en-US/about_SMTP.txt

TOPIC
   about_SMTP


Network Working Group J. Klensin
Request for Comments: 5321 October 2008
Obsoletes: 2821
Updates: 1123
Category: Standards Track


                     Simple Mail Transfer Protocol

Status of This Memo

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

Abstract

   This document is a specification of the basic protocol for Internet
   electronic mail transport. It consolidates, updates, and clarifies
   several previous documents, making all or parts of most of them
   obsolete. It covers the SMTP extension mechanisms and best practices
   for the contemporary Internet, but does not provide details about
   particular extensions. Although SMTP was designed as a mail
   transport and delivery protocol, this specification also contains
   information that is important to its use as a "mail submission"
   protocol for "split-UA" (User Agent) mail reading systems and mobile
   environments.






















Klensin Standards Track [Page 1]
 
RFC 5321 SMTP October 2008


Table of Contents

   1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 5
     1.1. Transport of Electronic Mail . . . . . . . . . . . . . . . 5
     1.2. History and Context for This Document . . . . . . . . . . 5
     1.3. Document Conventions . . . . . . . . . . . . . . . . . . . 6
   2. The SMTP Model . . . . . . . . . . . . . . . . . . . . . . . . 7
     2.1. Basic Structure . . . . . . . . . . . . . . . . . . . . . 7
     2.2. The Extension Model . . . . . . . . . . . . . . . . . . . 9
       2.2.1. Background . . . . . . . . . . . . . . . . . . . . . . 9
       2.2.2. Definition and Registration of Extensions . . . . . . 10
       2.2.3. Special Issues with Extensions . . . . . . . . . . . . 11
     2.3. SMTP Terminology . . . . . . . . . . . . . . . . . . . . . 11
       2.3.1. Mail Objects . . . . . . . . . . . . . . . . . . . . . 11
       2.3.2. Senders and Receivers . . . . . . . . . . . . . . . . 12
       2.3.3. Mail Agents and Message Stores . . . . . . . . . . . . 12
       2.3.4. Host . . . . . . . . . . . . . . . . . . . . . . . . . 13
       2.3.5. Domain Names . . . . . . . . . . . . . . . . . . . . . 13
       2.3.6. Buffer and State Table . . . . . . . . . . . . . . . . 14
       2.3.7. Commands and Replies . . . . . . . . . . . . . . . . . 14
       2.3.8. Lines . . . . . . . . . . . . . . . . . . . . . . . . 14
       2.3.9. Message Content and Mail Data . . . . . . . . . . . . 15
       2.3.10. Originator, Delivery, Relay, and Gateway Systems . . . 15
       2.3.11. Mailbox and Address . . . . . . . . . . . . . . . . . 15
     2.4. General Syntax Principles and Transaction Model . . . . . 16
   3. The SMTP Procedures: An Overview . . . . . . . . . . . . . . . 17
     3.1. Session Initiation . . . . . . . . . . . . . . . . . . . . 18
     3.2. Client Initiation . . . . . . . . . . . . . . . . . . . . 18
     3.3. Mail Transactions . . . . . . . . . . . . . . . . . . . . 19
     3.4. Forwarding for Address Correction or Updating . . . . . . 21
     3.5. Commands for Debugging Addresses . . . . . . . . . . . . . 22
       3.5.1. Overview . . . . . . . . . . . . . . . . . . . . . . . 22
       3.5.2. VRFY Normal Response . . . . . . . . . . . . . . . . . 24
       3.5.3. Meaning of VRFY or EXPN Success Response . . . . . . . 25
       3.5.4. Semantics and Applications of EXPN . . . . . . . . . . 26
     3.6. Relaying and Mail Routing . . . . . . . . . . . . . . . . 26
       3.6.1. Source Routes and Relaying . . . . . . . . . . . . . . 26
       3.6.2. Mail eXchange Records and Relaying . . . . . . . . . . 26
       3.6.3. Message Submission Servers as Relays . . . . . . . . . 27
     3.7. Mail Gatewaying . . . . . . . . . . . . . . . . . . . . . 28
       3.7.1. Header Fields in Gatewaying . . . . . . . . . . . . . 28
       3.7.2. Received Lines in Gatewaying . . . . . . . . . . . . . 29
       3.7.3. Addresses in Gatewaying . . . . . . . . . . . . . . . 29
       3.7.4. Other Header Fields in Gatewaying . . . . . . . . . . 29
       3.7.5. Envelopes in Gatewaying . . . . . . . . . . . . . . . 30
     3.8. Terminating Sessions and Connections . . . . . . . . . . . 30
     3.9. Mailing Lists and Aliases . . . . . . . . . . . . . . . . 31
       3.9.1. Alias . . . . . . . . . . . . . . . . . . . . . . . . 31



Klensin Standards Track [Page 2]
 
RFC 5321 SMTP October 2008


       3.9.2. List . . . . . . . . . . . . . . . . . . . . . . . . . 31
   4. The SMTP Specifications . . . . . . . . . . . . . . . . . . . 32
     4.1. SMTP Commands . . . . . . . . . . . . . . . . . . . . . . 32
       4.1.1. Command Semantics and Syntax . . . . . . . . . . . . . 32
       4.1.2. Command Argument Syntax . . . . . . . . . . . . . . . 41
       4.1.3. Address Literals . . . . . . . . . . . . . . . . . . . 43
       4.1.4. Order of Commands . . . . . . . . . . . . . . . . . . 44
       4.1.5. Private-Use Commands . . . . . . . . . . . . . . . . . 46
     4.2. SMTP Replies . . . . . . . . . . . . . . . . . . . . . . . 46
       4.2.1. Reply Code Severities and Theory . . . . . . . . . . . 48
       4.2.2. Reply Codes by Function Groups . . . . . . . . . . . . 50
       4.2.3. Reply Codes in Numeric Order . . . . . . . . . . . . . 52
       4.2.4. Reply Code 502 . . . . . . . . . . . . . . . . . . . . 53
       4.2.5. Reply Codes after DATA and the Subsequent
               <CRLF>.<CRLF> . . . . . . . . . . . . . . . . . . . . 53
     4.3. Sequencing of Commands and Replies . . . . . . . . . . . . 54
       4.3.1. Sequencing Overview . . . . . . . . . . . . . . . . . 54
       4.3.2. Command-Reply Sequences . . . . . . . . . . . . . . . 55
     4.4. Trace Information . . . . . . . . . . . . . . . . . . . . 57
     4.5. Additional Implementation Issues . . . . . . . . . . . . . 61
       4.5.1. Minimum Implementation . . . . . . . . . . . . . . . . 61
       4.5.2. Transparency . . . . . . . . . . . . . . . . . . . . . 62
       4.5.3. Sizes and Timeouts . . . . . . . . . . . . . . . . . . 62
         4.5.3.1. Size Limits and Minimums . . . . . . . . . . . . . 62
           4.5.3.1.1. Local-part . . . . . . . . . . . . . . . . . . 63
           4.5.3.1.2. Domain . . . . . . . . . . . . . . . . . . . . 63
           4.5.3.1.3. Path . . . . . . . . . . . . . . . . . . . . . 63
           4.5.3.1.4. Command Line . . . . . . . . . . . . . . . . . 63
           4.5.3.1.5. Reply Line . . . . . . . . . . . . . . . . . . 63
           4.5.3.1.6. Text Line . . . . . . . . . . . . . . . . . . 63
           4.5.3.1.7. Message Content . . . . . . . . . . . . . . . 63
           4.5.3.1.8. Recipients Buffer . . . . . . . . . . . . . . 64
           4.5.3.1.9. Treatment When Limits Exceeded . . . . . . . . 64
           4.5.3.1.10. Too Many Recipients Code . . . . . . . . . . . 64
         4.5.3.2. Timeouts . . . . . . . . . . . . . . . . . . . . . 65
           4.5.3.2.1. Initial 220 Message: 5 Minutes . . . . . . . . 65
           4.5.3.2.2. MAIL Command: 5 Minutes . . . . . . . . . . . 65
           4.5.3.2.3. RCPT Command: 5 Minutes . . . . . . . . . . . 65
           4.5.3.2.4. DATA Initiation: 2 Minutes . . . . . . . . . . 66
           4.5.3.2.5. Data Block: 3 Minutes . . . . . . . . . . . . 66
           4.5.3.2.6. DATA Termination: 10 Minutes. . . . . . . . . 66
           4.5.3.2.7. Server Timeout: 5 Minutes. . . . . . . . . . . 66
       4.5.4. Retry Strategies . . . . . . . . . . . . . . . . . . . 66
       4.5.5. Messages with a Null Reverse-Path . . . . . . . . . . 68
   5. Address Resolution and Mail Handling . . . . . . . . . . . . . 69
     5.1. Locating the Target Host . . . . . . . . . . . . . . . . . 69
     5.2. IPv6 and MX Records . . . . . . . . . . . . . . . . . . . 71
   6. Problem Detection and Handling . . . . . . . . . . . . . . . . 71



Klensin Standards Track [Page 3]
 
RFC 5321 SMTP October 2008


     6.1. Reliable Delivery and Replies by Email . . . . . . . . . . 71
     6.2. Unwanted, Unsolicited, and "Attack" Messages . . . . . . . 72
     6.3. Loop Detection . . . . . . . . . . . . . . . . . . . . . . 73
     6.4. Compensating for Irregularities . . . . . . . . . . . . . 73
   7. Security Considerations . . . . . . . . . . . . . . . . . . . 75
     7.1. Mail Security and Spoofing . . . . . . . . . . . . . . . . 75
     7.2. "Blind" Copies . . . . . . . . . . . . . . . . . . . . . . 76
     7.3. VRFY, EXPN, and Security . . . . . . . . . . . . . . . . . 76
     7.4. Mail Rerouting Based on the 251 and 551 Response Codes . . 77
     7.5. Information Disclosure in Announcements . . . . . . . . . 77
     7.6. Information Disclosure in Trace Fields . . . . . . . . . . 78
     7.7. Information Disclosure in Message Forwarding . . . . . . . 78
     7.8. Resistance to Attacks . . . . . . . . . . . . . . . . . . 78
     7.9. Scope of Operation of SMTP Servers . . . . . . . . . . . . 78
   8. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 79
   9. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 80
   10. References . . . . . . . . . . . . . . . . . . . . . . . . . . 81
     10.1. Normative References . . . . . . . . . . . . . . . . . . . 81
     10.2. Informative References . . . . . . . . . . . . . . . . . . 82
   Appendix A. TCP Transport Service . . . . . . . . . . . . . . . . 85
   Appendix B. Generating SMTP Commands from RFC 822 Header
                Fields . . . . . . . . . . . . . . . . . . . . . . . 85
   Appendix C. Source Routes . . . . . . . . . . . . . . . . . . . . 86
   Appendix D. Scenarios . . . . . . . . . . . . . . . . . . . . . . 87
     D.1. A Typical SMTP Transaction Scenario . . . . . . . . . . . 88
     D.2. Aborted SMTP Transaction Scenario . . . . . . . . . . . . 89
     D.3. Relayed Mail Scenario . . . . . . . . . . . . . . . . . . 90
     D.4. Verifying and Sending Scenario . . . . . . . . . . . . . . 92
   Appendix E. Other Gateway Issues . . . . . . . . . . . . . . . . 92
   Appendix F. Deprecated Features of RFC 821 . . . . . . . . . . . 93
     F.1. TURN . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
     F.2. Source Routing . . . . . . . . . . . . . . . . . . . . . . 93
     F.3. HELO . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
     F.4. #-literals . . . . . . . . . . . . . . . . . . . . . . . . 94
     F.5. Dates and Years . . . . . . . . . . . . . . . . . . . . . 94
     F.6. Sending versus Mailing . . . . . . . . . . . . . . . . . . 94















Klensin Standards Track [Page 4]
 
RFC 5321 SMTP October 2008


1. Introduction

1.1. Transport of Electronic Mail

   The objective of the Simple Mail Transfer Protocol (SMTP) is to
   transfer mail reliably and efficiently.

   SMTP is independent of the particular transmission subsystem and
   requires only a reliable ordered data stream channel. While this
   document specifically discusses transport over TCP, other transports
   are possible. Appendices to RFC 821 [1] describe some of them.

   An important feature of SMTP is its capability to transport mail
   across multiple networks, usually referred to as "SMTP mail relaying"
   (see Section 3.6). A network consists of the mutually-TCP-accessible
   hosts on the public Internet, the mutually-TCP-accessible hosts on a
   firewall-isolated TCP/IP Intranet, or hosts in some other LAN or WAN
   environment utilizing a non-TCP transport-level protocol. Using
   SMTP, a process can transfer mail to another process on the same
   network or to some other network via a relay or gateway process
   accessible to both networks.

   In this way, a mail message may pass through a number of intermediate
   relay or gateway hosts on its path from sender to ultimate recipient.
   The Mail eXchanger mechanisms of the domain name system (RFC 1035
   [2], RFC 974 [12], and Section 5 of this document) are used to
   identify the appropriate next-hop destination for a message being
   transported.

