idnits 2.17.1 draft-ietf-spfbis-4408bis-12.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- == There are 1 instance of lines with non-RFC2606-compliant FQDNs in the document. == There are 6 instances of lines with non-RFC6890-compliant IPv4 addresses in the document. If these are example addresses, they should be changed. == There are 1 instance of lines with private range IPv4 addresses in the document. If these are generic example addresses, they should be changed to use any of the ranges defined in RFC 6890 (or successor): 192.0.2.x, 198.51.100.x or 203.0.113.x. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year == Line 1665 has weird spacing: '...pe-from the e...' == The document seems to contain a disclaimer for pre-RFC5378 work, but was first submitted on or after 10 November 2008. The disclaimer is usually necessary only for documents that revise or obsolete older RFCs, and that take significant amounts of text from those RFCs. If you can contact all authors of the source material and they are willing to grant the BCP78 rights to the IETF Trust, you can and should remove the disclaimer. Otherwise, the disclaimer is needed and you can ignore this comment. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (March 16, 2013) is 4058 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: 'CFWS' is mentioned on line 2318, but not defined ** Obsolete normative reference: RFC 5451 (Obsoleted by RFC 7001) ** Downref: Normative reference to an Informational RFC: RFC 5598 -- Possible downref: Non-RFC (?) normative reference: ref. 'US-ASCII' -- Obsolete informational reference (is this intentional?): RFC 4408 (Obsoleted by RFC 7208) -- Obsolete informational reference (is this intentional?): RFC 5751 (Obsoleted by RFC 8551) Summary: 2 errors (**), 0 flaws (~~), 7 warnings (==), 4 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group S. Kitterman 3 Internet-Draft Kitterman Technical Services 4 Obsoletes: 4408 (if approved) March 16, 2013 5 Intended status: Standards Track 6 Expires: September 17, 2013 8 Sender Policy Framework (SPF) for Authorizing Use of Domains in Email, 9 Version 1 10 draft-ietf-spfbis-4408bis-12.txt 12 Abstract 14 Email on the Internet can be forged in a number of ways. In 15 particular, existing protocols place no restriction on what a sending 16 host can use as the "MAIL FROM" of a message or the domain given on 17 the SMTP HELO/EHLO commands. This document describes version 1 of 18 the Sender Policy Framework (SPF) protocol, whereby an ADMD can 19 explicitly authorize the hosts that are allowed to use its domain 20 names, and a receiving host can check such authorization. 22 This document obsoletes RFC4408. 24 Status of this Memo 26 This Internet-Draft is submitted in full conformance with the 27 provisions of BCP 78 and BCP 79. 29 Internet-Drafts are working documents of the Internet Engineering 30 Task Force (IETF). Note that other groups may also distribute 31 working documents as Internet-Drafts. The list of current Internet- 32 Drafts is at http://datatracker.ietf.org/drafts/current/. 34 Internet-Drafts are draft documents valid for a maximum of six months 35 and may be updated, replaced, or obsoleted by other documents at any 36 time. It is inappropriate to use Internet-Drafts as reference 37 material or to cite them other than as "work in progress." 39 This Internet-Draft will expire on September 17, 2013. 41 Copyright Notice 43 Copyright (c) 2013 IETF Trust and the persons identified as the 44 document authors. All rights reserved. 46 This document is subject to BCP 78 and the IETF Trust's Legal 47 Provisions Relating to IETF Documents 48 (http://trustee.ietf.org/license-info) in effect on the date of 49 publication of this document. Please review these documents 50 carefully, as they describe your rights and restrictions with respect 51 to this document. Code Components extracted from this document must 52 include Simplified BSD License text as described in Section 4.e of 53 the Trust Legal Provisions and are provided without warranty as 54 described in the Simplified BSD License. 56 This document may contain material from IETF Documents or IETF 57 Contributions published or made publicly available before November 58 10, 2008. The person(s) controlling the copyright in some of this 59 material may not have granted the IETF Trust the right to allow 60 modifications of such material outside the IETF Standards Process. 61 Without obtaining an adequate license from the person(s) controlling 62 the copyright in such materials, this document may not be modified 63 outside the IETF Standards Process, and derivative works of it may 64 not be created outside the IETF Standards Process, except to format 65 it for publication as an RFC or to translate it into languages other 66 than English. 68 Table of Contents 70 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 6 71 1.1. Terminology . . . . . . . . . . . . . . . . . . . . . . . 6 72 1.1.1. Keywords . . . . . . . . . . . . . . . . . . . . . . . 6 73 1.1.2. Imported Definitions . . . . . . . . . . . . . . . . . 6 74 1.1.3. MAIL FROM Definition . . . . . . . . . . . . . . . . . 7 75 1.1.4. HELO Definition . . . . . . . . . . . . . . . . . . . 7 76 1.2. check_host() . . . . . . . . . . . . . . . . . . . . . . . 7 77 2. Operational Overview . . . . . . . . . . . . . . . . . . . . . 8 78 2.1. The "HELO" Identity . . . . . . . . . . . . . . . . . . . 8 79 2.2. The "MAIL FROM" Identity . . . . . . . . . . . . . . . . . 8 80 2.3. Publishing Authorization . . . . . . . . . . . . . . . . . 8 81 2.4. Checking Authorization . . . . . . . . . . . . . . . . . . 9 82 2.5. Location of Checks . . . . . . . . . . . . . . . . . . . . 10 83 2.6. Results of Evaluation . . . . . . . . . . . . . . . . . . 10 84 2.6.1. None . . . . . . . . . . . . . . . . . . . . . . . . . 11 85 2.6.2. Neutral . . . . . . . . . . . . . . . . . . . . . . . 11 86 2.6.3. Pass . . . . . . . . . . . . . . . . . . . . . . . . . 11 87 2.6.4. Fail . . . . . . . . . . . . . . . . . . . . . . . . . 11 88 2.6.5. Softfail . . . . . . . . . . . . . . . . . . . . . . . 11 89 2.6.6. Temperror . . . . . . . . . . . . . . . . . . . . . . 11 90 2.6.7. Permerror . . . . . . . . . . . . . . . . . . . . . . 11 91 3. SPF Records . . . . . . . . . . . . . . . . . . . . . . . . . 12 92 3.1. DNS Resource Records . . . . . . . . . . . . . . . . . . . 12 93 3.2. Multiple DNS Records . . . . . . . . . . . . . . . . . . . 13 94 3.3. Multiple Strings in a Single DNS record . . . . . . . . . 13 95 3.4. Record Size . . . . . . . . . . . . . . . . . . . . . . . 13 96 3.5. Wildcard Records . . . . . . . . . . . . . . . . . . . . . 13 97 4. The check_host() Function . . . . . . . . . . . . . . . . . . 15 98 4.1. Arguments . . . . . . . . . . . . . . . . . . . . . . . . 15 99 4.2. Results . . . . . . . . . . . . . . . . . . . . . . . . . 15 100 4.3. Initial Processing . . . . . . . . . . . . . . . . . . . . 16 101 4.4. Record Lookup . . . . . . . . . . . . . . . . . . . . . . 16 102 4.5. Selecting Records . . . . . . . . . . . . . . . . . . . . 16 103 4.6. Record Evaluation . . . . . . . . . . . . . . . . . . . . 17 104 4.6.1. Term Evaluation . . . . . . . . . . . . . . . . . . . 17 105 4.6.2. Mechanisms . . . . . . . . . . . . . . . . . . . . . . 17 106 4.6.3. Modifiers . . . . . . . . . . . . . . . . . . . . . . 18 107 4.6.4. DNS Lookup Limits . . . . . . . . . . . . . . . . . . 18 108 4.7. Default Result . . . . . . . . . . . . . . . . . . . . . . 19 109 4.8. Domain Specification . . . . . . . . . . . . . . . . . . . 19 110 5. Mechanism Definitions . . . . . . . . . . . . . . . . . . . . 21 111 5.1. "all" . . . . . . . . . . . . . . . . . . . . . . . . . . 22 112 5.2. "include" . . . . . . . . . . . . . . . . . . . . . . . . 22 113 5.3. "a" . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 114 5.4. "mx" . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 115 5.5. "ptr" (do not use) . . . . . . . . . . . . . . . . . . . . 24 116 5.6. "ip4" and "ip6" . . . . . . . . . . . . . . . . . . . . . 26 117 5.7. "exists" . . . . . . . . . . . . . . . . . . . . . . . . . 26 118 6. Modifier Definitions . . . . . . . . . . . . . . . . . . . . . 28 119 6.1. redirect: Redirected Query . . . . . . . . . . . . . . . . 28 120 6.2. exp: Explanation . . . . . . . . . . . . . . . . . . . . . 29 121 7. Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 122 7.1. Formal Specification . . . . . . . . . . . . . . . . . . . 31 123 7.2. Macro Definitions . . . . . . . . . . . . . . . . . . . . 31 124 7.3. Notes . . . . . . . . . . . . . . . . . . . . . . . . . . 32 125 7.4. Expansion Examples . . . . . . . . . . . . . . . . . . . . 34 126 8. Result Handling . . . . . . . . . . . . . . . . . . . . . . . 36 127 8.1. None . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 128 8.2. Neutral . . . . . . . . . . . . . . . . . . . . . . . . . 36 129 8.3. Pass . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 130 8.4. Fail . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 131 8.5. Softfail . . . . . . . . . . . . . . . . . . . . . . . . . 37 132 8.6. Temperror . . . . . . . . . . . . . . . . . . . . . . . . 38 133 8.7. Permerror . . . . . . . . . . . . . . . . . . . . . . . . 38 134 9. Recording The Result . . . . . . . . . . . . . . . . . . . . . 39 135 9.1. The Received-SPF Header Field . . . . . . . . . . . . . . 39 136 9.2. SPF Results in the Authentication-Results Header Field . . 41 137 10. Effects on Infrastructure . . . . . . . . . . . . . . . . . . 43 138 10.1. Sending Domains . . . . . . . . . . . . . . . . . . . . . 43 139 10.1.1. DNS Resource Considerations . . . . . . . . . . . . . 43 140 10.1.2. Administrator's Considerations . . . . . . . . . . . . 44 141 10.1.3. Bounces . . . . . . . . . . . . . . . . . . . . . . . 45 142 10.2. Receivers . . . . . . . . . . . . . . . . . . . . . . . . 45 143 10.3. Mediators . . . . . . . . . . . . . . . . . . . . . . . . 46 144 11. Security Considerations . . . . . . . . . . . . . . . . . . . 47 145 11.1. Processing Limits . . . . . . . . . . . . . . . . . . . . 47 146 11.2. SPF-Authorized Email May Contain Other False Identities . 47 147 11.3. Spoofed DNS and IP Data . . . . . . . . . . . . . . . . . 48 148 11.4. Cross-User Forgery . . . . . . . . . . . . . . . . . . . . 48 149 11.5. Untrusted Information Sources . . . . . . . . . . . . . . 48 150 11.5.1. Recorded Results . . . . . . . . . . . . . . . . . . . 48 151 11.5.2. External Explanations . . . . . . . . . . . . . . . . 49 152 11.5.3. Macro Expansion . . . . . . . . . . . . . . . . . . . 49 153 11.6. Privacy Exposure . . . . . . . . . . . . . . . . . . . . . 49 154 11.7. Delivering Mail Producing a 'Fail' Result . . . . . . . . 49 155 12. Contributors and Acknowledgements . . . . . . . . . . . . . . 51 156 13. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 52 157 13.1. The SPF DNS Record Type . . . . . . . . . . . . . . . . . 52 158 13.2. The Received-SPF Mail Header Field . . . . . . . . . . . . 52 159 13.3. SPF Modifier Registration . . . . . . . . . . . . . . . . 52 160 14. References . . . . . . . . . . . . . . . . . . . . . . . . . . 53 161 14.1. Normative References . . . . . . . . . . . . . . . . . . . 53 162 14.2. Informative References . . . . . . . . . . . . . . . . . . 54 163 Appendix A. Collected ABNF . . . . . . . . . . . . . . . . . . . 56 164 Appendix B. Extended Examples . . . . . . . . . . . . . . . . . . 59 165 B.1. Simple Examples . . . . . . . . . . . . . . . . . . . . . 59 166 B.2. Multiple Domain Example . . . . . . . . . . . . . . . . . 60 167 B.3. DNSBL Style Example . . . . . . . . . . . . . . . . . . . 61 168 B.4. Multiple Requirements Example . . . . . . . . . . . . . . 61 169 Appendix C. Further Testing Advice . . . . . . . . . . . . . . . 62 170 Appendix D. SPF/Mediator Interactions . . . . . . . . . . . . . . 63 171 Appendix E. Mail Services . . . . . . . . . . . . . . . . . . . . 65 172 Appendix F. MTA Relays . . . . . . . . . . . . . . . . . . . . . 66 173 Appendix G. Local Policy Considerations . . . . . . . . . . . . . 67 174 G.1. Policy For SPF Pass . . . . . . . . . . . . . . . . . . . 67 175 G.2. Policy For SPF Fail . . . . . . . . . . . . . . . . . . . 67 176 G.3. Policy For SPF Permerror . . . . . . . . . . . . . . . . . 68 177 Appendix H. Protocol Status . . . . . . . . . . . . . . . . . . . 69 178 Appendix I. Change History . . . . . . . . . . . . . . . . . . . 70 179 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . . 73 181 1. Introduction 183 The current email infrastructure has the property that any host 184 injecting mail into the system can use any DNS domain name it wants 185 in each of the various identifiers specified by [RFC5321] and 186 [RFC5322]. Although this feature is desirable in some circumstances, 187 it is a major obstacle to reducing Unsolicited Bulk Email (UBE, aka 188 spam). Furthermore, many domain owning ADMDs (ADministrative 189 Management Domains, see [RFC5598]) are understandably concerned about 190 the ease with which other entities can make use of their domain 191 names, often with malicious intent. 193 This document defines a protocol by which ADMDs can authorize hosts 194 to use their domain names in the "MAIL FROM" or "HELO" identities. 195 Compliant ADMDs publish Sender Policy Framework (SPF) records in the 196 DNS specifying which hosts are permitted to use their names, and 197 compliant mail receivers use the published SPF records to test the 198 authorization of sending Mail Transfer Agents (MTAs) using a given 199 "HELO" or "MAIL FROM" identity during a mail transaction. 201 An additional benefit to mail receivers is that after the use of an 202 identity is verified, local policy decisions about the mail can be 203 made based on the sender's domain, rather than the host's IP address. 204 This is advantageous because reputation of domain names is likely to 205 be more accurate than reputation of host IP addresses. Furthermore, 206 if a claimed identity fails verification, local policy can take 207 stronger action against such email, such as rejecting it. 209 1.1. Terminology 211 1.1.1. Keywords 213 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 214 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 215 "OPTIONAL" in this document are to be interpreted as described in 216 [RFC2119]. 218 1.1.2. Imported Definitions 220 The ABNF tokens "ALPHA", "DIGIT", and "SP" are defined in [RFC5234]. 222 The token "local-part" is defined in [RFC5321]. 224 "dot-atom", "quoted-string", "comment", "CFWS", "FWS", and "CRLF" are 225 defined in [RFC5322]. 227 1.1.3. MAIL FROM Definition 229 This document is concerned with the portion of a mail message 230 commonly called "envelope sender", "return path", "reverse path", 231 "bounce address", "5321 FROM", "MAIL FROM", or RFC5321.MailFrom. 232 Since these terms are either not well defined or often used casually, 233 this document uses "MAIL FROM" for consistency. This means the 234 RFC5321.MailFrom as defined in [RFC5598]. Note that other terms that 235 might superficially look like the common terms, such as "reverse- 236 path", are used only with the defined meanings from normative 237 documents. 239 1.1.4. HELO Definition 241 This document also makes use of the HELO/EHLO identity. The "HELO" 242 identity derives from either the SMTP HELO or EHLO command (see 243 [RFC5321]). Since HELO and EHLO can, in many cases, be used 244 interchangeably, they are identified commonly as "HELO" in this 245 document. This means RFC5321.HELO/.EHLO as defined in [RFC5598]. 246 These commands supply the identity of the SMTP client (sending host) 247 for the SMTP session. 