages it sends to another agent.Section 15.5 describes the placement and value of this attribute.When the agent receives what it believes is a STUN message, then, i addition to other basic checks, the agent also checks that th message contains a FINGERPRINT attribute and that the attribut contains the correct value. Section 7.3 describes when in th overall processing of a STUN message the FINGERPRINT check i performed. This additional check helps the agent detect messages o other protocols that might otherwise seem to be STUN messages.

9. DNS Discovery of a Server

This section describes an optional procedure for STUN that allows client to use DNS to determine the IP address and port of a server.A STUN usage must describe if and when this extension is used. T use this procedure, the client must know a server's domain name and service name; the usage must also describe how the client obtain these. Hard-coding the domain name of the server into software i NOT RECOMMENDED in case the domain name is lost or needs to chang for legal or other reasons.

When a client wishes to locate a STUN server in the public Interne that accepts Binding request/response transactions, the SRV servic name is "stun". When it wishes to locate a STUN server that accept Binding request/response transactions over a TLS session, the SR service name is "stuns". STUN usages MAY define additional DNS SR service names.

The domain name is resolved to a transport address using the SR procedures specified in [RFC2782]. The DNS SRV service name is th service name provided as input to this procedure. The protocol i the SRV lookup is the transport protocol the client will run STU over: "udp" for UDP and "tcp" for TCP. Note that only "tcp" i defined with "stuns" at this time.

The procedures of RFC 2782 are followed to determine the server t contact. RFC 2782 spells out the details of how a set of SRV record is sorted and then tried. However, RFC 2782 only states that th client should "try to connect to the (protocol, address, service)"
without giving any details on what happens in the event of failure.When following these procedures, if the STUN transaction times ou without receipt of a response, the client SHOULD retry the request to

the next server in the ordered defined by RFC 2782. Such a retry i only possible for request/response transmissions, since indicatio transactions generate no response or timeout.

The default port for STUN requests is 3478, for both TCP and UDP.

Administrators of STUN servers SHOULD use this port in their SR records for UDP and TCP. In all cases, the port in DNS MUST reflec the one on which the server is listening. The default port for STU over TLS is 5349. Servers can run STUN over TLS on the same port a STUN over TCP if the server software supports determining whether th initial message is a TLS or STUN message.

If no SRV records were found, the client performs an A or AAAA recor lookup of the domain name. The result will be a list of I addresses, each of which can be contacted at the default port usin UDP or TCP, independent of the STUN usage. For usages that requir TLS, the client connects to one of the IP addresses using the defaul STUN over TLS port.

10. Authentication and Message-Integrity Mechanisms

This section defines two mechanisms for STUN that a client and serve can use to provide authentication and message integrity; these tw mechanisms are known as the short-term credential mechanism and th long-term credential mechanism. These two mechanisms are optional,and each usage must specify if and when these mechanisms are used.Consequently, both clients and servers will know which mechanism (i any) to follow based on knowledge of which usage applies. Fo example, a STUN server on the public Internet supporting ICE woul have no authentication, whereas the STUN server functionality in a agent supporting connectivity checks would utilize short-ter credentials. An overview of these two mechanisms is given i Section 3.

Each mechanism specifies the additional processing required to us that mechanism, extending the processing specified in Section 7. Th additional processing occurs in three different places: when formin a message, when receiving a message immediately after the basi checks have been performed, and when doing the detailed processing o error responses.

10.1. Short-Term Credential Mechanism

The short-term credential mechanism assumes that, prior to the STU transaction, the client and server have used some other protocol t exchange a credential in the form of a username and password. Thi credential is time-limited. The time limit is defined by the usage.

As an example, in the ICE usage [MMUSIC-ICE], the two endpoints us out-of-band signaling to agree on a username and password, and thi username and password are applicable for the duration of the medi session.

This credential is used to form a message-integrity check in eac request and in many responses. There is no challenge and response a in the long-term mechanism; consequently, replay is prevented b virtue of the time-limited nature of the credential.

10.1.1. Forming a Request or Indication

For a request or indication message, the agent MUST include th USERNAME and MESSAGE-INTEGRITY attributes in the message. The HMA for the MESSAGE-INTEGRITY attribute is computed as described i Section 15.4. Note that the password is never included in th request or indication.

