idnits 2.17.1 draft-ietf-sasl-scram-08.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** The document seems to lack a License Notice according IETF Trust Provisions of 28 Dec 2009, Section 6.b.i or Provisions of 12 Sep 2009 Section 6.b -- however, there's a paragraph with a matching beginning. Boilerplate error? (You're using the IETF Trust Provisions' Section 6.b License Notice from 12 Feb 2009 rather than one of the newer Notices. See https://trustee.ietf.org/license-info/.) Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- == There are 1 instance of lines with non-RFC2606-compliant FQDNs in the document. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (October 2, 2009) is 5319 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: 'I-D.ietf-sasl-gs2' is mentioned on line 1166, but not defined == Missing Reference: 'RFC2743' is mentioned on line 1171, but not defined == Missing Reference: 'RFC4121' is mentioned on line 1180, but not defined == Missing Reference: 'RFC3962' is mentioned on line 1177, but not defined == Missing Reference: 'RFC3961' is mentioned on line 1174, but not defined == Missing Reference: 'RFC4401' is mentioned on line 1185, but not defined == Missing Reference: 'RFC4402' is mentioned on line 1189, but not defined ** Obsolete undefined reference: RFC 4402 (Obsoleted by RFC 7802) == Missing Reference: 'RFCXXXX' is mentioned on line 976, but not defined ** Downref: Normative reference to an Informational RFC: RFC 2104 ** Downref: Normative reference to an Informational RFC: RFC 3174 ** Obsolete normative reference: RFC 3454 (Obsoleted by RFC 7564) ** Obsolete normative reference: RFC 4013 (Obsoleted by RFC 7613) == Outdated reference: A later version (-10) exists of draft-altman-tls-channel-bindings-06 == Outdated reference: A later version (-04) exists of draft-melnikov-sasl-scram-ldap-02 -- Obsolete informational reference (is this intentional?): RFC 2898 (Obsoleted by RFC 8018) -- Obsolete informational reference (is this intentional?): RFC 5226 (Obsoleted by RFC 8126) -- Obsolete informational reference (is this intentional?): RFC 5246 (Obsoleted by RFC 8446) Summary: 6 errors (**), 0 flaws (~~), 12 warnings (==), 4 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 NETWORK WORKING GROUP A. Menon-Sen 3 Internet-Draft Oryx Mail Systems GmbH 4 Intended status: Standards Track A. Melnikov 5 Expires: April 5, 2010 Isode Ltd 6 C. Newman 7 N. Williams 8 Sun Microsystems 9 October 2, 2009 11 Salted Challenge Response (SCRAM) SASL and GSS-API Mechanism 12 draft-ietf-sasl-scram-08.txt 14 Status of this Memo 16 This Internet-Draft is submitted to IETF in full conformance with the 17 provisions of BCP 78 and BCP 79. 19 Internet-Drafts are working documents of the Internet Engineering 20 Task Force (IETF), its areas, and its working groups. Note that 21 other groups may also distribute working documents as Internet- 22 Drafts. 24 Internet-Drafts are draft documents valid for a maximum of six months 25 and may be updated, replaced, or obsoleted by other documents at any 26 time. It is inappropriate to use Internet-Drafts as reference 27 material or to cite them other than as "work in progress." 29 The list of current Internet-Drafts can be accessed at 30 http://www.ietf.org/ietf/1id-abstracts.txt. 32 The list of Internet-Draft Shadow Directories can be accessed at 33 http://www.ietf.org/shadow.html. 35 This Internet-Draft will expire on April 5, 2010. 37 Copyright Notice 39 Copyright (c) 2009 IETF Trust and the persons identified as the 40 document authors. All rights reserved. 42 This document is subject to BCP 78 and the IETF Trust's Legal 43 Provisions Relating to IETF Documents in effect on the date of 44 publication of this document (http://trustee.ietf.org/license-info). 45 Please review these documents carefully, as they describe your rights 46 and restrictions with respect to this document. 48 Abstract 50 The secure authentication mechanism most widely deployed and used by 51 Internet application protocols is the transmission of clear-text 52 passwords over a channel protected by Transport Layer Security (TLS). 53 There are some significant security concerns with that mechanism, 54 which could be addressed by the use of a challenge response 55 authentication mechanism protected by TLS. Unfortunately, the 56 challenge response mechanisms presently on the standards track all 57 fail to meet requirements necessary for widespread deployment, and 58 have had success only in limited use. 60 This specification describes a family of Simple Authentication and 61 Security Layer (SASL, RFC 4422) authentication mechanisms called the 62 Salted Challenge Response Authentication Mechanism (SCRAM), which 63 addresses the security concerns and meets the deployability 64 requirements. When used in combination with TLS or an equivalent 65 security layer, a mechanism from this family could improve the 66 status-quo for application protocol authentication and provide a 67 suitable choice for a mandatory-to-implement mechanism for future 68 application protocol standards. 70 Table of Contents 72 1. Conventions Used in This Document . . . . . . . . . . 4 73 1.1. Terminology . . . . . . . . . . . . . . . . . . . . . 4 74 1.2. Notation . . . . . . . . . . . . . . . . . . . . . . . 5 75 2. Introduction . . . . . . . . . . . . . . . . . . . . . 7 76 3. SCRAM Algorithm Overview . . . . . . . . . . . . . . . 9 77 4. SCRAM Mechanism Names . . . . . . . . . . . . . . . . 11 78 5. SCRAM Authentication Exchange . . . . . . . . . . . . 12 79 5.1. SCRAM Attributes . . . . . . . . . . . . . . . . . . . 13 80 5.2. Compliance with SASL mechanism requirements . . . . . 16 81 6. Channel Binding . . . . . . . . . . . . . . . . . . . 17 82 6.1. Default Channel Binding . . . . . . . . . . . . . . . 18 83 7. Formal Syntax . . . . . . . . . . . . . . . . . . . . 19 84 8. SCRAM as a GSS-API Mechanism . . . . . . . . . . . . . 22 85 8.1. GSS-API Principal Name Types for SCRAM . . . . . . . . 22 86 8.2. GSS-API Per-Message Tokens for SCRAM . . . . . . . . . 22 87 8.3. GSS_Pseudo_random() for SCRAM . . . . . . . . . . . . 23 88 9. Security Considerations . . . . . . . . . . . . . . . 24 89 10. IANA Considerations . . . . . . . . . . . . . . . . . 26 90 11. Acknowledgements . . . . . . . . . . . . . . . . . . . 28 91 Appendix A. Other Authentication Mechanisms . . . . . . . . . . . 29 92 Appendix B. Design Motivations . . . . . . . . . . . . . . . . . . 30 93 Appendix C. Internet-Draft Change History . . . . . . . . . . . . 31 94 12. References . . . . . . . . . . . . . . . . . . . . . . 33 95 12.1. Normative References . . . . . . . . . . . . . . . . . 33 96 12.2. Normative References for GSS-API implementors . . . . 33 97 12.3. Informative References . . . . . . . . . . . . . . . . 34 98 Authors' Addresses . . . . . . . . . . . . . . . . . . 