249 1.2. check_host() 251 Section 4 introduces an algorithm to evaluate an SPF policy against 252 an arriving email transaction. In an early implementation, this 253 algorithm was encoded in a function called check_host(). That name 254 is used in this document as symbolic of the SPF evaluation algorithm, 255 but of course implementers are not required to use this name. 257 2. Operational Overview 259 2.1. The "HELO" Identity 261 It is RECOMMENDED that SPF verifiers not only check the "MAIL FROM" 262 identity, but also separately check the "HELO" identity by applying 263 the check_host() function (Section 4) to the "HELO" identity as the 264 . Checking "HELO" promotes consistency of results and can 265 reduce DNS resource usage. Additionally, since SPF records published 266 for "HELO" identities refer to a single host, when available, they 267 are a very reliable source of host authorization status. 269 Note that requirements for the domain presented in the EHLO or HELO 270 command are not always clear to the sending party, and SPF verifiers 271 MUST be prepared for the "HELO" identity to be malformed or an IP 272 address literal. This SPF check can only be performed when the 273 "HELO" string is a valid fully qualified domain. 275 2.2. The "MAIL FROM" Identity 277 SPF verifiers MUST check the ""MAIL FROM" identity if a completed 278 "HELO" check has not reached a definitive policy result by applying 279 the check_host() function to the "MAIL FROM" identity as the 280 . 282 [RFC5321] allows the reverse-path to be null (see Section 4.5.5 in 283 [RFC5321]). In this case, there is no explicit sender mailbox, and 284 such a message can be assumed to be a notification message from the 285 mail system itself. When the reverse-path is null, this document 286 defines the "MAIL FROM" identity to be the mailbox composed of the 287 local-part "postmaster" and the "HELO" identity (which might or might 288 not have been checked separately before). 290 2.3. Publishing Authorization 292 An SPF-compliant domain MUST have valid SPF records as described in 293 Section 3. These records authorize the use of the relevant domain 294 names in the "HELO" and "MAIL FROM" identities by the MTAs specified 295 therein. 297 SPF results can be used to make both positive (source is authorized) 298 and negative (source is not authorized) determinations. If domain 299 owners choose to publish SPF records and want to support receivers 300 making negative authorization determinations, then they MUST publish 301 records that end in "-all", or redirect to other records that do, 302 otherwise, no definitive determination of authorization can be made. 303 Potential issues and mitigations associated with negative 304 determinations are discussed in Section 10. 306 ADMDs can publish SPF records that explicitly authorize no hosts for 307 domain names that are neither used in the domain part of email 308 addresses nor expected to originate mail. 310 When changing SPF records, care has to be taken to ensure that there 311 is a transition period so that the old policy remains valid until all 312 legitimate email can reasonably expect to have been checked. 313 [RFC5321] Section 4.5.4.1 discusses how long a message might be in 314 transit. While offline checks are possible, the closer to the 315 original transmission time checks are performed, the more likely they 316 are to get an SPF result that matches the sending ADMD intent at the 317 time the message was sent. 319 2.4. Checking Authorization 321 A mail receiver can perform a set of SPF checks for each mail message 322 it receives. An SPF check tests the authorization of a client host 323 to emit mail with a given identity. Typically, such checks are done 324 by a receiving MTA, but can be performed elsewhere in the mail 325 processing chain so long as the required information is available and 326 reliable. At least the "MAIL FROM" identity MUST be checked, but it 327 is RECOMMENDED that the "HELO" identity also be checked beforehand. 329 Without explicit approval of the domain owner, checking other 330 identities against SPF version 1 records is NOT RECOMMENDED because 331 there are cases that are known to give incorrect results. For 332 example, almost all mailing lists rewrite the "MAIL FROM" identity 333 (see Section 10.3), but some do not change any other identities in 334 the message. Documents that define other identities will have to 335 define the method for explicit approval. 337 It is possible that mail receivers will use the SPF check as part of 338 a larger set of tests on incoming mail. The results of other tests 339 might influence whether or not a particular SPF check is performed. 340 For example, finding the sending host's IP address on a local white 341 list might cause all other tests to be skipped and all mail from that 342 host to be accepted. 344 When a mail receiver decides to perform an SPF check, it MUST use a 345 correctly-implemented check_host() function (Section 4) evaluated 346 with the correct parameters. Although the test as a whole is 347 optional, once it has been decided to perform a test it has to be 348 performed as specified so that the correct semantics are preserved 349 between publisher and receiver. 351 To make the test, the mail receiver MUST evaluate the check_host() 352 function with the arguments set as follows: 354 - the IP address of the SMTP client that is emitting the 355 mail, either IPv4 or IPv6. 357 - the domain portion of the "MAIL FROM" or "HELO" identity. 359 - the "MAIL FROM" or "HELO" identity. 361 Although invalid, malformed, or non-existent domains cause SPF checks 362 to return "none" because no SPF record can be found, it has long been 363 the policy of many MTAs to reject email from such domains, especially 364 in the case of invalid "MAIL FROM". Rejecting email will prevent one 365 method of circumventing of SPF records. 367 Implementations have to take care to correctly extract the 368 from the data given with the SMTP MAIL FROM command as many MTAs will 369 still accept such things as source routes (see [RFC5321], Appendix 370 C), the %-hack (see [RFC1123]), and bang paths (see [RFC1983]). 371 These archaic features have been maliciously used to bypass security 372 systems. 374 2.5. Location of Checks 376 The authorization check SHOULD be performed during the processing of 377 the SMTP transaction that sends the mail. This reduces the 378 complexity of determining the correct IP address to use as an input 379 to check_host() and allows errors to be returned directly to the 380 sending MTA by way of SMTP replies. 382 Performing the authorization other than using the return-path and 383 client address at the time of the MAIL command during the SMTP 384 transaction can cause problems, such as the following: (1) It might 385 be difficult to accurately extract the required information from 386 potentially deceptive headers; (2) legitimate email might fail 387 because the sender's policy had since changed. 389 Generating non-delivery notifications to forged identities that have 390 failed the authorization check is a source of backscatter and SHOULD 391 be avoided. Section 2 of [RFC3834] describes backscatter and the 392 problems it causes. 394 2.6. Results of Evaluation 396 Section 4 defines check_host(), a model function definition that uses 397 the inputs defined above and the sender's policy published in the DNS 398 to reach a conclusion about client authorization. An SPF verifier 399 implements something semantically identical to the function defined 400 there. 402 This section enumerates and briefly defines the possible outputs of 403 that function. Information about how to handle these outputs is in 404 Section 8. 406 2.6.1. None 408 A result of "none" means either (a) no syntactically valid DNS domain 409 name was extracted from the SMTP session that could be used as the 410 one to be authorized, or (b) no TXT records were retrieved from the 411 DNS that appeared to be intended for use by SPF verifiers. 413 2.6.2. Neutral 415 The domain owner has explicitly stated that it is not asserting 416 whether the IP address is authorized. This result MUST be treated 417 exactly like the "none" result; the distinction exists only for 418 informational purposes. 420 2.6.3. Pass 422 A "pass" result means that the client is authorized to inject mail 423 with the given identity. The domain can now, in the sense of 424 reputation, be considered responsible for sending the message. 425 Further policy checks can now proceed with confidence in the 426 legitimate use of the identity. This is further discussed in 427 Appendix G.1. 429 2.6.4. Fail 431 A "fail" result is an explicit statement that the client is not 432 authorized to use the domain in the given identity. 434 2.6.5. Softfail 436 The domain owner has published a weak statement that the host is 437 probably not authorized. It has not published a stronger, more 438 definitive policy that results in a "fail" 440 2.6.6. Temperror 442 A "temperror" result means the SPF verifier encountered a transient 443 (generally DNS) error while performing the check. 445 2.6.7. Permerror 447 A "permerror" result means the domain's published records could not 448 be correctly interpreted. This signals an error condition that 449 definitely requires manual intervention to be resolved. 451 3. SPF Records 453 An SPF record is a DNS record that declares which hosts are, and are 454 not, authorized to use a domain name for the "HELO" and "MAIL FROM" 455 identities. Loosely, the record partitions all hosts into permitted 456 and not-permitted sets (though some hosts might fall into neither 457 category). 459 The SPF record is a single string of text. The record format is 460 described below in Section 4. An example record is the following: 462 v=spf1 +mx a:colo.example.com/28 -all 464 This record has a version of "spf1" and three directives: "+mx", 465 "a:colo.example.com/28" (the + is implied), and "-all". 467 Each SPF record is placed in the DNS tree at the host name it 468 pertains to, not a subdomain under it, such as is done with SRV 469 records [RFC2782]. 471 The example in this section might be published via these lines in a 472 domain zone file: 474 example.com. TXT "v=spf1 +mx a:colo.example.com/28 -all" 475 smtp-out.example.com. TXT "v=spf1 a -all" 477 Since TXT records have multiple uses, beware of other TXT records 478 published there for other purposes. They might cause problems with 479 size limits (see Section 3.4) and care has to be taken to ensure only 480 SPF records are used for SPF processing. 482 ADMDs publishing SPF records SHOULD try to keep the number of 483 "include" mechanisms and chained "redirect" modifiers to a minimum. 484 ADMDs SHOULD also try to minimize the amount of other DNS information 485 needed to evaluate a record. Section 4.6.4 and Section 10.1.1 486 provide some suggestions on how to achieve this. 488 3.1. DNS Resource Records 490 SPF records MUST be published as a DNS TXT (type 16) Resource Record 491 (RR) [RFC1035] only. The character content of the record is encoded 492 as [US-ASCII]. Use of alternate DNS RR types was supported in SPF's 493 experimental phase, but has been discontinued. See Appendix A of 494 [RFC6686] for further information. 496 3.2. Multiple DNS Records 498 A domain name MUST NOT have multiple records that would cause an 499 authorization check to select more than one record. See Section 4.5 500 for the selection rules. 502 3.3. Multiple Strings in a Single DNS record 504 As defined in [RFC1035] sections 3.3.14 and 3.3, a single text DNS 505 record can be composed of more than one string. If a published 506 record contains multiple character-strings, then the record MUST be 507 treated as if those strings are concatenated together without adding 508 spaces. For example: 510 IN TXT "v=spf1 .... first" "second string..." 512 MUST be treated as equivalent to: 514 IN TXT "v=spf1 .... firstsecond string..." 516 TXT records containing multiple strings are useful in constructing 517 records that would exceed the 255-byte maximum length of a character- 518 string within a single TXT record. 520 3.4. Record Size 522 The published SPF record for a given domain name SHOULD remain small 523 enough that the results of a query for it will fit within 512 octets. 524 This UDP limit is defined in [RFC1035] section 2.3.4. This will keep 525 even older DNS implementations from falling over to TCP. Since the 526 answer size is dependent on many things outside the scope of this 527 document, it is only possible to give this guideline: If the combined 528 length of the DNS name and the text of all the records of a given 529 type is under 450 octets, then DNS answers ought to fit in UDP 530 packets. Records that are too long to fit in a single UDP packet 531 could be silently ignored by SPF verifiers due to firewall and other 532 issues that cause DNS over TCP to be less reliable than DNS over UDP. 534 Note that when computing the sizes for replies to queries of the TXT 535 format, one has to take into account any other TXT records published 536 at the domain name. Similarly, the sizes for replies to all queries 537 related to SPF have to be evaluated to fit in a single UDP packet. 539 3.5. Wildcard Records 541 Use of wildcard records for publishing is discouraged and care has to 542 be taken if they are used. If a zone includes wildcard MX records, 543 it might want to publish wildcard declarations, subject to the same 544 requirements and problems. In particular, the declaration MUST be 545 repeated for any host that has any RR records at all, and for 546 subdomains thereof. Consider the example in [RFC1034], Section 547 4.3.3. Based on that, we can do the following: 549 EXAMPLE.COM. MX 10 A.EXAMPLE.COM 550 EXAMPLE.COM. TXT "v=spf1 a:A.EXAMPLE.COM -all" 552 *.EXAMPLE.COM. MX 10 A.EXAMPLE.COM 553 *.EXAMPLE.COM. TXT "v=spf1 a:A.EXAMPLE.COM -all" 555 A.EXAMPLE.COM. A 203.0.113.1 556 A.EXAMPLE.COM. MX 10 A.EXAMPLE.COM 557 A.EXAMPLE.COM. TXT "v=spf1 a:A.EXAMPLE.COM -all" 559 *.A.EXAMPLE.COM. MX 10 A.EXAMPLE.COM 560 *.A.EXAMPLE.COM. TXT "v=spf1 a:A.EXAMPLE.COM -all" 562 SPF records have to be listed twice for every name within the zone: 563 once for the name, and once with a wildcard to cover the tree under 564 the name, in order to cover all domains in use in outgoing mail. 566 4. The check_host() Function 568 This description is not an API (Application Program Interface) 569 definition, but rather a function description used to illustrate the 570 algorithm. A compliant SPF implementation MUST do something 571 semantically equivalent to this description. 573 The check_host() function fetches SPF records, parses them, and 574 evaluates them to determine whether a particular host is or is not 575 permitted to send mail with a given identity. Mail receivers that 576 perform this check MUST correctly evaluate the check_host() function 577 as described here. 579 Implementations MAY use a different algorithm than the canonical 580 algorithm defined here, so long as the results are the same in all 581 cases. 583 4.1. Arguments 585 The check_host() function takes these arguments: 587 - the IP address of the SMTP client that is emitting the 588 mail, either IPv4 or IPv6. 590 - the domain that provides the sought-after authorization 591 information; initially, the domain portion of the "MAIL 592 FROM" or "HELO" identity. 594 - the "MAIL FROM" or "HELO" identity. 596 For recursive evaluations, the domain portion of might not 597 be the same as the argument when check_host() is initially 598 evaluated. In most other cases it will be the same. (See 599 Section 5.2 below). 601 Note that the argument might not be a well-formed domain 602 name. For example, if the reverse-path was null, then the EHLO/HELO 603 domain is used, with its associated problems (see Section 2.1). In 604 these cases, check_host() is defined in Section 4.3 to return a 605 "none" result. 