en-US/about_SPF.help.txt

TOPIC
   about_SPF


Internet Engineering Task Force (IETF) S. Kitterman
Request for Comments: 7208 Kitterman Technical Services
Obsoletes: 4408 April 2014
Category: Standards Track
ISSN: 2070-1721


                     Sender Policy Framework (SPF)
           for Authorizing Use of Domains in Email, Version 1

Abstract

   Email on the Internet can be forged in a number of ways. In
   particular, existing protocols place no restriction on what a sending
   host can use as the "MAIL FROM" of a message or the domain given on
   the SMTP HELO/EHLO commands. This document describes version 1 of
   the Sender Policy Framework (SPF) protocol, whereby ADministrative
   Management Domains (ADMDs) can explicitly authorize the hosts that
   are allowed to use their domain names, and a receiving host can check
   such authorization.

   This document obsoletes RFC 4408.

Status of This Memo

   This is an Internet Standards Track document.

   This document is a product of the Internet Engineering Task Force
   (IETF). It represents the consensus of the IETF community. It has
   received public review and has been approved for publication by the
   Internet Engineering Steering Group (IESG). Further information on
   Internet Standards is available in Section 2 of RFC 5741.

   Information about the current status of this document, any errata,
   and how to provide feedback on it may be obtained at
   http://www.rfc-editor.org/info/rfc7208.















Kitterman Standards Track [Page 1]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


Copyright Notice

   Copyright (c) 2014 IETF Trust and the persons identified as the
   document authors. All rights reserved.

   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents
   (http://trustee.ietf.org/license-info) in effect on the date of
   publication of this document. Please review these documents
   carefully, as they describe your rights and restrictions with respect
   to this document. Code Components extracted from this document must
   include Simplified BSD License text as described in Section 4.e of
   the Trust Legal Provisions and are provided without warranty as
   described in the Simplified BSD License.

   This document may contain material from IETF Documents or IETF
   Contributions published or made publicly available before November
   10, 2008. The person(s) controlling the copyright in some of this
   material may not have granted the IETF Trust the right to allow
   modifications of such material outside the IETF Standards Process.
   Without obtaining an adequate license from the person(s) controlling
   the copyright in such materials, this document may not be modified
   outside the IETF Standards Process, and derivative works of it may
   not be created outside the IETF Standards Process, except to format
   it for publication as an RFC or to translate it into languages other
   than English.

Table of Contents

   1. Introduction ....................................................5
      1.1. Terminology ................................................5
           1.1.1. Key Words ...........................................5
           1.1.2. Imported Definitions ................................5
           1.1.3. MAIL FROM Definition ................................6
           1.1.4. HELO Definition .....................................6
      1.2. check_host() ...............................................6
   2. Operational Overview ............................................6
      2.1. Publishing Authorization ...................................6
      2.2. Checking Authorization .....................................7
      2.3. The "HELO" Identity ........................................8
      2.4. The "MAIL FROM" Identity ...................................9
      2.5. Location of Checks .........................................9
      2.6. Results of Evaluation ......................................9
           2.6.1. None ...............................................10
           2.6.2. Neutral ............................................10
           2.6.3. Pass ...............................................10
           2.6.4. Fail ...............................................10




Kitterman Standards Track [Page 2]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


           2.6.5. Softfail ...........................................10
           2.6.6. Temperror ..........................................10
           2.6.7. Permerror ..........................................10
   3. SPF Records ....................................................11
      3.1. DNS Resource Records ......................................11
      3.2. Multiple DNS Records ......................................12
      3.3. Multiple Strings in a Single DNS Record ...................12
      3.4. Record Size ...............................................13
      3.5. Wildcard Records ..........................................13
   4. The check_host() Function ......................................14
      4.1. Arguments .................................................14
      4.2. Results ...................................................15
      4.3. Initial Processing ........................................15
      4.4. Record Lookup .............................................15
      4.5. Selecting Records .........................................15
      4.6. Record Evaluation .........................................16
           4.6.1. Term Evaluation ....................................16
           4.6.2. Mechanisms .........................................16
           4.6.3. Modifiers ..........................................17
           4.6.4. DNS Lookup Limits ..................................17
      4.7. Default Result ............................................18
      4.8. Domain Specification ......................................19
   5. Mechanism Definitions ..........................................20
      5.1. "all" .....................................................21
      5.2. "include" .................................................21
      5.3. "a" .......................................................23
      5.4. "mx" ......................................................23
      5.5. "ptr" (do not use) ........................................23
      5.6. "ip4" and "ip6" ...........................................25
      5.7. "exists" ..................................................25
   6. Modifier Definitions ...........................................26
      6.1. redirect: Redirected Query ................................26
      6.2. exp: Explanation ..........................................27
   7. Macros .........................................................28
      7.1. Formal Specification ......................................29
      7.2. Macro Definitions .........................................29
      7.3. Macro Processing Details ..................................30
      7.4. Expansion Examples ........................................32
   8. Result Handling ................................................33
      8.1. None ......................................................34
      8.2. Neutral ...................................................34
      8.3. Pass ......................................................34
      8.4. Fail ......................................................35
      8.5. Softfail ..................................................35
      8.6. Temperror .................................................36
      8.7. Permerror .................................................36





Kitterman Standards Track [Page 3]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   9. Recording the Result ...........................................36
      9.1. The Received-SPF Header Field .............................37
      9.2. SPF Results in the Authentication-Results Header Field ....39
   10. Effects on Infrastructure .....................................39
      10.1. Sending Domains ..........................................40
           10.1.1. DNS Resource Considerations .......................40
           10.1.2. Administrator's Considerations ....................41
           10.1.3. Bounces ...........................................41
      10.2. Receivers ................................................42
      10.3. Mediators ................................................42
   11. Security Considerations .......................................43
      11.1. Processing Limits ........................................43
      11.2. SPF-Authorized Email May Contain Other False Identities ..44
      11.3. Spoofed DNS and IP Data ..................................44
      11.4. Cross-User Forgery .......................................44
      11.5. Untrusted Information Sources ............................45
           11.5.1. Recorded Results ..................................45
           11.5.2. External Explanations .............................45
           11.5.3. Macro Expansion ...................................46
      11.6. Privacy Exposure .........................................46
      11.7. Delivering Mail Producing a "Fail" Result ................46
   12. Collected ABNF ................................................46
   13. Contributors and Acknowledgements .............................48
   14. IANA Considerations ...........................................49
      14.1. The SPF DNS Record Type ..................................49
      14.2. The Received-SPF Mail Header Field .......................50
      14.3. SPF Modifier Registry ....................................50
   15. References ....................................................50
      15.1. Normative References .....................................50
      15.2. Informative References ...................................51
   Appendix A. Extended Examples .....................................54
     A.1. Simple Examples ............................................55
     A.2. Multiple Domain Example ....................................56
     A.3. DNS Blacklist (DNSBL) Style Example ........................56
     A.4. Multiple Requirements Example ..............................57
   Appendix B. Changes in Implementation Requirements from RFC 4408 ..57
   Appendix C. Further Testing Advice ................................58
   Appendix D. SPF/Mediator Interactions .............................59
     D.1. Originating ADMDs ..........................................59
     D.2. Mediators ..................................................60
     D.3. Receiving ADMDs ............................................60
   Appendix E. Mail Services .........................................61
   Appendix F. MTA Relays ............................................61
   Appendix G. Local Policy Considerations ...........................62
     G.1. Policy for SPF Pass ........................................62
     G.2. Policy for SPF Fail ........................................62
     G.3. Policy for SPF Permerror ...................................63
     G.4. Policy for SPF Temperror ...................................63



Kitterman Standards Track [Page 4]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


1. Introduction

   The current email infrastructure has the property that any host
   injecting mail into the system can use any DNS domain name it wants
   in each of the various identifiers specified by [RFC5321] and
   [RFC5322]. Although this feature is desirable in some circumstances,
   it is a major obstacle to reducing Unsolicited Bulk Email (UBE, aka
   spam). Furthermore, ADMDs (as described in [RFC5598]) are
   understandably concerned about the ease with which other entities can
   make use of their domain names, often with malicious intent.

   This document defines a protocol by which ADMDs can authorize hosts
   to use their domain names in the "MAIL FROM" or "HELO" identities.
   Compliant ADMDs publish Sender Policy Framework (SPF) records in the
   DNS specifying which hosts are permitted to use their names, and
   compliant mail receivers use the published SPF records to test the
   authorization of sending Mail Transfer Agents (MTAs) using a given
   "HELO" or "MAIL FROM" identity during a mail transaction.

   An additional benefit to mail receivers is that after the use of an
   identity is verified, local policy decisions about the mail can be
   made based on the sender's domain, rather than the host's IP address.
   This is advantageous because reputation of domain names is likely to
   be more accurate than reputation of host IP addresses since domains
   are likely to be more stable over a longer period. Furthermore, if a
   claimed identity fails verification, local policy can take stronger
   action against such email, such as rejecting it.

1.1. Terminology

1.1.1. Key Words

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
   "OPTIONAL" in this document are to be interpreted as described in
   [RFC2119].

1.1.2. Imported Definitions

   ABNF (Augmented Backus-Naur Form) ABNF is defined in [RFC5234], as
   are the tokens "ALPHA", "DIGIT", and "SP" (space).

   The tokens "Local-part", "Domain", and "Mailbox" are defined in
   [RFC5321].

   "dot-atom", "quoted-string", "comment", "CFWS" (comment folded white
   space), "FWS" (folded white space), and "CRLF" (carriage-return/
   line-feed) are defined in [RFC5322].



Kitterman Standards Track [Page 5]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


1.1.3. MAIL FROM Definition

   This document is concerned with the identity of the sender of a mail
   message, as referred to in [RFC5321]:

      The transaction starts with a MAIL command that gives the sender
      identification.

   Since there are many other names for this identity, it is important
   to choose a name that is:

   1. commonly used

   2. well defined

   As such, throughout this document the term "MAIL FROM" will be used,
   which is defined as the RFC5321.MailFrom (reverse-path) identity
   described in [RFC5598].

1.1.4. HELO Definition

   This document also makes use of the HELO/EHLO identity. The "HELO"
   identity derives from either the SMTP HELO or EHLO command (see
   [RFC5321]). Since HELO and EHLO can, in many cases, be used
   interchangeably, they are identified commonly as "HELO" in this
   document. This means RFC5321.HELO/.EHLO as defined in [RFC5598].
   These commands supply the identity of the SMTP client (sending host)
   for the SMTP session.

1.2. check_host()

   Section 4 introduces an algorithm to evaluate an SPF policy against
   an arriving email transaction. In an early implementation, this
   algorithm was encoded in a function called check_host(). That name
   is used in this document as symbolic of the SPF evaluation algorithm,
   but of course implementers are not required to use this name.

2. Operational Overview

2.1. Publishing Authorization

   An SPF-compliant domain publishes valid SPF records as described in
   Section 3. These records authorize the use of the relevant domain
   names in the "HELO" and "MAIL FROM" identities by the MTAs specified
   therein.






Kitterman Standards Track [Page 6]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   SPF results can be used to make both positive (source is authorized)
   and negative (source is not authorized) determinations. If ADMDs
   choose to publish SPF records and want to support receivers making
   negative authorization determinations, it is necessary for them to
   publish records that end in "-all", or redirect to other records that
   do; otherwise, no definitive determination of authorization can be
   made. Potential issues and mitigations associated with negative
   determinations are discussed in Section 10.

   ADMDs that wish to declare that no hosts are authorized to use their
   DNS domain names in the HELO or MAIL FROM commands during SMTP
   sessions can publish SPF records that say so for domain names that
   are neither used in the domain part of email addresses nor expected
   to originate mail.

   When changing SPF records, care has to be taken to ensure that there
   is a transition period so that the old policy remains valid until all
   legitimate email can reasonably expect to have been checked.
   [RFC5321], Section 4.5.4.1 discusses how long a message might be in
   transit. While offline checks are possible, the closer to the
   original transmission time checks are performed, the more likely they
   are to get an SPF result that matches the sending ADMD intent at the
   time the message was sent.

2.2. Checking Authorization

   A mail receiver can perform a set of SPF checks for each mail message
   it receives. An SPF check tests the authorization of a client host
   to emit mail with a given identity. Typically, such checks are done
   by a receiving MTA, but can be performed elsewhere in the mail
   processing chain so long as the required information is available and
   reliable. The "MAIL FROM" and "HELO" identities are checked as
   described in Sections 2.4 and 2.3, respectively.

   Without explicit approval of the publishing ADMD, checking other
   identities against SPF version 1 records is NOT RECOMMENDED because
   there are cases that are known to give incorrect results. For
   example, almost all mailing lists rewrite the "MAIL FROM" identity
   (see Section 10.3), but some do not change any other identities in
   the message. Documents that define other identities will have to
   define the method for explicit approval.

   It is possible that mail receivers will use the SPF check as part of
   a larger set of tests on incoming mail. The results of other tests
   might influence whether or not a particular SPF check is performed.
   For example, finding the sending host's IP address on a local
   whitelist might cause all other tests to be skipped and all mail from
   that host to be accepted.



Kitterman Standards Track [Page 7]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   When a mail receiver decides to perform an SPF check, it has to use a
   correctly implemented check_host() function (Section 4) evaluated
   with the correct parameters. Although the test as a whole is
   optional, once it has been decided to perform a test it has to be
   performed as specified so that the correct semantics are preserved
   between publisher and receiver.

   To make the test, the mail receiver MUST evaluate the check_host()
   function with the arguments described in Section 4.1.

   Although invalid, malformed, or non-existent domains cause SPF checks
   to return "none" because no SPF record can be found, it has long been
   the policy of many MTAs to reject email from such domains, especially
   in the case of invalid "MAIL FROM". Rejecting email will prevent one
   method of circumventing of SPF records.

   Implementations have to take care to correctly extract the <domain>
   from the data given with the SMTP MAIL FROM command as many MTAs will
   still accept such things as source routes (see Appendix C of
   [RFC5321]), the %-hack (see [RFC1123]), and bang paths (see
   [RFC1983]). These archaic features have been maliciously used to
   bypass security systems.