1.2. History and Context for This Document

   This document is a specification of the basic protocol for the
   Internet electronic mail transport. It consolidates, updates and
   clarifies, but does not add new or change existing functionality of
   the following:

   o the original SMTP (Simple Mail Transfer Protocol) specification of
      RFC 821 [1],

   o domain name system requirements and implications for mail
      transport from RFC 1035 [2] and RFC 974 [12],

   o the clarifications and applicability statements in RFC 1123 [3],
      and

   o material drawn from the SMTP Extension mechanisms in RFC 1869
      [13].




Klensin Standards Track [Page 5]
 
RFC 5321 SMTP October 2008


   o Editorial and clarification changes to RFC 2821 [14] to bring that
      specification to Draft Standard.

   It obsoletes RFC 821, RFC 974, RFC 1869, and RFC 2821 and updates RFC
   1123 (replacing the mail transport materials of RFC 1123). However,
   RFC 821 specifies some features that were not in significant use in
   the Internet by the mid-1990s and (in appendices) some additional
   transport models. Those sections are omitted here in the interest of
   clarity and brevity; readers needing them should refer to RFC 821.

   It also includes some additional material from RFC 1123 that required
   amplification. This material has been identified in multiple ways,
   mostly by tracking flaming on various lists and newsgroups and
   problems of unusual readings or interpretations that have appeared as
   the SMTP extensions have been deployed. Where this specification
   moves beyond consolidation and actually differs from earlier
   documents, it supersedes them technically as well as textually.

   Although SMTP was designed as a mail transport and delivery protocol,
   this specification also contains information that is important to its
   use as a "mail submission" protocol, as recommended for Post Office
   Protocol (POP) (RFC 937 [15], RFC 1939 [16]) and IMAP (RFC 3501
   [17]). In general, the separate mail submission protocol specified
   in RFC 4409 [18] is now preferred to direct use of SMTP; more
   discussion of that subject appears in that document.

   Section 2.3 provides definitions of terms specific to this document.
   Except when the historical terminology is necessary for clarity, this
   document uses the current 'client' and 'server' terminology to
   identify the sending and receiving SMTP processes, respectively.

   A companion document, RFC 5322 [4], discusses message header sections
   and bodies and specifies formats and structures for them.

1.3. Document Conventions

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
   document are to be interpreted as described in RFC 2119 [5]. As each
   of these terms was intentionally and carefully chosen to improve the
   interoperability of email, each use of these terms is to be treated
   as a conformance requirement.

   Because this document has a long history and to avoid the risk of
   various errors and of confusing readers and documents that point to
   this one, most examples and the domain names they contain are
   preserved from RFC 2821. Readers are cautioned that these are




Klensin Standards Track [Page 6]
 
RFC 5321 SMTP October 2008


   illustrative examples that should not actually be used in either code
   or configuration files.

2. The SMTP Model

2.1. Basic Structure

   The SMTP design can be pictured as:

                  +----------+ +----------+
      +------+ | | | |
      | User |<-->| | SMTP | |
      +------+ | Client- |Commands/Replies| Server- |
      +------+ | SMTP |<-------------->| SMTP | +------+
      | File |<-->| | and Mail | |<-->| File |
      |System| | | | | |System|
      +------+ +----------+ +----------+ +------+
                   SMTP client SMTP server

   When an SMTP client has a message to transmit, it establishes a two-
   way transmission channel to an SMTP server. The responsibility of an
   SMTP client is to transfer mail messages to one or more SMTP servers,
   or report its failure to do so.

   The means by which a mail message is presented to an SMTP client, and
   how that client determines the identifier(s) ("names") of the
   domain(s) to which mail messages are to be transferred, is a local
   matter, and is not addressed by this document. In some cases, the
   designated domain(s), or those determined by an SMTP client, will
   identify the final destination(s) of the mail message. In other
   cases, common with SMTP clients associated with implementations of
   the POP (RFC 937 [15], RFC 1939 [16]) or IMAP (RFC 3501 [17])
   protocols, or when the SMTP client is inside an isolated transport
   service environment, the domain determined will identify an
   intermediate destination through which all mail messages are to be
   relayed. SMTP clients that transfer all traffic regardless of the
   target domains associated with the individual messages, or that do
   not maintain queues for retrying message transmissions that initially
   cannot be completed, may otherwise conform to this specification but
   are not considered fully-capable. Fully-capable SMTP
   implementations, including the relays used by these less capable
   ones, and their destinations, are expected to support all of the
   queuing, retrying, and alternate address functions discussed in this
   specification. In many situations and configurations, the less-
   capable clients discussed above SHOULD be using the message
   submission protocol (RFC 4409 [18]) rather than SMTP.





Klensin Standards Track [Page 7]
 
RFC 5321 SMTP October 2008


   The means by which an SMTP client, once it has determined a target
   domain, determines the identity of an SMTP server to which a copy of
   a message is to be transferred, and then performs that transfer, is
   covered by this document. To effect a mail transfer to an SMTP
   server, an SMTP client establishes a two-way transmission channel to
   that SMTP server. An SMTP client determines the address of an
   appropriate host running an SMTP server by resolving a destination
   domain name to either an intermediate Mail eXchanger host or a final
   target host.

   An SMTP server may be either the ultimate destination or an
   intermediate "relay" (that is, it may assume the role of an SMTP
   client after receiving the message) or "gateway" (that is, it may
   transport the message further using some protocol other than SMTP).
   SMTP commands are generated by the SMTP client and sent to the SMTP
   server. SMTP replies are sent from the SMTP server to the SMTP
   client in response to the commands.

   In other words, message transfer can occur in a single connection
   between the original SMTP-sender and the final SMTP-recipient, or can
   occur in a series of hops through intermediary systems. In either
   case, once the server has issued a success response at the end of the
   mail data, a formal handoff of responsibility for the message occurs:
   the protocol requires that a server MUST accept responsibility for
   either delivering the message or properly reporting the failure to do
   so (see Sections 6.1, 6.2, and 7.8, below).

   Once the transmission channel is established and initial handshaking
   is completed, the SMTP client normally initiates a mail transaction.
   Such a transaction consists of a series of commands to specify the
   originator and destination of the mail and transmission of the
   message content (including any lines in the header section or other
   structure) itself. When the same message is sent to multiple
   recipients, this protocol encourages the transmission of only one
   copy of the data for all recipients at the same destination (or
   intermediate relay) host.

   The server responds to each command with a reply; replies may
   indicate that the command was accepted, that additional commands are
   expected, or that a temporary or permanent error condition exists.
   Commands specifying the sender or recipients may include server-
   permitted SMTP service extension requests, as discussed in
   Section 2.2. The dialog is purposely lock-step, one-at-a-time,
   although this can be modified by mutually agreed upon extension
   requests such as command pipelining (RFC 2920 [19]).

   Once a given mail message has been transmitted, the client may either
   request that the connection be shut down or may initiate other mail



Klensin Standards Track [Page 8]
 
RFC 5321 SMTP October 2008


   transactions. In addition, an SMTP client may use a connection to an
   SMTP server for ancillary services such as verification of email
   addresses or retrieval of mailing list subscriber addresses.

   As suggested above, this protocol provides mechanisms for the
   transmission of mail. Historically, this transmission normally
   occurred directly from the sending user's host to the receiving
   user's host when the two hosts are connected to the same transport
   service. When they are not connected to the same transport service,
   transmission occurs via one or more relay SMTP servers. A very
   common case in the Internet today involves submission of the original
   message to an intermediate, "message submission" server, which is
   similar to a relay but has some additional properties; such servers
   are discussed in Section 2.3.10 and at some length in RFC 4409 [18].
   An intermediate host that acts as either an SMTP relay or as a
   gateway into some other transmission environment is usually selected
   through the use of the domain name service (DNS) Mail eXchanger
   mechanism.

   Usually, intermediate hosts are determined via the DNS MX record, not
   by explicit "source" routing (see Section 5 and Appendix C and
   Appendix F.2).

2.2. The Extension Model

2.2.1. Background

   In an effort that started in 1990, approximately a decade after RFC
   821 was completed, the protocol was modified with a "service
   extensions" model that permits the client and server to agree to
   utilize shared functionality beyond the original SMTP requirements.
   The SMTP extension mechanism defines a means whereby an extended SMTP
   client and server may recognize each other, and the server can inform
   the client as to the service extensions that it supports.

   Contemporary SMTP implementations MUST support the basic extension
   mechanisms. For instance, servers MUST support the EHLO command even
   if they do not implement any specific extensions and clients SHOULD
   preferentially utilize EHLO rather than HELO. (However, for
   compatibility with older conforming implementations, SMTP clients and
   servers MUST support the original HELO mechanisms as a fallback.)
   Unless the different characteristics of HELO must be identified for
   interoperability purposes, this document discusses only EHLO.

   SMTP is widely deployed and high-quality implementations have proven
   to be very robust. However, the Internet community now considers
   some services to be important that were not anticipated when the
   protocol was first designed. If support for those services is to be



Klensin Standards Track [Page 9]
 
RFC 5321 SMTP October 2008


   added, it must be done in a way that permits older implementations to
   continue working acceptably. The extension framework consists of:

   o The SMTP command EHLO, superseding the earlier HELO,

   o a registry of SMTP service extensions,

   o additional parameters to the SMTP MAIL and RCPT commands, and

   o optional replacements for commands defined in this protocol, such
      as for DATA in non-ASCII transmissions (RFC 3030 [20]).

   SMTP's strength comes primarily from its simplicity. Experience with
   many protocols has shown that protocols with few options tend towards
   ubiquity, whereas protocols with many options tend towards obscurity.

   Each and every extension, regardless of its benefits, must be
   carefully scrutinized with respect to its implementation, deployment,
   and interoperability costs. In many cases, the cost of extending the
   SMTP service will likely outweigh the benefit.

2.2.2. Definition and Registration of Extensions

   The IANA maintains a registry of SMTP service extensions. A
   corresponding EHLO keyword value is associated with each extension.
   Each service extension registered with the IANA must be defined in a
   formal Standards-Track or IESG-approved Experimental protocol
   document. The definition must include:

   o the textual name of the SMTP service extension;

   o the EHLO keyword value associated with the extension;

   o the syntax and possible values of parameters associated with the
      EHLO keyword value;

   o any additional SMTP verbs associated with the extension
      (additional verbs will usually be, but are not required to be, the
      same as the EHLO keyword value);

   o any new parameters the extension associates with the MAIL or RCPT
      verbs;

   o a description of how support for the extension affects the
      behavior of a server and client SMTP; and






Klensin Standards Track [Page 10]
 
RFC 5321 SMTP October 2008


   o the increment by which the extension is increasing the maximum
      length of the commands MAIL and/or RCPT, over that specified in
      this Standard.

   In addition, any EHLO keyword value starting with an upper or lower
   case "X" refers to a local SMTP service extension used exclusively
   through bilateral agreement. Keywords beginning with "X" MUST NOT be
   used in a registered service extension. Conversely, keyword values
   presented in the EHLO response that do not begin with "X" MUST
   correspond to a Standard, Standards-Track, or IESG-approved
   Experimental SMTP service extension registered with IANA. A
   conforming server MUST NOT offer non-"X"-prefixed keyword values that
   are not described in a registered extension.

   Additional verbs and parameter names are bound by the same rules as
   EHLO keywords; specifically, verbs beginning with "X" are local
   extensions that may not be registered or standardized. Conversely,
   verbs not beginning with "X" must always be registered.

2.2.3. Special Issues with Extensions

   Extensions that change fairly basic properties of SMTP operation are
   permitted. The text in other sections of this document must be
   understood in that context. In particular, extensions can change the
   minimum limits specified in Section 4.5.3, can change the ASCII
   character set requirement as mentioned above, or can introduce some
   optional modes of message handling.

   In particular, if an extension implies that the delivery path
   normally supports special features of that extension, and an
   intermediate SMTP system finds a next hop that does not support the
   required extension, it MAY choose, based on the specific extension
   and circumstances, to requeue the message and try later and/or try an
   alternate MX host. If this strategy is employed, the timeout to fall
   back to an unextended format (if one is available) SHOULD be less
   than the normal timeout for bouncing as undeliverable (e.g., if
   normal timeout is three days, the requeue timeout before attempting
   to transmit the mail without the extension might be one day).

2.3. SMTP Terminology

2.3.1. Mail Objects

   SMTP transports a mail object. A mail object contains an envelope
   and content.

   The SMTP envelope is sent as a series of SMTP protocol units
   (described in Section 3). It consists of an originator address (to



Klensin Standards Track [Page 11]
 
RFC 5321 SMTP October 2008


   which error reports should be directed), one or more recipient
   addresses, and optional protocol extension material. Historically,
   variations on the reverse-path (originator) address specification
   command (MAIL) could be used to specify alternate delivery modes,
   such as immediate display; those variations have now been deprecated
   (see Appendix F and Appendix F.6).

   The SMTP content is sent in the SMTP DATA protocol unit and has two
   parts: the header section and the body. If the content conforms to
   other contemporary standards, the header section consists of a
   collection of header fields, each consisting of a header name, a
   colon, and data, structured as in the message format specification
   (RFC 5322 [4]); the body, if structured, is defined according to MIME
   (RFC 2045 [21]). The content is textual in nature, expressed using
   the US-ASCII repertoire [6]. Although SMTP extensions (such as
   "8BITMIME", RFC 1652 [22]) may relax this restriction for the content
   body, the content header fields are always encoded using the US-ASCII
   repertoire. Two MIME extensions (RFC 2047 [23] and RFC 2231 [24])
   define an algorithm for representing header values outside the US-
   ASCII repertoire, while still encoding them using the US-ASCII
   repertoire.