36 100 1. Conventions Used in This Document 102 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 103 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 104 document are to be interpreted as described in [RFC2119]. 106 Formal syntax is defined by [RFC5234] including the core rules 107 defined in Appendix B of [RFC5234]. 109 Example lines prefaced by "C:" are sent by the client and ones 110 prefaced by "S:" by the server. If a single "C:" or "S:" label 111 applies to multiple lines, then the line breaks between those lines 112 are for editorial clarity only, and are not part of the actual 113 protocol exchange. 115 1.1. Terminology 117 This document uses several terms defined in [RFC4949] ("Internet 118 Security Glossary") including the following: authentication, 119 authentication exchange, authentication information, brute force, 120 challenge-response, cryptographic hash function, dictionary attack, 121 eavesdropping, hash result, keyed hash, man-in-the-middle, nonce, 122 one-way encryption function, password, replay attack and salt. 123 Readers not familiar with these terms should use that glossary as a 124 reference. 126 Some clarifications and additional definitions follow: 128 o Authentication information: Information used to verify an identity 129 claimed by a SCRAM client. The authentication information for a 130 SCRAM identity consists of salt, iteration count, the "StoredKey" 131 and "ServerKey" (as defined in the algorithm overview) for each 132 supported cryptographic hash function. 134 o Authentication database: The database used to look up the 135 authentication information associated with a particular identity. 136 For application protocols, LDAPv3 (see [RFC4510]) is frequently 137 used as the authentication database. For network-level protocols 138 such as PPP or 802.11x, the use of RADIUS [RFC2865] is more 139 common. 141 o Base64: An encoding mechanism defined in [RFC4648] which converts 142 an octet string input to a textual output string which can be 143 easily displayed to a human. The use of base64 in SCRAM is 144 restricted to the canonical form with no whitespace. 146 o Octet: An 8-bit byte. 148 o Octet string: A sequence of 8-bit bytes. 150 o Salt: A random octet string that is combined with a password 151 before applying a one-way encryption function. This value is used 152 to protect passwords that are stored in an authentication 153 database. 155 1.2. Notation 157 The pseudocode description of the algorithm uses the following 158 notations: 160 o ":=": The variable on the left hand side represents the octet 161 string resulting from the expression on the right hand side. 163 o "+": Octet string concatenation. 165 o "[ ]": A portion of an expression enclosed in "[" and "]" may not 166 be included in the result under some circumstances. See the 167 associated text for a description of those circumstances. 169 o Normalize(str): Apply a Unicode normalization algorithm to a UTF-8 170 [RFC3629] encoded "str". The resulting string is also in UTF-8. 171 Implementations SHOULD use the SASLPrep profile [RFC4013] of the 172 "stringprep" algorithm [RFC3454] as the normalization algorithm. 174 o HMAC(key, str): Apply the HMAC keyed hash algorithm (defined in 175 [RFC2104]) using the octet string represented by "key" as the key 176 and the octet string "str" as the input string. The size of the 177 result is the hash result size for the hash function in use. For 178 example, it is 20 octets for SHA-1 (see [RFC3174]). 180 o H(str): Apply the cryptographic hash function to the octet string 181 "str", producing an octet string as a result. The size of the 182 result depends on the hash result size for the hash function in 183 use. 185 o XOR: Apply the exclusive-or operation to combine the octet string 186 on the left of this operator with the octet string on the right of 187 this operator. The length of the output and each of the two 188 inputs will be the same for this use. 190 o Hi(str, salt): 192 U0 := HMAC(str, salt + INT(1)) 193 U1 := HMAC(str, U0) 194 U2 := HMAC(str, U1) 195 ... 196 Ui-1 := HMAC(str, Ui-2) 197 Ui := HMAC(str, Ui-1) 199 Hi := U0 XOR U1 XOR U2 XOR ... XOR Ui 201 where "i" is the iteration count, "+" is the string concatenation 202 operator and INT(g) is a four-octet encoding of the integer g, 203 most significant octet first. 205 Hi() is, essentially, PBKDF2 [RFC2898] with HMAC() as the PRF and 206 with dkLen == output length of HMAC() == output length of H(). 208 2. Introduction 210 This specification describes a family of authentication mechanisms 211 called the Salted Challenge Response Authentication Mechanism (SCRAM) 212 which addresses the requirements necessary to deploy a challenge- 213 response mechanism more widely than past attempts (see Appendix A and 214 Appendix B). When used in combination with Transport Layer Security 215 (TLS, see [RFC5246]) or an equivalent security layer, a mechanism 216 from this family could improve the status-quo for application 217 protocol authentication and provide a suitable choice for a 218 mandatory-to-implement mechanism for future application protocol 219 standards. 221 For simplicity, this family of mechanisms does not presently include 222 negotiation of a security layer [RFC4422]. It is intended to be used 223 with an external security layer such as that provided by TLS or SSH, 224 with optional channel binding [RFC5056] to the external security 225 layer. 227 SCRAM is specified herein as a pure Simple Authentication and 228 Security Layer (SASL) [RFC4422] mechanism, but it conforms to the new 229 bridge between SASL and the Generic Security Services Application 230 Programming Interface (GSS-API) called "GS2" [I-D.ietf-sasl-gs2]. 231 This means that this document defines both, a SASL mechanism and a 232 GSS-API mechanism. 234 SCRAM provides the following protocol features: 236 o The authentication information stored in the authentication 237 database is not sufficient by itself to impersonate the client. 238 The information is salted to prevent a pre-stored dictionary 239 attack if the database is stolen. 241 o The server does not gain the ability to impersonate the client to 242 other servers (with an exception for server-authorized proxies). 244 o The mechanism permits the use of a server-authorized proxy without 245 requiring that proxy to have super-user rights with the back-end 246 server. 248 o Mutual authentication is supported, but only the client is named 249 (i.e., the server has no name). 251 o When used as a SASL mechanism, SCRAM is capable of transporting 252 authorization identities (see [RFC4422], Section 2) from the 253 client to the server. 255 A separate document defines a standard LDAPv3 [RFC4510] attribute 256 that enables storage of the SCRAM authentication information in LDAP. 257 See [I-D.melnikov-sasl-scram-ldap]. 