10.1.2. Receiving a Request or Indication

After the agent has done the basic processing of a message, the agen performs the checks listed below in order specified:

o If the message does not contain both a MESSAGE-INTEGRITY and USERNAME attribute:

* If the message is a request, the server MUST reject the reques with an error response. This response MUST use an error cod of 400 (Bad Request).

* If the message is an indication, the agent MUST silentl discard the indication.

o If the USERNAME does not contain a username value currently vali within the server:

* If the message is a request, the server MUST reject the reques with an error response. This response MUST use an error cod of 401 (Unauthorized).

* If the message is an indication, the agent MUST silentl discard the indication.

o Using the password associated with the username, compute the valu for the message integrity as described in Section 15.4. If th resulting value does not match the contents of the MESSAGE-INTEGRITY attribute:

* If the message is a request, the server MUST reject the reques with an error response. This response MUST use an error cod of 401 (Unauthorized).

* If the message is an indication, the agent MUST silentl discard the indication.

If these checks pass, the agent continues to process the request o indication. Any response generated by a server MUST include th MESSAGE-INTEGRITY attribute, computed using the password utilized t authenticate the request. The response MUST NOT contain the USERNAM attribute.

If any of the checks fail, a server MUST NOT include a MESSAGE-INTEGRITY or USERNAME attribute in the error response. This i because, in these failure cases, the server cannot determine th shared secret necessary to compute MESSAGE-INTEGRITY.

10.1.3. Receiving a Response

The client looks for the MESSAGE-INTEGRITY attribute in the response.If present, the client computes the message integrity over th response as defined in Section 15.4, using the same password i utilized for the request. If the resulting value matches th contents of the MESSAGE-INTEGRITY attribute, the response i considered authenticated. If the value does not match, or i MESSAGE-INTEGRITY was absent, the response MUST be discarded, as i it was never received. This means that retransmits, if applicable,will continue.

10.2. Long-Term Credential Mechanism

The long-term credential mechanism relies on a long-term credential,in the form of a username and password that are shared between clien and server. The credential is considered long-term since it i assumed that it is provisioned for a user, and remains in effec until the user is no longer a subscriber of the system, or i changed. This is basically a traditional "log-in" username an password given to users.

Because these usernames and passwords are expected to be valid fo extended periods of time, replay prevention is provided in the for of a digest challenge. In this mechanism, the client initially send a request, without offering any credentials or any integrity checks.The server rejects this request, providing the user a realm (used t guide the user or agent in selection of a username and password) an a nonce. The nonce provides the replay protection. It is a cookie,selected by the server, and encoded in such a way as to indicate a

duration of validity or client identity from which it is valid. Th client retries the request, this time including its username and th realm, and echoing the nonce provided by the server. The client als includes a message-integrity, which provides an HMAC over the entir request, including the nonce. The server validates the nonce an checks the message integrity. If they match, the request i authenticated. If the nonce is no longer valid, it is considere "stale", and the server rejects the request, providing a new nonce.

In subsequent requests to the same server, the client reuses th nonce, username, realm, and password it used previously. In thi way, subsequent requests are not rejected until the nonce become invalid by the server, in which case the rejection provides a ne nonce to the client.

Note that the long-term credential mechanism cannot be used t protect indications, since indications cannot be challenged. Usage utilizing indications must either use a short-term credential or omi authentication and message integrity for them.

Since the long-term credential mechanism is susceptible to offlin dictionary attacks, deployments SHOULD utilize passwords that ar difficult to guess. In cases where the credentials are not entere by the user, but are rather placed on a client device during devic provisioning, the password SHOULD have at least 128 bits o randomness. In cases where the credentials are entered by the user,they should follow best current practices around password structure.

10.2.1. Forming a Request

There are two cases when forming a request. In the first case, thi is the first request from the client to the server (as identified b its IP address and port). In the second case, the client i submitting a subsequent request once a previous request/respons transaction has completed successfully. Forming a request as consequence of a 401 or 438 error response is covered i Section 10.2.3 and is not considered a "subsequent request" and thu does not utilize the rules described in Section 10.2.1.2.