2.3.2. Senders and Receivers

   In RFC 821, the two hosts participating in an SMTP transaction were
   described as the "SMTP-sender" and "SMTP-receiver". This document
   has been changed to reflect current industry terminology and hence
   refers to them as the "SMTP client" (or sometimes just "the client")
   and "SMTP server" (or just "the server"), respectively. Since a
   given host may act both as server and client in a relay situation,
   "receiver" and "sender" terminology is still used where needed for
   clarity.

2.3.3. Mail Agents and Message Stores

   Additional mail system terminology became common after RFC 821 was
   published and, where convenient, is used in this specification. In
   particular, SMTP servers and clients provide a mail transport service
   and therefore act as "Mail Transfer Agents" (MTAs). "Mail User
   Agents" (MUAs or UAs) are normally thought of as the sources and
   targets of mail. At the source, an MUA might collect mail to be
   transmitted from a user and hand it off to an MTA; the final
   ("delivery") MTA would be thought of as handing the mail off to an
   MUA (or at least transferring responsibility to it, e.g., by
   depositing the message in a "message store"). However, while these
   terms are used with at least the appearance of great precision in
   other environments, the implied boundaries between MUAs and MTAs
   often do not accurately match common, and conforming, practices with



Klensin Standards Track [Page 12]
 
RFC 5321 SMTP October 2008


   Internet mail. Hence, the reader should be cautious about inferring
   the strong relationships and responsibilities that might be implied
   if these terms were used elsewhere.

2.3.4. Host

   For the purposes of this specification, a host is a computer system
   attached to the Internet (or, in some cases, to a private TCP/IP
   network) and supporting the SMTP protocol. Hosts are known by names
   (see the next section); they SHOULD NOT be identified by numerical
   addresses, i.e., by address literals as described in Section 4.1.2.

2.3.5. Domain Names

   A domain name (or often just a "domain") consists of one or more
   components, separated by dots if more than one appears. In the case
   of a top-level domain used by itself in an email address, a single
   string is used without any dots. This makes the requirement,
   described in more detail below, that only fully-qualified domain
   names appear in SMTP transactions on the public Internet,
   particularly important where top-level domains are involved. These
   components ("labels" in DNS terminology, RFC 1035 [2]) are restricted
   for SMTP purposes to consist of a sequence of letters, digits, and
   hyphens drawn from the ASCII character set [6]. Domain names are
   used as names of hosts and of other entities in the domain name
   hierarchy. For example, a domain may refer to an alias (label of a
   CNAME RR) or the label of Mail eXchanger records to be used to
   deliver mail instead of representing a host name. See RFC 1035 [2]
   and Section 5 of this specification.

   The domain name, as described in this document and in RFC 1035 [2],
   is the entire, fully-qualified name (often referred to as an "FQDN").
   A domain name that is not in FQDN form is no more than a local alias.
   Local aliases MUST NOT appear in any SMTP transaction.

   Only resolvable, fully-qualified domain names (FQDNs) are permitted
   when domain names are used in SMTP. In other words, names that can
   be resolved to MX RRs or address (i.e., A or AAAA) RRs (as discussed
   in Section 5) are permitted, as are CNAME RRs whose targets can be
   resolved, in turn, to MX or address RRs. Local nicknames or
   unqualified names MUST NOT be used. There are two exceptions to the
   rule requiring FQDNs:

   o The domain name given in the EHLO command MUST be either a primary
      host name (a domain name that resolves to an address RR) or, if
      the host has no name, an address literal, as described in
      Section 4.1.3 and discussed further in the EHLO discussion of
      Section 4.1.4.



Klensin Standards Track [Page 13]
 
RFC 5321 SMTP October 2008


   o The reserved mailbox name "postmaster" may be used in a RCPT
      command without domain qualification (see Section 4.1.1.3) and
      MUST be accepted if so used.

2.3.6. Buffer and State Table

   SMTP sessions are stateful, with both parties carefully maintaining a
   common view of the current state. In this document, we model this
   state by a virtual "buffer" and a "state table" on the server that
   may be used by the client to, for example, "clear the buffer" or
   "reset the state table", causing the information in the buffer to be
   discarded and the state to be returned to some previous state.

2.3.7. Commands and Replies

   SMTP commands and, unless altered by a service extension, message
   data, are transmitted from the sender to the receiver via the
   transmission channel in "lines".

   An SMTP reply is an acknowledgment (positive or negative) sent in
   "lines" from receiver to sender via the transmission channel in
   response to a command. The general form of a reply is a numeric
   completion code (indicating failure or success) usually followed by a
   text string. The codes are for use by programs and the text is
   usually intended for human users. RFC 3463 [25], specifies further
   structuring of the reply strings, including the use of supplemental
   and more specific completion codes (see also RFC 5248 [26]).

2.3.8. Lines

   Lines consist of zero or more data characters terminated by the
   sequence ASCII character "CR" (hex value 0D) followed immediately by
   ASCII character "LF" (hex value 0A). This termination sequence is
   denoted as <CRLF> in this document. Conforming implementations MUST
   NOT recognize or generate any other character or character sequence
   as a line terminator. Limits MAY be imposed on line lengths by
   servers (see Section 4).

   In addition, the appearance of "bare" "CR" or "LF" characters in text
   (i.e., either without the other) has a long history of causing
   problems in mail implementations and applications that use the mail
   system as a tool. SMTP client implementations MUST NOT transmit
   these characters except when they are intended as line terminators
   and then MUST, as indicated above, transmit them only as a <CRLF>
   sequence.






Klensin Standards Track [Page 14]
 
RFC 5321 SMTP October 2008


2.3.9. Message Content and Mail Data

   The terms "message content" and "mail data" are used interchangeably
   in this document to describe the material transmitted after the DATA
   command is accepted and before the end of data indication is
   transmitted. Message content includes the message header section and
   the possibly structured message body. The MIME specification (RFC
   2045 [21]) provides the standard mechanisms for structured message
   bodies.

2.3.10. Originator, Delivery, Relay, and Gateway Systems

   This specification makes a distinction among four types of SMTP
   systems, based on the role those systems play in transmitting
   electronic mail. An "originating" system (sometimes called an SMTP
   originator) introduces mail into the Internet or, more generally,
   into a transport service environment. A "delivery" SMTP system is
   one that receives mail from a transport service environment and
   passes it to a mail user agent or deposits it in a message store that
   a mail user agent is expected to subsequently access. A "relay" SMTP
   system (usually referred to just as a "relay") receives mail from an
   SMTP client and transmits it, without modification to the message
   data other than adding trace information, to another SMTP server for
   further relaying or for delivery.

   A "gateway" SMTP system (usually referred to just as a "gateway")
   receives mail from a client system in one transport environment and
   transmits it to a server system in another transport environment.
   Differences in protocols or message semantics between the transport
   environments on either side of a gateway may require that the gateway
   system perform transformations to the message that are not permitted
   to SMTP relay systems. For the purposes of this specification,
   firewalls that rewrite addresses should be considered as gateways,
   even if SMTP is used on both sides of them (see RFC 2979 [27]).

2.3.11. Mailbox and Address

   As used in this specification, an "address" is a character string
   that identifies a user to whom mail will be sent or a location into
   which mail will be deposited. The term "mailbox" refers to that
   depository. The two terms are typically used interchangeably unless
   the distinction between the location in which mail is placed (the
   mailbox) and a reference to it (the address) is important. An
   address normally consists of user and domain specifications. The
   standard mailbox naming convention is defined to be
   "local-part@domain"; contemporary usage permits a much broader set of
   applications than simple "user names". Consequently, and due to a
   long history of problems when intermediate hosts have attempted to



Klensin Standards Track [Page 15]
 
RFC 5321 SMTP October 2008


   optimize transport by modifying them, the local-part MUST be
   interpreted and assigned semantics only by the host specified in the
   domain part of the address.

2.4. General Syntax Principles and Transaction Model

   SMTP commands and replies have a rigid syntax. All commands begin
   with a command verb. All replies begin with a three digit numeric
   code. In some commands and replies, arguments are required following
   the verb or reply code. Some commands do not accept arguments (after
   the verb), and some reply codes are followed, sometimes optionally,
   by free form text. In both cases, where text appears, it is
   separated from the verb or reply code by a space character. Complete
   definitions of commands and replies appear in Section 4.

   Verbs and argument values (e.g., "TO:" or "to:" in the RCPT command
   and extension name keywords) are not case sensitive, with the sole
   exception in this specification of a mailbox local-part (SMTP
   Extensions may explicitly specify case-sensitive elements). That is,
   a command verb, an argument value other than a mailbox local-part,
   and free form text MAY be encoded in upper case, lower case, or any
   mixture of upper and lower case with no impact on its meaning. The
   local-part of a mailbox MUST BE treated as case sensitive.
   Therefore, SMTP implementations MUST take care to preserve the case
   of mailbox local-parts. In particular, for some hosts, the user
   "smith" is different from the user "Smith". However, exploiting the
   case sensitivity of mailbox local-parts impedes interoperability and
   is discouraged. Mailbox domains follow normal DNS rules and are
   hence not case sensitive.

   A few SMTP servers, in violation of this specification (and RFC 821)
   require that command verbs be encoded by clients in upper case.
   Implementations MAY wish to employ this encoding to accommodate those
   servers.

   The argument clause consists of a variable-length character string
   ending with the end of the line, i.e., with the character sequence
   <CRLF>. The receiver will take no action until this sequence is
   received.

   The syntax for each command is shown with the discussion of that
   command. Common elements and parameters are shown in Section 4.1.2.

   Commands and replies are composed of characters from the ASCII
   character set [6]. When the transport service provides an 8-bit byte
   (octet) transmission channel, each 7-bit character is transmitted,
   right justified, in an octet with the high-order bit cleared to zero.
   More specifically, the unextended SMTP service provides 7-bit



Klensin Standards Track [Page 16]
 
RFC 5321 SMTP October 2008


   transport only. An originating SMTP client that has not successfully
   negotiated an appropriate extension with a particular server (see the
   next paragraph) MUST NOT transmit messages with information in the
   high-order bit of octets. If such messages are transmitted in
   violation of this rule, receiving SMTP servers MAY clear the high-
   order bit or reject the message as invalid. In general, a relay SMTP
   SHOULD assume that the message content it has received is valid and,
   assuming that the envelope permits doing so, relay it without
   inspecting that content. Of course, if the content is mislabeled and
   the data path cannot accept the actual content, this may result in
   the ultimate delivery of a severely garbled message to the recipient.
   Delivery SMTP systems MAY reject such messages, or return them as
   undeliverable, rather than deliver them. In the absence of a server-
   offered extension explicitly permitting it, a sending SMTP system is
   not permitted to send envelope commands in any character set other
   than US-ASCII. Receiving systems SHOULD reject such commands,
   normally using "500 syntax error - invalid character" replies.

   8-bit message content transmission MAY be requested of the server by
   a client using extended SMTP facilities, notably the "8BITMIME"
   extension, RFC 1652 [22]. 8BITMIME SHOULD be supported by SMTP
   servers. However, it MUST NOT be construed as authorization to
   transmit unrestricted 8-bit material, nor does 8BITMIME authorize
   transmission of any envelope material in other than ASCII. 8BITMIME
   MUST NOT be requested by senders for material with the high bit on
   that is not in MIME format with an appropriate content-transfer
   encoding; servers MAY reject such messages.

   The metalinguistic notation used in this document corresponds to the
   "Augmented BNF" used in other Internet mail system documents. The
   reader who is not familiar with that syntax should consult the ABNF
   specification in RFC 5234 [7]. Metalanguage terms used in running
   text are surrounded by pointed brackets (e.g., <CRLF>) for clarity.
   The reader is cautioned that the grammar expressed in the
   metalanguage is not comprehensive. There are many instances in which
   provisions in the text constrain or otherwise modify the syntax or
   semantics implied by the grammar.

3. The SMTP Procedures: An Overview

   This section contains descriptions of the procedures used in SMTP:
   session initiation, mail transaction, forwarding mail, verifying
   mailbox names and expanding mailing lists, and opening and closing
   exchanges. Comments on relaying, a note on mail domains, and a
   discussion of changing roles are included at the end of this section.
   Several complete scenarios are presented in Appendix D.





Klensin Standards Track [Page 17]
 
RFC 5321 SMTP October 2008


3.1. Session Initiation

   An SMTP session is initiated when a client opens a connection to a
   server and the server responds with an opening message.

   SMTP server implementations MAY include identification of their
   software and version information in the connection greeting reply
   after the 220 code, a practice that permits more efficient isolation
   and repair of any problems. Implementations MAY make provision for
   SMTP servers to disable the software and version announcement where
   it causes security concerns. While some systems also identify their
   contact point for mail problems, this is not a substitute for
   maintaining the required "postmaster" address (see Section 4).