259 For an in-depth discussion of why other challenge response mechanisms 260 are not considered sufficient, see appendix A. For more information 261 about the motivations behind the design of this mechanism, see 262 appendix B. 264 3. SCRAM Algorithm Overview 266 The following is a description of a full, uncompressed SASL SCRAM 267 authentication exchange. Nothing in SCRAM prevents either sending 268 the client-first message with the SASL authentication request defined 269 by an application protocol ("initial client response"), nor sending 270 the server-final message as additional data of the SASL outcome of 271 authentication exchange defined by an application protocol. See 272 [RFC4422] for more details. 274 Note that this section omits some details, such as client and server 275 nonces. See Section 5 for more details. 277 To begin with, the SCRAM client is in possession of a username and 278 password (*) (or a ClientKey/ServerKey, or SaltedPassword). It sends 279 the username to the server, which retrieves the corresponding 280 authentication information, i.e. a salt, StoredKey, ServerKey and the 281 iteration count i. (Note that a server implementation may choose to 282 use the same iteration count for all accounts.) The server sends the 283 salt and the iteration count to the client, which then computes the 284 following values and sends a ClientProof to the server: 286 (*) - Note that both the username and the password MUST be encoded in 287 UTF-8 [RFC3629]. 289 Informative Note: Implementors are encouraged to create test cases 290 that use both username passwords with non-ASCII characters. In 291 particular, it's useful to test characters whose "Unicode 292 Normalization Form C" and "Unicode Normalization Form KC" are 293 different. Some examples of such characters include Vulgar Fraction 294 One Half (U+00BD) and Acute Accent (U+00B4). 296 SaltedPassword := Hi(Normalize(password), salt) 297 ClientKey := HMAC(SaltedPassword, "Client Key") 298 StoredKey := H(ClientKey) 299 AuthMessage := client-first-message-bare + "," + 300 server-first-message + "," + 301 client-final-message-without-proof 302 ClientSignature := HMAC(StoredKey, AuthMessage) 303 ClientProof := ClientKey XOR ClientSignature 304 ServerKey := HMAC(SaltedPassword, "Server Key") 305 ServerSignature := HMAC(ServerKey, AuthMessage) 307 The server authenticates the client by computing the ClientSignature, 308 exclusive-ORing that with the ClientProof to recover the ClientKey 309 and verifying the correctness of the ClientKey by applying the hash 310 function and comparing the result to the StoredKey. If the ClientKey 311 is correct, this proves that the client has access to the user's 312 password. 314 Similarly, the client authenticates the server by computing the 315 ServerSignature and comparing it to the value sent by the server. If 316 the two are equal, it proves that the server had access to the user's 317 ServerKey. 319 The AuthMessage is computed by concatenating messages from the 320 authentication exchange. The format of these messages is defined in 321 Section 7. 323 4. SCRAM Mechanism Names 325 A SCRAM mechanism name is a string "SCRAM-" followed by the 326 uppercased name of the underlying hash function taken from the IANA 327 "Hash Function Textual Names" registry (see http://www.iana.org), 328 optionally followed by the suffix "-PLUS" (see below). Note that 329 SASL mechanism names are limited to 20 characters, which means that 330 only hash function names with lengths shorter or equal to 9 331 characters (20-length("SCRAM-")-length("-PLUS") can be used. For 332 cases when the underlying hash function name is longer than 9 333 characters, an alternative 9 character (or shorter) name can be used 334 to construct the corresponding SCRAM mechanism name, as long as this 335 alternative name doesn't conflict with any other hash function name 336 from the IANA "Hash Function Textual Names" registry. In order to 337 prevent future conflict, such alternative name SHOULD be registered 338 in the IANA "Hash Function Textual Names" registry. 340 For interoperability, all SCRAM clients and servers MUST implement 341 the SCRAM-SHA-1 authentication mechanism, i.e. an authentication 342 mechanism from the SCRAM family that uses the SHA-1 hash function as 343 defined in [RFC3174]. 345 The "-PLUS" suffix is used only when the server supports channel 346 binding to the external channel. If the server supports channel 347 binding, it will advertise both the "bare" and "plus" versions of 348 whatever mechanisms it supports (e.g., if the server supports only 349 SCRAM with SHA-1 then it will advertise support for both SCRAM-SHA-1 350 and SCRAM-SHA-1-PLUS); if the server does not support channel 351 binding, then it will advertise only the "bare" version of the 352 mechanism (e.g., only SCRAM-SHA-1). The "-PLUS" exists to allow 353 negotiation of the use of channel binding. See Section 6. 355 5. SCRAM Authentication Exchange 357 SCRAM is a SASL mechanism whose client response and server challenge 358 messages are text-based messages containing one or more attribute- 359 value pairs separated by commas. Each attribute has a one-letter 360 name. The messages and their attributes are described in 361 Section 5.1, and defined in Section 7. 363 SCRAM is a client-first SASL mechanism (See [RFC4422], Section 5, 364 item 2a), and returns additional data together with a server's 365 indication of a successful outcome. 367 This is a simple example of a SCRAM-SHA-1 authentication exchange 368 when the client doesn't support channel bindings: 370 C: n,,n=Chris Newman,r=ClientNonce 371 S: r=ClientNonceServerNonce,s=PxR/wv+epq,i=128 372 C: c=biwsCg==,r=ClientNonceServerNonce,p=WxPv/siO5l+qxN4 373 S: v=WxPv/siO5l+qxN4 375 [[anchor5: Note that the all hashes above are fake and will be fixed 376 during AUTH48.]] 378 With channel-binding data sent by the client this might look like 379 this (see [tls-server-end-point] for the definition of tls-server- 380 end-point TLS channel binding): 382 C: p=tls-server-end-point,,n=Chris Newman,r=ClientNonce 383 S: r=ClientNonceServerNonce,s=PxR/wv+epq,i=128 384 C: c=cD10bHMtc2VydmVyLWVuZC1wb2ludCwsy1hFtXOnZ+ySrQM6srFp 385 l/77uqvtxrg7nBY1BetEr/g=,r=ClientNonceServerNonce,p=Wx 386 Pv/siO5l+qxN4 387 S: v=WxPv/siO5l+qxN4 389 [[anchor6: Note that all hashes above are fake and will be fixed 390 during AUTH48.]] 392 First, the client sends the "client-first-message" containing: 394 o a GS2 header consisting of a flag indicating whether channel 395 binding is supported-but-not-used, not supported, or used, and an 396 optional SASL authorization identity; 398 o SCRAM username and a random, unique nonce attributes. 400 Note that the client's first message will always start with "n", "y" 401 or "p", otherwise the message is invalid and authentication MUST 402 fail. This is important, as it allows for GS2 extensibility (e.g., 403 to add support for security layers). 