2.3. The "HELO" Identity

   It is RECOMMENDED that SPF verifiers not only check the "MAIL FROM"
   identity but also separately check the "HELO" identity by applying
   the check_host() function (Section 4) to the "HELO" identity as the
   <sender>. Checking "HELO" promotes consistency of results and can
   reduce DNS resource usage. If a conclusive determination about the
   message can be made based on a check of "HELO", then the use of DNS
   resources to process the typically more complex "MAIL FROM" can be
   avoided. Additionally, since SPF records published for "HELO"
   identities refer to a single host, when available, they are a very
   reliable source of host authorization status. Checking "HELO" before
   "MAIL FROM" is the RECOMMENDED sequence if both are checked.

   Note that requirements for the domain presented in the EHLO or HELO
   command are not always clear to the sending party, and SPF verifiers
   have to be prepared for the identity to be an IP address literal (see
   [RFC5321], Section 4.1.3) or simply be malformed. This SPF check can
   only be performed when the "HELO" string is a valid, multi-label
   domain name.








Kitterman Standards Track [Page 8]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


2.4. The "MAIL FROM" Identity

   SPF verifiers MUST check the "MAIL FROM" identity if a "HELO" check
   either has not been performed or has not reached a definitive policy
   result by applying the check_host() function to the "MAIL FROM"
   identity as the <sender>.

   [RFC5321] allows the reverse-path to be null (see Section 4.5.5 in
   [RFC5321]). In this case, there is no explicit sender mailbox, and
   such a message can be assumed to be a notification message from the
   mail system itself. When the reverse-path is null, this document
   defines the "MAIL FROM" identity to be the mailbox composed of the
   local-part "postmaster" and the "HELO" identity (which might or might
   not have been checked separately before).

2.5. Location of Checks

   The authorization check SHOULD be performed during the processing of
   the SMTP transaction that receives the mail. This reduces the
   complexity of determining the correct IP address to use as an input
   to check_host() and allows errors to be returned directly to the
   sending MTA by way of SMTP replies. Appendix D of [RFC7001] provides
   a more thorough discussion of this topic.

   The authorization check is performed during the SMTP transaction at
   the time of the MAIL command, and uses the MAIL FROM value and the
   client IP address. Performing the check at later times or with other
   input can cause problems such as the following:

   o It might be difficult to accurately extract the required
      information from potentially deceptive headers.

   o Legitimate email might fail the authorization check because the
      sender's policy has since changed.

   Generating non-delivery notifications to forged identities that have
   failed the authorization check often constitutes backscatter, i.e.,
   nuisance rejection notices that are not actionable. Operators are
   strongly advised to avoid such practices. Section 2 of [RFC3834]
   describes backscatter and the problems it causes.

2.6. Results of Evaluation

   Section 4 defines check_host(), a model function definition that uses
   the inputs defined above and the sender's policy published in the DNS
   to reach a conclusion about client authorization. An SPF verifier
   implements something semantically equivalent to the function defined
   there.



Kitterman Standards Track [Page 9]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   This section enumerates and briefly defines the possible outputs of
   that function. Note, however, that the protocol establishes no
   normative requirements for handling any particular result.
   Discussion of handling options for each result can be found in
   Section 8.

2.6.1. None

   A result of "none" means either (a) no syntactically valid DNS domain
   name was extracted from the SMTP session that could be used as the
   one to be authorized, or (b) no SPF records were retrieved from
   the DNS.

2.6.2. Neutral

   A "neutral" result means the ADMD has explicitly stated that it is
   not asserting whether the IP address is authorized.

2.6.3. Pass

   A "pass" result is an explicit statement that the client is
   authorized to inject mail with the given identity.

2.6.4. Fail

   A "fail" result is an explicit statement that the client is not
   authorized to use the domain in the given identity.

2.6.5. Softfail

   A "softfail" result is a weak statement by the publishing ADMD that
   the host is probably not authorized. It has not published a
   stronger, more definitive policy that results in a "fail".

2.6.6. Temperror

   A "temperror" result means the SPF verifier encountered a transient
   (generally DNS) error while performing the check. A later retry may
   succeed without further DNS operator action.

2.6.7. Permerror

   A "permerror" result means the domain's published records could not
   be correctly interpreted. This signals an error condition that
   definitely requires DNS operator intervention to be resolved.






Kitterman Standards Track [Page 10]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


3. SPF Records

   An SPF record is a DNS record that declares which hosts are, and are
   not, authorized to use a domain name for the "HELO" and "MAIL FROM"
   identities. Loosely, the record partitions hosts into permitted and
   not-permitted sets (though some hosts might fall into neither
   category).

   The SPF record is expressed as a single string of text found in the
   RDATA of a single DNS TXT resource record; multiple SPF records are
   not permitted for the same owner name. The record format and the
   process for selecting records are described below in Section 4. An
   example record is the following:

      v=spf1 +mx a:colo.example.com/28 -all

   This record has a version of "spf1" and three directives: "+mx",
   "a:colo.example.com/28" (the "+" is implied), and "-all".

   Each SPF record is placed in the DNS tree at the owner name it
   pertains to, not in a subdomain under the owner name. This is
   similar to how SRV records [RFC2782] are done.

   The example in this section might be published via these lines in a
   domain zone file:

      example.com. TXT "v=spf1 +mx a:colo.example.com/28 -all"

   Since TXT records have multiple uses, beware of other TXT records
   published there for other purposes. They might cause problems with
   size limits (see Section 3.4), and care has to be taken to ensure
   that only SPF records are used for SPF processing.

   ADMDs publishing SPF records ought to keep the amount of DNS
   information needed to evaluate a record to a minimum. Sections 4.6.4
   and 10.1.1 provide some suggestions about "include" mechanisms and
   chained "redirect" modifiers.

3.1. DNS Resource Records

   SPF records MUST be published as a DNS TXT (type 16) Resource Record
   (RR) [RFC1035] only. The character content of the record is encoded
   as [US-ASCII]. Use of alternative DNS RR types was supported in
   SPF's experimental phase but has been discontinued.

   In 2003, when SPF was first being developed, the requirements for
   assignment of a new DNS RR type were considerably more stringent than
   they are now. Additionally, support for easy deployment of new DNS



Kitterman Standards Track [Page 11]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   RR types was not widely deployed in DNS servers and provisioning
   systems. As a result, developers of SPF found it easier and more
   practical to use the TXT RR type for SPF records.

   In its review of [RFC4408], the SPFbis working group concluded that
   its dual RR type transition model was fundamentally flawed since it
   contained no common RR type that implementers were required to serve
   and required to check. Many alternatives were considered to resolve
   this issue, but ultimately the working group concluded that
   significant migration to the SPF RR type in the foreseeable future
   was very unlikely and that the best solution for resolving this
   interoperability issue was to drop support for the SPF RR type from
   SPF version 1. See Appendix A of [RFC6686] for further information.

   The circumstances surrounding SPF's initial deployment a decade ago
   are unique. If a future update to SPF were developed that did not
   reuse existing SPF records, it could use the SPF RR type. SPF's use
   of the TXT RR type for structured data should in no way be taken as
   precedent for future protocol designers. Further discussion of
   design considerations when using new DNS RR types can be found in
   [RFC5507].

3.2. Multiple DNS Records

   A domain name MUST NOT have multiple records that would cause an
   authorization check to select more than one record. See Section 4.5
   for the selection rules.

3.3. Multiple Strings in a Single DNS Record

   As defined in [RFC1035], Sections 3.3 and 3.3.14, a single text DNS
   record can be composed of more than one string. If a published
   record contains multiple character-strings, then the record MUST be
   treated as if those strings are concatenated together without adding
   spaces. For example:

      IN TXT "v=spf1 .... first" "second string..."

   is equivalent to:

      IN TXT "v=spf1 .... firstsecond string..."

   TXT records containing multiple strings are useful in constructing
   records that would exceed the 255-octet maximum length of a
   character-string within a single TXT record.






Kitterman Standards Track [Page 12]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


3.4. Record Size

   The published SPF record for a given domain name SHOULD remain small
   enough that the results of a query for it will fit within 512 octets.
   Otherwise, there is a possibility of exceeding a DNS protocol limit.
   This UDP limit is defined in [RFC1035], Section 2.3.4, although it
   was raised by [RFC2671]. Staying below 512 octets ought to prevent
   older DNS implementations from failing over to TCP and will work with
   UDP in the absence of EDNS0 [RFC6891] support. Since the answer size
   is dependent on many things outside the scope of this document, it is
   only possible to give this guideline: If the size of the DNS message,
   the combined length of the DNS name and the text of all the records
   of a given type is under 450 octets, then DNS answers ought to fit in
   UDP packets. Records that are too long to fit in a single UDP packet
   could be silently ignored by SPF verifiers due to firewall and other
   issues that interfere with the operation of DNS over TCP or using
   ENDS0.

   Note that when computing the sizes for replies to queries of the TXT
   format, one has to take into account any other TXT records published
   at the domain name. Similarly, the sizes for replies to all queries
   related to SPF have to be evaluated to fit in a single 512-octet UDP
   packet (i.e., DNS message size limited to 450 octets).

3.5. Wildcard Records

   Use of wildcard records for publishing is discouraged, and care has
   to be taken if they are used. If a zone includes wildcard MX
   records, it might want to publish wildcard declarations, subject to
   the same requirements and problems. In particular, the declaration
   MUST be repeated for any host that has any RR records at all, and for
   subdomains thereof. Consider the example in [RFC1034],
   Section 4.3.3. Based on that, we can do the following:

      EXAMPLE.COM. MX 10 A.EXAMPLE.COM
      EXAMPLE.COM. TXT "v=spf1 a:A.EXAMPLE.COM -all"

      *.EXAMPLE.COM. MX 10 A.EXAMPLE.COM
      *.EXAMPLE.COM. TXT "v=spf1 a:A.EXAMPLE.COM -all"

      A.EXAMPLE.COM. A 203.0.113.1
      A.EXAMPLE.COM. MX 10 A.EXAMPLE.COM
      A.EXAMPLE.COM. TXT "v=spf1 a:A.EXAMPLE.COM -all"

      *.A.EXAMPLE.COM. MX 10 A.EXAMPLE.COM
      *.A.EXAMPLE.COM. TXT "v=spf1 a:A.EXAMPLE.COM -all"





Kitterman Standards Track [Page 13]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   SPF records have to be listed twice for every name within the zone:
   once for the name, and once with a wildcard to cover the tree under
   the name, in order to cover all domains in use in outgoing mail.

4. The check_host() Function

   This description is not an application programming interface
   definition, but rather a function description used to illustrate the
   algorithm. A compliant SPF implementation MUST produce results
   semantically equivalent to this description.

   The check_host() function fetches SPF records, parses them, and
   evaluates them to determine whether a particular host is or is not
   permitted to send mail with a given identity. Receiving ADMDs that
   perform this check MUST correctly evaluate the check_host() function
   as described here.

   Implementations MAY use a different algorithm than the canonical
   algorithm defined here, so long as the results are the same in all
   cases.

4.1. Arguments

   The check_host() function takes these arguments:

   <ip> - the IP address of the SMTP client that is emitting
              the mail, either IPv4 or IPv6.

   <domain> - the domain that provides the sought-after authorization
              information; initially, the domain portion of the
              "MAIL FROM" or "HELO" identity.

   <sender> - the "MAIL FROM" or "HELO" identity.

   For recursive evaluations, the domain portion of <sender> might not
   be the same as the <domain> argument when check_host() is initially
   evaluated. In most other cases it will be the same (see Section 5.2
   below). The overall DNS lookup limit for SPF terms described below
   in Section 4.6.4 must be tracked as a single global limit for all
   evaluations, not just for a single instance of a recursive
   evaluation.

   Note that the <domain> argument might not be a well-formed domain
   name. For example, if the reverse-path was null, then the EHLO/HELO
   domain is used, with its associated problems (see Section 2.3). In
   these cases, check_host() is defined in Section 4.3 to return a
   "none" result.




Kitterman Standards Track [Page 14]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


4.2. Results

   The check_host() function can return one of several results described
   in Section 2.6. Based on the result, the action to be taken is
   determined by the local policies of the receiver. This is discussed
   in Section 8.

4.3. Initial Processing

   If the <domain> is malformed (e.g., label longer than 63 characters,
   zero-length label not at the end, etc.) or is not a multi-label
   domain name, or if the DNS lookup returns "Name Error" (RCODE 3, also
   known as "NXDOMAIN" [RFC2308]), check_host() immediately returns the
   result "none". DNS RCODEs are defined in [RFC1035]. Properly formed
   domains are fully qualified domains as defined in [RFC1983]. That
   is, in the DNS they are implicitly qualified relative to the root
   (see Section 3.1 of [RFC1034]). Internationalized domain names MUST
   be encoded as A-labels, as described in Section 2.3 of [RFC5890].

   If the <sender> has no local-part, substitute the string "postmaster"
   for the local-part.

4.4. Record Lookup

   In accordance with how the records are published (see Section 3
   above), a DNS query needs to be made for the <domain> name, querying
   for type TXT only.

   If the DNS lookup returns a server failure (RCODE 2) or some other
   error (RCODE other than 0 or 3), or if the lookup times out, then
   check_host() terminates immediately with the result "temperror".

4.5. Selecting Records

   Records begin with a version section:

   record = version terms *SP
   version = "v=spf1"

   Starting with the set of records that were returned by the lookup,
   discard records that do not begin with a version section of exactly
   "v=spf1". Note that the version section is terminated by either an
   SP character or the end of the record. As an example, a record with
   a version section of "v=spf10" does not match and is discarded.

   If the resultant record set includes no records, check_host()
   produces the "none" result. If the resultant record set includes
   more than one record, check_host() produces the "permerror" result.



Kitterman Standards Track [Page 15]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


4.6. Record Evaluation

   The check_host() function parses and interprets the SPF record to
   find a result for the current test. The syntax of the record is
   validated first, and if there are any syntax errors anywhere in the
   record, check_host() returns immediately with the result "permerror",
   without further interpretation or evaluation.

4.6.1. Term Evaluation

   There are two types of terms: mechanisms (defined in Section 5) and
   modifiers (defined in Section 6). A record contains an ordered list
   of these as specified in the following Augmented Backus-Naur Form
   (ABNF).

   terms = *( 1*SP ( directive / modifier ) )

   directive = [ qualifier ] mechanism
   qualifier = "+" / "-" / "?" / "~"
   mechanism = ( all / include
                      / a / mx / ptr / ip4 / ip6 / exists )
   modifier = redirect / explanation / unknown-modifier
   unknown-modifier = name "=" macro-string
                      ; where name is not any known modifier

   name = ALPHA *( ALPHA / DIGIT / "-" / "_" / "." )

   Most mechanisms allow a ":" or "/" character after the name.