   The SMTP protocol allows a server to formally reject a mail session
   while still allowing the initial connection as follows: a 554
   response MAY be given in the initial connection opening message
   instead of the 220. A server taking this approach MUST still wait
   for the client to send a QUIT (see Section 4.1.1.10) before closing
   the connection and SHOULD respond to any intervening commands with
   "503 bad sequence of commands". Since an attempt to make an SMTP
   connection to such a system is probably in error, a server returning
   a 554 response on connection opening SHOULD provide enough
   information in the reply text to facilitate debugging of the sending
   system.

3.2. Client Initiation

   Once the server has sent the greeting (welcoming) message and the
   client has received it, the client normally sends the EHLO command to
   the server, indicating the client's identity. In addition to opening
   the session, use of EHLO indicates that the client is able to process
   service extensions and requests that the server provide a list of the
   extensions it supports. Older SMTP systems that are unable to
   support service extensions, and contemporary clients that do not
   require service extensions in the mail session being initiated, MAY
   use HELO instead of EHLO. Servers MUST NOT return the extended EHLO-
   style response to a HELO command. For a particular connection
   attempt, if the server returns a "command not recognized" response to
   EHLO, the client SHOULD be able to fall back and send HELO.

   In the EHLO command, the host sending the command identifies itself;
   the command may be interpreted as saying "Hello, I am <domain>" (and,
   in the case of EHLO, "and I support service extension requests").







Klensin Standards Track [Page 18]
 
RFC 5321 SMTP October 2008


3.3. Mail Transactions

   There are three steps to SMTP mail transactions. The transaction
   starts with a MAIL command that gives the sender identification. (In
   general, the MAIL command may be sent only when no mail transaction
   is in progress; see Section 4.1.4.) A series of one or more RCPT
   commands follows, giving the receiver information. Then, a DATA
   command initiates transfer of the mail data and is terminated by the
   "end of mail" data indicator, which also confirms the transaction.

   The first step in the procedure is the MAIL command.

      MAIL FROM:<reverse-path> [SP <mail-parameters> ] <CRLF>

   This command tells the SMTP-receiver that a new mail transaction is
   starting and to reset all its state tables and buffers, including any
   recipients or mail data. The <reverse-path> portion of the first or
   only argument contains the source mailbox (between "<" and ">"
   brackets), which can be used to report errors (see Section 4.2 for a
   discussion of error reporting). If accepted, the SMTP server returns
   a "250 OK" reply. If the mailbox specification is not acceptable for
   some reason, the server MUST return a reply indicating whether the
   failure is permanent (i.e., will occur again if the client tries to
   send the same address again) or temporary (i.e., the address might be
   accepted if the client tries again later). Despite the apparent
   scope of this requirement, there are circumstances in which the
   acceptability of the reverse-path may not be determined until one or
   more forward-paths (in RCPT commands) can be examined. In those
   cases, the server MAY reasonably accept the reverse-path (with a 250
   reply) and then report problems after the forward-paths are received
   and examined. Normally, failures produce 550 or 553 replies.

   Historically, the <reverse-path> was permitted to contain more than
   just a mailbox; however, contemporary systems SHOULD NOT use source
   routing (see Appendix C).

   The optional <mail-parameters> are associated with negotiated SMTP
   service extensions (see Section 2.2).

   The second step in the procedure is the RCPT command. This step of
   the procedure can be repeated any number of times.

      RCPT TO:<forward-path> [ SP <rcpt-parameters> ] <CRLF>

   The first or only argument to this command includes a forward-path
   (normally a mailbox and domain, always surrounded by "<" and ">"
   brackets) identifying one recipient. If accepted, the SMTP server
   returns a "250 OK" reply and stores the forward-path. If the



Klensin Standards Track [Page 19]
 
RFC 5321 SMTP October 2008


   recipient is known not to be a deliverable address, the SMTP server
   returns a 550 reply, typically with a string such as "no such user -
   " and the mailbox name (other circumstances and reply codes are
   possible).

   The <forward-path> can contain more than just a mailbox.
   Historically, the <forward-path> was permitted to contain a source
   routing list of hosts and the destination mailbox; however,
   contemporary SMTP clients SHOULD NOT utilize source routes (see
   Appendix C). Servers MUST be prepared to encounter a list of source
   routes in the forward-path, but they SHOULD ignore the routes or MAY
   decline to support the relaying they imply. Similarly, servers MAY
   decline to accept mail that is destined for other hosts or systems.
   These restrictions make a server useless as a relay for clients that
   do not support full SMTP functionality. Consequently, restricted-
   capability clients MUST NOT assume that any SMTP server on the
   Internet can be used as their mail processing (relaying) site. If a
   RCPT command appears without a previous MAIL command, the server MUST
   return a 503 "Bad sequence of commands" response. The optional
   <rcpt-parameters> are associated with negotiated SMTP service
   extensions (see Section 2.2).

   Since it has been a common source of errors, it is worth noting that
   spaces are not permitted on either side of the colon following FROM
   in the MAIL command or TO in the RCPT command. The syntax is exactly
   as given above.

   The third step in the procedure is the DATA command (or some
   alternative specified in a service extension).

      DATA <CRLF>

   If accepted, the SMTP server returns a 354 Intermediate reply and
   considers all succeeding lines up to but not including the end of
   mail data indicator to be the message text. When the end of text is
   successfully received and stored, the SMTP-receiver sends a "250 OK"
   reply.

   Since the mail data is sent on the transmission channel, the end of
   mail data must be indicated so that the command and reply dialog can
   be resumed. SMTP indicates the end of the mail data by sending a
   line containing only a "." (period or full stop). A transparency
   procedure is used to prevent this from interfering with the user's
   text (see Section 4.5.2).

   The end of mail data indicator also confirms the mail transaction and
   tells the SMTP server to now process the stored recipients and mail




Klensin Standards Track [Page 20]
 
RFC 5321 SMTP October 2008


   data. If accepted, the SMTP server returns a "250 OK" reply. The
   DATA command can fail at only two points in the protocol exchange:

   If there was no MAIL, or no RCPT, command, or all such commands were
   rejected, the server MAY return a "command out of sequence" (503) or
   "no valid recipients" (554) reply in response to the DATA command.
   If one of those replies (or any other 5yz reply) is received, the
   client MUST NOT send the message data; more generally, message data
   MUST NOT be sent unless a 354 reply is received.

   If the verb is initially accepted and the 354 reply issued, the DATA
   command should fail only if the mail transaction was incomplete (for
   example, no recipients), if resources were unavailable (including, of
   course, the server unexpectedly becoming unavailable), or if the
   server determines that the message should be rejected for policy or
   other reasons.

   However, in practice, some servers do not perform recipient
   verification until after the message text is received. These servers
   SHOULD treat a failure for one or more recipients as a "subsequent
   failure" and return a mail message as discussed in Section 6 and, in
   particular, in Section 6.1. Using a "550 mailbox not found" (or
   equivalent) reply code after the data are accepted makes it difficult
   or impossible for the client to determine which recipients failed.

   When the RFC 822 format ([28], [4]) is being used, the mail data
   include the header fields such as those named Date, Subject, To, Cc,
   and From. Server SMTP systems SHOULD NOT reject messages based on
   perceived defects in the RFC 822 or MIME (RFC 2045 [21]) message
   header section or message body. In particular, they MUST NOT reject
   messages in which the numbers of Resent-header fields do not match or
   Resent-to appears without Resent-from and/or Resent-date.

   Mail transaction commands MUST be used in the order discussed above.

3.4. Forwarding for Address Correction or Updating

   Forwarding support is most often required to consolidate and simplify
   addresses within, or relative to, some enterprise and less frequently
   to establish addresses to link a person's prior address with a
   current one. Silent forwarding of messages (without server
   notification to the sender), for security or non-disclosure purposes,
   is common in the contemporary Internet.

   In both the enterprise and the "new address" cases, information
   hiding (and sometimes security) considerations argue against exposure
   of the "final" address through the SMTP protocol as a side effect of
   the forwarding activity. This may be especially important when the



Klensin Standards Track [Page 21]
 
RFC 5321 SMTP October 2008


   final address may not even be reachable by the sender. Consequently,
   the "forwarding" mechanisms described in Section 3.2 of RFC 821, and
   especially the 251 (corrected destination) and 551 reply codes from
   RCPT must be evaluated carefully by implementers and, when they are
   available, by those configuring systems (see also Section 7.4).

   In particular:

   o Servers MAY forward messages when they are aware of an address
      change. When they do so, they MAY either provide address-updating
      information with a 251 code, or may forward "silently" and return
      a 250 code. However, if a 251 code is used, they MUST NOT assume
      that the client will actually update address information or even
      return that information to the user.

   Alternately,

   o Servers MAY reject messages or return them as non-deliverable when
      they cannot be delivered precisely as addressed. When they do so,
      they MAY either provide address-updating information with a 551
      code, or may reject the message as undeliverable with a 550 code
      and no address-specific information. However, if a 551 code is
      used, they MUST NOT assume that the client will actually update
      address information or even return that information to the user.

   SMTP server implementations that support the 251 and/or 551 reply
   codes SHOULD provide configuration mechanisms so that sites that
   conclude that they would undesirably disclose information can disable
   or restrict their use.

3.5. Commands for Debugging Addresses

3.5.1. Overview

   SMTP provides commands to verify a user name or obtain the content of
   a mailing list. This is done with the VRFY and EXPN commands, which
   have character string arguments. Implementations SHOULD support VRFY
   and EXPN (however, see Section 3.5.2 and Section 7.3).

   For the VRFY command, the string is a user name or a user name and
   domain (see below). If a normal (i.e., 250) response is returned,
   the response MAY include the full name of the user and MUST include
   the mailbox of the user. It MUST be in either of the following
   forms:

      User Name <local-part@domain>
      local-part@domain




Klensin Standards Track [Page 22]
 
RFC 5321 SMTP October 2008


   When a name that is the argument to VRFY could identify more than one
   mailbox, the server MAY either note the ambiguity or identify the
   alternatives. In other words, any of the following are legitimate
   responses to VRFY:

      553 User ambiguous

   or

      553- Ambiguous; Possibilities are
      553-Joe Smith <jsmith@foo.com>
      553-Harry Smith <hsmith@foo.com>
      553 Melvin Smith <dweep@foo.com>

   or

      553-Ambiguous; Possibilities
      553- <jsmith@foo.com>
      553- <hsmith@foo.com>
      553 <dweep@foo.com>

   Under normal circumstances, a client receiving a 553 reply would be
   expected to expose the result to the user. Use of exactly the forms
   given, and the "user ambiguous" or "ambiguous" keywords, possibly
   supplemented by extended reply codes, such as those described in RFC
   3463 [25], will facilitate automated translation into other languages
   as needed. Of course, a client that was highly automated or that was
   operating in another language than English might choose to try to
   translate the response to return some other indication to the user
   than the literal text of the reply, or to take some automated action
   such as consulting a directory service for additional information
   before reporting to the user.

   For the EXPN command, the string identifies a mailing list, and the
   successful (i.e., 250) multiline response MAY include the full name
   of the users and MUST give the mailboxes on the mailing list.

   In some hosts, the distinction between a mailing list and an alias
   for a single mailbox is a bit fuzzy, since a common data structure
   may hold both types of entries, and it is possible to have mailing
   lists containing only one mailbox. If a request is made to apply
   VRFY to a mailing list, a positive response MAY be given if a message
   so addressed would be delivered to everyone on the list, otherwise an
   error SHOULD be reported (e.g., "550 That is a mailing list, not a
   user" or "252 Unable to verify members of mailing list"). If a
   request is made to expand a user name, the server MAY return a





Klensin Standards Track [Page 23]
 
RFC 5321 SMTP October 2008


   positive response consisting of a list containing one name, or an
   error MAY be reported (e.g., "550 That is a user name, not a mailing
   list").

   In the case of a successful multiline reply (normal for EXPN),
   exactly one mailbox is to be specified on each line of the reply.
   The case of an ambiguous request is discussed above.

   "User name" is a fuzzy term and has been used deliberately. An
   implementation of the VRFY or EXPN commands MUST include at least
   recognition of local mailboxes as "user names". However, since
   current Internet practice often results in a single host handling
   mail for multiple domains, hosts, especially hosts that provide this
   functionality, SHOULD accept the "local-part@domain" form as a "user
   name"; hosts MAY also choose to recognize other strings as "user
   names".

   The case of expanding a mailbox list requires a multiline reply, such
   as:

      C: EXPN Example-People
      S: 250-Jon Postel <Postel@isi.edu>
      S: 250-Fred Fonebone <Fonebone@physics.foo-u.edu>
      S: 250 Sam Q. Smith <SQSmith@specific.generic.com>

   or

      C: EXPN Executive-Washroom-List
      S: 550 Access Denied to You.

   The character string arguments of the VRFY and EXPN commands cannot
   be further restricted due to the variety of implementations of the
   user name and mailbox list concepts. On some systems, it may be
   appropriate for the argument of the EXPN command to be a file name
   for a file containing a mailing list, but again there are a variety
   of file naming conventions in the Internet. Similarly, historical
   variations in what is returned by these commands are such that the
   response SHOULD be interpreted very carefully, if at all, and SHOULD
   generally only be used for diagnostic purposes.