405 In response, the server sends a "server-first-message" containing the 406 user's iteration count i, the user's salt, and appends its own nonce 407 to the client-specified one. 409 The client then responds by sending "client-final-message" with the 410 same nonce and a ClientProof computed using the selected hash 411 function as explained earlier. 413 The server verifies the nonce and the proof, verifies that the 414 authorization identity (if supplied by the client in the first 415 message) is authorized to act as the authentication identity, and, 416 finally, it responds with a "server-final-message", concluding the 417 authentication exchange. 419 The client then authenticates the server by computing the 420 ServerSignature and comparing it to the value sent by the server. If 421 the two are different, the client MUST consider the authentication 422 exchange to be unsuccessful and it might have to drop the connection. 424 5.1. SCRAM Attributes 426 This section describes the permissible attributes, their use, and the 427 format of their values. All attribute names are single US-ASCII 428 letters and are case-sensitive. 430 Note that the order of attributes in client or server messages is 431 fixed, with the exception of extension attributes (described by the 432 "extensions" ABNF production), which can appear in any order in the 433 designated positions. See the ABNF section for authoritative 434 reference. 436 o a: This is an optional attribute, and is part of the GS2 437 [I-D.ietf-sasl-gs2] bridge between the GSS-API and SASL. This 438 attribute specifies an authorization identity. A client may 439 include it in its first message to the server if it wants to 440 authenticate as one user, but subsequently act as a different 441 user. This is typically used by an administrator to perform some 442 management task on behalf of another user, or by a proxy in some 443 situations. 445 Upon the receipt of this value the server verifies its 446 correctness according to the used SASL protocol profile. 447 Failed verification results in failed authentication exchange. 449 If this attribute is omitted (as it normally would be), the 450 authorization identity is assumed to be derived from the 451 username specified with the (required) "n" attribute. 453 The server always authenticates the user specified by the "n" 454 attribute. If the "a" attribute specifies a different user, 455 the server associates that identity with the connection after 456 successful authentication and authorization checks. 458 The syntax of this field is the same as that of the "n" field 459 with respect to quoting of '=' and ','. 461 o n: This attribute specifies the name of the user whose password is 462 used for authentication (a.k.a. "authentication identity" 463 [RFC4422]). A client MUST include it in its first message to the 464 server. If the "a" attribute is not specified (which would 465 normally be the case), this username is also the identity which 466 will be associated with the connection subsequent to 467 authentication and authorization. 469 Before sending the username to the server, the client SHOULD 470 prepare the username using the "SASLPrep" profile [RFC4013] of 471 the "stringprep" algorithm [RFC3454] treating it as a query 472 string (i.e., unassigned Unicode code points are allowed). If 473 the preparation of the username fails or results in an empty 474 string, the client SHOULD abort the authentication exchange 475 (*). 477 (*) An interactive client can request a repeated entry of the 478 username value. 480 Upon receipt of the username by the server, the server SHOULD 481 prepare it using the "SASLPrep" profile [RFC4013] of the 482 "stringprep" algorithm [RFC3454] treating it as a query string 483 (i.e., unassigned Unicode code points are allowed). If the 484 preparation of the username fails or results in an empty 485 string, the server SHOULD abort the authentication exchange. 486 Whether or not the server prepares the username using 487 "SASLPrep", it MUST use it as received in hash calculations. 489 The characters ',' or '=' in usernames are sent as '=2C' and 490 '=3D' respectively. If the server receives a username which 491 contains '=' not followed by either '2C' or '3D', then the 492 server MUST fail the authentication. 494 o m: This attribute is reserved for future extensibility. In this 495 version of SCRAM, its presence in a client or a server message 496 MUST cause authentication failure when the attribute is parsed by 497 the other end. 499 o r: This attribute specifies a sequence of random printable ASCII 500 characters excluding ',' which forms the nonce used as input to 501 the hash function. No quoting is applied to this string. As 502 described earlier, the client supplies an initial value in its 503 first message, and the server augments that value with its own 504 nonce in its first response. It is important that this value be 505 different for each authentication (see [RFC4086] for more details 506 on how to achieve this). The client MUST verify that the initial 507 part of the nonce used in subsequent messages is the same as the 508 nonce it initially specified. The server MUST verify that the 509 nonce sent by the client in the second message is the same as the 510 one sent by the server in its first message. 512 o c: This REQUIRED attribute specifies the base64-encoded GS2 header 513 and channel-binding data. It is sent by the client in its second 514 authentication message. The attribute data consist of: 516 * the GS2 header from the client's first message (recall that the 517 GS2 header contains a channel binding flag and an optional 518 authzid). This header is going to include channel binding type 519 prefix (see [RFC5056]), if and only if the client is using 520 channel binding; 522 * followed by the external channel's channel binding data, if and 523 only if the client is using channel binding. 525 o s: This attribute specifies the base64-encoded salt used by the 526 server for this user. It is sent by the server in its first 527 message to the client. 529 o i: This attribute specifies an iteration count for the selected 530 hash function and user, and MUST be sent by the server along with 531 the user's salt. 533 For SCRAM-SHA-1/SCRAM-SHA-1-PLUS SASL mechanism servers SHOULD 534 announce a hash iteration-count of at least 4096. Note that a 535 client implementation MAY cache ClientKey&ServerKey (or just 536 SaltedPassword) for later reauthentication to the same service, 537 as it is likely that the server is going to advertise the same 538 salt value upon reauthentication. This might be useful for 539 mobile clients where CPU usage is a concern. 541 o p: This attribute specifies a base64-encoded ClientProof. The 542 client computes this value as described in the overview and sends 543 it to the server. 