607 4.2. Results 609 The function check_host() can return one of several results described 610 in Section 2.6. Based on the result, the action to be taken is 611 determined by the local policies of the receiver. This is discussed 612 in Section 8. 614 4.3. Initial Processing 616 If the is malformed (e.g. label longer than 63 characters, 617 zero-length label not at the end, etc.) or is not a fully qualified 618 domain name, or if the DNS lookup returns "domain does not exist" 619 (RCODE 3), check_host() immediately returns the result "none". 620 Properly formed domains are fully qualified email domains as 621 described in [RFC5321] Section 2.3.5. Internationalized domain names 622 MUST be encoded as A-labels, as described in Section 2.3 of 623 [RFC5890].on 2.3 of [RFC5890]. 625 If the has no local-part, substitute the string "postmaster" 626 for the local-part. 628 4.4. Record Lookup 630 In accordance with how the records are published (see Section 3 631 above), a DNS query needs to be made for the name, querying 632 for type TXT only. 634 If all DNS lookups that are made return a server failure (RCODE 2), 635 or other error (RCODE other than 0 or 3), or time out, then 636 check_host() terminates immediately with the result "temperror". 637 Alternatively, for a server failure (RCODE 2) result, check_host() 638 MAY track failures and treat multiple failures within 24 hours for 639 the same domain as "permerror". 641 This alternative is intended to shorten the queue time of messages 642 that cannot be accepted, by returning a permanent negative completion 643 reply code to the client, instead of a transient one. [RFC2308] 644 suggests on an algorithm for doing such tracking and handling of 645 server failure codes. 647 4.5. Selecting Records 649 Records begin with a version section: 651 record = version terms *SP 652 version = "v=spf1" 654 Starting with the set of records that were returned by the lookup, 655 discard records that do not begin with a version section of exactly 656 "v=spf1". Note that the version section is terminated either by an 657 SP character or the end of the record. A record with a version 658 section of "v=spf10" does not match and is discarded. 660 If the resultant record set includes no records, check_host() 661 produces the "none" result. If the resultant record set includes 662 more than one record, check_host() produces the "permerror" result. 664 4.6. Record Evaluation 666 The check_host() function parses and interprets the SPF record to 667 find a result for the current test. If there are any syntax errors 668 anywhere in the record, check_host() returns immediately with the 669 result "permerror", without further interpretation. 671 4.6.1. Term Evaluation 673 There are two types of terms: mechanisms and modifiers. A record 674 contains an ordered list of these as specified in the following 675 Augmented Backus-Naur Form (ABNF). 677 terms = *( 1*SP ( directive / modifier ) ) 679 directive = [ qualifier ] mechanism 680 qualifier = "+" / "-" / "?" / "~" 681 mechanism = ( all / include 682 / A / MX / PTR / IP4 / IP6 / exists ) 683 modifier = redirect / explanation / unknown-modifier 684 unknown-modifier = name "=" macro-string 685 ; where name is not any known modifier 687 name = ALPHA *( ALPHA / DIGIT / "-" / "_" / "." ) 689 Most mechanisms allow a ":" or "/" character after the name. 691 Modifiers always contain an equals ('=') character immediately after 692 the name, and before any ":" or "/" characters that might be part of 693 the macro-string. 695 Terms that do not contain any of "=", ":", or "/" are mechanisms, as 696 defined in Section 5. 698 As per the definition of the ABNF notation in [RFC5234], mechanism 699 and modifier names are case-insensitive. 701 4.6.2. Mechanisms 703 Each mechanism is considered in turn from left to right. If there 704 are no more mechanisms, the result is specified in Section 4.7. 706 When a mechanism is evaluated, one of three things can happen: it can 707 match, not match, or return an exception. 709 If it matches, processing ends and the qualifier value is returned as 710 the result of that record. If it does not match, processing 711 continues with the next mechanism. If it returns an exception, 712 mechanism processing ends and the exception value is returned. 714 The possible qualifiers, and the results they cause check_host() to 715 return are as follows: 717 "+" pass 718 "-" fail 719 "~" softfail 720 "?" neutral 722 The qualifier is optional and defaults to "+". 724 When a mechanism matches and the qualifier is "-", then a "fail" 725 result is returned and the explanation string is computed as 726 described in Section 6.2. 728 The specific mechanisms are described in Section 5. 730 4.6.3. Modifiers 732 Modifiers are not mechanisms. They do not return match or not-match. 733 Instead, they provide additional information. Although modifiers do 734 not directly affect the evaluation of the record, the "redirect" 735 modifier has an effect after all the mechanisms have been evaluated. 737 4.6.4. DNS Lookup Limits 739 SPF implementations MUST limit the number of mechanisms and modifiers 740 ("terms") that cause any DNS query to at most 10 during SPF 741 evaluation. Specifically, the "include", "a", "mx", "ptr", and 742 "exists" mechanisms as well as the "redirect" modifier count against 743 this limit. The "all", "ip4", and "ip6" mechanisms do not count 744 against this limit. If this number is exceeded during a check, a 745 permerror MUST be returned. The "exp" modifier does not count 746 against this limit because the DNS lookup to fetch the explanation 747 string occurs after the SPF record evaluation has been completed. 749 When evaluating the "mx" mechanism, the number of "MX" resource 750 records queried is included in the overall limit of 10 mechanisms/ 751 modifiers that cause DNS look ups described above. The evaluation of 752 each "MX" record MUST NOT result in querying more than 10 "A" 753 resource records. If this limit is exceeded, the "mx" mechanism MUST 754 produce a "permerror" result. 756 When evaluating the "ptr" mechanism or the %{p} macro, the number of 757 "PTR" resource records queried is included in the overall limit of 10 758 mechanisms/modifiers that cause DNS look ups described above. The 759 evaluation of each "PTR" record MUST NOT result in querying more than 760 10 "A" resource records. If this limit is exceeded, all records 761 other than the first 10 MUST be ignored. 763 The reason for the disparity is that the set of and contents of the 764 MX record are under control of the domain owner, while the set of and 765 contents of PTR records are under control of the owner of the IP 766 address actually making the connection. 768 These limits are per mechanism or macro in the record, and are in 769 addition to the lookup limits specified above. 771 MTAs or other processors SHOULD impose a limit on the maximum amount 772 of elapsed time to evaluate check_host(). Such a limit SHOULD allow 773 at least 20 seconds. If such a limit is exceeded, the result of 774 authorization SHOULD be "temperror". 776 4.7. Default Result 778 If none of the mechanisms match and there is no "redirect" modifier, 779 then the check_host() returns a result of "neutral", just as if 780 "?all" were specified as the last directive. If there is a 781 "redirect" modifier, check_host() proceeds as defined in Section 6.1. 783 Note that records SHOULD always use either a "redirect" modifier or 784 an "all" mechanism to explicitly terminate processing. Although the 785 latter has default (specifically "?all"), it aids debugging efforts 786 if it is explicitly included. 788 For example: 790 v=spf1 +mx -all 791 or 792 v=spf1 +mx redirect=_spf.example.com 794 4.8. Domain Specification 796 Several of these mechanisms and modifiers have a domain-spec section. 797 The domain-spec string is subject to macro expansion (see Section 7). 798 The resulting string is the common presentation form of a fully- 799 qualified DNS name: a series of labels separated by periods. This 800 domain is called the in the rest of this document. 802 Note: The result of the macro expansion is not subject to any further 803 escaping. Hence, this facility cannot produce all characters that 804 are legal in a DNS label (e.g., the control characters). However, 805 this facility is powerful enough to express legal host names and 806 common utility labels (such as "_spf") that are used in DNS. 808 For several mechanisms, the domain-spec is optional. If it is not 809 provided, the from the check_host() arguments (see 810 Section 4.1) is used as the . Domain and domain-spec 811 are syntactically identical after macro expansion. Domain is an 812 input value for check_host() while domain-spec is computed by 813 check_host(). 815 Note: Historically, this document has made no provisions for how to 816 handle domain-specs, or macro-expansions thereof, that are 817 syntactically invalid per [RFC1035], such as names with empty labels 818 (e.g., "foo..example.com") or overlong labels (more than 63 819 characters). Some implementations choose to treat as a no-match 820 mechanisms, and ignore modifiers with such names, whereas others 821 return a "permerror" exception. The outcome for an unexpected 822 domain-spec without macros might even differ from that for an 823 unexpected after macro expansion. 825 5. Mechanism Definitions 827 This section defines two types of mechanisms. 829 Basic mechanisms contribute to the language framework. They do not 830 specify a particular type of authorization scheme. 832 all 833 include 835 Designated sender mechanisms are used to designate a set of 836 addresses as being permitted or not permitted to use the for 837 sending mail. 839 a 840 mx 841 ptr (do not use) 842 ip4 843 ip6 844 exists 846 The following conventions apply to all mechanisms that perform a 847 comparison between and an IP address at any point: 849 If no CIDR prefix length is given in the directive, then and the 850 IP address are compared for equality. (Here, CIDR is Classless 851 Inter-Domain Routing, described in [RFC4632].) 853 If a CIDR prefix length is specified, then only the specified number 854 of high-order bits of and the IP address are compared for 855 equality. 857 When any mechanism fetches host addresses to compare with , when 858 is an IPv4, "A" records are fetched; when is an IPv6 859 address, "AAAA" records are fetched. SPF implementations on IPv6 860 servers need to handle both "AAAA" and "A" secords, for clients on 861 IPv4 mapped IPv6 addresses [RFC4291]. IPv4 addresses are only 862 listed in an SPF record using the "ip4" mechanism. 864 Several mechanisms rely on information fetched from the DNS. For 865 these DNS queries, except where noted, if the DNS server returns an 866 error (RCODE other than 0 or 3) or the query times out, the mechanism 867 stops and the topmost check_host() returns "temperror". If the 868 server returns "domain does not exist" (RCODE 3), then evaluation of 869 the mechanism continues as if the server returned no error (RCODE 0) 870 and zero answer records. 872 5.1. "all" 874 all = "all" 876 The "all" mechanism is a test that always matches. It is used as the 877 rightmost mechanism in a record to provide an explicit default. 879 For example: 881 v=spf1 a mx -all 883 Mechanisms after "all" will never be tested. Mechanisms listed after 884 "all" MUST be ignored. Any "redirect" modifier (Section 6.1) MUST be 885 ignored when there is an "all" mechanism in the record. 887 5.2. "include" 889 include = "include" ":" domain-spec 891 The "include" mechanism triggers a recursive evaluation of 892 check_host(). 894 1. The domain-spec is expanded as per Section 7. 896 2. Check_host() is evaluated with the resulting string as the 897 . The and arguments remain the same as in 898 the current evaluation of check_host(). 900 3. The recursive evaluation returns either match, not match, or an 901 error. If it matches, then the appropriate result for the 902 include: mechanism is used (e.g. include or +include gives a 903 "pass" result and -include gives "fail). 905 4. If there is no match, the parent check_host() resumes processing 906 as per the table below, with the previous value of 907 restored. 909 In hindsight, the name "include" was poorly chosen. Only the 910 evaluated result of the referenced SPF record is used, rather than 911 acting as if the referenced SPF record was literally included in the 912 first. For example, evaluating a "-all" directive in the referenced 913 record does not terminate the overall processing and does not 914 necessarily result in an overall "fail". (Better names for this 915 mechanism would have been "if-match", "on-match", etc.) 917 The "include" mechanism makes it possible for one domain to designate 918 multiple administratively-independent domains. For example, a vanity 919 domain "example.net" might send mail using the servers of 920 administratively-independent domains example.com and example.org. 922 Example.net could say 924 IN TXT "v=spf1 include:example.com include:example.org -all" 926 This would direct check_host() to, in effect, check the records of 927 example.com and example.org for a "pass" result. Only if the host 928 were not permitted for either of those domains would the result be 929 "fail". 931 Whether this mechanism matches, does not match, or returns an 932 exception depends on the result of the recursive evaluation of 933 check_host(): 935 +---------------------------------+---------------------------------+ 936 | A recursive check_host() result | Causes the "include" mechanism | 937 | of: | to: | 938 +---------------------------------+---------------------------------+ 939 | pass | match | 940 | | | 941 | fail | not match | 942 | | | 943 | softfail | not match | 944 | | | 945 | neutral | not match | 946 | | | 947 | temperror | return temperror | 948 | | | 949 | permerror | return permerror | 950 | | | 951 | none | return permerror | 952 +---------------------------------+---------------------------------+ 954 The "include" mechanism is intended for crossing administrative 955 boundaries. For example, if example.com and example.org were managed 956 by the same entity, and if the permitted set of hosts for both 957 domains was 958 "mx:example.com", it would be possible for example.org to specify 959 "include:example.com", but it would be preferable to specify 960 "redirect=example.com" or even "mx:example.com". 962 With the "include" mechanism an administratively external set of 963 hosts can be authorized, but determination of sender policy is still 964 a function of the original domain's SPF record (as determined by the 965 "all" mechanism in that record). The redirect modifier is more 966 suitable for consolidating both authorizations and policy into a 967 common set to be shared within an ADMD. Redirect is much more like a 968 common code element to be shared among records in a single ADMD. It 969 is possible to control both authorized hosts and policy for an 970 arbitrary number of domains from a single record. 972 5.3. "a" 974 This mechanism matches if is one of the 's IP 975 addresses. 977 a = "a" [ ":" domain-spec ] [ dual-cidr-length ] 979 An address lookup is done on the . The is compared 980 to the returned address(es). If any address matches, the mechanism 981 matches. 983 5.4. "mx" 985 This mechanism matches if is one of the MX hosts for a domain 986 name. 988 mx = "mx" [ ":" domain-spec ] [ dual-cidr-length ] 990 check_host() first performs an MX lookup on the . Then 991 it performs an address lookup on each MX name returned. The is 992 compared to each returned IP address. To prevent Denial of Service 993 (DoS) attacks, more than 10 MX names MUST NOT be looked up during the 994 evaluation of an "mx" mechanism. If there are more than 10 MX names 995 then permerror is returned and the evaluation terminated (see 996 Section 4.6.4). If any address matches, the mechanism matches. 