   Modifiers always contain an equals ('=') character immediately after
   the name, and before any ":" or "/" characters that might be part of
   the macro-string.

   Terms that do not contain any of "=", ":", or "/" are mechanisms, as
   defined in Section 5.

   As per the definition of the ABNF notation in [RFC5234], mechanism
   and modifier names are case-insensitive.

4.6.2. Mechanisms

   Each mechanism is considered in turn from left to right. If there
   are no more mechanisms, the result is the default result as described
   in Section 4.7.

   When a mechanism is evaluated, one of three things can happen: it can
   match, not match, or return an exception.




Kitterman Standards Track [Page 16]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   If it matches, processing ends and the qualifier value is returned as
   the result of that record. If it does not match, processing
   continues with the next mechanism. If it returns an exception,
   mechanism processing ends and the exception value is returned.

   The possible qualifiers, and the results they cause check_host() to
   return, are as follows:

      "+" pass
      "-" fail
      "~" softfail
      "?" neutral

   The qualifier is optional and defaults to "+".

   When a mechanism matches and the qualifier is "-", then a "fail"
   result is returned and the explanation string is computed as
   described in Section 6.2.

   The specific mechanisms are described in Section 5.

4.6.3. Modifiers

   Modifiers are not mechanisms. They do not return match or not-match.
   Instead, they provide additional information. Although modifiers do
   not directly affect the evaluation of the record, the "redirect"
   modifier has an effect after all the mechanisms have been evaluated.

4.6.4. DNS Lookup Limits

   Some mechanisms and modifiers (collectively, "terms") cause DNS
   queries at the time of evaluation, and some do not. The following
   terms cause DNS queries: the "include", "a", "mx", "ptr", and
   "exists" mechanisms, and the "redirect" modifier. SPF
   implementations MUST limit the total number of those terms to 10
   during SPF evaluation, to avoid unreasonable load on the DNS. If
   this limit is exceeded, the implementation MUST return "permerror".
   The other terms -- the "all", "ip4", and "ip6" mechanisms, and the
   "exp" modifier -- do not cause DNS queries at the time of SPF
   evaluation (the "exp" modifier only causes a lookup at a later time),
   and their use is not subject to this limit.

   When evaluating the "mx" mechanism, the number of "MX" resource
   records queried is included in the overall limit of 10 mechanisms/
   modifiers that cause DNS lookups as described above. In addition to
   that limit, the evaluation of each "MX" record MUST NOT result in





Kitterman Standards Track [Page 17]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   querying more than 10 address records -- either "A" or "AAAA"
   resource records. If this limit is exceeded, the "mx" mechanism MUST
   produce a "permerror" result.

   When evaluating the "ptr" mechanism or the %{p} macro, the number of
   "PTR" resource records queried is included in the overall limit of 10
   mechanisms/modifiers that cause DNS lookups as described above. In
   addition to that limit, the evaluation of each "PTR" record MUST NOT
   result in querying more than 10 address records -- either "A" or
   "AAAA" resource records. If this limit is exceeded, all records
   other than the first 10 MUST be ignored.

   The reason for the disparity is that the set of and contents of the
   MX record are under control of the publishing ADMD, while the set of
   and contents of PTR records are under control of the owner of the IP
   address actually making the connection.

   These limits are per mechanism or macro in the record, and are in
   addition to the lookup limits specified above.

   MTAs or other processors SHOULD impose a limit on the maximum amount
   of elapsed time to evaluate check_host(). Such a limit SHOULD allow
   at least 20 seconds. If such a limit is exceeded, the result of
   authorization SHOULD be "temperror".

   As described at the end of Section 11.1, there may be cases where it
   is useful to limit the number of "terms" for which DNS queries return
   either a positive answer (RCODE 0) with an answer count of 0, or a
   "Name Error" (RCODE 3) answer. These are sometimes collectively
   referred to as "void lookups". SPF implementations SHOULD limit
   "void lookups" to two. An implementation MAY choose to make such a
   limit configurable. In this case, a default of two is RECOMMENDED.
   Exceeding the limit produces a "permerror" result.

4.7. Default Result

   If none of the mechanisms match and there is no "redirect" modifier,
   then the check_host() returns a result of "neutral", just as if
   "?all" were specified as the last directive. If there is a
   "redirect" modifier, check_host() proceeds as defined in Section 6.1.

   It is better to use either a "redirect" modifier or an "all"
   mechanism to explicitly terminate processing. Although there is an
   implicit "?all" at the end of every record that is not explicitly
   terminated, it aids debugging efforts when it is explicitly provided.






Kitterman Standards Track [Page 18]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   For example:

      v=spf1 +mx -all

   or

      v=spf1 +mx redirect=_spf.example.com

4.8. Domain Specification

   Several of these mechanisms and modifiers have a <domain-spec>
   section. The <domain-spec> string is subject to macro expansion (see
   Section 7). The resulting string is the common presentation form of
   a fully qualified DNS name: a series of labels separated by periods.
   This domain is called the <target-name> in the rest of this document.

   Note: The result of the macro expansion is not subject to any further
   escaping. Hence, this facility cannot produce all characters that
   are legal in a DNS label (e.g., the control characters). However,
   this facility is powerful enough to express legal host names and
   common utility labels (such as "_spf") that are used in DNS.

   For several mechanisms, the <domain-spec> is optional. If it is not
   provided, the <domain> from the check_host() arguments (see
   Section 4.1) is used as the <target-name>. "domain" and
   <domain-spec> are syntactically identical after macro expansion.
   "domain" is an input value for check_host(), while <domain-spec> is
   computed by check_host().

   The result of evaluating check_host() with a syntactically invalid
   domain is undefined.

   Note: This document and its predecessors make no provisions for
   defining correct handling of a syntactically invalid <domain-spec>
   (which might be the result of macro expansion), per [RFC1035].
   Examples include names with empty labels, such as "foo..example.com",
   and labels that are longer than 63 characters. Some implementations
   choose to treat such errors as not-match and therefore ignore such
   names, while others return a "permerror" exception.












Kitterman Standards Track [Page 19]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


5. Mechanism Definitions

   This section defines two types of mechanisms: basic language
   framework mechanisms and designated sender mechanisms.

   Basic mechanisms contribute to the language framework. They do not
   specify a particular type of authorization scheme. The basic
   mechanisms are as follows:

      all
      include

   Designated sender mechanisms are used to identify a set of <ip>
   addresses as being permitted or not permitted to use the <domain> for
   sending mail. The designated sender mechanisms are as follows:

      a
      mx
      ptr (do not use)
      ip4
      ip6
      exists

   The following conventions apply to all mechanisms that perform a
   comparison between <ip> and an IP address at any point:

   If no CIDR prefix length is given in the directive, then <ip> and the
   IP address are compared for equality. (Here, CIDR is Classless
   Inter-Domain Routing, described in [RFC4632].)

   If a CIDR prefix length is specified, then only the specified number
   of high-order bits of <ip> and the IP address are compared for
   equality.

   When any mechanism fetches host addresses to compare with <ip>, when
   <ip> is an IPv4, "A" records are fetched; when <ip> is an IPv6
   address, "AAAA" records are fetched. SPF implementations on IPv6
   servers need to handle both "AAAA" and "A" records, for clients on
   IPv4-mapped IPv6 addresses [RFC4291]. IPv4 <ip> addresses are only
   listed in an SPF record using the "ip4" mechanism.

   Several mechanisms rely on information fetched from the DNS. For
   these DNS queries, except where noted, if the DNS server returns an
   error (RCODE other than 0 or 3) or the query times out, the mechanism
   stops and the topmost check_host() returns "temperror". If the
   server returns "Name Error" (RCODE 3), then evaluation of the
   mechanism continues as if the server returned no error (RCODE 0) and
   zero answer records.



Kitterman Standards Track [Page 20]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


5.1. "all"

   all = "all"

   The "all" mechanism is a test that always matches. It is used as the
   rightmost mechanism in a record to provide an explicit default.

   For example:

      v=spf1 a mx -all

   Mechanisms after "all" will never be tested. Mechanisms listed after
   "all" MUST be ignored. Any "redirect" modifier (Section 6.1) MUST be
   ignored when there is an "all" mechanism in the record, regardless of
   the relative ordering of the terms.

5.2. "include"

   include = "include" ":" domain-spec

   The "include" mechanism triggers a recursive evaluation of
   check_host().

   1. The <domain-spec> is expanded as per Section 7.

   2. check_host() is evaluated with the resulting string as the
       <domain>. The <ip> and <sender> arguments remain the same as in
       the current evaluation of check_host().

   3. The recursive evaluation returns match, not-match, or an error.

   4. If it returns match, then the appropriate result for the
       "include" mechanism is used (e.g., include or +include produces a
       "pass" result and -include produces "fail").

   5. If it returns not-match or an error, the parent check_host()
       resumes processing as per the table below, with the previous
       value of <domain> restored.

   In hindsight, the name "include" was poorly chosen. Only the
   evaluated result of the referenced SPF record is used, rather than
   literally including the mechanisms of the referenced record in the
   first. For example, evaluating a "-all" directive in the referenced
   record does not terminate the overall processing and does not
   necessarily result in an overall "fail". (Better names for this
   mechanism would have been "if-match", "on-match", etc.)





Kitterman Standards Track [Page 21]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   The "include" mechanism makes it possible for one domain to designate
   multiple administratively independent domains. For example, a vanity
   domain "example.net" might send mail using the servers of
   administratively independent domains example.com and example.org.

   Example.net could say

      IN TXT "v=spf1 include:example.com include:example.org -all"

   This would direct check_host() to, in effect, check the records of
   example.com and example.org for a "pass" result. Only if the host
   were not permitted for either of those domains would the result be
   "fail".

   Whether this mechanism matches, does not match, or returns an
   exception depends on the result of the recursive evaluation of
   check_host():

   +---------------------------------+---------------------------------+
   | A recursive check_host() result | Causes the "include" mechanism |
   | of: | to: |
   +---------------------------------+---------------------------------+
   | pass | match |
   | | |
   | fail | not match |
   | | |
   | softfail | not match |
   | | |
   | neutral | not match |
   | | |
   | temperror | return temperror |
   | | |
   | permerror | return permerror |
   | | |
   | none | return permerror |
   +---------------------------------+---------------------------------+

   The "include" mechanism is intended for crossing administrative
   boundaries. When remaining within one administrative authority,
   "include" is usually not the best choice. For example, if
   example.com and example.org were managed by the same entity, and if
   the permitted set of hosts for both domains was "mx:example.com", it
   would be possible for example.org to specify "include:example.com",
   but it would be preferable to specify "redirect=example.com" or even
   "mx:example.com".






Kitterman Standards Track [Page 22]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   With the "include" mechanism, an administratively external set of
   hosts can be authorized, but determination of sender policy is still
   a function of the original domain's SPF record (as determined by the
   "all" mechanism in that record). The "redirect" modifier is more
   suitable for consolidating both authorizations and policy into a
   common set to be shared within an ADMD. Redirect is much more like a
   common code element to be shared among records in a single ADMD. It
   is possible to control both authorized hosts and policy for an
   arbitrary number of domains from a single record.

5.3. "a"

   This mechanism matches if <ip> is one of the <target-name>'s IP
   addresses. For clarity, this means the "a" mechanism also matches
   AAAA records.

   a = "a" [ ":" domain-spec ] [ dual-cidr-length ]

   An address lookup is done on the <target-name> using the type of
   lookup (A or AAAA) appropriate for the connection type (IPv4 or
   IPv6). The <ip> is compared to the returned address(es). If any
   address matches, the mechanism matches.

5.4. "mx"

   This mechanism matches if <ip> is one of the MX hosts for a domain
   name.

   mx = "mx" [ ":" domain-spec ] [ dual-cidr-length ]

   check_host() first performs an MX lookup on the <target-name>. Then
   it performs an address lookup on each MX name returned. The <ip> is
   compared to each returned IP address. To prevent denial-of-service
   (DoS) attacks, the processing limits defined in Section 4.6.4 MUST be
   followed. If the MX lookup limit is exceeded, then "permerror" is
   returned and the evaluation is terminated. If any address matches,
   the mechanism matches.

   Note regarding implicit MXes: If the <target-name> has no MX record,
   check_host() MUST NOT apply the implicit MX rules of [RFC5321] by
   querying for an A or AAAA record for the same name.

5.5. "ptr" (do not use)

   This mechanism tests whether the DNS reverse-mapping for <ip> exists
   and correctly points to a domain name within a particular domain.
   This mechanism SHOULD NOT be published. See the note at the end of
   this section for more information.



Kitterman Standards Track [Page 23]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   ptr = "ptr" [ ":" domain-spec ]

   The <ip>'s name is looked up using this procedure:

   o Perform a DNS reverse-mapping for <ip>: Look up the corresponding
      PTR record in "in-addr.arpa." if the address is an IPv4 address
      and in "ip6.arpa." if it is an IPv6 address.

   o For each record returned, validate the domain name by looking up
      its IP addresses. To prevent DoS attacks, the PTR processing
      limits defined in Section 4.6.4 MUST be applied. If they are
      exceeded, processing is terminated and the mechanism does not
      match.

   o If <ip> is among the returned IP addresses, then that domain name
      is validated.

   Check all validated domain names to see if they either match the
   <target-name> domain or are a subdomain of the <target-name> domain.
   If any do, this mechanism matches. If no validated domain name can
   be found, or if none of the validated domain names match or are a
   subdomain of the <target-name>, this mechanism fails to match. If a
   DNS error occurs while doing the PTR RR lookup, then this mechanism
   fails to match. If a DNS error occurs while doing an A RR lookup,
   then that domain name is skipped and the search continues.

   This mechanism matches if

   o the <target-name> is a subdomain of a validated domain name, or

   o the <target-name> and a validated domain name are the same.

   For example, "mail.example.com" is within the domain "example.com",
   but "mail.bad-example.com" is not.