545 o v: This attribute specifies a base64-encoded ServerSignature. It 546 is sent by the server in its final message, and is used by the 547 client to verify that the server has access to the user's 548 authentication information. This value is computed as explained 549 in the overview. 551 5.2. Compliance with SASL mechanism requirements 553 This section describes compliance with SASL mechanism requirements 554 specified in Section 5 of [RFC4422]. 556 1) "SCRAM-SHA-1" and "SCRAM-SHA-1-PLUS". 558 2a) SCRAM is a client-first mechanism. 560 2b) SCRAM sends additional data with success. 562 3) SCRAM is capable of transferring authorization identities from the 563 client to the server. 565 4) SCRAM does not offer any security layers (SCRAM offers channel 566 binding instead). 568 5) SCRAM has a hash protecting the authorization identity. 570 6. Channel Binding 572 SCRAM supports channel binding to external secure channels, such as 573 TLS. Clients and servers may or may not support channel binding, 574 therefore the use of channel binding is negotiable. SCRAM does not 575 provide security layers, however, therefore it is imperative that 576 SCRAM provide integrity protection for the negotiation of channel 577 binding. 579 Use of channel binding is negotiated as follows: 581 o Servers SHOULD advertise both non-PLUS (SCRAM-) and 582 the PLUS-variant (SCRAM--PLUS) SASL mechanism 583 names. If the server cannot support channel binding, it MAY 584 advertise only the non-PLUS variant. If the server would never 585 succeed authentication of the non-PLUS variant due to policy 586 reasons, it MAY advertise only the PLUS-variant. 588 o If the client negotiates mechanisms then the client MUST select 589 SCRAM--PLUS if offered by the server and the client 590 wants to select SCRAM with the given hash function. Otherwise 591 (the client does not negotiate mechanisms), if the client has no 592 prior knowledge about mechanisms supported by the server and 593 wasn't explicitly configured to use a particular variant of the 594 SCRAM mechanism, then it MUST select only SCRAM- 595 (not suffixed with "-PLUS"). 597 o If the client supports channel binding and the server appears to 598 support it (i.e., the client sees SCRAM--PLUS), or 599 if the client wishes to use channel binding but the client does 600 not negotiate mechanisms, then the client MUST set the GS2 channel 601 binding flag to "p" in order to indicate the channel binding type 602 it is using and it MUST include the channel binding data for the 603 external channel in the computation of the "c=" attribute (see 604 Section 5.1). 606 o If the client supports channel binding but the server does not 607 appear to (i.e., the client did not see SCRAM-- 608 PLUS) then the client MUST either fail authentication or it MUST 609 choose the non-PLUS mechanism and set the GS2 channel binding flag 610 to "y" and MUST NOT include channel binding data for the external 611 channel in the computation of the "c=" attribute (see 612 Section 5.1). 614 o If the client does not support channel binding then the client 615 MUST set the GS2 channel binding flag to "n" and MUST NOT include 616 channel binding data for the external channel in the computation 617 of the "c=" attribute (see Section 5.1). 619 o Upon receipt of the client first message the server checks the GS2 620 channel binding flag (gs2-cb-flag). 622 * If the flag is set to "y" and the server supports channel 623 binding the server MUST fail authentication. This is because 624 if the client sets the GS2 channel binding flag set to "y" then 625 the client must have believed that the server did not support 626 channel binding -- if the server did in fact support channel 627 binding then this is an indication that there has been a 628 downgrade attack (e.g., an attacker changed the server's 629 mechanism list to exclude the -PLUS suffixed SCRAM mechanism 630 name(s)). 632 * If the channel binding flag was "p" and the server does not 633 support the indicated channel binding type then the server MUST 634 fail authentication. 636 The server MUST always validate the client's "c=" field. The server 637 does this by constructing the value of the "c=" attribute and then 638 checking that it matches the client's c= attribute value. 640 For more discussions of channel bindings, and the syntax of the 641 channel binding data for various security protocols, see [RFC5056]. 643 6.1. Default Channel Binding 645 A default channel binding type agreement process for all SASL 646 application protocols that do not provide their own channel binding 647 type agreement is provided as follows. 649 'tls-unique' is the default channel binding type for any application 650 that doesn't specify one. 652 Servers MUST implement the "tls-unique" [tls-unique] 653 [I-D.altman-tls-channel-bindings] channel binding type, if they 654 implement any channel binding. Clients SHOULD implement the "tls- 655 unique" [tls-unique] [I-D.altman-tls-channel-bindings] channel 656 binding type, if they implement any channel binding. Clients and 657 servers SHOULD choose the highest- layer/innermost end-to-end TLS 658 channel as the channel to bind to. 660 Servers MUST choose the channel binding type indicated by the client, 661 or fail authentication if they don't support it. 663 7. Formal Syntax 665 The following syntax specification uses the Augmented Backus-Naur 666 Form (ABNF) notation as specified in [RFC5234]. "UTF8-2", "UTF8-3" 667 and "UTF8-4" non-terminal are defined in [RFC3629]. 669 ALPHA = 670 DIGIT = 671 UTF8-2 = 672 UTF8-3 = 673 UTF8-4 = 675 attr-val = ALPHA "=" value 676 ;; Generic syntax of any attribute sent 677 ;; by server or client 679 value = 1*value-char 681 value-safe-char = %x01-2B / %x2D-3C / %x3E-7F / 682 UTF8-2 / UTF8-3 / UTF8-4 683 ;; UTF8-char except NUL, "=", and ",". 685 value-char = value-safe-char / "=" 687 printable = %x21-2B / %x2D-7E 688 ;; Printable ASCII except ",". 689 ;; Note that any "printable" is also 690 ;; a valid "value". 692 base64-char = ALPHA / DIGIT / "/" / "+" 694 base64-4 = 4base64-char 696 base64-3 = 3base64-char "=" 698 base64-2 = 2base64-char "==" 700 base64 = *base64-4 [base64-3 / base64-2] 702 posit-number = %x31-39 *DIGIT 703 ;; A positive number 705 saslname = 1*(value-safe-char / "=2C" / "=3D") 706 ;; Conforms to 708 authzid = "a=" saslname 709 ;; Protocol specific. 711 cb-name = 1*(ALPHA / DIGIT / "." / "-") 712 ;; See RFC 5056 section 7. 713 ;; E.g. "tls-server-end-point" or 714 ;; "tls-unique" 716 gs2-cbind-flag = "p=" cb-name / "n" / "y" 717 ;; "n" -> client doesn't support channel binding 718 ;; "y" -> client does support channel binding 719 ;; but thinks the server does not. 720 ;; "p" -> client requires channel binding. 721 ;; The selected channel binding follows "p=". 723 gs2-header = gs2-cbind-flag "," [ authzid ] "," 724 ;; GS2 header for SCRAM 725 ;; (the actual GS2 header includes an optional 726 ;; flag to indicate that the GSS mechanism is not 727 ;; "standard" but since SCRAM is "standard" we 728 ;; don't include that flag). 