10.2.1.1. First Request

If the client has not completed a successful request/respons transaction with the server (as identified by hostname, if the DN procedures of Section 9 are used, else IP address if not), it SHOUL omit the USERNAME, MESSAGE-INTEGRITY, REALM, and NONCE attributes.In other words, the very first request is sent as if there were n authentication or message integrity applied.

10.2.1.2. Subsequent Requests

Once a request/response transaction has completed successfully, th client will have been presented a realm and nonce by the server, an selected a username and password with which it authenticated. Th client SHOULD cache the username, password, realm, and nonce fo subsequent communications with the server. When the client sends subsequent request, it SHOULD include the USERNAME, REALM, and NONC attributes with these cached values. It SHOULD include a MESSAGE-INTEGRITY attribute, computed as described in Section 15.4 using th cached password.

10.2.2. Receiving a Request

After the server has done the basic processing of a request, i performs the checks listed below in the order specified:

o If the message does not contain a MESSAGE-INTEGRITY attribute, th server MUST generate an error response with an error code of 40 (Unauthorized). This response MUST include a REALM value. It i RECOMMENDED that the REALM value be the domain name of th provider of the STUN server. The response MUST include a NONCE,selected by the server. The response SHOULD NOT contain USERNAME or MESSAGE-INTEGRITY attribute.

o If the message contains a MESSAGE-INTEGRITY attribute, but i missing the USERNAME, REALM, or NONCE attribute, the server MUS generate an error response with an error code of 400 (Ba Request). This response SHOULD NOT include a USERNAME, NONCE,REALM, or MESSAGE-INTEGRITY attribute.

o If the NONCE is no longer valid, the server MUST generate an erro response with an error code of 438 (Stale Nonce). This respons MUST include NONCE and REALM attributes and SHOULD NOT include th USERNAME or MESSAGE-INTEGRITY attribute. Servers can invalidat nonces in order to provide additional security. See Section 4. of [RFC2617] for guidelines.

o If the username in the USERNAME attribute is not valid, the serve MUST generate an error response with an error code of 40 (Unauthorized). This response MUST include a REALM value. It i RECOMMENDED that the REALM value be the domain name of th provider of the STUN server. The response MUST include a NONCE,selected by the server. The response SHOULD NOT contain USERNAME or MESSAGE-INTEGRITY attribute.

o Using the password associated with the username in the USERNAM attribute, compute the value for the message integrity a described in Section 15.4. If the resulting value does not matc the contents of the MESSAGE-INTEGRITY attribute, the server MUS reject the request with an error response. This response MUST us an error code of 401 (Unauthorized). It MUST include REALM an NONCE attributes and SHOULD NOT include the USERNAME or MESSAGE-INTEGRITY attribute.

If these checks pass, the server continues to process the request.Any response generated by the server (excepting the cases describe above) MUST include the MESSAGE-INTEGRITY attribute, computed usin the username and password utilized to authenticate the request. Th REALM, NONCE, and USERNAME attributes SHOULD NOT be included.

10.2.3. Receiving a Response

If the response is an error response with an error code of 40 (Unauthorized), the client SHOULD retry the request with a ne transaction. This request MUST contain a USERNAME, determined by th client as the appropriate username for the REALM from the erro response. The request MUST contain the REALM, copied from the erro response. The request MUST contain the NONCE, copied from the erro response. The request MUST contain the MESSAGE-INTEGRITY attribute,computed using the password associated with the username in th USERNAME attribute. The client MUST NOT perform this retry if it i not changing the USERNAME or REALM or its associated password, fro the previous attempt.

If the response is an error response with an error code of 438 (Stal Nonce), the client MUST retry the request, using the new NONC supplied in the 438 (Stale Nonce) response. This retry MUST als include the USERNAME, REALM, and MESSAGE-INTEGRITY.

The client looks for the MESSAGE-INTEGRITY attribute in the respons (either success or failure). If present, the client computes th message integrity over the response as defined in Section 15.4, usin the same password it utilized for the request. If the resultin value matches the contents of the MESSAGE-INTEGRITY attribute, th response is considered authenticated. If the value does not match,or if MESSAGE-INTEGRITY was absent, the response MUST be discarded,as if it was never received. This means that retransmits, i applicable, will continue.