   Note: This mechanism is slow, it is not as reliable as other
   mechanisms in cases of DNS errors, and it places a large burden on
   the .arpa name servers. If used, proper PTR records have to be in
   place for the domain's hosts and the "ptr" mechanism SHOULD be one of
   the last mechanisms checked. After many years of SPF deployment
   experience, it has been concluded that it is unnecessary and more
   reliable alternatives should be used instead. It is, however, still
   in use as part of the SPF protocol, so compliant check_host()
   implementations MUST support it.







Kitterman Standards Track [Page 24]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


5.6. "ip4" and "ip6"

   These mechanisms test whether <ip> is contained within a given
   IP network.

   ip4 = "ip4" ":" ip4-network [ ip4-cidr-length ]
   ip6 = "ip6" ":" ip6-network [ ip6-cidr-length ]

   ip4-cidr-length = "/" ("0" / %x31-39 0*1DIGIT) ; value range 0-32
   ip6-cidr-length = "/" ("0" / %x31-39 0*2DIGIT) ; value range 0-128
   dual-cidr-length = [ ip4-cidr-length ] [ "/" ip6-cidr-length ]

   ip4-network = qnum "." qnum "." qnum "." qnum
   qnum = DIGIT ; 0-9
                      / %x31-39 DIGIT ; 10-99
                      / "1" 2DIGIT ; 100-199
                      / "2" %x30-34 DIGIT ; 200-249
                      / "25" %x30-35 ; 250-255
            ; as per conventional dotted-quad notation, e.g., 192.0.2.0

   ip6-network = <as per Section 2.2 of [RFC4291]>
            ; e.g., 2001:db8::cd30

   The <ip> is compared to the given network. If CIDR prefix length
   high-order bits match, the mechanism matches.

   If ip4-cidr-length is omitted, it is taken to be "/32". If
   ip6-cidr-length is omitted, it is taken to be "/128". It is not
   permitted to omit parts of the IP address instead of using CIDR
   notations. That is, use 192.0.2.0/24 instead of 192.0.2.

5.7. "exists"

   This mechanism is used to construct an arbitrary domain name that is
   used for a DNS A record query. It allows for complicated schemes
   involving arbitrary parts of the mail envelope to determine what is
   permitted.

   exists = "exists" ":" domain-spec

   The <domain-spec> is expanded as per Section 7. The resulting domain
   name is used for a DNS A RR lookup (even when the connection type is
   IPv6). If any A record is returned, this mechanism matches.

   Domains can use this mechanism to specify arbitrarily complex
   queries. For example, suppose example.com publishes the record:

      v=spf1 exists:%{ir}.%{l1r+-}._spf.%{d} -all



Kitterman Standards Track [Page 25]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   The <target-name> might expand to
   "1.2.0.192.someuser._spf.example.com". This makes fine-grained
   decisions possible at the level of the user and client IP address.

6. Modifier Definitions

   Modifiers are name/value pairs that provide additional information.
   Modifiers always have an "=" separating the name and the value.

   The modifiers defined in this document ("redirect" and "exp") SHOULD
   appear at the end of the record, after all mechanisms, though
   syntactically they can appear anywhere in the record. Ordering of
   these two modifiers does not matter. These two modifiers MUST NOT
   appear in a record more than once each. If they do, then
   check_host() exits with a result of "permerror".

   Unrecognized modifiers MUST be ignored no matter where, or how often,
   they appear in a record. This allows implementations conforming to
   this document to gracefully handle records with modifiers that are
   defined in other specifications.

6.1. redirect: Redirected Query

   The "redirect" modifier is intended for consolidating both
   authorizations and policy into a common set to be shared within a
   single ADMD. It is possible to control both authorized hosts and
   policy for an arbitrary number of domains from a single record.

   redirect = "redirect" "=" domain-spec

   If all mechanisms fail to match, and a "redirect" modifier is
   present, then processing proceeds as follows:

   The <domain-spec> portion of the redirect section is expanded as per
   the macro rules in Section 7. Then check_host() is evaluated with
   the resulting string as the <domain>. The <ip> and <sender>
   arguments remain the same as in the current evaluation of
   check_host().

   The result of this new evaluation of check_host() is then considered
   the result of the current evaluation with the exception that if no
   SPF record is found, or if the <target-name> is malformed, the result
   is a "permerror" rather than "none".

   Note that the newly queried domain can itself specify redirect
   processing.





Kitterman Standards Track [Page 26]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   This facility is intended for use by organizations that wish to apply
   the same record to multiple domains. For example:

     la.example.com. TXT "v=spf1 redirect=_spf.example.com"
     ny.example.com. TXT "v=spf1 redirect=_spf.example.com"
     sf.example.com. TXT "v=spf1 redirect=_spf.example.com"
   _spf.example.com. TXT "v=spf1 mx:example.com -all"

   In this example, mail from any of the three domains is described by
   the same record. This can be an administrative advantage.

   Note: In general, the domain "A" cannot reliably use a redirect to
   another domain "B" not under the same administrative control. Since
   the <sender> stays the same, there is no guarantee that the record at
   domain "B" will correctly work for mailboxes in domain "A",
   especially if domain "B" uses mechanisms involving local-parts. An
   "include" directive will generally be more appropriate.

   For clarity, any "redirect" modifier SHOULD appear as the very last
   term in a record. Any "redirect" modifier MUST be ignored if there
   is an "all" mechanism anywhere in the record.

6.2. exp: Explanation

   explanation = "exp" "=" domain-spec

   If check_host() results in a "fail" due to a mechanism match (such as
   "-all"), and the "exp" modifier is present, then the explanation
   string returned is computed as described below. If no "exp" modifier
   is present, then either a default explanation string or an empty
   explanation string MUST be returned to the calling application.

   The <domain-spec> is macro expanded (see Section 7) and becomes the
   <target-name>. The DNS TXT RRset for the <target-name> is fetched.

   If there are any DNS processing errors (any RCODE other than 0), or
   if no records are returned, or if more than one record is returned,
   or if there are syntax errors in the explanation string, then proceed
   as if no "exp" modifier was given.

   The fetched TXT record's strings are concatenated with no spaces, and
   then treated as an explain-string, which is macro-expanded. This
   final result is the explanation string. Implementations MAY limit
   the length of the resulting explanation string to allow for other
   protocol constraints and/or reasonable processing limits. Since the
   explanation string is intended for an SMTP response and Section 2.4
   of [RFC5321] says that responses are in [US-ASCII], the explanation
   string MUST be limited to [US-ASCII].



Kitterman Standards Track [Page 27]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   Software evaluating check_host() can use this string to communicate
   information from the publishing domain in the form of a short message
   or URL. Software SHOULD make it clear that the explanation string
   comes from a third party. For example, it can prepend the macro
   string "%{o} explains: " to the explanation, as shown in the example
   in Section 8.4.

   Suppose example.com has this record:

      v=spf1 mx -all exp=explain._spf.%{d}

   Here are some examples of possible explanation TXT records at
   explain._spf.example.com:

      "Mail from example.com should only be sent by its own servers."

         -- a simple, constant message

      "%{i} is not one of %{d}'s designated mail servers."

         -- a message with a little more information, including the
            IP address that failed the check

      "See http://%{d}/why.html?s=%{S}&i=%{I}"

         -- a complicated example that constructs a URL with the
            arguments to check_host() so that a web page can be
            generated with detailed, custom instructions

   Note: During recursion into an "include" mechanism, an "exp" modifier
   from the <target-name> MUST NOT be used. In contrast, when executing
   a "redirect" modifier, an "exp" modifier from the original domain
   MUST NOT be used. This is because "include" is meant to cross
   administrative boundaries and the explanation provided should be the
   one from the receiving ADMD, while "redirect" is meant to operate as
   a tool to consolidate policy records within an ADMD so the redirected
   explanation is the one that ought to have priority.

7. Macros

   When evaluating an SPF policy record, certain character sequences are
   intended to be replaced by parameters of the message or of the
   connection. These character sequences are referred to as "macros".








Kitterman Standards Track [Page 28]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


7.1. Formal Specification

   The ABNF description for a macro is as follows:

   domain-spec = macro-string domain-end
   domain-end = ( "." toplabel [ "." ] ) / macro-expand

   toplabel = ( *alphanum ALPHA *alphanum ) /
                      ( 1*alphanum "-" *( alphanum / "-" ) alphanum )
   alphanum = ALPHA / DIGIT

   explain-string = *( macro-string / SP )

   macro-string = *( macro-expand / macro-literal )
   macro-expand = ( "%{" macro-letter transformers *delimiter "}" )
                      / "%%" / "%_" / "%-"
   macro-literal = %x21-24 / %x26-7E
                      ; visible characters except "%"
   macro-letter = "s" / "l" / "o" / "d" / "i" / "p" / "h" /
                      "c" / "r" / "t" / "v"
   transformers = *DIGIT [ "r" ]
   delimiter = "." / "-" / "+" / "," / "/" / "_" / "="

   The "toplabel" construction is subject to the letter-digit-hyphen
   (LDH) rule plus additional top-level domain (TLD) restrictions. See
   Section 2 of [RFC3696] for background.

   Some special cases:

   o A literal "%" is expressed by "%%".

   o "%_" expands to a single " " space.

   o "%-" expands to a URL-encoded space, viz., "%20".

7.2. Macro Definitions

   The following macro letters are expanded in term arguments:

      s = <sender>
      l = local-part of <sender>
      o = domain of <sender>
      d = <domain>
      i = <ip>
      p = the validated domain name of <ip> (do not use)
      v = the string "in-addr" if <ip> is ipv4, or "ip6" if <ip> is ipv6
      h = HELO/EHLO domain




Kitterman Standards Track [Page 29]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   <domain>, <sender>, and <ip> are defined in Section 4.1.

   The following macro letters are allowed only in "exp" text:

      c = SMTP client IP (easily readable format)
      r = domain name of host performing the check
      t = current timestamp

7.3. Macro Processing Details

   A '%' character not followed by a '{', '%', '-', or '_' character is
   a syntax error. So:

      -exists:%(ir).sbl.example.org

   is incorrect and will cause check_host() to yield a "permerror".
   Instead, the following is legal:

      -exists:%{ir}.sbl.example.org

   Optional transformers are the following:

      *DIGIT = zero or more digits

      'r' = reverse value, splitting on dots by default

   If transformers or delimiters are provided, the replacement value for
   a macro letter is split into parts separated by one or more of the
   specified delimiter characters. After performing any reversal
   operation and/or removal of left-hand parts, the parts are rejoined
   using "." and not the original splitting characters.

   By default, strings are split on "." (dots). Note that no special
   treatment is given to leading, trailing, or consecutive delimiters in
   input strings, and so the list of parts might contain empty strings.
   Some older implementations of SPF prohibit trailing dots in domain
   names, so trailing dots SHOULD NOT be published, although they MUST
   be accepted by implementations conforming to this document. Macros
   can specify delimiter characters that are used instead of ".".

   The "r" transformer indicates a reversal operation: if the client IP
   address were 192.0.2.1, the macro %{i} would expand to "192.0.2.1"
   and the macro %{ir} would expand to "1.2.0.192".

   The DIGIT transformer indicates the number of right-hand parts to
   use, after optional reversal. If a DIGIT is specified, the value
   MUST be nonzero. If no DIGITs are specified, or if the value
   specifies more parts than are available, all the available parts are



Kitterman Standards Track [Page 30]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   used. If the DIGIT was 5, and only 3 parts were available, the macro
   interpreter would pretend the DIGIT was 3. Implementations MUST
   support at least a value of 127, as that is the maximum number of
   labels in a domain name (less the zero-length label at the end).

   The "s" macro expands to the <sender> argument. It is an email
   address with a local-part, an "@" character, and a domain. The "l"
   macro expands to just the local-part. The "o" macro expands to just
   the domain part. Note that these values remain the same during
   recursive and chained evaluations due to "include" and/or "redirect".
   Note also that if the original <sender> had no local-part, the
   local-part was set to "postmaster" in initial processing (see
   Section 4.3).

   For IPv4 addresses, both the "i" and "c" macros expand to the
   standard dotted-quad format.

   For IPv6 addresses, the "i" macro expands to a dot-format address; it
   is intended for use in %{ir}. The "c" macro can expand to any of the
   hexadecimal colon-format addresses specified in Section 2.2 of
   [RFC4291]. It is intended for humans to read.

   The "p" macro expands to the validated domain name of <ip>. The
   procedure for finding the validated domain name is defined in
   Section 5.5. If the <domain> is present in the list of validated
   domains, it SHOULD be used. Otherwise, if a subdomain of the
   <domain> is present, it SHOULD be used. Otherwise, any name from the
   list can be used. If there are no validated domain names or if a DNS
   error occurs, the string "unknown" is used.

   This macro SHOULD NOT be published (see Section 5.5 for the
   discussion).

   The "h" macro expands to the parameter that was provided to the SMTP
   server via the HELO or EHLO SMTP verb. For sessions where that verb
   was provided more than once, the most recent instance is used.

   The "r" macro expands to the name of the receiving MTA. This SHOULD
   be a fully qualified domain name, but if one does not exist (as when
   the checking is done by a Mail User Agent (MUA)) or if policy
   restrictions dictate otherwise, the word "unknown" SHOULD be
   substituted. The domain name can be different from the name found in
   the MX record that the client MTA used to locate the receiving MTA.








Kitterman Standards Track [Page 31]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   The "t" macro expands to the decimal representation of the
   approximate number of seconds since the Epoch (Midnight, January 1,
   1970, UTC) at the time of the evaluation. This is the same value as
   the value that is returned by the Portable Operating System Interface
   (POSIX) time() function in most standards-compliant libraries.

   When the result of macro expansion is used in a domain name query, if
   the expanded domain name exceeds 253 characters (the maximum length
   of a domain name in this format), the left side is truncated to fit,
   by removing successive domain labels (and their following dots) until
   the total length does not exceed 253 characters.

   Uppercase macros expand exactly as their lowercase equivalents, and
   are then URL escaped. URL escaping MUST be performed for characters
   not in the "unreserved" set, which is defined in [RFC3986].

   Care has to be taken by the sending ADMD so that macro expansion for
   legitimate email does not exceed the 63-character limit on DNS
   labels. The local-part of email addresses, in particular, can have
   more than 63 characters between dots.

   To minimize DNS lookup resource requirements, it is better if sending
   ADMDs avoid using the "s", "l", "o", or "h" macros in conjunction
   with any mechanism directive. Although these macros are powerful and
   allow per-user records to be published, they severely limit the
   ability of implementations to cache results of check_host() and they
   reduce the effectiveness of DNS caches.