998 Note regarding implicit MXs: If the has no MX records, 999 check_host() MUST NOT pretend the target is its single MX, and MUST 1000 NOT default to an A or AAAA lookup on the directly. 1001 This behavior diverges from the legacy "implicit MX" rule, (See 1002 [RFC5321], Section 5. If such behavior is desired, the publisher 1003 will have to specify an "a" directive). 1005 5.5. "ptr" (do not use) 1007 This mechanism tests whether the DNS reverse-mapping for exists 1008 and correctly points to a domain name within a particular domain. 1009 This mechanism SHOULD NOT be used. See below for discussion. 1011 ptr = "ptr" [ ":" domain-spec ] 1013 The 's name is looked up using this procedure: 1015 o Perform a DNS reverse-mapping for : Look up the corresponding 1016 PTR record in "in-addr.arpa." if the address is an IPv4 one and in 1017 "ip6.arpa." if it is an IPv6 address. 1019 o For each record returned, validate the domain name by looking up 1020 its IP addresses. To prevent DoS attacks, more than 10 PTR names 1021 MUST NOT be looked up during the evaluation of a "ptr" mechanism 1022 (see Section 4.6.4). 1024 o If is among the returned IP addresses, then that domain name 1025 is validated. 1027 Check all validated domain names to see if they either match the 1028 domain or are a subdomain of the domain. 1029 If any do, this mechanism matches. If no validated domain name can 1030 be found, or if none of the validated domain names match or are a 1031 subdomain of the , this mechanism fails to match. If a 1032 DNS error occurs while doing the PTR RR lookup, then this mechanism 1033 fails to match. If a DNS error occurs while doing an A RR lookup, 1034 then that domain name is skipped and the search continues. 1036 Pseudocode: 1038 sending-domain_names := ptr_lookup(sending-host_IP); 1039 if more than 10 sending-domain_names are found, use at most 10. 1040 for each name in (sending-domain_names) { 1041 IP_addresses := a_lookup(name); 1042 if the sending-domain_IP is one of the IP_addresses { 1043 validated-sending-domain_names += name; 1044 } 1045 } 1047 for each name in (validated-sending-domain_names) { 1048 if name ends in , return match. 1049 if name is , return match. 1050 } 1051 return no-match. 1053 This mechanism matches if the is either a subdomain of 1054 a validated domain name or if the and a validated 1055 domain name are the same. For example: "mail.example.com" is within 1056 the domain "example.com", but "mail.bad-example.com" is not. 1058 Note: This mechanism is slow, it is not as reliable as other 1059 mechanisms in cases of DNS errors, and it places a large burden on 1060 the .arpa name servers. If used, proper PTR records MUST be in place 1061 for the domain's hosts and the "ptr" mechanism SHOULD be one of the 1062 last mechanisms checked. After many years of SPF deployment 1063 experience it has been concluded it is unnecessary and more reliable 1064 alternatives used instead. It is, however, still in use and part of 1065 the SPF protocol, so compliant check_host() implementations MUST 1066 support it. 1068 5.6. "ip4" and "ip6" 1070 These mechanisms test whether is contained within a given IP 1071 network. 1073 ip4 = "ip4" ":" ip4-network [ ip4-cidr-length ] 1074 ip6 = "ip6" ":" ip6-network [ ip6-cidr-length ] 1076 ip4-cidr-length = "/" 1*DIGIT 1077 ip6-cidr-length = "/" 1*DIGIT 1078 dual-cidr-length = [ ip4-cidr-length ] [ "/" ip6-cidr-length ] 1080 ip4-network = qnum "." qnum "." qnum "." qnum 1081 qnum = DIGIT ; 0-9 1082 / %x31-39 DIGIT ; 10-99 1083 / "1" 2DIGIT ; 100-199 1084 / "2" %x30-34 DIGIT ; 200-249 1085 / "25" %x30-35 ; 250-255 1086 ; as per conventional dotted quad notation. e.g., 192.0.2.0 1087 ip6-network = 1088 ; e.g., 2001:DB8::CD30 1090 The is compared to the given network. If CIDR prefix length 1091 high-order bits match, the mechanism matches. 1093 If ip4-cidr-length is omitted, it is taken to be "/32". If 1094 ip6-cidr-length is omitted, it is taken to be "/128". It is not 1095 permitted to omit parts of the IP address instead of using CIDR 1096 notations. That is, use 192.0.2.0/24 instead of 192.0.2. 1098 5.7. "exists" 1100 This mechanism is used to construct an arbitrary domain name that is 1101 used for a DNS A record query. It allows for complicated schemes 1102 involving arbitrary parts of the mail envelope to determine what is 1103 permitted. 1105 exists = "exists" ":" domain-spec 1107 The domain-spec is expanded as per Section 7. The resulting domain 1108 name is used for a DNS A RR lookup (even when the connection type is 1109 IPv6). If any A record is returned, this mechanism matches. 1111 Domains can use this mechanism to specify arbitrarily complex 1112 queries. For example, suppose example.com publishes the record: 1114 v=spf1 exists:%{ir}.%{l1r+-}._spf.%{d} -all 1116 The might expand to 1117 "1.2.0.192.someuser._spf.example.com". This makes fine-grained 1118 decisions possible at the level of the user and client IP address. 1120 6. Modifier Definitions 1122 Modifiers are name/value pairs that provide additional information. 1123 Modifiers always have an "=" separating the name and the value. 1125 The modifiers defined in this document ("redirect" and "exp") MAY 1126 appear anywhere in the record, but SHOULD appear at the end, after 1127 all mechanisms. Ordering of these two modifiers does not matter. 1128 These two modifiers MUST NOT appear in a record more than once each. 1129 If they do, then check_host() exits with a result of "permerror". 1131 Unrecognized modifiers MUST be ignored no matter where in a record, 1132 or how often. This allows implementations of this document to 1133 gracefully handle records with modifiers that are defined in other 1134 specifications. 1136 6.1. redirect: Redirected Query 1138 The redirect modifier is intended for consolidating both 1139 authorizations and policy into a common set to be shared within a 1140 single ADMD. Redirect is like a common code element to be shared 1141 among records in a single ADMD. It is possible to control both 1142 authorized hosts and policy for an arbitrary number of domains from a 1143 single record. 1145 redirect = "redirect" "=" domain-spec 1147 If all mechanisms fail to match, and a "redirect" modifier is 1148 present, then processing proceeds as follows: 1150 The domain-spec portion of the redirect section is expanded as per 1151 the macro rules in Section 7. Then check_host() is evaluated with 1152 the resulting string as the . The and 1153 arguments remain the same as in the current evaluation of 1154 check_host(). 1156 The result of this new evaluation of check_host() is then considered 1157 the result of the current evaluation with the exception that if no 1158 SPF record is found, or if the is malformed, the result 1159 is a "permerror" rather than "none". 1161 Note that the newly-queried domain can itself specify redirect 1162 processing. 1164 This facility is intended for use by organizations that wish to apply 1165 the same record to multiple domains. For example: 1167 la.example.com. TXT "v=spf1 redirect=_spf.example.com" 1168 ny.example.com. TXT "v=spf1 redirect=_spf.example.com" 1169 sf.example.com. TXT "v=spf1 redirect=_spf.example.com" 1170 _spf.example.com. TXT "v=spf1 mx:example.com -all" 1172 In this example, mail from any of the three domains is described by 1173 the same record. This can be an administrative advantage. 1175 Note: In general, the domain "A" cannot reliably use a redirect to 1176 another domain "B" not under the same administrative control. Since 1177 the stays the same, there is no guarantee that the record at 1178 domain "B" will correctly work for mailboxes in domain "A", 1179 especially if domain "B" uses mechanisms involving local-parts. An 1180 "include" directive is generally be more appropriate. 1182 For clarity, it is RECOMMENDED that any "redirect" modifier appear as 1183 the very last term in a record. 1185 6.2. exp: Explanation 1187 explanation = "exp" "=" domain-spec 1189 If check_host() results in a "fail" due to a mechanism match (such as 1190 "-all"), and the "exp" modifier is present, then the explanation 1191 string returned is computed as described below. If no "exp" modifier 1192 is present, then either a default explanation string or an empty 1193 explanation string MUST be returned. 1195 The domain-spec is macro expanded (see Section 7) and becomes the 1196 . The DNS TXT record for the is fetched. 1198 If there are any DNS processing errors (any RCODE other than 0), or 1199 if no records are returned, or if more than one record is returned, 1200 or if there are syntax errors in the explanation string, then proceed 1201 as if no exp modifier was given. 1203 The fetched TXT record's strings are concatenated with no spaces, and 1204 then treated as an explain-string, which is macro-expanded. This 1205 final result is the explanation string. Implementations MAY limit 1206 the length of the resulting explanation string to allow for other 1207 protocol constraints and/or reasonable processing limits. Since the 1208 explanation string is intended for an SMTP response and [RFC5321] 1209 Section 2.4 says that responses are in [US-ASCII], the explanation 1210 string MUST be limited to US-ASCII. 1212 Software evaluating check_host() can use this string to communicate 1213 information from the publishing domain in the form of a short message 1214 or URL. Software SHOULD make it clear that the explanation string 1215 comes from a third party. For example, it can prepend the macro 1216 string "%{o} explains: " to the explanation, such as shown in 1217 Section 2.6.4. 1219 Suppose example.com has this record: 1221 v=spf1 mx -all exp=explain._spf.%{d} 1223 Here are some examples of possible explanation TXT records at 1224 explain._spf.example.com: 1226 "Mail from example.com should only be sent by its own servers." 1227 -- a simple, constant message 1229 "%{i} is not one of %{d}'s designated mail servers." 1230 -- a message with a little more information, including the IP 1231 address that failed the check 1233 "See http://%{d}/why.html?s=%{S}&i=%{I}" 1234 -- a complicated example that constructs a URL with the 1235 arguments to check_host() so that a web page can be 1236 generated with detailed, custom instructions 1238 Note: During recursion into an "include" mechanism, an exp= modifier 1239 from the MUST NOT be used. In contrast, when executing 1240 a "redirect" modifier, an exp= modifier from the original domain MUST 1241 NOT be used. 1243 7. Macros 1245 When evaluating an SPF policy record, certain character sequences are 1246 intended to be replaced by parameters of the message or of the 1247 connection. These character sequences are referred to as "macros". 1249 7.1. Formal Specification 1251 The ABNF description for a macro is as follows: 1253 domain-spec = macro-string domain-end 1254 domain-end = ( "." toplabel [ "." ] ) / macro-expand 1256 toplabel = ( *alphanum ALPHA *alphanum ) / 1257 ( 1*alphanum "-" *( alphanum / "-" ) alphanum ) 1258 alphanum = ALPHA / DIGIT 1260 explain-string = *( macro-string / SP ) 1262 macro-string = *( macro-expand / macro-literal ) 1263 macro-expand = ( "%{" macro-letter transformers *delimiter "}" ) 1264 / "%%" / "%_" / "%-" 1265 macro-literal = %x21-24 / %x26-7E 1266 ; visible characters except "%" 1267 macro-letter = "s" / "l" / "o" / "d" / "i" / "p" / "h" / 1268 "c" / "r" / "t" / "v" 1269 transformers = *DIGIT [ "r" ] 1270 delimiter = "." / "-" / "+" / "," / "/" / "_" / "=" 1272 The "toplabel" construction is subject to the LDH rule plus 1273 additional top-level domain (TLD) restrictions. See Section 2 of 1274 [RFC3696] for background. 1276 Some special cases: 1278 o A literal "%" is expressed by "%%". 1280 o "%_" expands to a single " " space. 1282 o "%-" expands to a URL-encoded space, viz., "%20". 1284 7.2. Macro Definitions 1286 The following macro letters are expanded in term arguments: 1288 s = 1289 l = local-part of 1290 o = domain of 1291 d = 1292 i = 1293 p = the validated domain name of (do not use) 1294 v = the string "in-addr" if is ipv4, or "ip6" if is ipv6 1295 h = HELO/EHLO domain 1297 , , and are defined in Section 2.4. 1299 The following macro letters are allowed only in "exp" text: 1301 c = SMTP client IP (easily readable format) 1302 r = domain name of host performing the check 1303 t = current timestamp 1305 7.3. Notes 1307 A '%' character not followed by a '{', '%', '-', or '_' character is 1308 a syntax error. So: 1310 -exists:%(ir).sbl.spamhaus.example.org 1312 is incorrect and will cause check_host() to yield a "permerror". 1313 Instead, the following is legal: 1315 -exists:%{ir}.sbl.spamhaus.example.org 1317 Optional transformers are the following: 1319 *DIGIT = zero or more digits 1320 r = reverse value, splitting on dots by default 1322 If transformers or delimiters are provided, the replacement value for 1323 a macro letter is split into parts separated by one or more of the 1324 specified delimiter characters. After performing any reversal 1325 operation and/or removal of left-hand parts, the parts are rejoined 1326 using "." and not the original splitting characters. 1328 By default, strings are split on "." (dots). Note that no special 1329 treatment is given to leading, trailing, or consecutive delimiters in 1330 input strings, and so the list of parts might contain empty strings. 1331 Some older implementations of SPF prohibit trailing dots in domain 1332 names, so trailing dots SHOULD NOT be published by domain owners, 1333 although they MUST be accepted by implementations conforming to this 1334 document. Macros can specify delimiter characters that are used 1335 instead of ".". 1337 The "r" transformer indicates a reversal operation: if the client IP 1338 address were 192.0.2.1, the macro %{i} would expand to "192.0.2.1" 1339 and the macro %{ir} would expand to "1.2.0.192". 1341 The DIGIT transformer indicates the number of right-hand parts to 1342 use, after optional reversal. If a DIGIT is specified, the value 1343 MUST be nonzero. If no DIGITs are specified, or if the value 1344 specifies more parts than are available, all the available parts are 1345 used. If the DIGIT was 5, and only 3 parts were available, the macro 1346 interpreter would pretend the DIGIT was 3. Implementations MUST 1347 support at least a value of 128, as that is the maximum number of 1348 labels in a domain name. 1350 The "s" macro expands to the argument. It is an email 1351 address with a local-part, an "@" character, and a domain. The "l" 1352 macro expands to just the local-part. The "o" macro expands to just 1353 the domain part. Note that these values remain the same during 1354 recursive and chained evaluations due to "include" and/or "redirect". 1355 Note also that if the original had no local-part, the local- 1356 part was set to "postmaster" in initial processing (see Section 4.3). 1358 For IPv4 addresses, both the "i" and "c" macros expand to the 1359 standard dotted-quad format. 