730 username = "n=" saslname 731 ;; Usernames are prepared using SASLPrep. 733 reserved-mext = "m=" 1*(value-char) 734 ;; Reserved for signalling mandatory extensions. 735 ;; The exact syntax will be defined in 736 ;; the future. 738 channel-binding = "c=" base64 739 ;; base64 encoding of cbind-input 741 proof = "p=" base64 743 nonce = "r=" c-nonce [s-nonce] 744 ;; Second part provided by server. 746 c-nonce = printable 748 s-nonce = printable 750 salt = "s=" base64 752 verifier = "v=" base64 753 ;; base-64 encoded ServerSignature. 755 iteration-count = "i=" posit-number 756 ;; A positive number 758 client-first-message-bare = 760 [reserved-mext ","] 761 username "," nonce ["," extensions] 763 client-first-message = 764 gs2-header client-first-message-bare 766 server-first-message = 767 [reserved-mext ","] nonce "," salt "," 768 iteration-count ["," extensions] 770 client-final-message-without-proof = 771 channel-binding "," nonce ["," 772 extensions] 774 client-final-message = 775 client-final-message-without-proof "," proof 777 gss-server-error = "e=" value 778 server-final-message = gss-server-error / 779 verifier ["," extensions] 780 ;; The error message is only for the GSS-API 781 ;; form of SCRAM, and it is OPTIONAL to 782 ;; implement it. 784 extensions = attr-val *("," attr-val) 785 ;; All extensions are optional, 786 ;; i.e. unrecognized attributes 787 ;; not defined in this document 788 ;; MUST be ignored. 790 cbind-data = 1*OCTET 792 cbind-input = gs2-header [ cbind-data ] 793 ;; cbind-data MUST be present for 794 ;; gs2-cbind-flag of "p" and MUST be absent 795 ;; for "y" or "n". 797 8. SCRAM as a GSS-API Mechanism 799 This section and its sub-sections and all normative references of it 800 not referenced elsewhere in this document are INFORMATIONAL for SASL 801 implementors, but they are NORMATIVE for GSS-API implementors. 803 SCRAM is actually also GSS-API mechanism. The messages are the same, 804 but a) the GS2 header on the client's first message and channel 805 binding data is excluded when SCRAM is used as a GSS-API mechanism, 806 and b) the RFC2743 section 3.1 initial context token header is 807 prefixed to the client's first authentication message (context 808 token). 810 The GSS-API mechanism OID for SCRAM-SHA-1 is (see Section 10). 812 8.1. GSS-API Principal Name Types for SCRAM 814 SCRAM does not name acceptors. Therefore only GSS_C_NO_NAME and 815 names of type GSS_C_NT_ANONYMOUS shall be allowed as the target name 816 input of GSS_Init_sec_context() when using a SCRAM mechanism. 818 SCRAM supports only a single name type for initiators: 819 GSS_C_NT_USER_NAME. GSS_C_NT_USER_NAME is the default name type for 820 SCRAM. 822 There is no name canonicalization procedure for SCRAM beyond applying 823 SASLprep as described in Section 5.1. 825 The query, display and exported name syntax for SCRAM principal names 826 is the same: there is no syntax -- SCRAM principal names are free- 827 form. (The exported name token does, of course, conform to [RFC2743] 828 section 3.2, but the "NAME" part of the token is just a SCRAM user 829 name.) 831 8.2. GSS-API Per-Message Tokens for SCRAM 833 The per-message tokens for SCRAM as a GSS-API mechanism SHALL be the 834 same as those for the Kerberos V GSS-API mechanism [RFC4121] (see 835 Section 4.2 and sub-sections), using the Kerberos V "aes128-cts-hmac- 836 sha1-96" enctype [RFC3962]. 838 The 128-bit session "protocol key" SHALL be derived by using the 839 least significant (right-most) 128 bits of HMAC(StoredKey, "GSS-API 840 session key" || ClientKey || AuthMessage). "Specific keys" are then 841 derived as usual as described in Section 2 of [RFC4121], [RFC3961] 842 and [RFC3962]. 844 The terms "protocol key" and "specific key" are Kerberos V5 terms 846 [RFC3961]. 848 SCRAM does support PROT_READY, and is PROT_READY on the initiator 849 side first upon receipt of the server's reply to the initial security 850 context token. 852 8.3. GSS_Pseudo_random() for SCRAM 854 The GSS_Pseudo_random() [RFC4401] for SCRAM SHALL be the same as for 855 the Kerberos V GSS-API mechanism [RFC4402]. There is no acceptor- 856 asserted sub-session key for SCRAM, thus GSS_C_PRF_KEY_FULL and 857 GSS_C_PRF_KEY_PARTIAL are equivalent for SCRAM's GSS_Pseudo_random(). 858 The protocol key to be used for the GSS_Pseudo_random() SHALL be the 859 same as the key defined in Section 8.2. 861 9. Security Considerations 863 If the authentication exchange is performed without a strong security 864 layer (such as TLS with data confidentiality), then a passive 865 eavesdropper can gain sufficient information to mount an offline 866 dictionary or brute-force attack which can be used to recover the 867 user's password. The amount of time necessary for this attack 868 depends on the cryptographic hash function selected, the strength of 869 the password and the iteration count supplied by the server. An 870 external security layer with strong encryption will prevent this 871 attack. 873 If the external security layer used to protect the SCRAM exchange 874 uses an anonymous key exchange, then the SCRAM channel binding 875 mechanism can be used to detect a man-in-the-middle attack on the 876 security layer and cause the authentication to fail as a result. 877 However, the man-in-the-middle attacker will have gained sufficient 878 information to mount an offline dictionary or brute-force attack. 879 For this reason, SCRAM includes the ability to increase the iteration 880 count over time. 882 If the authentication information is stolen from the authentication 883 database, then an offline dictionary or brute-force attack can be 884 used to recover the user's password. The use of salt mitigates this 885 attack somewhat by requiring a separate attack on each password. 886 Authentication mechanisms which protect against this attack are 887 available (e.g., the EKE class of mechanisms). RFC 2945 [RFC2945] is 888 an example of such technology. The WG selected not to use EKE like 889 mechanisms as basis for SCRAM. 891 If an attacker obtains the authentication information from the 892 authentication repository and either eavesdrops on one authentication 893 exchange or impersonates a server, the attacker gains the ability to 894 impersonate that user to all servers providing SCRAM access using the 895 same hash function, password, iteration count and salt. For this 896 reason, it is important to use randomly-generated salt values. 898 SCRAM does not negotiate a hash function to use. Hash function 899 negotiation is left to the SASL mechanism negotiation. It is 900 important that clients be able to sort a locally available list of 901 mechanisms by preference so that the client may pick the most 902 preferred of a server's advertised mechanism list. This preference 903 order is not specified here as it is a local matter. The preference 904 order should include objective and subjective notions of mechanism 905 cryptographic strength (e.g., SCRAM with a successor to SHA-1 may be 906 preferred over SCRAM with SHA-1). 