   If no directive processed during the evaluation of check_host()
   contains an "s", "l", "o", or "h" macro, then the results of the
   evaluation can be cached on the basis of <domain> and <ip> alone for
   as long as the DNS record involved with the shortest Time to Live
   (TTL) has not expired.

7.4. Expansion Examples

   The <sender> is strong-bad@email.example.com. The IPv4 SMTP client
   IP is 192.0.2.3. The IPv6 SMTP client IP is 2001:db8::cb01. The PTR
   domain name of the client IP is mx.example.org.












Kitterman Standards Track [Page 32]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   macro expansion
   ------- ----------------------------
   %{s} strong-bad@email.example.com
   %{o} email.example.com
   %{d} email.example.com
   %{d4} email.example.com
   %{d3} email.example.com
   %{d2} example.com
   %{d1} com
   %{dr} com.example.email
   %{d2r} example.email
   %{l} strong-bad
   %{l-} strong.bad
   %{lr} strong-bad
   %{lr-} bad.strong
   %{l1r-} strong

   macro-string expansion
   --------------------------------------------------------------------
   %{ir}.%{v}._spf.%{d2} 3.2.0.192.in-addr._spf.example.com
   %{lr-}.lp._spf.%{d2} bad.strong.lp._spf.example.com

   %{lr-}.lp.%{ir}.%{v}._spf.%{d2}
                       bad.strong.lp.3.2.0.192.in-addr._spf.example.com

   %{ir}.%{v}.%{l1r-}.lp._spf.%{d2}
                           3.2.0.192.in-addr.strong.lp._spf.example.com

   %{d2}.trusted-domains.example.net
                                example.com.trusted-domains.example.net

   IPv6:
   %{ir}.%{v}._spf.%{d2} 1.0.b.c.0.0.0.0.
   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

8. Result Handling

   This section provides guidance for SPF verifier operators in response
   to the various possible outputs of check_host() on a message.
   Definitions of SPF results are presented in Section 2.6; this section
   provides more detail on each for use in developing local policy for
   message handling.

   Every operating environment is different. There are some receivers
   for whom strict adherence to SPF is appropriate, and definitive
   treatment of messages that are evaluated to be explicitly
   unauthorized ("fail" and sometimes "softfail") is the norm. There
   are others for which the "false negative" cases are more of a



Kitterman Standards Track [Page 33]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   concern. This concern is typically handled by merely recording the
   result in the header and allowing the message to pass on for
   additional processing. There are still others where SPF is one of
   several inputs to the message-handling decision. As such, there is
   no comprehensive normative requirement for message handling in
   response to any particular result. This section is provided to
   present a complete picture of the likely cause of each result and,
   where available, the experience gained during experimental
   deployment.

   There are essentially two classes of handling choices:

   o Handling within the SMTP session that attempted to deliver the
      message, such as by returning a permanent SMTP error (rejection)
      or temporary SMTP error ("try again later");

   o Permitting the message to pass (a successful SMTP reply code) and
      adding an additional header field that indicates the result
      returned by check_host() and other salient details; this is
      discussed in more detail in Section 9.

8.1. None

   With a "none" result, the SPF verifier has no information at all
   about the authorization or lack thereof of the client to use the
   checked identity or identities. The check_host() function completed
   without errors but was not able to reach any conclusion.

8.2. Neutral

   A "neutral" result indicates that although a policy for the identity
   was discovered, there is no definite assertion (positive or negative)
   about the client.

   A "neutral" result MUST be treated exactly like the "none" result;
   the distinction exists only for informational purposes. Treating
   "neutral" more harshly than "none" would discourage ADMDs from
   testing the use of SPF records (see Section 10.1).

8.3. Pass

   A "pass" result means the client is authorized to inject mail with
   the given identity. The domain can now, in the sense of reputation,
   be considered responsible for sending the message. Further policy
   checks can now proceed with confidence in the legitimate use of the
   identity. This is further discussed in Appendix G.1.





Kitterman Standards Track [Page 34]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


8.4. Fail

   A "fail" result is an explicit statement that the client is not
   authorized to use the domain in the given identity. Disposition of
   SPF fail messages is a matter of local policy. See Appendix G.2 for
   considerations on developing local policy.

   If the checking software chooses to reject the mail during the SMTP
   transaction, then it SHOULD use an SMTP reply code of 550 (see
   [RFC5321]) and, if supported, the 5.7.1 enhanced status code (see
   [RFC3463], Section 3.8), in addition to an appropriate reply text.
   The check_host() function will return either a default explanation
   string or one from the domain that published the SPF records (see
   Section 6.2). If the information does not originate with the
   checking software, it is good to make it clear that the text is
   provided by the sender's domain. For example:

       550 5.7.1 SPF MAIL FROM check failed:
       550 5.7.1 The domain example.com explains:
       550 5.7.1 Please see http://www.example.com/mailpolicy.html

   If the checking software chooses not to reject the mail during the
   SMTP transaction, then it SHOULD add a Received-SPF or
   Authentication-Results header field (see Section 9) to communicate
   this result to downstream message processors. While this is true for
   all SPF results, it is of particular importance for "fail" results
   since the message is explicitly not authorized by the ADMD.

8.5. Softfail

   A "softfail" result ought to be treated as somewhere between "fail"
   and "neutral"/"none". The ADMD believes the host is not authorized
   but is not willing to make a strong policy statement. Receiving
   software SHOULD NOT reject the message based solely on this result,
   but MAY subject the message to closer scrutiny than normal.

   The ADMD wants to discourage the use of this host and thus desires
   limited feedback when a "softfail" result occurs. For example, the
   recipient's MUA could highlight the "softfail" status, or the
   receiving MTA could give the sender a message using greylisting
   [RFC6647], with a note the first time the message is received, but
   accept it on a later attempt based on receiver policy.









Kitterman Standards Track [Page 35]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


8.6. Temperror

   A "temperror" result means the SPF verifier encountered a transient
   (generally DNS) error while performing the check. Checking software
   can choose to accept or temporarily reject the message. If the
   message is rejected during the SMTP transaction for this reason, the
   software SHOULD use an SMTP reply code of 451 and, if supported, the
   4.4.3 enhanced status code (see Section 3.5 of [RFC3463]). These
   errors can be caused by problems in either the sender's or receiver's
   DNS software. See Appendix G.4 for considerations on developing
   local policy.

8.7. Permerror

   A "permerror" result means the domain's published records could not
   be correctly interpreted. This signals an error condition that
   definitely requires DNS operator intervention to be resolved. If the
   message is rejected during the SMTP transaction for this reason, the
   software SHOULD use an SMTP reply code of 550 and, if supported, the
   5.5.2 enhanced status code (see [RFC3463], Section 3.6). Be aware
   that if the ADMD uses macros (Section 7), it is possible that this
   result is due to the checked identities having an unexpected format.
   It is also possible that this result is generated by certain SPF
   verifiers due to the input arguments having an unexpected format; see
   Section 4.8. See Appendix G.3 for considerations on developing local
   policy.

9. Recording the Result

   To provide downstream agents, such as MUAs, with the information they
   might need in terms of evaluating or representing the apparent safety
   of the message content, it is RECOMMENDED that SMTP receivers record
   the result of SPF processing in the message header. For SPF verifier
   operators that choose to record SPF results in the header of the
   message for processing by internal filters or MUAs, two methods are
   presented: Section 9.1 defines the Received-SPF field, which is the
   results field originally defined for SPF use. Section 9.2 discusses
   the Authentication-Results header field [RFC7001], which was
   specified more recently and is designed for use by SPF and other
   authentication methods.

   Both are in common use, and hence both are included here. However,
   it is important to note that they were designed to serve slightly
   different purposes. Received-SPF is intended to include enough
   information to enable reconstruction of the SPF evaluation of the
   message, while Authentication-Results is designed only to relay the
   result itself and related output details of likely use to end users
   (e.g., what property of the message was actually authenticated and



Kitterman Standards Track [Page 36]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   what it contained), leaving reconstructive work to the purview of
   system logs and the Received field contents. Also, Received-SPF
   relies on compliance of agents within the receiving ADMD to adhere to
   the header field ordering rules of [RFC5321] and [RFC5322], while
   Authentication-Results includes some provisions to protect against
   non-compliant implementations.

   An SPF verifier operator could choose to use both to serve different
   downstream agents. In such cases, care needs to be taken to ensure
   that both fields are conveying the same details, or unexpected
   results can occur.

9.1. The Received-SPF Header Field

   The Received-SPF header field is a trace field (see [RFC5322],
   Section 3.6.7) and SHOULD be prepended to the existing header, above
   the Received: field that is generated by the SMTP receiver. It MUST
   appear above all other Received-SPF fields in the message. The
   header field has the following format:

   header-field = "Received-SPF:" [CFWS] result FWS [comment FWS]
                      [ key-value-list ] CRLF

   result = "pass" / "fail" / "softfail" / "neutral" /
                      "none" / "temperror" / "permerror"

   key-value-list = key-value-pair *( ";" [CFWS] key-value-pair )
                      [";"]

   key-value-pair = key [CFWS] "=" ( dot-atom / quoted-string )

   key = "client-ip" / "envelope-from" / "helo" /
                      "problem" / "receiver" / "identity" /
                       "mechanism" / name

   identity = "mailfrom" ; for the "MAIL FROM" identity
                      / "helo" ; for the "HELO" identity
                      / name ; other identities

   dot-atom = <unquoted word as per [RFC5322]>
   quoted-string = <quoted string as per [RFC5322]>
   comment = <comment string as per [RFC5322]>
   CFWS = <comment or folding white space as per [RFC5322]>
   FWS = <folding white space as per [RFC5322]>
   CRLF = <standard end-of-line token as per [RFC5322]>






Kitterman Standards Track [Page 37]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   The header field SHOULD include a "(...)" style comment after the
   result, conveying supporting information for the result, such as
   <ip>, <sender>, and <domain>.

   The following key-value pairs are designed for later machine parsing.
   SPF verifiers SHOULD give enough information so that the SPF results
   can be verified -- that is, at least "client-ip", "helo", and, if the
   "MAIL FROM" identity was checked, "envelope-from".

   client-ip the IP address of the SMTP client

   envelope-from the envelope sender mailbox

   helo the host name given in the HELO or EHLO command

   mechanism the mechanism that matched (if no mechanisms matched,
                  substitute the word "default")

   problem if an error was returned, details about the error

   receiver the host name of the SPF verifier

   identity the identity that was checked; see the <identity>
                  ABNF rule

   Other keys MAY be defined by SPF verifiers.

   SPF verifiers MUST make sure that the Received-SPF header field does
   not contain invalid characters, is not excessively long (see
   [RFC5322], Section 2.1.1), and does not contain malicious data that
   has been provided by the sender.

   Examples of various header field styles that could be generated are
   the following:

   Received-SPF: pass (mybox.example.org: domain of
    myname@example.com designates 192.0.2.1 as permitted sender)
       receiver=mybox.example.org; client-ip=192.0.2.1;
       envelope-from="myname@example.com"; helo=foo.example.com;

   Received-SPF: fail (mybox.example.org: domain of
                     myname@example.com does not designate
                     192.0.2.1 as permitted sender)
                     identity=mailfrom; client-ip=192.0.2.1;
                     envelope-from="myname@example.com";






Kitterman Standards Track [Page 38]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   Received-SPF: pass (mybox.example.org: domain of
    myname@example.com designates 192.0.2.1 as permitted sender)
       receiver=mybox.example.org; client-ip=192.0.2.1;
       mechanism=ip4:192.0.2.1; envelope-from="myname@example.com";
       helo=foo.example.com;

9.2. SPF Results in the Authentication-Results Header Field

   As mentioned in Section 9, the Authentication-Results header field is
   designed to communicate lists of tests a border MTA did and their
   results. The specified elements of the field provide less
   information than the Received-SPF field:

   Authentication-Results: myhost.example.org; spf=pass
     smtp.mailfrom=example.net

   Received-SPF: pass (myhost.example.org: domain of
    myname@example.com designates 192.0.2.1 as permitted sender)
       receiver=mybox.example.org; client-ip=192.0.2.1;
       envelope-from="myname@example.com"; helo=foo.example.com;

   It is, however, possible to add CFWS in the "reason" part of an
   Authentication-Results header field and provide the equivalent
   information, if desired.

   As an example, an expanded Authentication-Results header field might
   look like (for a "MAIL FROM" check in this example):

   Authentication-Results: myhost.example.org; spf=pass
     reason="client-ip=192.0.2.1; smtp.helo=foo.example.com"
     smtp.mailfrom=user@example.net

10. Effects on Infrastructure

   This section outlines the major implications that adoption of this
   protocol will have on various entities involved in Internet email.
   It is intended to make clear to the reader where this protocol
   knowingly affects the operation of such entities. This section is
   not a "how-to" manual, or a "best practices" document, and it is not
   a comprehensive list of what such entities ought to do in light of
   this specification.

   This section provides operational advice and instruction only. It is
   non-normative.

   [RFC5598] describes the Internet email architecture. This section is
   organized based on the different segments of the architecture.




Kitterman Standards Track [Page 39]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


10.1. Sending Domains

   Originating ADMDs (ADministrative Management Domains --
   Sections 2.2.1 and 2.3 of [RFC5598]) that wish to be compliant with
   this specification will need to determine the list of relays
   ([RFC5598], Section 2.2.2) that they allow to use their domain name
   in the "HELO" and "MAIL FROM" identities when relaying to other
   ADMDs. It is recognized that forming such a list is not just a
   simple technical exercise, but involves policy decisions with both
   technical and administrative considerations.

10.1.1. DNS Resource Considerations

   Minimizing the DNS resources needed for SPF lookups can be done by
   choosing directives that require less DNS information and by placing
   lower-cost mechanisms earlier in the SPF record.

   Section 4.6.4 specifies the limits receivers have to use. It is
   essential to publish records that do not exceed these requirements.
   It is also required to carefully weigh the cost and the
   maintainability of licit solutions.