1361 For IPv6 addresses, the "i" macro expands to a dot-format address; it 1362 is intended for use in %{ir}. The "c" macro can expand to any of the 1363 hexadecimal colon-format addresses specified in [RFC4291], Section 1364 2.2. It is intended for humans to read. 1366 The "p" macro expands to the validated domain name of . The 1367 procedure for finding the validated domain name is defined in 1368 Section 5.5. If the is present in the list of validated 1369 domains, it SHOULD be used. Otherwise, if a subdomain of the 1370 is present, it SHOULD be used. Otherwise, any name from the 1371 list can be used. If there are no validated domain names or if a DNS 1372 error occurs, the string "unknown" is used. This macro SHOULD NOT be 1373 used. See Section 5.5 for the discussion about why not. 1375 The "h" macro expands to the parameter that was provided to the SMTP 1376 server via the HELO or EHLO SMTP verb. For sessions where that verb 1377 was provide more than once, the most recent instance is used. 1379 The "r" macro expands to the name of the receiving MTA. This SHOULD 1380 be a fully qualified domain name, but if one does not exist (as when 1381 the checking is done by a MUA) or if policy restrictions dictate 1382 otherwise, the word "unknown" SHOULD be substituted. The domain name 1383 can be different from the name found in the MX record that the client 1384 MTA used to locate the receiving MTA. 1386 The "t" macro expands to the decimal representation of the 1387 approximate number of seconds since the Epoch (Midnight, January 1, 1388 1970, UTC) at the time of the evaluation. This is the same value as 1389 is returned by the POSIX time() function in most standards-compliant 1390 libraries. 1392 When the result of macro expansion is used in a domain name query, if 1393 the expanded domain name exceeds 253 characters (the maximum length 1394 of a domain name), the left side is truncated to fit, by removing 1395 successive domain labels (and their following dots) until the total 1396 length does not exceed 253 characters. 1398 Uppercased macros expand exactly as their lowercased equivalents, and 1399 are then URL escaped. URL escaping MUST be performed for characters 1400 not in the "unreserved" set, which is defined in [RFC3986]. 1402 Note: Care has to be taken by the sending ADMD so that macro 1403 expansion for legitimate email does not exceed the 63-character limit 1404 on DNS labels. The local-part of email addresses, in particular, can 1405 have more than 63 characters between dots. 1407 Note: To minimize DNS lookup resource requirements, it is better if 1408 sending ADMDs avoid using the "s", "l", "o", or "h" macros in 1409 conjunction with any mechanism directive. Although these macros are 1410 powerful and allow per-user records to be published, they severely 1411 limit the ability of implementations to cache results of check_host() 1412 and they reduce the effectiveness of DNS caches. 1414 Note: If no directive processed during the evaluation of check_host() 1415 contains an "s", "l", "o", or "h" macro, then the results of the 1416 evaluation can be cached on the basis of and alone for 1417 as long as the shortest Time To Live (TTL) of all the DNS records 1418 involved. 1420 7.4. Expansion Examples 1422 The is strong-bad@email.example.com. 1423 The IPv4 SMTP client IP is 192.0.2.3. 1424 The IPv6 SMTP client IP is 2001:DB8::CB01. 1425 The PTR domain name of the client IP is mx.example.org. 1427 macro expansion 1428 ------- ---------------------------- 1429 %{s} strong-bad@email.example.com 1430 %{o} email.example.com 1431 %{d} email.example.com 1432 %{d4} email.example.com 1433 %{d3} email.example.com 1434 %{d2} example.com 1435 %{d1} com 1436 %{dr} com.example.email 1437 %{d2r} example.email 1438 %{l} strong-bad 1439 %{l-} strong.bad 1440 %{lr} strong-bad 1441 %{lr-} bad.strong 1442 %{l1r-} strong 1444 macro-string expansion 1445 -------------------------------------------------------------------- 1446 %{ir}.%{v}._spf.%{d2} 3.2.0.192.in-addr._spf.example.com 1447 %{lr-}.lp._spf.%{d2} bad.strong.lp._spf.example.com 1449 %{lr-}.lp.%{ir}.%{v}._spf.%{d2} 1450 bad.strong.lp.3.2.0.192.in-addr._spf.example.com 1452 %{ir}.%{v}.%{l1r-}.lp._spf.%{d2} 1453 3.2.0.192.in-addr.strong.lp._spf.example.com 1455 %{d2}.trusted-domains.example.net 1456 example.com.trusted-domains.example.net 1458 IPv6: 1459 %{ir}.%{v}._spf.%{d2} 1.0.B.C.0.0.0.0. 1460 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.8.B.D.0.1.0.0.2.ip6._spf.example.com 1462 8. Result Handling 1464 This section provides guidance for operators in response to the 1465 various possible outputs of check_host() on a message. Terse 1466 definitions of SPF results are presented in Section 2.6; this section 1467 provides more detail on each for use in developing local policy for 1468 message handling. 1470 Every operating environment is different. There are some receivers 1471 for whom strict adherence to SPF is appropriate, and definitive 1472 treatment of messages that are evaluated to be explicity unauthorized 1473 ("fail" and sometimes "softfail") is the norm. There are others for 1474 which the "false negative" cases are more of a concern. This concern 1475 is typically handled by merely recording the result in the header and 1476 allowing the message to pass on for additional processing. There are 1477 still others where SPF is one of several inputs to the message 1478 handling decision. As such, there is no normative requirement for 1479 message handling in response to any particular result. This section 1480 is provided to present a complete picture of the likely cause of each 1481 result, and where available, the experience gained during 1482 experimental deployment. 1484 There are essentially two classes of handling choices: 1486 o Handling within the SMTP session that attempted to deliver the 1487 message, such as by returning a permanent SMTP error (rejection) 1488 or temporary SMTP error ("try again later"); 1490 o Permitting the message to pass (a successful SMTP reply code) and 1491 adding an additional header field that indicates the result 1492 returned by check_host() and other salient details; this is 1493 discussed in more detail in Section 9. 1495 8.1. None 1497 With a "none" result, the SPF verifier has no information at all 1498 about the authorization or lack thereof of the client to use the 1499 checked idenity or identities. The check_host() function completed 1500 without errors but was not able to reach any conclusion. 1502 8.2. Neutral 1504 A "neutral" result indicates that although a policy for the identity 1505 was discovered, there is no definite assertion about the (positive or 1506 negative) about the client. 1508 A "neutral" result MUST be treated exactly like the "none" result; 1509 the distinction exists only for informational purposes. Treating 1510 "neutral" more harshly than "none" would discourage domain owners 1511 from testing the use of SPF records (see Section 10.1). 1513 8.3. Pass 1515 A "pass" result means that the client is authorized to inject mail 1516 with the given identity. The domain can now, in the sense of 1517 reputation, be considered responsible for sending the message. 1518 Further policy checks can now proceed with confidence in the 1519 legitimate use of the identity. This is further discussed in 1520 Appendix G.1. 1522 8.4. Fail 1524 A "fail" result is an explicit statement that the client is not 1525 authorized to use the domain in the given identity. Disposition of 1526 SPF fail messages is a matter of local policy. See Appendix G.2 for 1527 considerations on developing local policy. 1529 If the checking software chooses to reject the mail during the SMTP 1530 transaction, then it SHOULD use an SMTP reply code of 550 (see 1531 [RFC5321]) and, if supported, the 5.7.1 enhanced status code (see 1532 [RFC3463]), in addition to an appropriate reply text. The 1533 check_host() function will return either a default explanation string 1534 or one from the domain that published the SPF records (see 1535 Section 6.2). If the information does not originate with the 1536 checking software, it is good to make it clear that the text is 1537 provided by the sender's domain. For example: 1539 550-5.7.1 SPF MAIL FROM check failed: 1540 550-5.7.1 The domain example.com explains: 1541 550 5.7.1 Please see http://www.example.com/mailpolicy.html 1543 If the checking software chooses not to reject the mail during the 1544 SMTP transaction, then it SHOULD add a Received-SPF or 1545 Authentication-Results header field (see Section 9) to communicate 1546 this result to downstream message processors. While this is true for 1547 all SPF results, it is of particular importance for "fail" results 1548 since the message is explicitly not authorized by the domain owner. 1550 8.5. Softfail 1552 A "softfail" result ought to be treated as somewhere between "fail" 1553 and "neutral"/"none". The domain owner believes the host is not 1554 authorized but is not willing to make a strong policy statement. 1555 Receiving software SHOULD NOT reject the message based solely on this 1556 result, but MAY subject the message to closer scrutiny than normal. 1558 The domain owner wants to discourage the use of this host and thus 1559 desires limited feedback when a "softfail" result occurs. For 1560 example, the recipient's Mail User Agent (MUA) could highlight the 1561 "softfail" status, or the receiving MTA could give the sender a 1562 message using greylisting, [RFC6647], with a note the first time the 1563 message is received, but accept it on a later attempt based on 1564 receiver policy. 1566 8.6. Temperror 1568 A "temperror" result means the SPF verifier encountered a transient 1569 (generally DNS) error while performing the check. Checking software 1570 can choose to accept or temporarily reject the message. If the 1571 message is rejected during the SMTP transaction for this reason, the 1572 software SHOULD use an SMTP reply code of 451 and, if supported, the 1573 4.4.3 enhanced status code. These errors can be caused by problems 1574 in either the sender's or receiver's DNS software. 1576 8.7. Permerror 1578 A "permerror" result means the domain's published records could not 1579 be correctly interpreted. This signals an error condition that 1580 definitely requires manual intervention to be resolved. If the 1581 message is rejected during the SMTP transaction for this reason, the 1582 software SHOULD use an SMTP reply code of 550 and, if supported, the 1583 5.5.2 enhanced status code. Be aware that if the domain owner uses 1584 macros (Section 7), it is possible that this result is due to the 1585 checked identities having an unexpected format. It is also possible 1586 that this result is generated by certain SPF clients due to the input 1587 arguments having an unexpected format; see Section 4.8. 1589 9. Recording The Result 1591 To provide downstream agents, such as MUAs, with the information they 1592 might need in terms of evaluating or representing the apparent safety 1593 of the message content, it is RECOMMENDED that SMTP receivers record 1594 the result of SPF processing in the message header. For operators 1595 that choose to record SPF results in the header of the message for 1596 processing by internal filters or MUAs, two methods are presented. 1597 Section 9.1 defines the Received-SPF field, which is the results 1598 field originally defined for SPF use. Section 9.2 discusses 1599 Authentication-Results [RFC5451] which was specified more recently 1600 and is designed for use by SPF and other authentication methods. 1602 Both are in common use, and hence both are included here. However, 1603 it is important to note that they were designed to serve slightly 1604 different purposes. Received-SPF is intended to include enough 1605 forensic information to enable reconstruction of the SPF evaluation 1606 of the message, while Authentication-Results is designed only to 1607 relay the result itself and related output details of likely use to 1608 end users (e.g., what property of the message was actually 1609 authenticated and what it contained), leaving forensic work to the 1610 purview of system logs and the Received field contents. Also, 1611 Received-SPF relies on compliance of agents within the receiving ADMD 1612 to adhere to the header field ordering rules of [RFC5321] and 1613 [RFC5322], while Authentication-Results includes some provisions to 1614 protect against non-compliant implementations. 1616 An operator could choose to use both to serve different downstream 1617 agents. In such cases, care needs to be taken to ensure both fields 1618 are conveying the same details, or unexpected results can occur. 1620 9.1. The Received-SPF Header Field 1622 The Received-SPF header field is a trace field (see [RFC5322] Section 1623 3.6.7) and SHOULD be prepended to the existing header, above the 1624 Received: field that is generated by the SMTP receiver. It MUST 1625 appear above all other Received-SPF fields in the message. The 1626 header field has the following format: 1628 header-field = "Received-SPF:" [CFWS] result FWS [comment FWS] 1629 [ key-value-list ] CRLF 1631 result = "pass" / "fail" / "softfail" / "neutral" / 1632 "none" / "temperror" / "permerror" 1634 key-value-list = key-value-pair *( ";" [CFWS] key-value-pair ) 1635 [";"] 1637 key-value-pair = key [CFWS] "=" ( dot-atom / quoted-string ) 1639 key = "client-ip" / "envelope-from" / "helo" / 1640 "problem" / "receiver" / "identity" / 1641 "mechanism" / name 1643 identity = "mailfrom" ; for the "MAIL FROM" identity 1644 / "helo" ; for the "HELO" identity 1645 / name ; other identities 1647 dot-atom = 1648 quoted-string = 1649 comment = 1650 CFWS = 1651 FWS = 1652 CRLF = 1654 The header field SHOULD include a "(...)" style comment after the 1655 result, conveying supporting information for the result, such as 1656 , , and . 1658 The following key-value pairs are designed for later machine parsing. 1659 SPF verifiers SHOULD give enough information so that the SPF results 1660 can be verified. That is, at least "client-ip", "helo", and, if the 1661 "MAIL FROM" identity was checked, "envelope-from". 1663 client-ip the IP address of the SMTP client 1665 envelope-from the envelope sender mailbox 1667 helo the host name given in the HELO or EHLO command 1669 mechanism the mechanism that matched (if no mechanisms matched, 1670 substitute the word "default") 1672 problem if an error was returned, details about the error 1673 receiver the host name of the SPF verifier 1675 identity the identity that was checked; see the ABNF 1676 rule 1678 Other keys MAY be defined by SPF verifiers. 1680 SPF verifiers MUST make sure that the Received-SPF header field does 1681 not contain invalid characters, is not excessively long (See 1682 [RFC5322] Section 2.1.1), and does not contain malicious data that 1683 has been provided by the sender. 1685 Examples of various header field styles that could be generated are 1686 the following: 1688 Received-SPF: pass (mybox.example.org: domain of 1689 myname@example.com designates 192.0.2.1 as permitted sender) 1690 receiver=mybox.example.org; client-ip=192.0.2.1; 1691 envelope-from="myname@example.com"; helo=foo.example.com; 1693 Received-SPF: fail (mybox.example.org: domain of 1694 myname@example.com does not designate 1695 192.0.2.1 as permitted sender) 1696 identity=mailfrom; client-ip=192.0.2.1; 1697 envelope-from="myname@example.com"; 1699 Received-SPF: pass (mybox.example.org: domain of 1700 myname@example.com designates 192.0.2.1 as permitted sender) 1701 receiver=mybox.example.org; client-ip=192.0.2.1; 1702 mechanism=ip4:192.0.2.1; envelope-from="myname@example.com"; 1703 helo=foo.example.com; 1705 9.2. SPF Results in the Authentication-Results Header Field 1707 As mentioned in Section 9, the Authentication-Results header field is 1708 designed to communicate lists of tests a border MTA did and their 1709 results. The specified elements of the field provide less 1710 information than the Received-SPF field: 1712 Authentication-Results: myhost.example.org; spf=pass 1713 smtp.mailfrom=example.net 1715 Received-SPF: pass (myhost.example.org: domain of 1716 myname@example.com designates 192.0.2.1 as permitted sender) 1717 receiver=mybox.example.org; client-ip=192.0.2.1; 1718 envelope-from="myname@example.com"; helo=foo.example.com; 1720 It is, however, possible to add CFWS in the "reason" part of an 1721 Authentication-Results header field and provide the equivalent 1722 information, if desired. 