908 Note that to protect the SASL mechanism negotiation applications 909 normally must list the server mechs twice: once before and once after 910 authentication, the latter using security layers. Since SCRAM does 911 not provide security layers the only ways to protect the mechanism 912 negotiation are: a) use channel binding to an external channel, or b) 913 use an external channel that authenticates a user-provided server 914 name. 916 SCRAM does not protect against downgrade attacks of channel binding 917 types. The complexities of negotiation a channel binding type, and 918 handling down-grade attacks in that negotiation, was intentionally 919 left out of scope for this document. 921 A hostile server can perform a computational denial-of-service attack 922 on clients by sending a big iteration count value. 924 See [RFC4086] for more information about generating randomness. 926 10. IANA Considerations 928 IANA is requested to add the following family of SASL mechanisms to 929 the SASL Mechanism registry established by [RFC4422]: 931 To: iana@iana.org 932 Subject: Registration of a new SASL family SCRAM 934 SASL mechanism name (or prefix for the family): SCRAM-* 935 Security considerations: Section 7 of [RFCXXXX] 936 Published specification (optional, recommended): [RFCXXXX] 937 Person & email address to contact for further information: 938 IETF SASL WG 939 Intended usage: COMMON 940 Owner/Change controller: IESG 941 Note: Members of this family must be explicitly registered 942 using the "IETF Review" [RFC5226] registration procedure. 943 Reviews must be requested on the SASL WG mailing list. 945 "IETF Review" [RFC5226] registration procedure MUST be used for 946 registering new mechanisms in this family. The SASL mailing list 947 (or a successor designated by the responsible 948 Security AD) MUST be used for soliciting reviews on such 949 registrations. 951 Note to future SCRAM- mechanism designers: each new SCRAM- SASL 952 mechanism MUST be explicitly registered with IANA and MUST comply 953 with SCRAM- mechanism naming convention defined in Section 4 of this 954 document. 956 IANA is requested to add the following entries to the SASL Mechanism 957 registry established by [RFC4422]: 959 To: iana@iana.org 960 Subject: Registration of a new SASL mechanism SCRAM-SHA-1 962 SASL mechanism name (or prefix for the family): SCRAM-SHA-1 963 Security considerations: Section 7 of [RFCXXXX] 964 Published specification (optional, recommended): [RFCXXXX] 965 Person & email address to contact for further information: 966 IETF SASL WG 967 Intended usage: COMMON 968 Owner/Change controller: IESG 969 Note: 971 To: iana@iana.org 972 Subject: Registration of a new SASL mechanism SCRAM-SHA-1-PLUS 974 SASL mechanism name (or prefix for the family): SCRAM-SHA-1-PLUS 975 Security considerations: Section 7 of [RFCXXXX] 976 Published specification (optional, recommended): [RFCXXXX] 977 Person & email address to contact for further information: 978 IETF SASL WG 979 Intended usage: COMMON 980 Owner/Change controller: IESG 981 Note: 983 This document also requests IANA to assign a GSS-API mechanism OID 984 for SCRAM-SHA-1 from the iso.org.dod.internet.security.mechanisms 985 prefix (see "SMI Security for Mechanism Codes" registry). 987 11. Acknowledgements 989 This document benefited from discussions on the SASL WG mailing list. 990 The authors would like to specially thank Dave Cridland, Simon 991 Josefsson, Jeffrey Hutzelman, Kurt Zeilenga, Pasi Eronen and Peter 992 Saint-Andrefor their contributions to this document. 994 Appendix A. Other Authentication Mechanisms 996 The DIGEST-MD5 [I-D.ietf-sasl-digest-to-historic] mechanism has 997 proved to be too complex to implement and test, and thus has poor 998 interoperability. The security layer is often not implemented, and 999 almost never used; everyone uses TLS instead. For a more complete 1000 list of problems with DIGEST-MD5 which lead to the creation of SCRAM 1001 see [I-D.ietf-sasl-digest-to-historic]. 1003 The CRAM-MD5 SASL mechanism, while widely deployed has also some 1004 problems, in particular it is missing some modern SASL features such 1005 as support for internationalized usernames and passwords, support for 1006 passing of authorization identity, support for channel bindings. It 1007 also doesn't support server authentication. For a more complete list 1008 of problems with CRAM-MD5 see [I-D.ietf-sasl-crammd5-to-historic]. 1010 The PLAIN [RFC4616] SASL mechanism allows a malicious server or 1011 eavesdropper to impersonate the authenticating user to any other 1012 server for which the user has the same password. It also sends the 1013 password in the clear over the network, unless TLS is used. Server 1014 authentication is not supported. 1016 Appendix B. Design Motivations 1018 The following design goals shaped this document. Note that some of 1019 the goals have changed since the initial version of the document. 1021 o The SASL mechanism has all modern SASL features: support for 1022 internationalized usernames and passwords, support for passing of 1023 authorization identity, support for channel bindings. 1025 o The protocol supports mutual authentication. 1027 o The authentication information stored in the authentication 1028 database is not sufficient by itself to impersonate the client. 1030 o The server does not gain the ability to impersonate the client to 1031 other servers (with an exception for server-authorized proxies), 1032 unless such other servers allow SCRAM authentication and use the 1033 same salt and iteration count for the user. 1035 o The mechanism is extensible, but [hopefully] not overengineered in 1036 this respect. 1038 o Easier to implement than DIGEST-MD5 in both clients and servers. 1040 Appendix C. Internet-Draft Change History 1042 (RFC Editor: Please delete this section and all subsections.) 1044 Changes since -10 1046 o Converted the source for this I-D to XML. 1048 o Added text to make SCRAM compliant with the new GS2 design. 1050 o Added text on channel binding negotiation. 1052 o Added text on channel binding, including a reference to RFC5056. 1054 o Added text on SCRAM as a GSS-API mechanism. This noted as not 1055 relevant to SASL-only implementors -- the normative references for 1056 SCRAM as a GSS-API mechanism are segregated as well. 1058 Changes since -07 1060 o Updated References. 1062 o Clarified purpose of the m= attribute. 1064 o Fixed a problem with authentication/authorization identity's ABNF 1065 not allowing for some characters. 1067 o Updated ABNF for nonce to show client-generated and server- 1068 generated parts. 