3.5.2. VRFY Normal Response

   When normal (2yz or 551) responses are returned from a VRFY or EXPN
   request, the reply MUST include the <Mailbox> name using a
   "<local-part@domain>" construction, where "domain" is a fully-
   qualified domain name. In circumstances exceptional enough to
   justify violating the intent of this specification, free-form text
   MAY be returned. In order to facilitate parsing by both computers



Klensin Standards Track [Page 24]
 
RFC 5321 SMTP October 2008


   and people, addresses SHOULD appear in pointed brackets. When
   addresses, rather than free-form debugging information, are returned,
   EXPN and VRFY MUST return only valid domain addresses that are usable
   in SMTP RCPT commands. Consequently, if an address implies delivery
   to a program or other system, the mailbox name used to reach that
   target MUST be given. Paths (explicit source routes) MUST NOT be
   returned by VRFY or EXPN.

   Server implementations SHOULD support both VRFY and EXPN. For
   security reasons, implementations MAY provide local installations a
   way to disable either or both of these commands through configuration
   options or the equivalent (see Section 7.3). When these commands are
   supported, they are not required to work across relays when relaying
   is supported. Since they were both optional in RFC 821, but VRFY was
   made mandatory in RFC 1123 [3], if EXPN is supported, it MUST be
   listed as a service extension in an EHLO response. VRFY MAY be
   listed as a convenience but, since support for it is required, SMTP
   clients are not required to check for its presence on the extension
   list before using it.

3.5.3. Meaning of VRFY or EXPN Success Response

   A server MUST NOT return a 250 code in response to a VRFY or EXPN
   command unless it has actually verified the address. In particular,
   a server MUST NOT return 250 if all it has done is to verify that the
   syntax given is valid. In that case, 502 (Command not implemented)
   or 500 (Syntax error, command unrecognized) SHOULD be returned. As
   stated elsewhere, implementation (in the sense of actually validating
   addresses and returning information) of VRFY and EXPN are strongly
   recommended. Hence, implementations that return 500 or 502 for VRFY
   are not in full compliance with this specification.

   There may be circumstances where an address appears to be valid but
   cannot reasonably be verified in real time, particularly when a
   server is acting as a mail exchanger for another server or domain.
   "Apparent validity", in this case, would normally involve at least
   syntax checking and might involve verification that any domains
   specified were ones to which the host expected to be able to relay
   mail. In these situations, reply code 252 SHOULD be returned. These
   cases parallel the discussion of RCPT verification in Section 2.1.
   Similarly, the discussion in Section 3.4 applies to the use of reply
   codes 251 and 551 with VRFY (and EXPN) to indicate addresses that are
   recognized but that would be forwarded or rejected were mail received
   for them. Implementations generally SHOULD be more aggressive about
   address verification in the case of VRFY than in the case of RCPT,
   even if it takes a little longer to do so.





Klensin Standards Track [Page 25]
 
RFC 5321 SMTP October 2008


3.5.4. Semantics and Applications of EXPN

   EXPN is often very useful in debugging and understanding problems
   with mailing lists and multiple-target-address aliases. Some systems
   have attempted to use source expansion of mailing lists as a means of
   eliminating duplicates. The propagation of aliasing systems with
   mail on the Internet for hosts (typically with MX and CNAME DNS
   records), for mailboxes (various types of local host aliases), and in
   various proxying arrangements has made it nearly impossible for these
   strategies to work consistently, and mail systems SHOULD NOT attempt
   them.

3.6. Relaying and Mail Routing

3.6.1. Source Routes and Relaying

   In general, the availability of Mail eXchanger records in the domain
   name system (RFC 1035 [2], RFC 974 [12]) makes the use of explicit
   source routes in the Internet mail system unnecessary. Many
   historical problems with the interpretation of explicit source routes
   have made their use undesirable. SMTP clients SHOULD NOT generate
   explicit source routes except under unusual circumstances. SMTP
   servers MAY decline to act as mail relays or to accept addresses that
   specify source routes. When route information is encountered, SMTP
   servers MAY ignore the route information and simply send to the final
   destination specified as the last element in the route and SHOULD do
   so. There has been an invalid practice of using names that do not
   appear in the DNS as destination names, with the senders counting on
   the intermediate hosts specified in source routing to resolve any
   problems. If source routes are stripped, this practice will cause
   failures. This is one of several reasons why SMTP clients MUST NOT
   generate invalid source routes or depend on serial resolution of
   names.

   When source routes are not used, the process described in RFC 821 for
   constructing a reverse-path from the forward-path is not applicable
   and the reverse-path at the time of delivery will simply be the
   address that appeared in the MAIL command.

3.6.2. Mail eXchange Records and Relaying

   A relay SMTP server is usually the target of a DNS MX record that
   designates it, rather than the final delivery system. The relay
   server may accept or reject the task of relaying the mail in the same
   way it accepts or rejects mail for a local user. If it accepts the
   task, it then becomes an SMTP client, establishes a transmission
   channel to the next SMTP server specified in the DNS (according to
   the rules in Section 5), and sends it the mail. If it declines to



Klensin Standards Track [Page 26]
 
RFC 5321 SMTP October 2008


   relay mail to a particular address for policy reasons, a 550 response
   SHOULD be returned.

   This specification does not deal with the verification of return
   paths for use in delivery notifications. Recent work, such as that
   on SPF [29] and DKIM [30] [31], has been done to provide ways to
   ascertain that an address is valid or belongs to the person who
   actually sent the message. A server MAY attempt to verify the return
   path before using its address for delivery notifications, but methods
   of doing so are not defined here nor is any particular method
   recommended at this time.

3.6.3. Message Submission Servers as Relays

   Many mail-sending clients exist, especially in conjunction with
   facilities that receive mail via POP3 or IMAP, that have limited
   capability to support some of the requirements of this specification,
   such as the ability to queue messages for subsequent delivery
   attempts. For these clients, it is common practice to make private
   arrangements to send all messages to a single server for processing
   and subsequent distribution. SMTP, as specified here, is not ideally
   suited for this role. A standardized mail submission protocol has
   been developed that is gradually superseding practices based on SMTP
   (see RFC 4409 [18]). In any event, because these arrangements are
   private and fall outside the scope of this specification, they are
   not described here.

   It is important to note that MX records can point to SMTP servers
   that act as gateways into other environments, not just SMTP relays
   and final delivery systems; see Sections 3.7 and 5.

   If an SMTP server has accepted the task of relaying the mail and
   later finds that the destination is incorrect or that the mail cannot
   be delivered for some other reason, then it MUST construct an
   "undeliverable mail" notification message and send it to the
   originator of the undeliverable mail (as indicated by the reverse-
   path). Formats specified for non-delivery reports by other standards
   (see, for example, RFC 3461 [32] and RFC 3464 [33]) SHOULD be used if
   possible.

   This notification message must be from the SMTP server at the relay
   host or the host that first determines that delivery cannot be
   accomplished. Of course, SMTP servers MUST NOT send notification
   messages about problems transporting notification messages. One way
   to prevent loops in error reporting is to specify a null reverse-path
   in the MAIL command of a notification message. When such a message
   is transmitted, the reverse-path MUST be set to null (see




Klensin Standards Track [Page 27]
 
RFC 5321 SMTP October 2008


   Section 4.5.5 for additional discussion). A MAIL command with a null
   reverse-path appears as follows:

      MAIL FROM:<>

   As discussed in Section 6.4, a relay SMTP has no need to inspect or
   act upon the header section or body of the message data and MUST NOT
   do so except to add its own "Received:" header field (Section 4.4)
   and, optionally, to attempt to detect looping in the mail system (see
   Section 6.3). Of course, this prohibition also applies to any
   modifications of these header fields or text (see also Section 7.9).

3.7. Mail Gatewaying

   While the relay function discussed above operates within the Internet
   SMTP transport service environment, MX records or various forms of
   explicit routing may require that an intermediate SMTP server perform
   a translation function between one transport service and another. As
   discussed in Section 2.3.10, when such a system is at the boundary
   between two transport service environments, we refer to it as a
   "gateway" or "gateway SMTP".

   Gatewaying mail between different mail environments, such as
   different mail formats and protocols, is complex and does not easily
   yield to standardization. However, some general requirements may be
   given for a gateway between the Internet and another mail
   environment.

3.7.1. Header Fields in Gatewaying

   Header fields MAY be rewritten when necessary as messages are
   gatewayed across mail environment boundaries. This may involve
   inspecting the message body or interpreting the local-part of the
   destination address in spite of the prohibitions in Section 6.4.

   Other mail systems gatewayed to the Internet often use a subset of
   the RFC 822 header section or provide similar functionality with a
   different syntax, but some of these mail systems do not have an
   equivalent to the SMTP envelope. Therefore, when a message leaves
   the Internet environment, it may be necessary to fold the SMTP
   envelope information into the message header section. A possible
   solution would be to create new header fields to carry the envelope
   information (e.g., "X-SMTP-MAIL:" and "X-SMTP-RCPT:"); however, this
   would require changes in mail programs in foreign environments and
   might risk disclosure of private information (see Section 7.2).






Klensin Standards Track [Page 28]
 
RFC 5321 SMTP October 2008


3.7.2. Received Lines in Gatewaying

   When forwarding a message into or out of the Internet environment, a
   gateway MUST prepend a Received: line, but it MUST NOT alter in any
   way a Received: line that is already in the header section.

   "Received:" header fields of messages originating from other
   environments may not conform exactly to this specification. However,
   the most important use of Received: lines is for debugging mail
   faults, and this debugging can be severely hampered by well-meaning
   gateways that try to "fix" a Received: line. As another consequence
   of trace header fields arising in non-SMTP environments, receiving
   systems MUST NOT reject mail based on the format of a trace header
   field and SHOULD be extremely robust in the light of unexpected
   information or formats in those header fields.

   The gateway SHOULD indicate the environment and protocol in the "via"
   clauses of Received header field(s) that it supplies.

3.7.3. Addresses in Gatewaying

   From the Internet side, the gateway SHOULD accept all valid address
   formats in SMTP commands and in the RFC 822 header section, and all
   valid RFC 822 messages. Addresses and header fields generated by
   gateways MUST conform to applicable standards (including this one and
   RFC 5322 [4]). Gateways are, of course, subject to the same rules
   for handling source routes as those described for other SMTP systems
   in Section 3.3.

3.7.4. Other Header Fields in Gatewaying

   The gateway MUST ensure that all header fields of a message that it
   forwards into the Internet mail environment meet the requirements for
   Internet mail. In particular, all addresses in "From:", "To:",
   "Cc:", etc., header fields MUST be transformed (if necessary) to
   satisfy the standard header syntax of RFC 5322 [4], MUST reference
   only fully-qualified domain names, and MUST be effective and useful
   for sending replies. The translation algorithm used to convert mail
   from the Internet protocols to another environment's protocol SHOULD
   ensure that error messages from the foreign mail environment are
   delivered to the reverse-path from the SMTP envelope, not to an
   address in the "From:", "Sender:", or similar header fields of the
   message.








Klensin Standards Track [Page 29]
 
RFC 5321 SMTP October 2008


3.7.5. Envelopes in Gatewaying

   Similarly, when forwarding a message from another environment into
   the Internet, the gateway SHOULD set the envelope return path in
   accordance with an error message return address, if supplied by the
   foreign environment. If the foreign environment has no equivalent
   concept, the gateway must select and use a best approximation, with
   the message originator's address as the default of last resort.

3.8. Terminating Sessions and Connections

   An SMTP connection is terminated when the client sends a QUIT
   command. The server responds with a positive reply code, after which
   it closes the connection.

   An SMTP server MUST NOT intentionally close the connection under
   normal operational circumstances (see Section 7.8) except:

   o After receiving a QUIT command and responding with a 221 reply.

   o After detecting the need to shut down the SMTP service and
      returning a 421 response code. This response code can be issued
      after the server receives any command or, if necessary,
      asynchronously from command receipt (on the assumption that the
      client will receive it after the next command is issued).

   o After a timeout, as specified in Section 4.5.3.2, occurs waiting
      for the client to send a command or data.

   In particular, a server that closes connections in response to
   commands that are not understood is in violation of this
   specification. Servers are expected to be tolerant of unknown
   commands, issuing a 500 reply and awaiting further instructions from
   the client.

   An SMTP server that is forcibly shut down via external means SHOULD
   attempt to send a line containing a 421 response code to the SMTP
   client before exiting. The SMTP client will normally read the 421
   response code after sending its next command.

   SMTP clients that experience a connection close, reset, or other
   communications failure due to circumstances not under their control
   (in violation of the intent of this specification but sometimes
   unavoidable) SHOULD, to maintain the robustness of the mail system,
   treat the mail transaction as if a 451 response had been received and
   act accordingly.





Klensin Standards Track [Page 30]
 
RFC 5321 SMTP October 2008


3.9. Mailing Lists and Aliases

   An SMTP-capable host SHOULD support both the alias and the list
   models of address expansion for multiple delivery. When a message is
   delivered or forwarded to each address of an expanded list form, the
   return address in the envelope ("MAIL FROM:") MUST be changed to be
   the address of a person or other entity who administers the list.
   However, in this case, the message header section (RFC 5322 [4]) MUST
   be left unchanged; in particular, the "From" field of the header
   section is unaffected.