1724 As an example, an expanded Authentication-Results header field might 1725 look like (for a "MAIL FROM" check in this example): 1727 Authentication-Results: myhost.example.org; spf=pass 1728 reason="client-ip=192.0.2.1; smtp.helo=foo.example.com" 1729 smtp.mailfrom=user@example.net 1731 10. Effects on Infrastructure 1733 This section outlines the major implications that adoption of this 1734 document will have on various entities involved in Internet email. 1735 It is intended to make clear to the reader where this document 1736 knowingly affects the operation of such entities. This section is 1737 not a "how-to" manual, or a "best practices" document, and it is not 1738 a comprehensive list of what such entities SHOULD do in light of this 1739 document. 1741 This section provides operational advice and instruction only. It is 1742 non-normative. 1744 [RFC5598] describes the Internet email architecture. This section is 1745 organized based on the different segments of the architecture. 1747 10.1. Sending Domains 1749 Originating ADMDs (ADministrative Management Domains - [RFC5598] 1750 Section 2.2.1 and Section 2.3) that wish to be compliant with this 1751 specification will need to determine the list of relays ([RFC5598] 1752 Section 2.2.2) that they allow to use their domain name in the "HELO" 1753 and "MAIL FROM" identities when relaying to other ADMDs. It is 1754 recognized that forming such a list is not just a simple technical 1755 exercise, but involves policy decisions with both technical and 1756 administrative considerations. 1758 10.1.1. DNS Resource Considerations 1760 Minimizing the DNS resources required for SPF lookups can be done by 1761 choosing directives that require less DNS information and by placing 1762 lower-cost mechanisms earlier in the SPF record. 1764 +----------+--------+-----------------+ 1765 | term | cost | limit | 1766 +----------+--------+-----------------+ 1767 | ip4/ip6 | 0 | - | 1768 | a | 1 | 10 | 1769 | include | 1 | 10 | 1770 | redirect | 1 | 10 | 1771 | exists | 1 | 10 | 1772 | mx | 1 + N* | 10 and N* <= 10 | 1773 | ptr/%{p} | 1 + N* | 10 and N* <= 10 | 1774 | all | 0 | - | 1775 +----------+--------+-----------------+ 1776 * N is the number of RRs found during each term evaluation 1778 Section 4.6.4 specifies the limits receivers have to use. It is 1779 essential to publish records that do not exceed these requirements. 1780 It is also required to carefully weight the cost and the 1781 maintainability of licit solutions. 1783 For example, consider a domain set up as follows: 1785 example.com. IN MX 10 mx.example.com. 1786 IN MX 20 mx2.example.com. 1787 mx.example.com. IN A 192.0.2.1 1788 mx2.example.com. IN A 192.0.2.129 1790 Assume the administrative point is to authorize (pass) mx and mx2 1791 while failing every other host. Compare the following solutions: 1793 Best record: 1794 example.com. IN TXT "v=spf1 ip4:192.0.2.1 ip4:192.0.2.129 -all" 1796 Good record: 1797 $ORIGIN example.com. 1798 @ IN TXT "v=spf1 a:authorized-spf.example.com -all" 1799 authorized-spf IN A 192.0.2.1 1800 IN A 192.0.2.129 1802 Expensive record: 1803 example.com. IN TXT "v=spf1 mx:example.com -all" 1805 Wasteful, bad record: 1806 example.com. IN TXT "v=spf1 ip4:192.0.2.0/24 mx -all" 1808 10.1.2. Administrator's Considerations 1810 There might be administrative considerations: using "a" over "ip4" or 1811 "ip6" allows hosts to be renumbered easily. Using "mx" over "a" 1812 allows the set of mail hosts to be changed easily. Unless such 1813 changes are common, it is better to use the less resource intensive 1814 mechanisms like "ip4" and "ip6" over "a" or "a" or "mx". 1816 In some specific cases, standard advice on record content is 1817 appropriate. Publishing SPF records for domains that send no mail is 1818 a well established best practice. The record for a domain that sends 1819 no mail is: 1821 www.example.com. IN TXT "v=spf1 -all" 1823 Publishing SPF records for individual hosts is also best practice. 1825 The hostname is generally the identity used in the 5321.HELO/.EHLO 1826 command. In the case of messages with a null 5321.MailFrom, this is 1827 used as the domain for 5321.MailFrom SPF checks, in addition to being 1828 used in 5321.HELO/.EHLO based SPF checks. The standard SPF record 1829 for an individual host that is involved in mail processing is: 1831 relay.example.com. IN TXT "v=spf1 a -all" 1833 Validating correct deployment is difficult. [RFC6652] describes one 1834 mechanism for soliciting feedback on SPF failures. Another 1835 suggestion can be found in Appendix C. 1837 Regardless of the method used, understanding the ADMD's outbound mail 1838 architecture is essential to effective deployment. 1840 10.1.3. Bounces 1842 As explained in Section 1.1.3, [RFC5321] allows the reverse-path to 1843 be null, which is typical of some Delivery Status Notification 1844 [RFC3464], commonly called email bounces. In this case the only 1845 entity available for performing an SPF check is the "HELO" identity 1846 defined in Section 1.1.4. SPF functionality is enhanced by 1847 administrators ensuring this identity is set correctly and has an 1848 appropriate SPF record. It is normal to have the HELO identity set 1849 to hostname instead of domain. Zone file generation for significant 1850 numbers of hosts can be consolidated using the redirect modifier and 1851 scripted for initial deployment. Specific deployment advice is given 1852 above in Section 10.1.2. 1854 10.2. Receivers 1856 SPF results can be used in combination with other methods to 1857 determine the final local disposition (either positive or negative of 1858 a message. It can also be considered dispositive on its own. 1860 An attempt to have one organization (sender) direct the email 1861 handling policies of another (receiver) is inherently challenging and 1862 often controversial. As stated elsewhere in this document, there is 1863 no normative requirement for specific handling of a message based on 1864 any SPF result. The information presented in Section 8 and in 1865 Appendix G is offered for receiver consideration when forming local 1866 handling policies. 1868 The primary considerations are that SPF might return "pass" for mail 1869 that is ultimately harmful (e.g., spammers that arrange for SPF to 1870 pass using nonsense domain names, or virus or spam outbreaks from 1871 within trusted sources), and might also return "fail" for mail that 1872 is ultimately legitimate (e.g., legitimate mail that has traversed a 1873 mail alias). It is important take both of these cases under 1874 consideration when establishing local handling policy. 1876 10.3. Mediators 1878 Mediators are a type of User actor.[RFC5598]. That is, a mediator 1879 takes 'delivery' of a message and posts a 'submission' of a new 1880 message. The mediator can make the newly-posted message be as 1881 similar or as different from the original message as they wish. 1882 Examples include mailing lists (see [RFC5598] Section 5.3) and 1883 ReSenders ([RFC5598] Section 5.2). This is discussed in [RFC5321], 1884 Section 3.9. For the operation of SPF, the essential concern is the 1885 email address in the 5321.MailFrom command for the new message. 1887 Because SPF evaluation is based on the IP Address of the "last" 1888 sending SMTP server, the address of the mediator will be used, rather 1889 than the address of the SMTP server that sent the message to the 1890 mediator. Some mediators retain the email address from the original 1891 message, while some use a new address. 1893 If the address is the same as for the original message, and the 1894 original message had an associated SPF record, then the SPF 1895 evaluation will fail unless mitigations such as those described in 1896 Appendix D are used. 1898 11. Security Considerations 1900 11.1. Processing Limits 1902 As with most aspects of email, there are a number of ways that 1903 malicious parties could use the protocol as an avenue for a 1904 Denial-of-Service (DoS) attack. The processing limits outlined in 1905 Section 4.6.4 are designed to prevent attacks such as the following: 1907 o A malicious party could create an SPF record with many references 1908 to a victim's domain and send many emails to different SPF 1909 verifiers; those SPF verifiers would then create a DoS attack. In 1910 effect, the SPF verifiers are being used to amplify the attacker's 1911 bandwidth by using fewer bytes in the SMTP session than are used 1912 by the DNS queries. Using SPF clients also allows the attacker to 1913 hide the true source of the attack. 1915 o Whereas implementations of check_host() are supposed to limit the 1916 number of DNS lookups, malicious domains could publish records 1917 that exceed these limits in an attempt to waste computation effort 1918 at their targets when they send them mail. Malicious domains 1919 could also design SPF records that cause particular 1920 implementations to use excessive memory or CPU usage, or to 1921 trigger bugs. 1923 o Malicious parties could send a large volume of mail purporting to 1924 come from the intended target to a wide variety of legitimate mail 1925 hosts. These legitimate machines would then present a DNS load on 1926 the target as they fetched the relevant records. 1928 Of these, the case of a third party referenced in the SPF record is 1929 the easiest for a DoS attack to effectively exploit. As a result, 1930 limits that might seem reasonable for an individual mail server can 1931 still allow an unreasonable amount of bandwidth amplification. 1932 Therefore, the processing limits need to be quite low. 1934 11.2. SPF-Authorized Email May Contain Other False Identities 1936 Do not construe the "MAIL FROM" and "HELO" identity authorizations to 1937 provide more assurance than they do. It is entirely possible for a 1938 malicious sender to inject a message using his own domain in the 1939 identities used by SPF, to have that domain's SPF record authorize 1940 the sending host, and yet the message can easily list other 1941 identities in its header. Unless the user or the MUA takes care to 1942 note that the authorized identity does not match the other more 1943 commonly-presented identities (such as the From: header field), the 1944 user might be lulled into a false sense of security. 1946 11.3. Spoofed DNS and IP Data 1948 There are two aspects of this protocol that malicious parties could 1949 exploit to undermine the validity of the check_host() function: 1951 o The evaluation of check_host() relies heavily on DNS. A malicious 1952 attacker could attack the DNS infrastructure and cause 1953 check_host() to see spoofed DNS data, and then return incorrect 1954 results. This could include returning "pass" for an value 1955 where the actual domain's record would evaluate to "fail". See 1956 [RFC3833] for a description of DNS weaknesses. 1958 o The client IP address, , is assumed to be correct. In a 1959 modern, correctly configured system the risk of this not being 1960 true is nil. 1962 11.4. Cross-User Forgery 1964 By definition, SPF policies just map domain names to sets of 1965 authorized MTAs, not whole email addresses to sets of authorized 1966 users. Although the "l" macro (Section 7) provides a limited way to 1967 define individual sets of authorized MTAs for specific email 1968 addresses, it is generally impossible to verify, through SPF, the use 1969 of specific email addresses by individual users of the same MTA. 1971 It is up to mail services and their MTAs to directly prevent 1972 cross-user forgery: based on SMTP AUTH ([RFC4954]), users have to be 1973 restricted to using only those email addresses that are actually 1974 under their control (see [RFC6409], Section 6.1). Another means to 1975 verify the identity of individual users is message cryptography such 1976 as PGP ([RFC4880]) or S/MIME ([RFC5751]). 1978 11.5. Untrusted Information Sources 1980 An SPF compliant receiver gathers information from the SMTP commands 1981 it receives and from the published DNS records of the sending domain 1982 holder, (e.g., "HELO" domain name, the "MAIL FROM" address from the 1983 envelope, and SPF DNS records published by the domain holder). 1985 11.5.1. Recorded Results 1987 This information, passed to the receiver in the Received-SPF: or 1988 Authentication-Results: trace fields, may be returned to the client 1989 MTA as an SMTP rejection message. If such an SMTP rejection message 1990 is generated, the information from the trace fields has to be checked 1991 for such problems as invalid characters and excessively long lines. 1993 11.5.2. External Explanations 1995 When the authorization check fails, an explanation string could be 1996 included in the reject response. Both the sender and the rejecting 1997 receiver need to be aware that the explanation was determined by the 1998 publisher of the SPF record checked and, in general, not the 1999 receiver. The explanation can contain malicious URLs, or it might be 2000 offensive or misleading. 2002 Explanations returned to sender domains due to "exp" modifiers, 2003 (Section 6.2), were generated by the sender policy published by the 2004 domain holders themselves. As long as messages are only returned 2005 with non-delivery notification ([RFC3464]) to domains publishing the 2006 explanation strings from their own DNS SPF records, the only affected 2007 parties are the original publishers of the domain's SPF records. 2009 In practice, such non-delivery notifications can be misdirected, such 2010 as when an MTA accepts an email and only later generates the 2011 notification to a forged address, or when an email forwarder does not 2012 direct the bounce back to the original sender. 2014 11.5.3. Macro Expansion 2016 Macros (Section 7) allow senders to inject arbitrary text (any non- 2017 null [US-ASCII] character) into receiver DNS queries. It is necesary 2018 to be prepared for hostile or unexpected content. 2020 11.6. Privacy Exposure 2022 Checking SPF records causes DNS queries to be sent to the domain 2023 owner. These DNS queries, especially if they are caused by the 2024 "exists" mechanism, can contain information about who is sending 2025 email and likely to which MTA the email is being sent. This can 2026 introduce some privacy concerns, which are more or less of an issue 2027 depending on local laws and the relationship between the domain owner 2028 and the person sending the email. 2030 11.7. Delivering Mail Producing a 'Fail' Result 2032 Operators that choose to deliver mail for which SPF produces a "fail" 2033 result need to understand that they are admitting content that is 2034 explicitly not authorized by the purported sender. While there are 2035 known failure modes that can be considered "false negatives", the 2036 distinct choice to admit those messages increases end-user exposure 2037 to likely harm. This is especially true for domains belonging to 2038 known good actors that are typically well-behaved; unauthorized mail 2039 from those sources might well be subjected to much higher skepticism 2040 and content analysis. 2042 SPF does not, however, include the capacity for identifying good 2043 actors from bad ones, nor does it handle the concept of known actors 2044 versus unknown ones. Those notions are out of scope for this 2045 specification. 2047 12. Contributors and Acknowledgements 2049 This document is largely based on the work of Meng Weng Wong, Mark 2050 Lentczner, and Wayne Schlitt. Although, as this section 2051 acknowledges, many people have contributed to this document, a very 2052 large portion of the writing and editing are due to Meng, Mark, and 2053 Wayne. 