11. ALTERNATE-SERVER Mechanism

This section describes a mechanism in STUN that allows a server t redirect a client to another server. This extension is optional, an a usage must define if and when this extension is used.

A server using this extension redirects a client to another server b replying to a request message with an error response message with a error code of 300 (Try Alternate). The server MUST include a ALTERNATE-SERVER attribute in the error response. The error respons message MAY be authenticated; however, there are uses cases fo ALTERNATE-SERVER where authentication of the response is not possibl or practical.

A client using this extension handles a 300 (Try Alternate) erro code as follows. The client looks for an ALTERNATE-SERVER attribut in the error response. If one is found, then the client consider the current transaction as failed, and reattempts the request wit the server specified in the attribute, using the same transpor protocol used for the previous request. That request, i authenticated, MUST utilize the same credentials that the clien would have used in the request to the server that performed th redirection. If the client has been redirected to a server on whic it has already tried this request within the last five minutes, i MUST ignore the redirection and consider the transaction to hav failed. This prevents infinite ping-ponging between servers in cas of redirection loops.

12. Backwards Compatibility with RFC 3489

This section defines procedures that allow a degree of backward compatibility with the original protocol defined in RFC 348 [RFC3489]. This mechanism is optional, meant to be utilized only i cases where a new client can connect to an old server, or vice versa.A usage must define if and when this procedure is used.

Section 19 lists all the changes between this specification and RF 3489 [RFC3489]. However, not all of these differences are important,because "classic STUN" was only used in a few specific ways. For th purposes of this extension, the important changes are the following.In RFC 3489:

o UDP was the only supported transport.

o The field that is now the magic cookie field was a part of th transaction ID field, and transaction IDs were 128 bits long.

o The XOR-MAPPED-ADDRESS attribute did not exist, and the Bindin method used the MAPPED-ADDRESS attribute instead.

o There were three comprehension-required attributes, RESPONSE-ADDRESS, CHANGE-REQUEST, and CHANGED-ADDRESS, that have bee removed from this specification.

* CHANGE-REQUEST and CHANGED-ADDRESS are now part of the NA Behavior Discovery usage [BEHAVE-NAT], and the other i deprecated.

12.1. Changes to Client Processing

A client that wants to interoperate with an [RFC3489] server SHOUL send a request message that uses the Binding method, contains n attributes, and uses UDP as the transport protocol to the server. I successful, the success response received from the server wil contain a MAPPED-ADDRESS attribute rather than an XOR-MAPPED-ADDRES attribute. A client seeking to interoperate with an older serve MUST be prepared to receive either. Furthermore, the client MUS ignore any Reserved comprehension-required attributes that migh appear in the response. Of the Reserved attributes in Section 18.2,0x0002, 0x0004, 0x0005, and 0x000B may appear in Binding response from a server compliant to RFC 3489. Other than this change, th processing of the response is identical to the procedures describe above.

12.2. Changes to Server Processing

A STUN server can detect when a given Binding request message wa sent from an RFC 3489 [RFC3489] client by the absence of the correc value in the magic cookie field. When the server detects an RFC 348 client, it SHOULD copy the value seen in the magic cookie field i the Binding request to the magic cookie field in the Binding respons message, and insert a MAPPED-ADDRESS attribute instead of an XOR-MAPPED-ADDRESS attribute.

The client might, in rare situations, include either the RESPONSE-ADDRESS or CHANGE-REQUEST attributes. In these situations, th server will view these as unknown comprehension-required attribute and reply with an error response. Since the mechanisms utilizin those attributes are no longer supported, this behavior i acceptable.

The RFC 3489 version of STUN lacks both the magic cookie and th FINGERPRINT attribute that allows for a very high probability o correctly identifying STUN messages when multiplexed with othe protocols. Therefore, STUN implementations that are backwards

compatible with RFC 3489 SHOULD NOT be used in cases where STUN wil be multiplexed with another protocol. However, that should not be a issue as such multiplexing was not available in RFC 3489.

13. Basic Server Behavior

This section defines the behavior of a basic, stand-alone STU server. A basic STUN server provides clients with server reflexiv transport addresses by receiving and replying to STUN Bindin reques