   An important mail facility is a mechanism for multi-destination
   delivery of a single message, by transforming (or "expanding" or
   "exploding") a pseudo-mailbox address into a list of destination
   mailbox addresses. When a message is sent to such a pseudo-mailbox
   (sometimes called an "exploder"), copies are forwarded or
   redistributed to each mailbox in the expanded list. Servers SHOULD
   simply utilize the addresses on the list; application of heuristics
   or other matching rules to eliminate some addresses, such as that of
   the originator, is strongly discouraged. We classify such a pseudo-
   mailbox as an "alias" or a "list", depending upon the expansion
   rules.

3.9.1. Alias

   To expand an alias, the recipient mailer simply replaces the pseudo-
   mailbox address in the envelope with each of the expanded addresses
   in turn; the rest of the envelope and the message body are left
   unchanged. The message is then delivered or forwarded to each
   expanded address.

3.9.2. List

   A mailing list may be said to operate by "redistribution" rather than
   by "forwarding". To expand a list, the recipient mailer replaces the
   pseudo-mailbox address in the envelope with each of the expanded
   addresses in turn. The return (backward-pointing) address in the
   envelope is changed so that all error messages generated by the final
   deliveries will be returned to a list administrator, not to the
   message originator, who generally has no control over the contents of
   the list and will typically find error messages annoying. Note that
   the key difference between handling aliases (Section 3.9.1) and
   forwarding (this subsection) is the change to the backward-pointing
   address in this case. When a list constrains its processing to the
   very limited set of modifications and actions described here, it is
   attempting to emulate an MTA; such lists can be treated as a
   continuation in email transit.




Klensin Standards Track [Page 31]
 
RFC 5321 SMTP October 2008


   There exist mailing lists that perform additional, sometimes
   extensive, modifications to a message and its envelope. Such mailing
   lists need to be viewed as full MUAs, which accept a delivery and
   post a new message.

4. The SMTP Specifications

4.1. SMTP Commands

4.1.1. Command Semantics and Syntax

   The SMTP commands define the mail transfer or the mail system
   function requested by the user. SMTP commands are character strings
   terminated by <CRLF>. The commands themselves are alphabetic
   characters terminated by <SP> if parameters follow and <CRLF>
   otherwise. (In the interest of improved interoperability, SMTP
   receivers SHOULD tolerate trailing white space before the terminating
   <CRLF>.) The syntax of the local part of a mailbox MUST conform to
   receiver site conventions and the syntax specified in Section 4.1.2.
   The SMTP commands are discussed below. The SMTP replies are
   discussed in Section 4.2.

   A mail transaction involves several data objects that are
   communicated as arguments to different commands. The reverse-path is
   the argument of the MAIL command, the forward-path is the argument of
   the RCPT command, and the mail data is the argument of the DATA
   command. These arguments or data objects must be transmitted and
   held, pending the confirmation communicated by the end of mail data
   indication that finalizes the transaction. The model for this is
   that distinct buffers are provided to hold the types of data objects;
   that is, there is a reverse-path buffer, a forward-path buffer, and a
   mail data buffer. Specific commands cause information to be appended
   to a specific buffer, or cause one or more buffers to be cleared.

   Several commands (RSET, DATA, QUIT) are specified as not permitting
   parameters. In the absence of specific extensions offered by the
   server and accepted by the client, clients MUST NOT send such
   parameters and servers SHOULD reject commands containing them as
   having invalid syntax.

4.1.1.1. Extended HELLO (EHLO) or HELLO (HELO)

   These commands are used to identify the SMTP client to the SMTP
   server. The argument clause contains the fully-qualified domain name
   of the SMTP client, if one is available. In situations in which the
   SMTP client system does not have a meaningful domain name (e.g., when
   its address is dynamically allocated and no reverse mapping record is




Klensin Standards Track [Page 32]
 
RFC 5321 SMTP October 2008


   available), the client SHOULD send an address literal (see
   Section 4.1.3).

   RFC 2821, and some earlier informal practices, encouraged following
   the literal by information that would help to identify the client
   system. That convention was not widely supported, and many SMTP
   servers considered it an error. In the interest of interoperability,
   it is probably wise for servers to be prepared for this string to
   occur, but SMTP clients SHOULD NOT send it.

   The SMTP server identifies itself to the SMTP client in the
   connection greeting reply and in the response to this command.

   A client SMTP SHOULD start an SMTP session by issuing the EHLO
   command. If the SMTP server supports the SMTP service extensions, it
   will give a successful response, a failure response, or an error
   response. If the SMTP server, in violation of this specification,
   does not support any SMTP service extensions, it will generate an
   error response. Older client SMTP systems MAY, as discussed above,
   use HELO (as specified in RFC 821) instead of EHLO, and servers MUST
   support the HELO command and reply properly to it. In any event, a
   client MUST issue HELO or EHLO before starting a mail transaction.

   These commands, and a "250 OK" reply to one of them, confirm that
   both the SMTP client and the SMTP server are in the initial state,
   that is, there is no transaction in progress and all state tables and
   buffers are cleared.

   Syntax:

   ehlo = "EHLO" SP ( Domain / address-literal ) CRLF

   helo = "HELO" SP Domain CRLF

   Normally, the response to EHLO will be a multiline reply. Each line
   of the response contains a keyword and, optionally, one or more
   parameters. Following the normal syntax for multiline replies, these
   keywords follow the code (250) and a hyphen for all but the last
   line, and the code and a space for the last line. The syntax for a
   positive response, using the ABNF notation and terminal symbols of
   RFC 5234 [7], is:

   ehlo-ok-rsp = ( "250" SP Domain [ SP ehlo-greet ] CRLF )
                    / ( "250-" Domain [ SP ehlo-greet ] CRLF
                    *( "250-" ehlo-line CRLF )
                    "250" SP ehlo-line CRLF )





Klensin Standards Track [Page 33]
 
RFC 5321 SMTP October 2008


   ehlo-greet = 1*(%d0-9 / %d11-12 / %d14-127)
                    ; string of any characters other than CR or LF

   ehlo-line = ehlo-keyword *( SP ehlo-param )

   ehlo-keyword = (ALPHA / DIGIT) *(ALPHA / DIGIT / "-")
                    ; additional syntax of ehlo-params depends on
                    ; ehlo-keyword

   ehlo-param = 1*(%d33-126)
                    ; any CHAR excluding <SP> and all
                    ; control characters (US-ASCII 0-31 and 127
                    ; inclusive)

   Although EHLO keywords may be specified in upper, lower, or mixed
   case, they MUST always be recognized and processed in a case-
   insensitive manner. This is simply an extension of practices
   specified in RFC 821 and Section 2.4.

   The EHLO response MUST contain keywords (and associated parameters if
   required) for all commands not listed as "required" in Section 4.5.1
   excepting only private-use commands as described in Section 4.1.5.
   Private-use commands MAY be listed.

4.1.1.2. MAIL (MAIL)

   This command is used to initiate a mail transaction in which the mail
   data is delivered to an SMTP server that may, in turn, deliver it to
   one or more mailboxes or pass it on to another system (possibly using
   SMTP). The argument clause contains a reverse-path and may contain
   optional parameters. In general, the MAIL command may be sent only
   when no mail transaction is in progress, see Section 4.1.4.

   The reverse-path consists of the sender mailbox. Historically, that
   mailbox might optionally have been preceded by a list of hosts, but
   that behavior is now deprecated (see Appendix C). In some types of
   reporting messages for which a reply is likely to cause a mail loop
   (for example, mail delivery and non-delivery notifications), the
   reverse-path may be null (see Section 3.6).

   This command clears the reverse-path buffer, the forward-path buffer,
   and the mail data buffer, and it inserts the reverse-path information
   from its argument clause into the reverse-path buffer.

   If service extensions were negotiated, the MAIL command may also
   carry parameters associated with a particular service extension.





Klensin Standards Track [Page 34]
 
RFC 5321 SMTP October 2008


   Syntax:

   mail = "MAIL FROM:" Reverse-path
                                       [SP Mail-parameters] CRLF

4.1.1.3. RECIPIENT (RCPT)

   This command is used to identify an individual recipient of the mail
   data; multiple recipients are specified by multiple uses of this
   command. The argument clause contains a forward-path and may contain
   optional parameters.

   The forward-path normally consists of the required destination
   mailbox. Sending systems SHOULD NOT generate the optional list of
   hosts known as a source route. Receiving systems MUST recognize
   source route syntax but SHOULD strip off the source route
   specification and utilize the domain name associated with the mailbox
   as if the source route had not been provided.

   Similarly, relay hosts SHOULD strip or ignore source routes, and
   names MUST NOT be copied into the reverse-path. When mail reaches
   its ultimate destination (the forward-path contains only a
   destination mailbox), the SMTP server inserts it into the destination
   mailbox in accordance with its host mail conventions.

   This command appends its forward-path argument to the forward-path
   buffer; it does not change the reverse-path buffer nor the mail data
   buffer.

   For example, mail received at relay host xyz.com with envelope
   commands

      MAIL FROM:<userx@y.foo.org>
      RCPT TO:<@hosta.int,@jkl.org:userc@d.bar.org>

   will normally be sent directly on to host d.bar.org with envelope
   commands

      MAIL FROM:<userx@y.foo.org>
      RCPT TO:<userc@d.bar.org>

   As provided in Appendix C, xyz.com MAY also choose to relay the
   message to hosta.int, using the envelope commands

      MAIL FROM:<userx@y.foo.org>
      RCPT TO:<@hosta.int,@jkl.org:userc@d.bar.org>





Klensin Standards Track [Page 35]
 
RFC 5321 SMTP October 2008


   or to jkl.org, using the envelope commands

      MAIL FROM:<userx@y.foo.org>
      RCPT TO:<@jkl.org:userc@d.bar.org>

   Attempting to use relaying this way is now strongly discouraged.
   Since hosts are not required to relay mail at all, xyz.com MAY also
   reject the message entirely when the RCPT command is received, using
   a 550 code (since this is a "policy reason").

   If service extensions were negotiated, the RCPT command may also
   carry parameters associated with a particular service extension
   offered by the server. The client MUST NOT transmit parameters other
   than those associated with a service extension offered by the server
   in its EHLO response.

   Syntax:

      rcpt = "RCPT TO:" ( "<Postmaster@" Domain ">" / "<Postmaster>" /
                  Forward-path ) [SP Rcpt-parameters] CRLF

                  Note that, in a departure from the usual rules for
                  local-parts, the "Postmaster" string shown above is
                  treated as case-insensitive.

4.1.1.4. DATA (DATA)

   The receiver normally sends a 354 response to DATA, and then treats
   the lines (strings ending in <CRLF> sequences, as described in
   Section 2.3.7) following the command as mail data from the sender.
   This command causes the mail data to be appended to the mail data
   buffer. The mail data may contain any of the 128 ASCII character
   codes, although experience has indicated that use of control
   characters other than SP, HT, CR, and LF may cause problems and
   SHOULD be avoided when possible.

   The mail data are terminated by a line containing only a period, that
   is, the character sequence "<CRLF>.<CRLF>", where the first <CRLF> is
   actually the terminator of the previous line (see Section 4.5.2).
   This is the end of mail data indication. The first <CRLF> of this
   terminating sequence is also the <CRLF> that ends the final line of
   the data (message text) or, if there was no mail data, ends the DATA
   command itself (the "no mail data" case does not conform to this
   specification since it would require that neither the trace header
   fields required by this specification nor the message header section
   required by RFC 5322 [4] be transmitted). An extra <CRLF> MUST NOT
   be added, as that would cause an empty line to be added to the
   message. The only exception to this rule would arise if the message



Klensin Standards Track [Page 36]
 
RFC 5321 SMTP October 2008


   body were passed to the originating SMTP-sender with a final "line"
   that did not end in <CRLF>; in that case, the originating SMTP system
   MUST either reject the message as invalid or add <CRLF> in order to
   have the receiving SMTP server recognize the "end of data" condition.

   The custom of accepting lines ending only in <LF>, as a concession to
   non-conforming behavior on the part of some UNIX systems, has proven
   to cause more interoperability problems than it solves, and SMTP
   server systems MUST NOT do this, even in the name of improved
   robustness. In particular, the sequence "<LF>.<LF>" (bare line
   feeds, without carriage returns) MUST NOT be treated as equivalent to
   <CRLF>.<CRLF> as the end of mail data indication.

   Receipt of the end of mail data indication requires the server to
   process the stored mail transaction information. This processing
   consumes the information in the reverse-path buffer, the forward-path
   buffer, and the mail data buffer, and on the completion of this
   command these buffers are cleared. If the processing is successful,
   the receiver MUST send an OK reply. If the processing fails, the
   receiver MUST send a failure reply. The SMTP model does not allow
   for partial failures at this point: either the message is accepted by
   the server for delivery and a positive response is returned or it is
   not accepted and a failure reply is returned. In sending a positive
   "250 OK" completion reply to the end of data indication, the receiver
   takes full responsibility for the message (see Section 6.1). Errors
   that are diagnosed subsequently MUST be reported in a mail message,
   as discussed in Section 4.4.

   When the SMTP server accepts a message either for relaying or for
   final delivery, it inserts a trace record (also referred to
   interchangeably as a "time stamp line" or "Received" line) at the top
   of the mail data. This trace record indicates the identity of the
   host that sent the message, the identity of the host that received
   the message (and is inserting this time stamp), and the date and time
   the message was received. Relayed messages will have multiple time
   stamp lines. Details for formation of these lines, including their
   syntax, is specified in Section 4.4.