2055 This design owes a debt of parentage to [RMX] by Hadmut Danisch and 2056 to [DMP] by Gordon Fecyk. The idea of using a DNS record to check 2057 the legitimacy of an email address traces its ancestry further back 2058 through messages on the namedroppers mailing list by Paul Vixie 2059 [Vixie] (based on suggestion by Jim Miller) and by David Green 2060 [Green]. 2062 Philip Gladstone contributed the concept of macros to the 2063 specification, multiplying the expressiveness of the language and 2064 making per-user and per-IP lookups possible. 2066 The authors of both this document and [RFC4408] would also like to 2067 thank the literally hundreds of individuals who have participated in 2068 the development of this design. They are far too numerous to name, 2069 but they include the following: 2071 The participants in the SPFbis working group. 2072 The folks on the spf-discuss mailing list. 2073 The folks on the SPAM-L mailing list. 2074 The folks on the IRTF ASRG mailing list. 2075 The folks on the IETF MARID mailing list. 2076 The folks on #perl. 2078 13. IANA Considerations 2080 13.1. The SPF DNS Record Type 2082 Per [RFC4408], the IANA assigned the Resource Record Type and Qtype 2083 from the DNS Parameters Registry for the SPF RR type with code 99. 2084 The format of this type is identical to the TXT RR [RFC1035]. The 2085 character content of the record is encoded as [US-ASCII]. Use of 2086 this record type is obsolete for SPF Version 1. 2088 IANA is requested to add an annotation to the SPF RRTYPE saying 2089 "(OBSOLETE - use TXT)" in the DNS Parameters registry. 2091 [NOTE TO RFC EDITOR: (to be changed to " ... has added ..." upon 2092 publication)] 2094 13.2. The Received-SPF Mail Header Field 2096 Per [RFC3864], the "Received-SPF:" header field is added to the IANA 2097 Permanent Message Header Field Registry. The following is the 2098 registration template: 2100 Header field name: Received-SPF 2101 Applicable protocol: mail ([RFC5322]) 2102 Status: Standards Track 2103 Author/Change controller: IETF 2104 Specification document(s): RFC XXXX 2105 [NOTE TO RFC EDITOR: (this document)] 2107 13.3. SPF Modifier Registration 2109 [RFC6652] created a new SPF Modifier Registration. IANA is requested 2110 to change the reference for the exp and redirect modifiers from 2111 [RFC4408] to this document. Their status should not be changed. 2113 14. References 2115 14.1. Normative References 2117 [RFC1035] Mockapetris, P., "Domain names - implementation and 2118 specification", STD 13, RFC 1035, November 1987. 2120 [RFC1123] Braden, R., "Requirements for Internet Hosts - Application 2121 and Support", STD 3, RFC 1123, October 1989. 2123 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 2124 Requirement Levels", BCP 14, RFC 2119, March 1997. 2126 [RFC3463] Vaudreuil, G., "Enhanced Mail System Status Codes", 2127 RFC 3463, January 2003. 2129 [RFC3864] Klyne, G., Nottingham, M., and J. Mogul, "Registration 2130 Procedures for Message Header Fields", BCP 90, RFC 3864, 2131 September 2004. 2133 [RFC3986] Berners-Lee, T., Fielding, R., and L. Masinter, "Uniform 2134 Resource Identifier (URI): Generic Syntax", STD 66, 2135 RFC 3986, January 2005. 2137 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 2138 Architecture", RFC 4291, February 2006. 2140 [RFC5234] Crocker, D. and P. Overell, "Augmented BNF for Syntax 2141 Specifications: ABNF", STD 68, RFC 5234, January 2008. 2143 [RFC5321] Klensin, J., "Simple Mail Transfer Protocol", RFC 5321, 2144 October 2008. 2146 [RFC5322] Resnick, P., Ed., "Internet Message Format", RFC 5322, 2147 October 2008. 2149 [RFC5451] Kucherawy, M., "Message Header Field for Indicating 2150 Message Authentication Status", RFC 5451, April 2009. 2152 [RFC5598] Crocker, D., "Internet Mail Architecture", RFC 5598, 2153 July 2009. 2155 [RFC5890] Klensin, J., "Internationalized Domain Names for 2156 Applications (IDNA): Definitions and Document Framework", 2157 RFC 5890, August 2010. 2159 [US-ASCII] 2160 American National Standards Institute (formerly United 2161 States of America Standards Institute), "USA Code for 2162 Information Interchange, X3.4", 1968. 2164 ANSI X3.4-1968 has been replaced by newer versions with 2165 slight modifications, but the 1968 version remains 2166 definitive for the Internet. 2168 14.2. Informative References 2170 [DMP] Fecyk, G., "Designated Mailers Protocol". 2172 Work In Progress 2174 [Green] Green, D., "Domain-Authorized SMTP Mail", 2002. 2176 [RFC1034] Mockapetris, P., "Domain names - concepts and facilities", 2177 STD 13, RFC 1034, November 1987. 2179 [RFC1983] Malkin, G., "Internet Users' Glossary", RFC 1983, 2180 August 1996. 2182 [RFC2308] Andrews, M., "Negative Caching of DNS Queries (DNS 2183 NCACHE)", RFC 2308, March 1998. 2185 [RFC2782] Gulbrandsen, A., Vixie, P., and L. Esibov, "A DNS RR for 2186 specifying the location of services (DNS SRV)", RFC 2782, 2187 February 2000. 2189 [RFC3464] Moore, K. and G. Vaudreuil, "An Extensible Message Format 2190 for Delivery Status Notifications", RFC 3464, 2191 January 2003. 2193 [RFC3696] Klensin, J., "Application Techniques for Checking and 2194 Transformation of Names", RFC 3696, February 2004. 2196 [RFC3833] Atkins, D. and R. Austein, "Threat Analysis of the Domain 2197 Name System (DNS)", RFC 3833, August 2004. 2199 [RFC3834] Moore, K., "Recommendations for Automatic Responses to 2200 Electronic Mail", RFC 3834, August 2004. 2202 [RFC4408] Wong, M. and W. Schlitt, "Sender Policy Framework (SPF) 2203 for Authorizing Use of Domains in E-Mail, Version 1", 2204 RFC 4408, April 2006. 2206 [RFC4632] Fuller, V. and T. Li, "Classless Inter-domain Routing 2207 (CIDR): The Internet Address Assignment and Aggregation 2208 Plan", BCP 122, RFC 4632, August 2006. 2210 [RFC4880] Callas, J., Donnerhacke, L., Finney, H., Shaw, D., and R. 2211 Thayer, "OpenPGP Message Format", RFC 4880, November 2007. 2213 [RFC4954] Siemborski, R. and A. Melnikov, "SMTP Service Extension 2214 for Authentication", RFC 4954, July 2007. 2216 [RFC5751] Ramsdell, B. and S. Turner, "Secure/Multipurpose Internet 2217 Mail Extensions (S/MIME) Version 3.2 Message 2218 Specification", RFC 5751, January 2010. 2220 [RFC6409] Gellens, R. and J. Klensin, "Message Submission for Mail", 2221 STD 72, RFC 6409, November 2011. 2223 [RFC6647] Kucherawy, M. and D. Crocker, "Email Greylisting: An 2224 Applicability Statement for SMTP", RFC 6647, June 2012. 2226 [RFC6652] Kitterman, S., "Sender Policy Framework (SPF) 2227 Authentication Failure Reporting Using the Abuse Reporting 2228 Format", RFC 6652, June 2012. 2230 [RFC6686] Kucherawy, M., "Resolution of the Sender Policy Framework 2231 (SPF) and Sender ID Experiments", RFC 6686, July 2012. 2233 [RMX] Danisch, H., "The RMX DNS RR Type for light weight sender 2234 authentication". 2236 Work In Progress 2238 [Vixie] Vixie, P., "Repudiating MAIL FROM", 2002. 2240 Appendix A. Collected ABNF 2242 This section is normative and any discrepancies with the ABNF 2243 fragments in the preceding text are to be resolved in favor of this 2244 grammar. 2246 See [RFC5234] for ABNF notation. Please note that as per this ABNF 2247 definition, literal text strings (those in quotes) are case- 2248 insensitive. Hence, "mx" matches "mx", "MX", "mX", and "Mx". 2250 record = version terms *SP 2251 version = "v=spf1" 2253 terms = *( 1*SP ( directive / modifier ) ) 2255 directive = [ qualifier ] mechanism 2256 qualifier = "+" / "-" / "?" / "~" 2257 mechanism = ( all / include 2258 / A / MX / PTR / IP4 / IP6 / exists ) 2260 all = "all" 2261 include = "include" ":" domain-spec 2262 A = "a" [ ":" domain-spec ] [ dual-cidr-length ] 2263 MX = "mx" [ ":" domain-spec ] [ dual-cidr-length ] 2264 PTR = "ptr" [ ":" domain-spec ] 2265 IP4 = "ip4" ":" ip4-network [ ip4-cidr-length ] 2266 IP6 = "ip6" ":" ip6-network [ ip6-cidr-length ] 2267 exists = "exists" ":" domain-spec 2269 modifier = redirect / explanation / unknown-modifier 2270 redirect = "redirect" "=" domain-spec 2271 explanation = "exp" "=" domain-spec 2272 unknown-modifier = name "=" macro-string 2273 ; where name is not any known modifier 2275 ip4-cidr-length = "/" 1*DIGIT 2276 ip6-cidr-length = "/" 1*DIGIT 2277 dual-cidr-length = [ ip4-cidr-length ] [ "/" ip6-cidr-length ] 2279 ip4-network = qnum "." qnum "." qnum "." qnum 2280 qnum = DIGIT ; 0-9 2281 / %x31-39 DIGIT ; 10-99 2282 / "1" 2DIGIT ; 100-199 2283 / "2" %x30-34 DIGIT ; 200-249 2284 / "25" %x30-35 ; 250-255 2285 ; conventional dotted quad notation. e.g., 192.0.2.0 2286 ip6-network = 2287 ; e.g., 2001:DB8::CD30 2289 domain-spec = macro-string domain-end 2290 domain-end = ( "." toplabel [ "." ] ) / macro-expand 2292 toplabel = ( *alphanum ALPHA *alphanum ) / 2293 ( 1*alphanum "-" *( alphanum / "-" ) alphanum ) 2294 ; LDH rule plus additional TLD restrictions 2295 ; (see [RFC3696], Section 2 for background) 2296 alphanum = ALPHA / DIGIT 2298 explain-string = *( macro-string / SP ) 2300 macro-string = *( macro-expand / macro-literal ) 2301 macro-expand = ( "%{" macro-letter transformers *delimiter "}" ) 2302 / "%%" / "%_" / "%-" 2303 macro-literal = %x21-24 / %x26-7E 2304 ; visible characters except "%" 2305 macro-letter = "s" / "l" / "o" / "d" / "i" / "p" / "h" / 2306 "c" / "r" / "t" / "v" 2307 transformers = *DIGIT [ "r" ] 2308 delimiter = "." / "-" / "+" / "," / "/" / "_" / "=" 2310 name = ALPHA *( ALPHA / DIGIT / "-" / "_" / "." ) 2312 header-field = "Received-SPF:" [CFWS] result FWS [comment FWS] 2313 [ key-value-list ] CRLF 2315 result = "pass" / "fail" / "softfail" / "neutral" / 2316 "none" / "temperror" / "permerror" 2318 key-value-list = key-value-pair *( ";" [CFWS] key-value-pair ) 2319 [";"] 2321 key-value-pair = key [CFWS] "=" ( dot-atom / quoted-string ) 2323 key = "client-ip" / "envelope-from" / "helo" / 2324 "problem" / "receiver" / "identity" / 2325 "mechanism" / name 2327 identity = "mailfrom" ; for the "MAIL FROM" identity 2328 / "helo" ; for the "HELO" identity 2329 / name ; other identities 2331 ALPHA = 2332 DIGIT = <0-9 as per [RFC5234]> 2333 SP = 2334 domain = 2335 dot-atom = 2336 quoted-string = 2337 comment = 2338 CFWS = 2339 FWS = 2340 CRLF = 2341 authserv-id = 2342 reasonspec = 2344 Appendix B. Extended Examples 2346 These examples are based on the following DNS setup: 2348 ; A domain with two mail servers, two hosts 2349 ; and two servers at the domain name 2350 $ORIGIN example.com. 2351 @ MX 10 mail-a 2352 MX 20 mail-b 2353 A 192.0.2.10 2354 A 192.0.2.11 2355 amy A 192.0.2.65 2356 bob A 192.0.2.66 2357 mail-a A 192.0.2.129 2358 mail-b A 192.0.2.130 2359 www CNAME example.com. 2361 ; A related domain 2362 $ORIGIN example.org. 2363 @ MX 10 mail-c 2364 mail-c A 192.0.2.140 2366 ; The reverse IP for those addresses 2367 $ORIGIN 2.0.192.in-addr.arpa. 2368 10 PTR example.com. 2369 11 PTR example.com. 2370 65 PTR amy.example.com. 2371 66 PTR bob.example.com. 2372 129 PTR mail-a.example.com. 2373 130 PTR mail-b.example.com. 2374 140 PTR mail-c.example.org. 2376 ; A rogue reverse IP domain that claims to be 2377 ; something it's not 2378 $ORIGIN 0.0.10.in-addr.arpa. 2379 4 PTR bob.example.com. 2381 B.1. Simple Examples 2383 These examples show various possible published records for 2384 example.com and which values if would cause check_host() to 2385 return "pass". Note that is "example.com". 2387 v=spf1 +all 2388 -- any passes 2390 v=spf1 a -all 2391 -- hosts 192.0.2.10 and 192.0.2.11 pass 2393 v=spf1 a:example.org -all 2394 -- no sending hosts pass since example.org has no A records 2396 v=spf1 mx -all 2397 -- sending hosts 192.0.2.129 and 192.0.2.130 pass 2399 v=spf1 mx:example.org -all 2400 -- sending host 192.0.2.140 passes 2402 v=spf1 mx mx:example.org -all 2403 -- sending hosts 192.0.2.129, 192.0.2.130, and 192.0.2.140 pass 2405 v=spf1 mx/30 mx:example.org/30 -all 2406 -- any sending host in 192.0.2.128/30 or 192.0.2.140/30 passes 2408 v=spf1 ptr -all 2409 -- sending host 192.0.2.65 passes (reverse DNS is valid and is in 2410 example.com) 2411 -- sending host 192.0.2.140 fails (reverse DNS is valid, but not 2412 in example.com) 2413 -- sending host 10.0.0.4 fails (reverse IP is not valid) 2415 v=spf1 ip4:192.0.2.128/28 -all 2416 -- sending host 192.0.2.65 fails 2417 -- sending host 192.0.2.129 passes 2419 B.2. Multiple Domain Example 2421 These examples show the effect of related records: 2423 example.org: "v=spf1 include:example.com include:example.net -all" 2425 This record would be used if mail from example.org actually came 2426 through servers at example.com and example.net. Example.org's 2427 designated servers are the union of example.com's and example.net's 2428 designated servers. 2430 la.example.org: "v=spf1 redirect=example.org" 2431 ny.example.org: "v=spf1 redirect=example.org" 2432 sf.example.org: "v=spf1 redirect=example.org" 2434 These records allow a set of domains that all use the same mail 2435 system to make use of that mail system's record. In this way, only 2436 the mail system's record needs to be updated when the mail setup 2437 changes. These domains' records never have to change. 2439 B.3. DNSBL Style Example 2441 Imagine that, in addition to the domain records listed above, there 2442 are these: 2444 $ORIGIN _spf.example.com. 2445 mary.mobile-users A 127.0.0.2 2446 fred.mobile-users A 127.0.0.2 2447 15.15.168.192.joel.remote-users A 127.0.0.2 2448 16.15.168.192.joel.remote-users A 127.0.0.2 2450 The following records describe users at example.com who mail from 2451 arbitrary servers, or who mail from personal servers. 2453 example.com: 2455 v=spf1 mx 2456 include:mobile-users._spf.%{d} 2457 include:remote-users._spf.%{d} 2458 -all 2460 mobile-users._spf.example.com: 2462 v=spf1 exists:%{l1r+}.%{d} 2464 remote-users._spf.example.com: 2466 v=spf1 exists:%{ir}.%{l1r+}.%{d} 2468 B.4. Multiple Requirements Example 2470 Say that your sender policy requires both that the IP address is 2471 within a certain range and that the reverse DNS for the IP matches. 2472 This can be done several ways, including the following: 2474 example.com. SPF ( "v=spf1 " 2475 "-include:ip4._spf.%{d} " 2476 "-include:ptr._spf.%{d} " 2477 "+all" ) 2478 ip4._spf.example.com. SPF "v=spf1 -ip4:192.0.2.0/24 +all" 2479 ptr._spf.example.com. SPF "v=spf1 -ptr +all" 2481 This example shows how the "-include" mechanism can be useful, how an 2482 SPF record that ends in "+all" can be very restrictive, and the use 2483 of De Morgan's Law. 2485 Appendix C. Further Testing Advice 2487 Another approach that can be helpful to publish records that include 2488 a "tracking exists:" mechanism. By looking at the name server logs, 2489 a rough list can then be generated. For example: 2491 v=spf1 exists:_h.%{h}._l.%{l}._o.%{o}._i.%{i}._spf.%{d} ?all 2493 Appendix D. SPF/Mediator Interactions 2495 There are three places that techniques can be used to ameliorate 2496 unintended SPF failures with mediators. 2498 1. The beginning, when email is first sent (Originating ADMDs). 2500 * "Neutral" results could be given for IP addresses that might 2501 be forwarders, instead of "fail" results based on a list of 2502 known reliable forwarders. For example: 2504 "v=spf1 mx ?exists:%{ir}.whitlist.example.org -all" 2506 This would cause a lookup on an DNS white list (DNSWL) and 2507 cause a result of "fail" only for email not either coming from 2508 the domain's mx host(s) (SPF pass) or white listed sources 2509 (SPF neutral). This, in effect, outsources an element of 2510 sender policy to the maintainer of the whitelist. 2512 * The "MAIL FROM" identity could have additional information in 2513 the local-part that cryptographically identifies the mail as 2514 coming from an authorized source. In this case, such an SPF 2515 record could be used: 2517 "v=spf1 mx exists:%{l}._spf_verify.