   For example, consider a domain set up as follows:

      example.com. IN MX 10 mx.example.com.
                       IN MX 20 mx2.example.com.
      mx.example.com. IN A 192.0.2.1
      mx2.example.com. IN A 192.0.2.129

   Assume the administrative point is to authorize (pass) mx and mx2
   while failing every other host. Compare the following solutions:

   Best record:
      example.com. IN TXT "v=spf1 ip4:192.0.2.1 ip4:192.0.2.129 -all"

   Good record:
      $ORIGIN example.com.
      @ IN TXT "v=spf1 a:authorized-spf.example.com -all"
      authorized-spf IN A 192.0.2.1
                     IN A 192.0.2.129

   Expensive record:
      example.com. IN TXT "v=spf1 mx:example.com -all"

   Wasteful, bad record:
      example.com. IN TXT "v=spf1 ip4:192.0.2.0/24 mx -all"





Kitterman Standards Track [Page 40]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


10.1.2. Administrator's Considerations

   There might be administrative considerations: using "a" over "ip4" or
   "ip6" allows hosts to be renumbered easily at the cost of a DNS query
   per receiver. Using "mx" over "a" allows the set of mail hosts to be
   changed easily. Unless such changes are common, it is better to use
   the less resource-intensive mechanisms like "ip4" and "ip6" over "a"
   or "a" over "mx".

   In some specific cases, standard advice on record content is
   appropriate. Publishing SPF records for domains that send no mail is
   a well-established best practice. The record for a domain that sends
   no mail is:

      www.example.com. IN TXT "v=spf1 -all"

   Publishing SPF records for individual hosts is also best practice.
   The host name is generally the identity used in the 5321.HELO/.EHLO
   command. In the case of messages with a null 5321.MailFrom, this is
   used as the domain for 5321.MailFrom SPF checks, in addition to being
   used in 5321.HELO/.EHLO-based SPF checks. The standard SPF record
   for an individual host that is involved in mail processing is:

      relay.example.com. IN TXT "v=spf1 a -all"

   Validating correct deployment is difficult. [RFC6652] describes one
   mechanism for soliciting feedback on SPF failures. Another
   suggestion can be found in Appendix C.

   Regardless of the method used, understanding the ADMD's outbound mail
   architecture is essential to effective deployment.

10.1.3. Bounces

   As explained in Section 2.4, [RFC5321] allows the MAIL FROM to be
   null, which is typical of some Delivery Status Notifications
   [RFC3464], commonly called email bounces. In this case, the only
   entity available for performing an SPF check is the "HELO" identity
   defined in Section 1.1.4. SPF functionality is enhanced by
   administrators ensuring this identity is set correctly and has an
   appropriate SPF record. It is normal to have the "HELO" identity set
   to the host name instead of the domain. Zone file generation for
   significant numbers of hosts can be consolidated using the "redirect"
   modifier and scripted for initial deployment. Specific deployment
   advice is given above in Section 10.1.2.






Kitterman Standards Track [Page 41]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


10.2. Receivers

   SPF results can be used in combination with other methods to
   determine the final local disposition (either positive or negative)
   of a message. It can also be considered dispositive on its own.

   An attempt to have one organization (sender) direct the email-
   handling policies of another (receiver) is inherently challenging and
   often controversial. As stated elsewhere in this document, there is
   no comprehensive normative requirement for specific handling of a
   message based on SPF results. The information presented in Section 8
   and in Appendix G is offered for receiver consideration when forming
   local handling policies.

   The primary considerations are that SPF might return "pass" for mail
   that is ultimately harmful (e.g., spammers that arrange for SPF to
   pass using disposable domain names, or virus or spam outbreaks from
   within trusted sources), and might also return "fail" for mail that
   is ultimately legitimate (e.g., legitimate mail that has traversed a
   mail alias). It is important to take both of these cases under
   consideration when establishing local handling policy.

10.3. Mediators

   Mediators are a type of User Actor [RFC5598]. That is, a mediator
   takes 'delivery' of a message and posts a 'submission' of a new
   message. The mediator can make the newly posted message be as
   similar to or as different from the original message as they wish.
   Examples include mailing lists (see Section 5.3 of [RFC5598]) and
   ReSenders (Section 5.2 of [RFC5598]). This is discussed in
   [RFC5321], Section 3.9. For the operation of SPF, the essential
   concern is the email address in the 5321.MailFrom command for the new
   message.

   Because SPF evaluation is based on the IP address of the "last"
   sending SMTP server, the address of the mediator will be used, rather
   than the address of the SMTP server that sent the message to the
   mediator. Some mediators retain the email address from the original
   message, while some use a new address.

   If the address is the same as for the original message, and the
   original message had an associated SPF record, then the SPF
   evaluation will fail unless mitigations such as those described in
   Appendix D are used.







Kitterman Standards Track [Page 42]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


11. Security Considerations

11.1. Processing Limits

   As with most aspects of email, there are a number of ways that
   malicious parties could use the protocol as an avenue for a DoS
   attack. The processing limits outlined in Section 4.6.4 are designed
   to prevent attacks such as the following:

   o A malicious party could create an SPF record with many references
      to a victim's domain and send many emails to different SPF
      verifiers; those SPF verifiers would then create a DoS attack. In
      effect, the SPF verifiers are being used to amplify the attacker's
      bandwidth by using fewer octets in the SMTP session than are used
      by the DNS queries. Using SPF verifiers also allows the attacker
      to hide the true source of the attack. This potential attack is
      based on large volumes of mail being transmitted.

   o Whereas implementations of check_host() are supposed to limit the
      number of DNS lookups, malicious domains could publish records
      that exceed these limits in an attempt to waste computation effort
      at their targets when they send them mail. Malicious domains
      could also design SPF records that cause particular
      implementations to use excessive memory or CPU or to trigger bugs.
      If a receiver is configured to accept mail with an SPF result of
      "temperror", such an attack might result in mail that would
      otherwise have been rejected due to an SPF "fail" result being
      accepted. This potential attack is based on specially crafted SPF
      records being used to exhaust DNS resources of the victim.

   o Malicious parties could send a large volume of mail purporting to
      come from the intended target to a wide variety of legitimate mail
      hosts. These legitimate machines would then present a DNS load on
      the target as they fetched the relevant records.

   o Malicious parties could, in theory, use SPF records as a vehicle
      for DNS lookup amplification for a DoS attack. In this scenario,
      the attacker publishes an SPF record in its own DNS that uses "a"
      and "mx" mechanisms directed toward the intended victim, e.g.,
      "a:example.com a:foo.example.com a:bar.example.com ..." and then
      distributes mail with a MAIL FROM value including its own domain
      in large volume to a wide variety of destinations. Any such
      destination operating an SPF verifier will begin querying all of
      the names associated with the "a" mechanisms in that record. The
      names used in the record needn't exist for the attack to be
      effective. Operational experience since the publication of
      [RFC4408] suggests that mitigation of this class of attack can be
      accomplished with minimal impact on the deployed base by having



Kitterman Standards Track [Page 43]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


      the verifier abort processing and return "permerror"
      (Section 2.6.7) as soon as more than two "void lookups" have been
      encountered (defined in Section 4.6.4).

   Of these, the case of a third party referenced in the SPF record is
   the easiest for a DoS attack to effectively exploit. As a result,
   limits that might seem reasonable for an individual mail server can
   still allow an unreasonable amount of bandwidth amplification.
   Therefore, the processing limits need to be quite low.

11.2. SPF-Authorized Email May Contain Other False Identities

   The "MAIL FROM" and "HELO" identity authorizations do not provide
   assurance about the authorization/authenticity of other identities
   used in the message. It is entirely possible for a malicious sender
   to inject a message using his own domain in the identities used by
   SPF and have that domain's SPF record authorize the sending host, and
   yet the message can easily list other identities in its header.
   Unless the user or the MUA takes care to note that the authorized
   identity does not match the other more commonly presented identities
   (such as the From: header field), the user might be lulled into a
   false sense of security.

11.3. Spoofed DNS and IP Data

   There are two aspects of this protocol that malicious parties could
   exploit to undermine the validity of the check_host() function:

   o The evaluation of check_host() relies heavily on DNS. A malicious
      attacker could attack the DNS infrastructure and cause
      check_host() to see spoofed DNS data, and then return incorrect
      results. This could include returning "pass" for an <ip> value
      where the actual domain's record would evaluate to "fail". See
      [RFC3833] for a description of DNS weaknesses, and see [RFC4033]
      for a countermeasure.

   o The client IP address, <ip>, is assumed to be correct. In a
      modern, correctly configured system, the risk of this not being
      true is nil.

11.4. Cross-User Forgery

   By definition, SPF policies just map domain names to sets of
   authorized MTAs, not whole email addresses to sets of authorized
   users. Although the "l" macro (Section 7) provides a limited way to
   define individual sets of authorized MTAs for specific email
   addresses, it is generally impossible to verify, through SPF, the use
   of specific email addresses by individual users of the same MTA.



Kitterman Standards Track [Page 44]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   It is up to mail services and their MTAs to directly prevent
   cross-user forgery: based on SMTP AUTH ([RFC4954]), users have to be
   restricted to using only those email addresses that are actually
   under their control (see Section 6.1 of [RFC6409]). Another means to
   verify the identity of individual users is message cryptography, such
   as Pretty Good Privacy (PGP) ([RFC4880]) or S/MIME ([RFC5751]).

11.5. Untrusted Information Sources

   An SPF-compliant receiver gathers information from the SMTP commands
   it receives and from the published DNS records of the sending domain
   holder (e.g., "HELO" domain name, the "MAIL FROM" address from the
   envelope, and SPF DNS records published by the domain holder). These
   parameters are not validated in the SMTP process.

   All of these pieces of information are generated by actors outside of
   the authority of the receiver, and thus are not guaranteed to be
   accurate or legitimate.

11.5.1. Recorded Results

   This information, passed to the receiver in the Received-SPF: or
   Authentication-Results: trace fields, can be returned to the client
   MTA as an SMTP rejection message. If such an SMTP rejection message
   is generated, the information from the trace fields has to be checked
   for such problems as invalid characters and excessively long lines.

11.5.2. External Explanations

   When the authorization check fails, an explanation string could be
   included in the reject response. Both the sender and the rejecting
   receiver need to be aware that the explanation was determined by the
   publisher of the SPF record checked and, in general, not the
   receiver. The explanation can contain malicious URLs, or it might be
   offensive or misleading.

   Explanations returned to sender domains due to "exp" modifiers
   (Section 6.2) were generated by the sender policy published by the
   domain holders themselves. As long as messages are only returned
   with non-delivery notifications ([RFC3464]) to domains publishing the
   explanation strings from their own DNS SPF records, the only affected
   parties are the original publishers of the domain's SPF records.

   In practice, such non-delivery notifications can be misdirected, such
   as when an MTA accepts an email and only later generates the
   notification to a forged address, or when an email forwarder does not
   direct the bounce back to the original sender.




Kitterman Standards Track [Page 45]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


11.5.3. Macro Expansion

   Macros (Section 7) allow senders to inject arbitrary text (any
   non-null [US-ASCII] character) into receiver DNS queries. It is
   necessary to be prepared for hostile or unexpected content.

11.6. Privacy Exposure

   Checking SPF records causes DNS queries to be sent to the domain
   owner. These DNS queries, especially if they are caused by the
   "exists" mechanism, can contain information about who is sending
   email and likely to which MTA the email is being sent. This can
   introduce some privacy concerns, which are more or less of an issue
   depending on local laws and the relationship between the ADMD and the
   person sending the email.

11.7. Delivering Mail Producing a "Fail" Result

   Operators that choose to deliver mail for which SPF produces a "fail"
   result need to understand that they are admitting content that is
   explicitly not authorized by the purported sender. While there are
   known failure modes that can be considered "false negatives", the
   distinct choice to admit those messages increases end-user exposure
   to likely harm. This is especially true for domains belonging to
   known good actors that are typically well-behaved; unauthorized mail
   from those sources might well be subjected to much higher skepticism
   and content analysis.

   SPF does not, however, include the capacity to distinguish good
   actors from bad ones, nor does it handle the concept of known actors
   versus unknown ones. Those notions are out of scope for this
   specification.

12. Collected ABNF

   This section is normative, and any discrepancies with the ABNF
   fragments in the preceding text are to be resolved in favor of this
   grammar.

   See [RFC5234] for ABNF notation. Please note that as per this ABNF
   definition, literal text strings (those in quotes) are case-
   insensitive. Hence, "mx" matches "mx", "MX", "mX", and "Mx".

   record = version terms *SP
   version = "v=spf1"

   terms = *( 1*SP ( directive / modifier ) )




Kitterman Standards Track [Page 46]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   directive = [ qualifier ] mechanism
   qualifier = "+" / "-" / "?" / "~"
   mechanism = ( all / include
                      / a / mx / ptr / ip4 / ip6 / exists )

   all = "all"
   include = "include" ":" domain-spec
   a = "a" [ ":" domain-spec ] [ dual-cidr-length ]
   mx = "mx" [ ":" domain-spec ] [ dual-cidr-length ]
   ptr = "ptr" [ ":" domain-spec ]
   ip4 = "ip4" ":" ip4-network [ ip4-cidr-length ]
   ip6 = "ip6" ":" ip6-network [ ip6-cidr-length ]
   exists = "exists" ":" domain-spec

   modifier = redirect / explanation / unknown-modifier
   redirect = "redirect" "=" domain-spec
   explanation = "exp" "=" domain-spec
   unknown-modifier = name "=" macro-string
                      ; where name is not any known modifier

   ip4-cidr-length = "/" ("0" / %x31-39 0*1DIGIT) ; value range 0-32
   ip6-cidr-length = "/" ("0" / %x31-39 0*2DIGIT) ; value range 0-128
   dual-cidr-length = [ ip4-cidr-length ] [ "/" ip6-cidr-length ]

   ip4-network = qnum "." qnum "." qnum "." qnum
   qnum = DIGIT ; 0-9
                      / %x31-39 DIGIT ; 10-99
                      / "1" 2DIGIT ; 100-199
                      / "2" %x30-34 DIGIT ; 200-249
                      / "25" %x30-35 ; 250-255
            ; conventional dotted-quad notation, e.g., 192.0.2.0
   ip6-network = <as per Section 2.2 of [RFC4291]>
            ; e.g., 2001:db8::cd30

   domain-spec = macro-string domain-end
   domain-end = ( "." toplabel [ "." ] ) / macro-expand

   toplabel = ( *alphanum ALPHA *alphanum ) /
                      ( 1*alphanum "-" *( alphanum / "-" ) alphanum )
                      ; LDH rule plus additional TLD restrictions
                      ; (see Section 2 of [RFC3696] for background)
   alphanum = ALPHA / DIGIT

   explain-string = *( macro-string / SP )

   macro-string = *( macro-expand / macro-literal )
   macro-expand = ( "%{" macro-letter transformers *delimiter "}" )
                      / "%%" / "%_" / "%-"