   Additional discussion about the operation of the DATA command appears
   in Section 3.3.

   Syntax:

      data = "DATA" CRLF







Klensin Standards Track [Page 37]
 
RFC 5321 SMTP October 2008


4.1.1.5. RESET (RSET)

   This command specifies that the current mail transaction will be
   aborted. Any stored sender, recipients, and mail data MUST be
   discarded, and all buffers and state tables cleared. The receiver
   MUST send a "250 OK" reply to a RSET command with no arguments. A
   reset command may be issued by the client at any time. It is
   effectively equivalent to a NOOP (i.e., it has no effect) if issued
   immediately after EHLO, before EHLO is issued in the session, after
   an end of data indicator has been sent and acknowledged, or
   immediately before a QUIT. An SMTP server MUST NOT close the
   connection as the result of receiving a RSET; that action is reserved
   for QUIT (see Section 4.1.1.10).

   Since EHLO implies some additional processing and response by the
   server, RSET will normally be more efficient than reissuing that
   command, even though the formal semantics are the same.

   There are circumstances, contrary to the intent of this
   specification, in which an SMTP server may receive an indication that
   the underlying TCP connection has been closed or reset. To preserve
   the robustness of the mail system, SMTP servers SHOULD be prepared
   for this condition and SHOULD treat it as if a QUIT had been received
   before the connection disappeared.

   Syntax:

      rset = "RSET" CRLF

4.1.1.6. VERIFY (VRFY)

   This command asks the receiver to confirm that the argument
   identifies a user or mailbox. If it is a user name, information is
   returned as specified in Section 3.5.

   This command has no effect on the reverse-path buffer, the forward-
   path buffer, or the mail data buffer.

   Syntax:

      vrfy = "VRFY" SP String CRLF










Klensin Standards Track [Page 38]
 
RFC 5321 SMTP October 2008


4.1.1.7. EXPAND (EXPN)

   This command asks the receiver to confirm that the argument
   identifies a mailing list, and if so, to return the membership of
   that list. If the command is successful, a reply is returned
   containing information as described in Section 3.5. This reply will
   have multiple lines except in the trivial case of a one-member list.

   This command has no effect on the reverse-path buffer, the forward-
   path buffer, or the mail data buffer, and it may be issued at any
   time.

   Syntax:

      expn = "EXPN" SP String CRLF

4.1.1.8. HELP (HELP)

   This command causes the server to send helpful information to the
   client. The command MAY take an argument (e.g., any command name)
   and return more specific information as a response.

   This command has no effect on the reverse-path buffer, the forward-
   path buffer, or the mail data buffer, and it may be issued at any
   time.

   SMTP servers SHOULD support HELP without arguments and MAY support it
   with arguments.

   Syntax:

      help = "HELP" [ SP String ] CRLF



















Klensin Standards Track [Page 39]
 
RFC 5321 SMTP October 2008


4.1.1.9. NOOP (NOOP)

   This command does not affect any parameters or previously entered
   commands. It specifies no action other than that the receiver send a
   "250 OK" reply.

   This command has no effect on the reverse-path buffer, the forward-
   path buffer, or the mail data buffer, and it may be issued at any
   time. If a parameter string is specified, servers SHOULD ignore it.

   Syntax:

      noop = "NOOP" [ SP String ] CRLF

4.1.1.10. QUIT (QUIT)

   This command specifies that the receiver MUST send a "221 OK" reply,
   and then close the transmission channel.

   The receiver MUST NOT intentionally close the transmission channel
   until it receives and replies to a QUIT command (even if there was an
   error). The sender MUST NOT intentionally close the transmission
   channel until it sends a QUIT command, and it SHOULD wait until it
   receives the reply (even if there was an error response to a previous
   command). If the connection is closed prematurely due to violations
   of the above or system or network failure, the server MUST cancel any
   pending transaction, but not undo any previously completed
   transaction, and generally MUST act as if the command or transaction
   in progress had received a temporary error (i.e., a 4yz response).

   The QUIT command may be issued at any time. Any current uncompleted
   mail transaction will be aborted.

   Syntax:

      quit = "QUIT" CRLF

4.1.1.11. Mail-Parameter and Rcpt-Parameter Error Responses

   If the server SMTP does not recognize or cannot implement one or more
   of the parameters associated with a particular MAIL FROM or RCPT TO
   command, it will return code 555.

   If, for some reason, the server is temporarily unable to accommodate
   one or more of the parameters associated with a MAIL FROM or RCPT TO
   command, and if the definition of the specific parameter does not
   mandate the use of another code, it should return code 455.




Klensin Standards Track [Page 40]
 
RFC 5321 SMTP October 2008


   Errors specific to particular parameters and their values will be
   specified in the parameter's defining RFC.

4.1.2. Command Argument Syntax

   The syntax of the argument clauses of the above commands (using the
   syntax specified in RFC 5234 [7] where applicable) is given below.
   Some of the productions given below are used only in conjunction with
   source routes as described in Appendix C. Terminals not defined in
   this document, such as ALPHA, DIGIT, SP, CR, LF, CRLF, are as defined
   in the "core" syntax in Section 6 of RFC 5234 [7] or in the message
   format syntax in RFC 5322 [4].

   Reverse-path = Path / "<>"

   Forward-path = Path

   Path = "<" [ A-d-l ":" ] Mailbox ">"

   A-d-l = At-domain *( "," At-domain )
                  ; Note that this form, the so-called "source
                  ; route", MUST BE accepted, SHOULD NOT be
                  ; generated, and SHOULD be ignored.

   At-domain = "@" Domain

   Mail-parameters = esmtp-param *(SP esmtp-param)

   Rcpt-parameters = esmtp-param *(SP esmtp-param)

   esmtp-param = esmtp-keyword ["=" esmtp-value]

   esmtp-keyword = (ALPHA / DIGIT) *(ALPHA / DIGIT / "-")

   esmtp-value = 1*(%d33-60 / %d62-126)
                  ; any CHAR excluding "=", SP, and control
                  ; characters. If this string is an email address,
                  ; i.e., a Mailbox, then the "xtext" syntax [32]
                  ; SHOULD be used.

   Keyword = Ldh-str

   Argument = Atom

   Domain = sub-domain *("." sub-domain)






Klensin Standards Track [Page 41]
 
RFC 5321 SMTP October 2008


   sub-domain = Let-dig [Ldh-str]

   Let-dig = ALPHA / DIGIT

   Ldh-str = *( ALPHA / DIGIT / "-" ) Let-dig

   address-literal = "[" ( IPv4-address-literal /
                    IPv6-address-literal /
                    General-address-literal ) "]"
                    ; See Section 4.1.3

   Mailbox = Local-part "@" ( Domain / address-literal )

   Local-part = Dot-string / Quoted-string
                  ; MAY be case-sensitive


   Dot-string = Atom *("." Atom)

   Atom = 1*atext

   Quoted-string = DQUOTE *QcontentSMTP DQUOTE

   QcontentSMTP = qtextSMTP / quoted-pairSMTP

   quoted-pairSMTP = %d92 %d32-126
                    ; i.e., backslash followed by any ASCII
                    ; graphic (including itself) or SPace

   qtextSMTP = %d32-33 / %d35-91 / %d93-126
                  ; i.e., within a quoted string, any
                  ; ASCII graphic or space is permitted
                  ; without blackslash-quoting except
                  ; double-quote and the backslash itself.

   String = Atom / Quoted-string

   While the above definition for Local-part is relatively permissive,
   for maximum interoperability, a host that expects to receive mail
   SHOULD avoid defining mailboxes where the Local-part requires (or
   uses) the Quoted-string form or where the Local-part is case-
   sensitive. For any purposes that require generating or comparing
   Local-parts (e.g., to specific mailbox names), all quoted forms MUST
   be treated as equivalent, and the sending system SHOULD transmit the
   form that uses the minimum quoting possible.

   Systems MUST NOT define mailboxes in such a way as to require the use
   in SMTP of non-ASCII characters (octets with the high order bit set



Klensin Standards Track [Page 42]
 
RFC 5321 SMTP October 2008


   to one) or ASCII "control characters" (decimal value 0-31 and 127).
   These characters MUST NOT be used in MAIL or RCPT commands or other
   commands that require mailbox names.

   Note that the backslash, "\", is a quote character, which is used to
   indicate that the next character is to be used literally (instead of
   its normal interpretation). For example, "Joe\,Smith" indicates a
   single nine-character user name string with the comma being the
   fourth character of that string.

   To promote interoperability and consistent with long-standing
   guidance about conservative use of the DNS in naming and applications
   (e.g., see Section 2.3.1 of the base DNS document, RFC 1035 [2]),
   characters outside the set of alphabetic characters, digits, and
   hyphen MUST NOT appear in domain name labels for SMTP clients or
   servers. In particular, the underscore character is not permitted.
   SMTP servers that receive a command in which invalid character codes
   have been employed, and for which there are no other reasons for
   rejection, MUST reject that command with a 501 response (this rule,
   like others, could be overridden by appropriate SMTP extensions).

4.1.3. Address Literals

   Sometimes a host is not known to the domain name system and
   communication (and, in particular, communication to report and repair
   the error) is blocked. To bypass this barrier, a special literal
   form of the address is allowed as an alternative to a domain name.
   For IPv4 addresses, this form uses four small decimal integers
   separated by dots and enclosed by brackets such as [123.255.37.2],
   which indicates an (IPv4) Internet Address in sequence-of-octets
   form. For IPv6 and other forms of addressing that might eventually
   be standardized, the form consists of a standardized "tag" that
   identifies the address syntax, a colon, and the address itself, in a
   format specified as part of the relevant standards (i.e., RFC 4291
   [8] for IPv6).

   Specifically:

   IPv4-address-literal = Snum 3("." Snum)

   IPv6-address-literal = "IPv6:" IPv6-addr

   General-address-literal = Standardized-tag ":" 1*dcontent

   Standardized-tag = Ldh-str
                     ; Standardized-tag MUST be specified in a
                     ; Standards-Track RFC and registered with IANA




Klensin Standards Track [Page 43]
 
RFC 5321 SMTP October 2008


   dcontent = %d33-90 / ; Printable US-ASCII
                  %d94-126 ; excl. "[", "\", "]"

   Snum = 1*3DIGIT
                  ; representing a decimal integer
                  ; value in the range 0 through 255

   IPv6-addr = IPv6-full / IPv6-comp / IPv6v4-full / IPv6v4-comp

   IPv6-hex = 1*4HEXDIG

   IPv6-full = IPv6-hex 7(":" IPv6-hex)

   IPv6-comp = [IPv6-hex *5(":" IPv6-hex)] "::"
                  [IPv6-hex *5(":" IPv6-hex)]
                  ; The "::" represents at least 2 16-bit groups of
                  ; zeros. No more than 6 groups in addition to the
                  ; "::" may be present.

   IPv6v4-full = IPv6-hex 5(":" IPv6-hex) ":" IPv4-address-literal

   IPv6v4-comp = [IPv6-hex *3(":" IPv6-hex)] "::"
                  [IPv6-hex *3(":" IPv6-hex) ":"]
                  IPv4-address-literal
                  ; The "::" represents at least 2 16-bit groups of
                  ; zeros. No more than 4 groups in addition to the
                  ; "::" and IPv4-address-literal may be present.

4.1.4. Order of Commands

   There are restrictions on the order in which these commands may be
   used.

   A session that will contain mail transactions MUST first be
   initialized by the use of the EHLO command. An SMTP server SHOULD
   accept commands for non-mail transactions (e.g., VRFY or EXPN)
   without this initialization.

   An EHLO command MAY be issued by a client later in the session. If
   it is issued after the session begins and the EHLO command is
   acceptable to the SMTP server, the SMTP server MUST clear all buffers
   and reset the state exactly as if a RSET command had been issued. In
   other words, the sequence of RSET followed immediately by EHLO is
   redundant, but not harmful other than in the performance cost of
   executing unnecessary commands.

   If the EHLO command is not acceptable to the SMTP server, 501, 500,
   502, or 550 failure replies MUST be returned as appropriate. The



Klensin Standards Track [Page 44]
 
RFC 5321 SMTP October 2008


   SMTP server MUST stay in the same state after transmitting these
   replies that it was in before the EHLO was received.

   The SMTP client MUST, if possible, ensure that the domain parameter
   to the EHLO command is a primary host name as specified for this
   command in Section 2.3.5. If this is not possible (e.g., when the
   client's address is dynamically assigned and the client does not have
   an obvious name), an address literal SHOULD be substituted for the
   domain name.

   An SMTP server MAY verify that the domain name argument in the EHLO
   command actually corresponds to the IP address of the client.
   However, if the verification fails, the server MUST NOT refuse to
   accept a message on that basis. Information captured in the
   verification attempt is for logging and tracing purposes. Note that
   this prohibition applies to the matching of the parameter to its IP
   address only; see Section 7.9 for a more extensive discussion of
   rejecting incoming connections or mail messages.

   The NOOP, HELP, EXPN, VRFY, and RSET commands can be used at any time
   during a session, or without previously initializing a session. SMTP
   servers SHOULD process these normally (that is, not return a 503
   code) even if no EHLO command has yet been received; clients SHOULD
   open a session with EHLO before sending these commands.