%{d} -all" 2519 Then, a specialized DNS server can be set up to serve the 2520 _spf_verify subdomain that validates the local-part. Although 2521 this requires an extra DNS lookup, this happens only when the 2522 email would otherwise be rejected as not coming from a known 2523 good source. 2524 Note that due to the 63-character limit for domain labels, 2525 this approach only works reliably if the local-part signature 2526 scheme is guaranteed either to only produce local-parts with a 2527 maximum of 63 characters or to gracefully handle truncated 2528 local-parts. 2530 * Similarly, a specialized DNS server could be set up that will 2531 rate-limit the email coming from unexpected IP addresses. 2533 "v=spf1 mx exists:%{ir}._spf_rate.%{d} -all" 2535 * SPF allows the creation of per-user policies for special 2536 cases. For example, the following SPF record and appropriate 2537 wildcard DNS records can be used: 2539 "v=spf1 mx redirect=%{l1r+}._at_.%{o}._spf.%{d}" 2541 2. The middle, when email is forwarded (Mediators). 2543 * Mediators can solve the problem by rewriting the "MAIL FROM" 2544 to be in their own domain. This means mail rejected from the 2545 external mailbox will have to be forwarded back to the 2546 original sender by the forwarding service. Various schemes to 2547 do this exist though they vary widely in complexity and 2548 resource requirements on the part of the mediator. 2550 * Several popular MTAs can be forced from "alias" semantics to 2551 "mailing list" semantics by configuring an additional alias 2552 with "owner-" prepended to the original alias name (e.g., an 2553 alias of "friends: george@example.com, fred@example.org" would 2554 need another alias of the form "owner-friends: localowner"). 2556 * Mediators could reject mail that would "fail" SPF if forwarded 2557 using an SMTP reply code of 551, User not local, (see 2558 [RFC5321] section 3.4) to communicate the correct target 2559 address to resend the mail to. 2561 3. The end, when email is received (Receiving ADMDs). 2563 * If the owner of the external mailbox wishes to trust the 2564 mediator, he can direct the external mailbox's MTA to skip SPF 2565 tests when the client host belongs to the mediator. 2567 * Tests against other identities, such as the "HELO" identity, 2568 MAY be used to override a failed test against the "MAIL FROM" 2569 identity. 2571 * For larger domains, it might not be possible to have a 2572 complete or accurate list of forwarding services used by the 2573 owners of the domain's mailboxes. In such cases, whitelists 2574 of generally-recognized forwarding services could be employed. 2576 Appendix E. Mail Services 2578 MSPs (Mail Service Providers - [RFC5598] Section 2.3) that offer mail 2579 services to third-party domains, such as sending of bulk mail, might 2580 want to adjust their configurations in light of the authorization 2581 check described in this document. If the domain part of the "MAIL 2582 FROM" identity used for such email uses the domain of one of the MSPs 2583 domain, then the provider needs only to ensure that its sending host 2584 is authorized by its own SPF record, if any. 2586 If the "MAIL FROM" identity does not use the MSP's domain, then extra 2587 care has to be taken. The SPF record format has several options for 2588 the third-party domain to authorize the service provider's MTAs to 2589 send mail on its behalf. For MSPs, such as ISPs, that have a wide 2590 variety of customers using the same MTA, steps are required to 2591 mitiate the risk of cross-customer forgery (see Section 11.4). 2593 Appendix F. MTA Relays 2595 Relays are described in [RFC5598] Section 2.2.2. The authorization 2596 check generally precludes the use of arbitrary MTA relays between 2597 sender and receiver of an email message. 2599 Within an organization, MTA relays can be effectively deployed. 2600 However, for purposes of this document, such relays are effectively 2601 transparent. The SPF authorization check is a check between border 2602 MTAs of different ADMDs. 2604 For mail senders, this means that published SPF records have to 2605 authorize any MTAs that actually send across the Internet. Usually, 2606 these are just the border MTAs as internal MTAs simply forward mail 2607 to these MTAs for relaying. 2609 The receiving ADMD will generally want to perform the authorization 2610 check at the boundary MTAs, including all secondary MXs. Internal 2611 MTAs (including MTAs that might serve both as boundary MTAs and 2612 internal relays from secondary MXs when they are processing the 2613 relayed mail stream) then do not perform the authorization test. To 2614 perform the authorization test other than at the boundary, the host 2615 that first transferred the message to the receiving ADMD have to be 2616 determined, which can be difficult to extract from the message header 2617 because (a) header fields can be forged or malformed, and (b) there's 2618 no standard way to encode that information such that it can be 2619 reliably extracted. Testing other than at the boundary is likely to 2620 produce unreliable results. 2622 Appendix G. Local Policy Considerations 2624 SPF results can be used in combination with other methods to 2625 determine the final local disposition (either positive or negative of 2626 a message. It can also be considered dispositive on its own. 2628 G.1. Policy For SPF Pass 2630 SPF pass results can be used in combination with "white lists" of 2631 known "good" domains to bypass some or all additional pre-delivery 2632 email checks. Exactly which checks and how to determine appropriate 2633 white list entries has to be based on local conditions and 2634 requirements. 2636 G.2. Policy For SPF Fail 2638 SPF fail results can be used to reject messages during the SMTP 2639 transaction based on either "MAIL FROM" or "HELO" identity results. 2640 This reduces resource requirements for various content filtering 2641 methods and conserves bandwidth since rejection can be done before 2642 the SMTP content is transferred. It also gives immediate feedback to 2643 the sender who might then be able to resolve the issue. Due to some 2644 of the issues described above in this section (Section 10), SPF based 2645 rejection does present some risk of rejecting legitimate email when 2646 rejecting based on "MAIL FROM" results. 2648 SPF fail results can alternately be used as one input into a larger 2649 set of evaluations which might, based on a combination with other 2650 evaluation techniques, result in the email being marked negatively in 2651 some way (this might be via delivery to a special spam folder, 2652 modifying subject lines, or other locally determined means). 2653 Developing the details of such an approach have to be based on local 2654 conditions and requirements. Using SPF results in this way does not 2655 have the advantages of resource conservation and immediate feedback 2656 to the sender associated with SMTP rejection, but could produce fewer 2657 undesirable rejections in a well designed system. Such an approach 2658 might result in email that was not authorized by the sending ADMD 2659 being unknowingly delivered to end users. 2661 Either general approach can be used as they both leave a clear 2662 disposition of emails. They are either delivered in some manner or 2663 the sender is notified of the failure. Other dispositions such as 2664 "dropping" or deleting email after acceptance are inappropriate 2665 because they leave uncertainty and reduce the overall reliabilility 2666 and utility of email across the Internet. 2668 G.3. Policy For SPF Permerror 2670 The "permerror" result (see Section 2.6.7) indicates the SPF 2671 processing module at the receiver determined that the retrieved SPF 2672 policy record could not be interpreted. This gives no true 2673 indication about the authorized use of the data found in the 2674 envelope. 2676 As with all results, implementers have a choice to make regarding 2677 what to do with a message that yields this result. SMTP allows only 2678 a few basic options. 2680 Rejection of the message is an option, in that it is the one thing a 2681 receiver can do to draw attention to the difficulty encountered while 2682 protecting itself from messages that do not have a definite SPF 2683 result of some kind. However, if the SPF implementation is defective 2684 and returns spurious "permerror" results, only the sender is actively 2685 notified of the defect (in the form of rejected mail), and not the 2686 receiver making use of SPF. 2688 The less intrusive handling choice is to deliver the message, perhaps 2689 with some kind of annotation of the difficulty encountered and/or 2690 logging of a similar nature. However, this will not be desirable to 2691 operators that wish to implement SPF checking as strictly as 2692 possible, nor is this sort of passive problem reporting typically 2693 effective. 2695 There is of course the option placing this choice in the hands of the 2696 operator rather than the implementer since this kind of choice is 2697 often a matter of local policy rather than a condition with a 2698 universal solution, but this adds one more piece of complexity to an 2699 already non-trivial environment. 2701 Both implementers and operators need to be cautious of all choices 2702 and outcomes when handling SPF results. 2704 Appendix H. Protocol Status 2706 SPF has been in development since the summer of 2003 and has seen 2707 deployment beyond the developers beginning in December 2003. The 2708 design of SPF slowly evolved until the spring of 2004 and has since 2709 stabilized. There have been quite a number of forms of SPF, some 2710 written up as documents, some submitted as Internet Drafts, and many 2711 discussed and debated in development forums. The protocol was 2712 originally defined in [RFC4408], which this document replaces. 2714 [RFC4408] was designed to clearly document the protocol defined by 2715 earlier draft specifications of SPF as used in existing 2716 implementations. This updated specification is intended to clarify 2717 identified ambiguities in [RFC4408], resolve techincal issues 2718 identified in post-RFC 4408 deplyment experience, and document widely 2719 deployed extensions to SPF that have been developed since [RFC4408] 2720 was published. 2722 This document updates and replaces RFC 4408 that was part of a group 2723 of simultaneously published Experimental RFCs (RFC 4405, RFC 4406, 2724 RFC 4407, and RFC 4408) in 2006. At that time the IESG requested the 2725 community observe the success or failure of the two approaches 2726 documented in these RFCs during the two years following publication, 2727 in order that a community consensus could be reached in the future. 2729 SPF is widely deployed by large and small email providers alike. 2730 There are multiple, interoperable implementations. 2732 For SPF (as documented in RFC 4408) a careful effort was made to 2733 collect and document lessons learned and errata during the two year 2734 period. The errata list has been stable (no new submissions) and 2735 only minor protocol lessons learned were identified. Resolution of 2736 the IESG's experiment is documented in [RFC6686]. 2738 Appendix I. Change History 2740 Changes since RFC 4408 (to be removed prior to publication) 2742 Moved to standards track 2744 Authors updated 2746 IESG Note regarding experimental use replaced with discussion of 2747 results 2749 Process errata: 2751 Resolved Section 2.5.7 PermError on invalid domains after macro 2752 expansion errata in favor of documenting that different clients 2753 produce different results. 2755 Add %v macro to ABNF grammar 2757 Replace "uric" by "unreserved" 2759 Recommend an SMTP reply code for optional permerror rejections 2761 Correct syntax in Received-SPF examples 2763 Fix unknown-modifier clause is too greedy in ABNF 2765 Correct use of empty domain-spec on exp modifier 2767 Fix minor typo errata 2769 Convert to spfbis working group draft, 2770 draft-ietf-spfbis-4408bis-00 2772 Addressed Ticket #1, RFC 4408 Section 2.5.6 - Temporary errors by 2773 giving the option to turn repeated SERVFAIL into permerror and 2774 adding RFC 2308 reference. 2776 Clarified text about IPv4 mapped addresses to resolve test suite 2777 ambiguity 2779 Clarified ambiguity about result when more than 10 "mx" or "ptr" 2780 records are returned for lookup to specify permerror. This 2781 resolves one of the test suite ambiguities 2783 Made all references to result codes lower case per issue #7 2784 Adjusted section 2.2 Requirement to check mail from per issue #15 2786 Added missing "v" element in macro-letter in the collected ABNF 2787 per issue #16 - section 8.1 was already fixed in the pre-WG draft 2789 Marked ptr and "p" macro SHOULD NOT use per issue #27 2791 Expunged lower case may from the draft per issue #8 2793 Expunged "x-" name as an obsolete concept 2795 Updated obslete references: RFC2821 to RFC5321, RFC2822 to 2796 RFC5322, and RFC4234 to RFC5234 2798 Refer to RFC6647 to describe greylisting instead of trying to 2799 describe it directly. 2801 Updated informative references to the current versions. 2803 Start to rework section 9 with some RFC5598 terms. 2805 Added mention of RFC 6552 feedback reports in section 9. 2807 Added draft-ietf-spfbis-experiment as an informational reference. 2809 Drop Type SPF. 2811 Try and clarify informational nature of RFC3696 2813 Fix ABNF nits and add missing definitions per Bill's ABNF checker. 2815 Make DNS lookup time limit SHOULD instead of MAY. 2817 Reorganize and clarify processing limits. Move hard limits to new 2818 section 4.6.4, Evaluation Limits. Move advice to non-normative 2819 section 9. 2821 Removed paragraph in section 10.1 about limiting total data 2822 volumes as it is unused (and removable per the charter) and serves 2823 no purpose (it isn't something that actually can be implemented in 2824 any reasonable way). 2826 Added text and figures from Alessandro Vesely in section 9.1 to 2827 better explain DNS resource limits. 2829 Multiple editorial fixes from Murray Kucherawy's review. 2831 Also based on Murray's review, reworked SMTP identity definitions 2832 and made RFC 5598 a normative reference instead of informative. 2833 This is a downref that will have to be mentioned in the last call. 2835 Added RFC 3834 as an informative reference about backscatter. 2837 Added IDN requirements and normative reference to RFC 5890 to deal 2838 with the question "like DKIM did it.: 2840 Added informative reference to RFC 4632 for CIDR and use CIDR 2841 prefix length instead of CIDR-length to match its terminology. 2843 Simplified the exists description. 2845 Added text on creating a Authentication-Results header field that 2846 matches the Received-SPF header field information and added a 2847 normative reference to RFC 5451. 2849 Added informative reference to RFC 2782 due to SRV mention. 2851 Added informative reference to RFC 3464 due to DSN mention. 2853 Added informative reference to RFC 5617 for its DNS wildcard use. 2855 Clarified the intended match/no-match method for exists. 2857 Added new sections on Receiver policy for SPF pass, fail, and 2858 permerror. 2860 Added new section 9 discussion on treatment of bounces and the 2861 significance of HELO records. 2863 Added request to IANA to update the SPF modifier registry. 2865 Substantially reorganized the document for improved readability 2866 for new users based on WG consensus. 2868 Author's Address 2870 Scott Kitterman 2871 Kitterman Technical Services 2872 3611 Scheel Dr 2873 Ellicott City, MD 21042 2874 United States of America 2876 Email: scott@kitterman.com