Kitterman Standards Track [Page 47]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   macro-literal = %x21-24 / %x26-7E
                      ; visible characters except "%"
   macro-letter = "s" / "l" / "o" / "d" / "i" / "p" / "h" /
                      "c" / "r" / "t" / "v"
   transformers = *DIGIT [ "r" ]
   delimiter = "." / "-" / "+" / "," / "/" / "_" / "="

   name = ALPHA *( ALPHA / DIGIT / "-" / "_" / "." )

   header-field = "Received-SPF:" [CFWS] result FWS [comment FWS]
                      [ key-value-list ] CRLF

   result = "pass" / "fail" / "softfail" / "neutral" /
                      "none" / "temperror" / "permerror"

   key-value-list = key-value-pair *( ";" [CFWS] key-value-pair )
                      [";"]

   key-value-pair = key [CFWS] "=" ( dot-atom / quoted-string )

   key = "client-ip" / "envelope-from" / "helo" /
                      "problem" / "receiver" / "identity" /
                       "mechanism" / name

   identity = "mailfrom" ; for the "MAIL FROM" identity
                      / "helo" ; for the "HELO" identity
                      / name ; other identities

   sender = Mailbox
   ip = ip4-network / ip6-network
   ALPHA = <A-Z / a-z as per [RFC5234]>
   DIGIT = <0-9 as per [RFC5234]>
   SP = <space character as per [RFC5234]>
   dot-atom = <unquoted word as per [RFC5322]>
   quoted-string = <quoted string as per [RFC5322]>
   comment = <comment string as per [RFC5322]>
   CFWS = <comment or folding white space as per [RFC5322]>
   FWS = <folding white space as per [RFC5322]>
   CRLF = <standard end-of-line token as per [RFC5322]>

13. Contributors and Acknowledgements

   This document is largely based on the work of Meng Weng Wong, Mark
   Lentczner, and Wayne Schlitt. Although, as this section
   acknowledges, many people have contributed to this document, a very
   large portion of the writing and editing is due to Meng, Mark, and
   Wayne.




Kitterman Standards Track [Page 48]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   This design owes a debt of parentage to [RMX] by Hadmut Danisch and
   to [DMP] by Gordon Fecyk. The idea of using a DNS record to check
   the legitimacy of an email address traces its ancestry further back
   through messages on the namedroppers mailing list by Paul Vixie
   [Vixie] (based on suggestion by Jim Miller) and by David Green
   [Green].

   Philip Gladstone contributed the concept of macros to the
   specification, multiplying the expressiveness of the language and
   making per-user and per-IP lookups possible.

   The authors of both this document and [RFC4408] would also like to
   thank the literally hundreds of individuals who have participated in
   the development of this design. They are far too numerous to name,
   but they include the following:

      The participants in the SPFbis working group. The folks on the
      spf-discuss mailing list. The folks on the SPAM-L mailing list.
      The folks on the IRTF ASRG mailing list. The folks on the IETF
      MARID mailing list. The folks on #perl.

14. IANA Considerations

14.1. The SPF DNS Record Type

   Per [RFC4408], the IANA assigned the Resource Record Type and Qtype
   from the "Domain Name System (DNS) Parameters" registry for the SPF
   RR type with code 99. The format of this type is identical to the
   TXT RR [RFC1035]. The character content of the record is encoded as
   [US-ASCII].

   Studies have shown that RRTYPE 99 has not seen any substantial use,
   and in fact its existence and mechanism defined in [RFC4408] have led
   to some interoperability issues. Accordingly, its use is no longer
   appropriate for SPF version 1; implementations are not to use it.

   IANA has updated the "Resource Record (RR) TYPEs" registry to
   indicate that this document is the reference document for that
   RRTYPE.












Kitterman Standards Track [Page 49]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


14.2. The Received-SPF Mail Header Field

   Per [RFC3864], the "Received-SPF:" header field is added to the IANA
   "Permanent Message Header Field Names" registry. The following is
   the registration template:

      Header field name: Received-SPF Applicable protocol: mail
      ([RFC5322]) Status: standard Author/Change controller: IETF
      Specification document(s): RFC 7208

14.3. SPF Modifier Registry

   IANA has changed the reference for the "exp" and "redirect" modifiers
   in the "Modifier Names" registry, under Sender Policy Framework
   Parameters, from [RFC4408] to this document. Their status is
   unchanged.

15. References

15.1. Normative References

   [RFC1035] Mockapetris, P., "Domain names - implementation and
              specification", STD 13, RFC 1035, November 1987.

   [RFC1123] Braden, R., "Requirements for Internet Hosts - Application
              and Support", STD 3, RFC 1123, October 1989.

   [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
              Requirement Levels", BCP 14, RFC 2119, March 1997.

   [RFC3463] Vaudreuil, G., "Enhanced Mail System Status Codes",
              RFC 3463, January 2003.

   [RFC3864] Klyne, G., Nottingham, M., and J. Mogul, "Registration
              Procedures for Message Header Fields", BCP 90, RFC 3864,
              September 2004.

   [RFC3986] Berners-Lee, T., Fielding, R., and L. Masinter, "Uniform
              Resource Identifier (URI): Generic Syntax", STD 66,
              RFC 3986, January 2005.

   [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing
              Architecture", RFC 4291, February 2006.

   [RFC5234] Crocker, D. and P. Overell, "Augmented BNF for Syntax
              Specifications: ABNF", STD 68, RFC 5234, January 2008.





Kitterman Standards Track [Page 50]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   [RFC5321] Klensin, J., "Simple Mail Transfer Protocol", RFC 5321,
              October 2008.

   [RFC5322] Resnick, P., Ed., "Internet Message Format", RFC 5322,
              October 2008.

   [RFC5598] Crocker, D., "Internet Mail Architecture", RFC 5598,
              July 2009.

   [RFC5890] Klensin, J., "Internationalized Domain Names for
              Applications (IDNA): Definitions and Document Framework",
              RFC 5890, August 2010.

   [RFC7001] Kucherawy, M., "Message Header Field for Indicating
              Message Authentication Status", RFC 7001, September 2013.

   [US-ASCII]
              American National Standards Institute (formerly United
              States of America Standards Institute), "USA Code for
              Information Interchange, X3.4", 1968.

              ANSI X3.4-1968 has been replaced by newer versions with
              slight modifications, but the 1968 version remains
              definitive for the Internet.

15.2. Informative References

   [BATV] Levine, J., Crocker, D., Silberman, S., and T. Finch,
              "Bounce Address Tag Validation (BATV)", Work in Progress,
              May 2008.

   [DMP] Fecyk, G., "Designated Mailers Protocol", Work in
              Progress, May 2004.

   [Green] Green, D., "Domain-Authorized SMTP Mail", June 2002,
              <http://www.mhonarc.org/archive/html/ietf-asrg/2003-03/
              msg01525.html>.

   [RFC1034] Mockapetris, P., "Domain names - concepts and facilities",
              STD 13, RFC 1034, November 1987.

   [RFC1983] Malkin, G., "Internet Users' Glossary", RFC 1983,
              August 1996.

   [RFC2308] Andrews, M., "Negative Caching of DNS Queries (DNS
              NCACHE)", RFC 2308, March 1998.





Kitterman Standards Track [Page 51]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   [RFC2671] Vixie, P., "Extension Mechanisms for DNS (EDNS0)",
              RFC 2671, August 1999.

   [RFC2782] Gulbrandsen, A., Vixie, P., and L. Esibov, "A DNS RR for
              specifying the location of services (DNS SRV)", RFC 2782,
              February 2000.

   [RFC3464] Moore, K. and G. Vaudreuil, "An Extensible Message Format
              for Delivery Status Notifications", RFC 3464,
              January 2003.

   [RFC3696] Klensin, J., "Application Techniques for Checking and
              Transformation of Names", RFC 3696, February 2004.

   [RFC3833] Atkins, D. and R. Austein, "Threat Analysis of the Domain
              Name System (DNS)", RFC 3833, August 2004.

   [RFC3834] Moore, K., "Recommendations for Automatic Responses to
              Electronic Mail", RFC 3834, August 2004.

   [RFC4033] Arends, R., Austein, R., Larson, M., Massey, D., and S.
              Rose, "DNS Security Introduction and Requirements",
              RFC 4033, March 2005.

   [RFC4408] Wong, M. and W. Schlitt, "Sender Policy Framework (SPF)
              for Authorizing Use of Domains in E-Mail, Version 1",
              RFC 4408, April 2006.

   [RFC4632] Fuller, V. and T. Li, "Classless Inter-domain Routing
              (CIDR): The Internet Address Assignment and Aggregation
              Plan", BCP 122, RFC 4632, August 2006.

   [RFC4880] Callas, J., Donnerhacke, L., Finney, H., Shaw, D., and R.
              Thayer, "OpenPGP Message Format", RFC 4880, November 2007.

   [RFC4954] Siemborski, R. and A. Melnikov, "SMTP Service Extension
              for Authentication", RFC 4954, July 2007.

   [RFC5507] IAB, Faltstrom, P., Austein, R., and P. Koch, "Design
              Choices When Expanding the DNS", RFC 5507, April 2009.

   [RFC5751] Ramsdell, B. and S. Turner, "Secure/Multipurpose Internet
              Mail Extensions (S/MIME) Version 3.2 Message
              Specification", RFC 5751, January 2010.

   [RFC5782] Levine, J., "DNS Blacklists and Whitelists", RFC 5782,
              February 2010.




Kitterman Standards Track [Page 52]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   [RFC6409] Gellens, R. and J. Klensin, "Message Submission for Mail",
              STD 72, RFC 6409, November 2011.

   [RFC6647] Kucherawy, M. and D. Crocker, "Email Greylisting: An
              Applicability Statement for SMTP", RFC 6647, June 2012.

   [RFC6648] Saint-Andre, P., Crocker, D., and M. Nottingham,
              "Deprecating the "X-" Prefix and Similar Constructs in
              Application Protocols", BCP 178, RFC 6648, June 2012.

   [RFC6652] Kitterman, S., "Sender Policy Framework (SPF)
              Authentication Failure Reporting Using the Abuse Reporting
              Format", RFC 6652, June 2012.

   [RFC6686] Kucherawy, M., "Resolution of the Sender Policy Framework
              (SPF) and Sender ID Experiments", RFC 6686, July 2012.

   [RFC6891] Damas, J., Graff, M., and P. Vixie, "Extension Mechanisms
              for DNS (EDNS(0))", STD 75, RFC 6891, April 2013.

   [RMX] Danisch, H., "The RMX DNS RR and method for lightweight
              SMTP sender authorization", Work in Progress, May 2004.

   [Vixie] Vixie, P., "Repudiating MAIL FROM", 2002,
              <http://marc.info/?l=namedroppers&m=102298170127004&w=4>.


























Kitterman Standards Track [Page 53]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


Appendix A. Extended Examples

   These examples are based on the following DNS setup:

   ; A domain with two mail servers, two hosts, and two servers
   ; at the domain name
   $ORIGIN example.com.
   @ MX 10 mail-a
               MX 20 mail-b
               A 192.0.2.10
               A 192.0.2.11
   amy A 192.0.2.65
   bob A 192.0.2.66
   mail-a A 192.0.2.129
   mail-b A 192.0.2.130
   www CNAME example.com.

   ; A related domain
   $ORIGIN example.org.
   @ MX 10 mail-c
   mail-c A 192.0.2.140

   ; The reverse IP for those addresses
   $ORIGIN 2.0.192.in-addr.arpa.
   10 PTR example.com.
   11 PTR example.com.
   65 PTR amy.example.com.
   66 PTR bob.example.com.
   129 PTR mail-a.example.com.
   130 PTR mail-b.example.com.
   140 PTR mail-c.example.org.

   ; A rogue reverse IP domain that claims to be
   ; something it's not
   $ORIGIN 0.0.10.in-addr.arpa.
   4 PTR bob.example.com.















Kitterman Standards Track [Page 54]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


A.1. Simple Examples

   These examples show various possible published records for
   example.com and which values of <ip> would cause check_host() to
   return "pass". Note that <domain> is "example.com".

   v=spf1 +all

      -- any <ip> passes

   v=spf1 a -all

      -- hosts 192.0.2.10 and 192.0.2.11 pass

   v=spf1 a:example.org -all

      -- no sending hosts pass since example.org has no A records

   v=spf1 mx -all

      -- sending hosts 192.0.2.129 and 192.0.2.130 pass

   v=spf1 mx:example.org -all

      -- sending host 192.0.2.140 passes

   v=spf1 mx mx:example.org -all

      -- sending hosts 192.0.2.129, 192.0.2.130, and 192.0.2.140 pass

   v=spf1 mx/30 mx:example.org/30 -all

      -- any sending host in 192.0.2.128/30 or 192.0.2.140/30 passes

   v=spf1 ptr -all

      -- sending host 192.0.2.65 passes (reverse DNS is valid and is
         in example.com)

      -- sending host 192.0.2.140 fails (reverse DNS is valid, but not
         in example.com)

      -- sending host 10.0.0.4 fails (reverse IP is not valid)








Kitterman Standards Track [Page 55]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   v=spf1 ip4:192.0.2.128/28 -all

      -- sending host 192.0.2.65 fails

      -- sending host 192.0.2.129 passes

A.2. Multiple Domain Example

   These examples show the effect of related records:

      example.org: "v=spf1 include:example.com include:example.net -all"

   This record would be used if mail from example.org actually came
   through servers at example.com and example.net. Example.org's
   designated servers are the union of example.com's and example.net's
   designated servers.

      la.example.org: "v=spf1 redirect=example.org"

      ny.example.org: "v=spf1 redirect=example.org"

      sf.example.org: "v=spf1 redirect=example.org"

   These records allow a set of domains that all use the same mail
   system to make use of that mail system's record. In this way, only
   the mail system's record needs to be updated when the mail setup
   changes. These domains' records never have to change.