1070 o Only register SCRAM-SHA-1 with IANA and require explicit 1071 registrations of all other SCRAM- mechanisms. 1073 Changes since -06 1075 o Removed hash negotiation from SCRAM and turned it into a family of 1076 SASL mechanisms. 1078 o Start using "Hash Function Textual Names" IANA registry for SCRAM 1079 mechanism naming. 1081 o Fixed definition of Hi(str, salt) to be consistent with [RFC2898]. 1083 o Clarified extensibility of SCRAM: added m= attribute (for future 1084 mandatory extensions) and specified that all unrecognized 1085 attributes must be ignored. 1087 Changes since -05 1088 o Changed the mandatory to implement hash algorithm to SHA-1 (as per 1089 WG consensus). 1091 o Added text about use of SASLPrep for username canonicalization/ 1092 validation. 1094 o Clarified that authorization identity is canonicalized/verified 1095 according to SASL protocol profile. 1097 o Clarified that iteration count is per-user. 1099 o Clarified how clients select the authentication function. 1101 o Added IANA registration for the new mechanism. 1103 o Added missing normative references (UTF-8, SASLPrep). 1105 o Various editorial changes based on comments from Hallvard B 1106 Furuseth, Nico William and Simon Josefsson. 1108 Changes since -04 1110 o Update Base64 and Security Glossary references. 1112 o Add Formal Syntax section. 1114 o Don't bother with "v=". 1116 o Make MD5 mandatory to implement. Suggest i=128. 1118 Changes since -03 1120 o Seven years have passed, in which it became clear that DIGEST-MD5 1121 suffered from unacceptably bad interoperability, so SCRAM-MD5 is 1122 now back from the dead. 1124 o Be hash agnostic, so MD5 can be replaced more easily. 1126 o General simplification. 1128 12. References 1130 12.1. Normative References 1132 [RFC2104] Krawczyk, H., Bellare, M., and R. Canetti, "HMAC: Keyed- 1133 Hashing for Message Authentication", RFC 2104, 1134 February 1997. 1136 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1137 Requirement Levels", BCP 14, RFC 2119, March 1997. 1139 [RFC3174] Eastlake, D. and P. Jones, "US Secure Hash Algorithm 1 1140 (SHA1)", RFC 3174, September 2001. 1142 [RFC3454] Hoffman, P. and M. Blanchet, "Preparation of 1143 Internationalized Strings ("stringprep")", RFC 3454, 1144 December 2002. 1146 [RFC3629] Yergeau, F., "UTF-8, a transformation format of ISO 1147 10646", STD 63, RFC 3629, November 2003. 1149 [RFC4013] Zeilenga, K., "SASLprep: Stringprep Profile for User Names 1150 and Passwords", RFC 4013, February 2005. 1152 [RFC4422] Melnikov, A. and K. Zeilenga, "Simple Authentication and 1153 Security Layer (SASL)", RFC 4422, June 2006. 1155 [RFC4648] Josefsson, S., "The Base16, Base32, and Base64 Data 1156 Encodings", RFC 4648, October 2006. 1158 [RFC5056] Williams, N., "On the Use of Channel Bindings to Secure 1159 Channels", RFC 5056, November 2007. 1161 [RFC5234] Crocker, D. and P. Overell, "Augmented BNF for Syntax 1162 Specifications: ABNF", STD 68, RFC 5234, January 2008. 1164 12.2. Normative References for GSS-API implementors 1166 [I-D.ietf-sasl-gs2] 1167 Josefsson, S. and N. Williams, "Using GSS-API Mechanisms 1168 in SASL: The GS2 Mechanism Family", draft-ietf-sasl-gs2-12 1169 (work in progress), April 2009. 1171 [RFC2743] Linn, J., "Generic Security Service Application Program 1172 Interface Version 2, Update 1", RFC 2743, January 2000. 1174 [RFC3961] Raeburn, K., "Encryption and Checksum Specifications for 1175 Kerberos 5", RFC 3961, February 2005. 1177 [RFC3962] Raeburn, K., "Advanced Encryption Standard (AES) 1178 Encryption for Kerberos 5", RFC 3962, February 2005. 1180 [RFC4121] Zhu, L., Jaganathan, K., and S. Hartman, "The Kerberos 1181 Version 5 Generic Security Service Application Program 1182 Interface (GSS-API) Mechanism: Version 2", RFC 4121, 1183 July 2005. 1185 [RFC4401] Williams, N., "A Pseudo-Random Function (PRF) API 1186 Extension for the Generic Security Service Application 1187 Program Interface (GSS-API)", RFC 4401, February 2006. 1189 [RFC4402] Williams, N., "A Pseudo-Random Function (PRF) for the 1190 Kerberos V Generic Security Service Application Program 1191 Interface (GSS-API) Mechanism", RFC 4402, February 2006. 1193 [tls-unique] 1194 Zhu, L., "Registration of TLS unique channel binding 1195 (generic)", IANA http://www.iana.org/assignments/ 1196 channel-binding-types/tls-unique, July 2008. 1198 12.3. Informative References 1200 [I-D.altman-tls-channel-bindings] 1201 Altman, J., Williams, N., and L. Zhu, "Channel Bindings 1202 for TLS", draft-altman-tls-channel-bindings-06 (work in 1203 progress), August 2009. 1205 [I-D.ietf-sasl-crammd5-to-historic] 1206 Zeilenga, K., "CRAM-MD5 to Historic", 1207 draft-ietf-sasl-crammd5-to-historic-00 (work in progress), 1208 November 2008. 1210 [I-D.ietf-sasl-digest-to-historic] 1211 Melnikov, A., "Moving DIGEST-MD5 to Historic", 1212 draft-ietf-sasl-digest-to-historic-00 (work in progress), 1213 July 2008. 1215 [I-D.melnikov-sasl-scram-ldap] 1216 Melnikov, A., "LDAP schema for storing SCRAM secrets", 1217 draft-melnikov-sasl-scram-ldap-02 (work in progress), 1218 July 2009. 1220 [RFC2865] Rigney, C., Willens, S., Rubens, A., and W. Simpson, 1221 "Remote Authentication Dial In User Service (RADIUS)", 1222 RFC 2865, June 2000. 1224 [RFC2898] Kaliski, B., "PKCS #5: Password-Based Cryptography 1225 Specification Version 2.0", RFC 2898, September 2000. 1227 [RFC2945] Wu, T., "The SRP Authentication and Key Exchange System", 1228 RFC 2945, September 2000. 1230 [RFC4086] Eastlake, D., Schiller, J., and S. Crocker, "Randomness 1231 Requirements for Security", BCP 106, RFC 4086, June 2005. 1233 [RFC4510] Zeilenga, K., "Lightweight Directory Access Protocol 1234 (LDAP): Technical Specification Road Map", RFC 4510, 1235 June 2006. 1237 [RFC4616] Zeilenga, K., "The PLAIN Simple Authentication and 1238 Security Layer (SASL) Mechanism", RFC 4616, August 2006. 1240 [RFC4949] Shirey, R., "Internet Security Glossary, Version 2", 1241 RFC 4949, August 2007. 1243 [RFC5226] Narten, T. and H. Alvestrand, "Guidelines for Writing an 1244 IANA Considerations Section in RFCs", BCP 26, RFC 5226, 1245 May 2008. 1247 [RFC5246] Dierks, T. and E. Rescorla, "The Transport Layer Security 1248 (TLS) Protocol Version 1.2", RFC 5246, August 2008. 1250 [tls-server-end-point] 1251 Zhu, L., "Registration of TLS server end-point channel 1252 bindings", IANA http://www.iana.org/assignments/ 1253 channel-binding-types/tls-server-end-point, July 2008. 1255 Authors' Addresses 1257 Abhijit Menon-Sen 1258 Oryx Mail Systems GmbH 1260 Email: ams@oryx.com 1262 Alexey Melnikov 1263 Isode Ltd 1265 Email: Alexey.Melnikov@isode.com 1267 Chris Newman 1268 Sun Microsystems 1269 1050 Lakes Drive 1270 West Covina, CA 91790 1271 USA 1273 Email: chris.newman@sun.com 1275 Nicolas Williams 1276 Sun Microsystems 1277 5300 Riata Trace Ct 1278 Austin, TX 78727 1279 USA 1281 Email: Nicolas.Williams@sun.com