A.3. DNS Blacklist (DNSBL) Style Example

   Imagine that, in addition to the domain records listed above, there
   are these (see [RFC5782]):

   $ORIGIN _spf.example.com.
   mary.mobile-users A 127.0.0.2
   fred.mobile-users A 127.0.0.2
   15.15.168.192.joel.remote-users A 127.0.0.2
   16.15.168.192.joel.remote-users A 127.0.0.2













Kitterman Standards Track [Page 56]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   The following records describe users at example.com who mail from
   arbitrary servers, or who mail from personal servers.

   example.com:

   v=spf1 mx
          include:mobile-users._spf.%{d}
          include:remote-users._spf.%{d}
          -all

   mobile-users._spf.example.com:

   v=spf1 exists:%{l1r+}.%{d}

   remote-users._spf.example.com:

   v=spf1 exists:%{ir}.%{l1r+}.%{d}

A.4. Multiple Requirements Example

   Say that your sender policy requires both that the IP address is
   within a certain range and that the reverse DNS for the IP matches.
   This can be done several ways, including the following:

   example.com. SPF ( "v=spf1 "
                                 "-include:ip4._spf.%{d} "
                                 "-include:ptr._spf.%{d} "
                                 "+all" )
   ip4._spf.example.com. SPF "v=spf1 -ip4:192.0.2.0/24 +all"
   ptr._spf.example.com. SPF "v=spf1 -ptr +all"

   This example shows how the "-include" mechanism can be useful, how an
   SPF record that ends in "+all" can be very restrictive, and the use
   of De Morgan's Law.

Appendix B. Changes in Implementation Requirements from RFC 4408

   The modifications to implementation requirements from [RFC4408] are
   all either (a) corrections to errors in [RFC4408] or (b) additional
   documentation based on consensus of operational experience acquired
   since the publication of [RFC4408].

   o Use of DNS RR type SPF (99) has been removed from the protocol;
      see [RFC6686] for background.

   o A new DNS-related processing limit based on "void lookups" has
      been added (Section 4.6.4).




Kitterman Standards Track [Page 57]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   o Use of the ptr mechanism and the %p macro has been strongly
      discouraged (Sections 5.5 and 7.2). The ptr mechanism and the %p
      macro remain part of the protocol because they were found to be in
      use, but records ought to be updated to avoid them.

   o Use of the "Authentication-Results" header field [RFC7001] as a
      possible alternative to use of the "Received-SPF" header field is
      discussed (Section 9.2).

   o There have been a number of minor corrections to the ABNF to make
      it more clear and correct (Section 12). SPF library implementers
      should give the revised ABNF a careful review to determine if
      implementation changes are needed.

   o Use of X- fields in the ABNF has been removed; see [RFC6648] for
      background.

   o Ambiguity about how to deal with invalid <domain-spec> after macro
      expansion has been documented. Depending on one specific behavior
      has to be avoided (Section 4.8).

   o General operational information has been updated and expanded
      based on eight years of post-[RFC4408] operations experience. See
      Section 10 and Appendices D through G below.

   o Security considerations have been reviewed and updated
      (Section 11).

Appendix C. Further Testing Advice

   Another approach that can be helpful is to publish records that
   include a "tracking exists:" mechanism. By looking at the name
   server logs, a rough list can then be generated. For example:

      v=spf1 exists:_h.%{h}._l.%{l}._o.%{o}._i.%{i}._spf.%{d} ?all

   This associated macro expansion would cause the sending HELO domain,
   local-part of the sending email address, domain part of the sending
   email address, and the IP address from which the connection was
   received to be embedded in an SPF query and logged in the sender's
   DNS logs.

   This approach, which has been used since very early in the SPF
   project, allows senders to unilaterally collect data to evaluate the
   correctness of their SPF records. Unlike newer feedback mechanisms,
   it does not require any special cooperation from SPF verifiers. A
   similar example, one of the earliest SPF records published, can still
   be found as of this writing at altavista.net.



Kitterman Standards Track [Page 58]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


Appendix D. SPF/Mediator Interactions

   There are three places that techniques can be used to ameliorate
   unintended SPF failures with mediators.

D.1. Originating ADMDs

   The beginning, when email is first sent:

   o "Neutral" results could be given for IP addresses that might be
      forwarders, instead of "fail" results based on a list of known
      reliable forwarders. For example:

         "v=spf1 mx ?exists:%{ir}.whitelist.example.org -all"

      This would cause a lookup on a DNS White List (DNSWL) and cause a
      result of "fail" only for email not coming from either the
      domain's mx host(s) (SPF pass) or whitelisted sources (SPF
      neutral). This, in effect, outsources an element of sender policy
      to the maintainer of the whitelist.

   o The "MAIL FROM" identity could have additional information in the
      local-part that cryptographically identifies the mail as coming
      from an authorized source. In this case, an SPF record such as
      the following could be used:

         "v=spf1 mx exists:%{l}._spf_verify.%{d} -all"

      Then, a specialized DNS server can be set up to serve the
      _spf_verify subdomain that validates the local-part. Although
      this requires an extra DNS lookup, this happens only when the
      email would otherwise be rejected as not coming from a known good
      source.

      Note that due to the 63-character limit for domain labels, this
      approach only works reliably if the local-part signature scheme is
      guaranteed to either only produce local-parts with a maximum of
      63 characters or gracefully handle truncated local-parts. The
      method used to secure the local-part is a local implementation
      issue; it need not be standard. An example of one way to do it
      can be found in [BATV].

   o Similarly, a specialized DNS server could be set up that will
      rate-limit the email coming from unexpected IP addresses.

         "v=spf1 mx exists:%{ir}._spf_rate.%{d} -all"





Kitterman Standards Track [Page 59]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   o SPF allows the creation of per-user policies for special cases.
      For example, the following SPF record and appropriate wildcard DNS
      records can be used:

         "v=spf1 mx redirect=%{l1r+}._at_.%{o}._spf.%{d}"

D.2. Mediators

   The middle, when email is forwarded:

   o Mediators can solve the problem by rewriting the "MAIL FROM" to be
      in their own domain. This means mail rejected from the external
      mailbox will have to be forwarded back to the original sender by
      the forwarding service. Various schemes to do this exist, though
      they vary widely in complexity and resource requirements on the
      part of the mediator.

   o Several popular MTAs can be forced from "alias" semantics to
      "mailing list" semantics by configuring an additional alias with
      "owner-" prepended to the original alias name (e.g., an alias of
      "friends: george@example.com, fred@example.org" would need another
      alias of the form "owner-friends: localowner").

   o Mediators could reject mail that would "fail" SPF if forwarded
      using an SMTP reply code of 551, User not local (see Section 3.4
      of [RFC5321]) to communicate the correct target address to resend
      the mail to.

D.3. Receiving ADMDs

   The end, when email is received:

   o If the owner of the external mailbox wishes to trust the mediator,
      he can direct the external mailbox's MTA to skip SPF tests when
      the client host belongs to the mediator.

   o Tests against other identities, such as the "HELO" identity, can
      be used to override a failed test against the "MAIL FROM"
      identity.

   o For larger domains, it might not be possible to have a complete or
      accurate list of forwarding services used by the owners of the
      domain's mailboxes. In such cases, whitelists of generally
      recognized forwarding services could be employed.







Kitterman Standards Track [Page 60]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


Appendix E. Mail Services

   MSPs (Mail Service Providers -- Section 2.3 of [RFC5598]) that offer
   mail services to third-party domains, such as the sending of bulk
   mail, might want to adjust their configurations in light of the
   authorization check described in this document. If the domain part
   of the "MAIL FROM" identity used for such email uses one of the MSP's
   domains, then the provider needs only to ensure that its sending host
   is authorized by its own SPF record, if any.

   If the "MAIL FROM" identity does not use the MSP's domain, then extra
   care has to be taken. The SPF record format has several options for
   the third-party domain to authorize the service provider's MTAs to
   send mail on its behalf. For MSPs, such as ISPs, that have a wide
   variety of customers using the same MTA, steps are required to
   mitigate the risk of cross-customer forgery (see Section 11.4).

Appendix F. MTA Relays

   Relays are described in [RFC5598], Section 2.2.2. The authorization
   check generally precludes the use of arbitrary MTA relays between the
   sender and receiver of an email message.

   Within an organization, MTA relays can be effectively deployed.
   However, for the purposes of this document, such relays are
   effectively transparent. The SPF authorization check is a check
   between border MTAs of different ADMDs.

   For mail senders, this means published SPF records have to authorize
   any MTAs that actually send across the Internet. Usually, these are
   just the border MTAs as internal MTAs simply forward mail to these
   MTAs for relaying.

   The receiving ADMD will generally want to perform the authorization
   check at the boundary MTAs, including all secondary MXs. Internal
   MTAs (including MTAs that might serve as both boundary MTAs and
   internal relays from secondary MXs when they are processing the
   relayed mail stream) then do not perform the authorization test. To
   perform the authorization test other than at the boundary, the host
   that first transferred the message to the receiving ADMD has to be
   determined, which can be difficult to extract from the message header
   because (a) header fields can be forged or malformed, and (b) there's
   no standard way to encode that information such that it can be
   reliably extracted. Testing other than at the boundary is likely to
   produce unreliable results. This is described further in Appendix D
   of [RFC7001].





Kitterman Standards Track [Page 61]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


Appendix G. Local Policy Considerations

   SPF results can be used in combination with other methods to
   determine the final local disposition (either positive or negative)
   of a message. It can also be considered dispositive on its own.

G.1. Policy for SPF Pass

   SPF "pass" results can be used in combination with "whitelists" of
   known "good" domains to bypass some or all additional pre-delivery
   email checks. Exactly which checks and how to determine appropriate
   whitelist entries have to be based on local conditions and
   requirements.

G.2. Policy for SPF Fail

   SPF "fail" results can be used to reject messages during the SMTP
   transaction based on either "MAIL FROM" or "HELO" identity results.
   This reduces resource requirements for various content-filtering
   methods and conserves bandwidth since rejection can be done before
   the SMTP content is transferred. It also gives immediate feedback to
   the sender, who might then be able to resolve the issue. Due to some
   of the issues described in this section (Appendix G), SPF-based
   rejection does present some risk of rejecting legitimate email when
   rejecting email based on "MAIL FROM" results.

   SPF "fail" results can alternately be used as one input into a larger
   set of evaluations that might, based on a combination of SPF "fail"
   results with other evaluation techniques, result in the email being
   marked negatively in some way (this might be via delivery to a
   special spam folder, modifying subject lines, or other locally
   determined means). Developing the details of such an approach has to
   be based on local conditions and requirements. Using SPF results in
   this way does not have the advantages of resource conservation and
   immediate feedback to the sender associated with SMTP rejection, but
   could produce fewer undesirable rejections in a well-designed system.
   Such an approach might result in email that was not authorized by the
   sending ADMD being unknowingly delivered to end users.

   Either general approach can be used, as they both leave a clear
   disposition of emails; either they are delivered in some manner or
   the sender is notified of the failure. Other dispositions such as
   "dropping" or deleting email after acceptance are inappropriate
   because they leave uncertainty and reduce the overall reliability and
   utility of email across the Internet.






Kitterman Standards Track [Page 62]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


G.3. Policy for SPF Permerror

   The "permerror" result (see Section 2.6.7) indicates that the SPF
   processing module at the receiver determined that the retrieved SPF
   policy record could not be interpreted. This gives no true
   indication about the authorized use of the data found in the
   envelope.

   As with all results, implementers have a choice to make regarding
   what to do with a message that yields this result. SMTP allows only
   a few basic options.

   Rejection of the message is an option, in that it is the one thing a
   receiver can do to draw attention to the difficulty encountered while
   protecting itself from messages that do not have a definite SPF
   result of some kind. However, if the SPF implementation is defective
   and returns spurious "permerror" results, only the sender is actively
   notified of the defect (in the form of rejected mail), and not the
   receiver making use of SPF.

   The less intrusive handling choice is to deliver the message, perhaps
   with some kind of annotation of the difficulty encountered and/or
   logging of a similar nature. However, this will not be desirable to
   SPF verifier operators that wish to implement SPF checking as
   strictly as possible, nor is this sort of passive reporting of
   problems typically effective.

   There is of course the option of placing this choice in the hands of
   the SPF verifier operator rather than the implementer since this kind
   of choice is often a matter of local policy rather than a condition
   with a universal solution, but this adds one more piece of complexity
   to an already non-trivial environment.

   Both implementers and SPF verifier operators need to be cautious of
   all choices and outcomes when handling SPF results.

G.4. Policy for SPF Temperror

   The "temperror" result (see Section 2.6.6) indicates that the SPF
   processing module at the receiver could not retrieve an SPF policy
   record due to a (probably) transient condition. This gives no true
   indication about the authorized use of the data found in the
   envelope.

   As with all results, implementers have a choice to make regarding
   what to do with a message that yields this result. SMTP allows only
   a few basic options.




Kitterman Standards Track [Page 63]
 
RFC 7208 Sender Policy Framework (SPF) April 2014


   Deferring the message is an option, in that it is the one thing a
   receiver can do to draw attention to the difficulty encountered while
   protecting itself from messages that do not have a definite SPF
   result of some kind. However, if the SPF implementation is defective
   and returns spurious "temperror" results, only the sender is actively
   notified of the defect (in the form of mail rejected after it times
   out of the sending queue), and not the receiver making use of SPF.

   Because of long queue lifetimes, it is possible that mail will be
   repeatedly deferred for several days, and so any awareness that the
   sender may have regarding a problem could be quite delayed. If
   "temperrors" persist for multiple delivery attempts, it might be
   preferable to treat the error as permanent and reduce the amount of
   time the message is in transit.

   The less intrusive handling choice is to deliver the message, perhaps
   with some kind of annotation of the difficulty encountered and/or
   logging of a similar nature. However, this will not be desirable to
   SPF verifier operators that wish to implement SPF checking as
   strictly as possible, nor is this sort of passive reporting of
   problems typically effective.

   There is of course the option of placing this choice in the hands of
   the SPF verifier operator rather than the implementer since this kind
   of choice is often a matter of local policy rather than a condition
   with a universal solution, but this adds one more piece of complexity
   to an already non-trivial environment.

   Both implementers and SPF verifier operators need to be cautious of
   all choices and outcomes when handling SPF results.

Author's Address

   Scott Kitterman
   Kitterman Technical Services
   3611 Scheel Dr.
   Ellicott City, MD 21042
   United States of America

   EMail: scott@kitterman.com











Kitterman Standards Track [Page 64]