   If these rules are followed, the example in RFC 821 that shows "550
   access denied to you" in response to an EXPN command is incorrect
   unless an EHLO command precedes the EXPN or the denial of access is
   based on the client's IP address or other authentication or
   authorization-determining mechanisms.

   The MAIL command (or the obsolete SEND, SOML, or SAML commands)
   begins a mail transaction. Once started, a mail transaction consists
   of a transaction beginning command, one or more RCPT commands, and a
   DATA command, in that order. A mail transaction may be aborted by
   the RSET, a new EHLO, or the QUIT command. There may be zero or more
   transactions in a session. MAIL (or SEND, SOML, or SAML) MUST NOT be
   sent if a mail transaction is already open, i.e., it should be sent
   only if no mail transaction had been started in the session, or if
   the previous one successfully concluded with a successful DATA
   command, or if the previous one was aborted, e.g., with a RSET or new
   EHLO.

   If the transaction beginning command argument is not acceptable, a
   501 failure reply MUST be returned and the SMTP server MUST stay in
   the same state. If the commands in a transaction are out of order to
   the degree that they cannot be processed by the server, a 503 failure




Klensin Standards Track [Page 45]
 
RFC 5321 SMTP October 2008


   reply MUST be returned and the SMTP server MUST stay in the same
   state.

   The last command in a session MUST be the QUIT command. The QUIT
   command SHOULD be used by the client SMTP to request connection
   closure, even when no session opening command was sent and accepted.

4.1.5. Private-Use Commands

   As specified in Section 2.2.2, commands starting in "X" may be used
   by bilateral agreement between the client (sending) and server
   (receiving) SMTP agents. An SMTP server that does not recognize such
   a command is expected to reply with "500 Command not recognized". An
   extended SMTP server MAY list the feature names associated with these
   private commands in the response to the EHLO command.

   Commands sent or accepted by SMTP systems that do not start with "X"
   MUST conform to the requirements of Section 2.2.2.

4.2. SMTP Replies

   Replies to SMTP commands serve to ensure the synchronization of
   requests and actions in the process of mail transfer and to guarantee
   that the SMTP client always knows the state of the SMTP server.
   Every command MUST generate exactly one reply.

   The details of the command-reply sequence are described in
   Section 4.3.

   An SMTP reply consists of a three digit number (transmitted as three
   numeric characters) followed by some text unless specified otherwise
   in this document. The number is for use by automata to determine
   what state to enter next; the text is for the human user. The three
   digits contain enough encoded information that the SMTP client need
   not examine the text and may either discard it or pass it on to the
   user, as appropriate. Exceptions are as noted elsewhere in this
   document. In particular, the 220, 221, 251, 421, and 551 reply codes
   are associated with message text that must be parsed and interpreted
   by machines. In the general case, the text may be receiver dependent
   and context dependent, so there are likely to be varying texts for
   each reply code. A discussion of the theory of reply codes is given
   in Section 4.2.1. Formally, a reply is defined to be the sequence: a
   three-digit code, <SP>, one line of text, and <CRLF>, or a multiline
   reply (as defined in the same section). Since, in violation of this
   specification, the text is sometimes not sent, clients that do not
   receive it SHOULD be prepared to process the code alone (with or
   without a trailing space character). Only the EHLO, EXPN, and HELP
   commands are expected to result in multiline replies in normal



Klensin Standards Track [Page 46]
 
RFC 5321 SMTP October 2008


   circumstances; however, multiline replies are allowed for any
   command.

   In ABNF, server responses are:

   Greeting = ( "220 " (Domain / address-literal)
                  [ SP textstring ] CRLF ) /
                  ( "220-" (Domain / address-literal)
                  [ SP textstring ] CRLF
                  *( "220-" [ textstring ] CRLF )
                  "220" [ SP textstring ] CRLF )

   textstring = 1*(%d09 / %d32-126) ; HT, SP, Printable US-ASCII

   Reply-line = *( Reply-code "-" [ textstring ] CRLF )
                  Reply-code [ SP textstring ] CRLF

   Reply-code = %x32-35 %x30-35 %x30-39

   where "Greeting" appears only in the 220 response that announces that
   the server is opening its part of the connection. (Other possible
   server responses upon connection follow the syntax of Reply-line.)

   An SMTP server SHOULD send only the reply codes listed in this
   document. An SMTP server SHOULD use the text shown in the examples
   whenever appropriate.

   An SMTP client MUST determine its actions only by the reply code, not
   by the text (except for the "change of address" 251 and 551 and, if
   necessary, 220, 221, and 421 replies); in the general case, any text,
   including no text at all (although senders SHOULD NOT send bare
   codes), MUST be acceptable. The space (blank) following the reply
   code is considered part of the text. Whenever possible, a receiver-
   SMTP SHOULD test the first digit (severity indication) of the reply
   code.

   The list of codes that appears below MUST NOT be construed as
   permanent. While the addition of new codes should be a rare and
   significant activity, with supplemental information in the textual
   part of the response being preferred, new codes may be added as the
   result of new Standards or Standards-Track specifications.
   Consequently, a sender-SMTP MUST be prepared to handle codes not
   specified in this document and MUST do so by interpreting the first
   digit only.

   In the absence of extensions negotiated with the client, SMTP servers
   MUST NOT send reply codes whose first digits are other than 2, 3, 4,




Klensin Standards Track [Page 47]
 
RFC 5321 SMTP October 2008


   or 5. Clients that receive such out-of-range codes SHOULD normally
   treat them as fatal errors and terminate the mail transaction.

4.2.1. Reply Code Severities and Theory

   The three digits of the reply each have a special significance. The
   first digit denotes whether the response is good, bad, or incomplete.
   An unsophisticated SMTP client, or one that receives an unexpected
   code, will be able to determine its next action (proceed as planned,
   redo, retrench, etc.) by examining this first digit. An SMTP client
   that wants to know approximately what kind of error occurred (e.g.,
   mail system error, command syntax error) may examine the second
   digit. The third digit and any supplemental information that may be
   present is reserved for the finest gradation of information.

   There are four values for the first digit of the reply code:

   2yz Positive Completion reply
      The requested action has been successfully completed. A new
      request may be initiated.

   3yz Positive Intermediate reply
      The command has been accepted, but the requested action is being
      held in abeyance, pending receipt of further information. The
      SMTP client should send another command specifying this
      information. This reply is used in command sequence groups (i.e.,
      in DATA).

   4yz Transient Negative Completion reply
      The command was not accepted, and the requested action did not
      occur. However, the error condition is temporary, and the action
      may be requested again. The sender should return to the beginning
      of the command sequence (if any). It is difficult to assign a
      meaning to "transient" when two different sites (receiver- and
      sender-SMTP agents) must agree on the interpretation. Each reply
      in this category might have a different time value, but the SMTP
      client SHOULD try again. A rule of thumb to determine whether a
      reply fits into the 4yz or the 5yz category (see below) is that
      replies are 4yz if they can be successful if repeated without any
      change in command form or in properties of the sender or receiver
      (that is, the command is repeated identically and the receiver
      does not put up a new implementation).

   5yz Permanent Negative Completion reply
      The command was not accepted and the requested action did not
      occur. The SMTP client SHOULD NOT repeat the exact request (in
      the same sequence). Even some "permanent" error conditions can be
      corrected, so the human user may want to direct the SMTP client to



Klensin Standards Track [Page 48]
 
RFC 5321 SMTP October 2008


      reinitiate the command sequence by direct action at some point in
      the future (e.g., after the spelling has been changed, or the user
      has altered the account status).

   It is worth noting that the file transfer protocol (FTP) [34] uses a
   very similar code architecture and that the SMTP codes are based on
   the FTP model. However, SMTP uses a one-command, one-response model
   (while FTP is asynchronous) and FTP's 1yz codes are not part of the
   SMTP model.

   The second digit encodes responses in specific categories:

   x0z Syntax: These replies refer to syntax errors, syntactically
      correct commands that do not fit any functional category, and
      unimplemented or superfluous commands.

   x1z Information: These are replies to requests for information, such
      as status or help.

   x2z Connections: These are replies referring to the transmission
      channel.

   x3z Unspecified.

   x4z Unspecified.

   x5z Mail system: These replies indicate the status of the receiver
      mail system vis-a-vis the requested transfer or other mail system
      action.

   The third digit gives a finer gradation of meaning in each category
   specified by the second digit. The list of replies illustrates this.
   Each reply text is recommended rather than mandatory, and may even
   change according to the command with which it is associated. On the
   other hand, the reply codes must strictly follow the specifications
   in this section. Receiver implementations should not invent new
   codes for slightly different situations from the ones described here,
   but rather adapt codes already defined.

   For example, a command such as NOOP, whose successful execution does
   not offer the SMTP client any new information, will return a 250
   reply. The reply is 502 when the command requests an unimplemented
   non-site-specific action. A refinement of that is the 504 reply for
   a command that is implemented, but that requests an unimplemented
   parameter.






Klensin Standards Track [Page 49]
 
RFC 5321 SMTP October 2008


   The reply text may be longer than a single line; in these cases the
   complete text must be marked so the SMTP client knows when it can
   stop reading the reply. This requires a special format to indicate a
   multiple line reply.

   The format for multiline replies requires that every line, except the
   last, begin with the reply code, followed immediately by a hyphen,
   "-" (also known as minus), followed by text. The last line will
   begin with the reply code, followed immediately by <SP>, optionally
   some text, and <CRLF>. As noted above, servers SHOULD send the <SP>
   if subsequent text is not sent, but clients MUST be prepared for it
   to be omitted.

   For example:

      250-First line
      250-Second line
      250-234 Text beginning with numbers
      250 The last line

   In a multiline reply, the reply code on each of the lines MUST be the
   same. It is reasonable for the client to rely on this, so it can
   make processing decisions based on the code in any line, assuming
   that all others will be the same. In a few cases, there is important
   data for the client in the reply "text". The client will be able to
   identify these cases from the current context.

4.2.2. Reply Codes by Function Groups

   500 Syntax error, command unrecognized (This may include errors such
      as command line too long)

   501 Syntax error in parameters or arguments

   502 Command not implemented (see Section 4.2.4)

   503 Bad sequence of commands

   504 Command parameter not implemented


   211 System status, or system help reply

   214 Help message (Information on how to use the receiver or the
      meaning of a particular non-standard command; this reply is useful
      only to the human user)





Klensin Standards Track [Page 50]
 
RFC 5321 SMTP October 2008


   220 <domain> Service ready

   221 <domain> Service closing transmission channel

   421 <domain> Service not available, closing transmission channel
      (This may be a reply to any command if the service knows it must
      shut down)


   250 Requested mail action okay, completed

   251 User not local; will forward to <forward-path> (See Section 3.4)

   252 Cannot VRFY user, but will accept message and attempt delivery
      (See Section 3.5.3)

   455 Server unable to accommodate parameters

   555 MAIL FROM/RCPT TO parameters not recognized or not implemented

   450 Requested mail action not taken: mailbox unavailable (e.g.,
      mailbox busy or temporarily blocked for policy reasons)

   550 Requested action not taken: mailbox unavailable (e.g., mailbox
      not found, no access, or command rejected for policy reasons)

   451 Requested action aborted: error in processing

   551 User not local; please try <forward-path> (See Section 3.4)

   452 Requested action not taken: insufficient system storage

   552 Requested mail action aborted: exceeded storage allocation

   553 Requested action not taken: mailbox name not allowed (e.g.,
      mailbox syntax incorrect)

   354 Start mail input; end with <CRLF>.<CRLF>

   554 Transaction failed (Or, in the case of a connection-opening
      response, "No SMTP service here")










Klensin Standards Track [Page 51]
 
RFC 5321 SMTP October 2008


4.2.3. Reply Codes in Numeric Order

   211 System status, or system help reply

   214 Help message (Information on how to use the receiver or the
      meaning of a particular non-standard command; this reply is useful
      only to the human user)

   220 <domain> Service ready

   221 <domain> Service closing transmission channel

   250 Requested mail action okay, completed

   251 User not local; will forward to <forward-path> (See Section 3.4)

   252 Cannot VRFY user, but will accept message and attempt delivery
      (See Section 3.5.3)

   354 Start mail input; end with <CRLF>.<CRLF>

   421 <domain> Service not available, closing transmission channel
      (This may be a reply to any command if the service knows it must
      shut down)

   450 Requested mail action not taken: mailbox unavailable (e.g.,
      mailbox busy or temporarily blocked for policy reasons)

   451 Requested action aborted: local error in processing

   452 Requested action not taken: insufficient system storage

   455 Server unable to accommodate parameters

   500 Syntax error, command unrecognized (This may include errors such
      as command line too long)

   501 Syntax error in parameters or arguments

   502 Command not implemented (see Section 4.2.4)

   503 Bad sequence of commands

   504 Command parameter not implemented

   550 Requested action not taken: mailbox unavailable (e.g., mailbox
      not found, no access, or command rejected for policy reasons)




Klensin Standards Track [Page 52]
 
RFC 5321 SMTP October 2008


   551 User not local; please try <forward-path> (See Section 3.4)

   552 Requested mail action aborted: exceeded storage allocation

   553 Requested action not taken: mailbox name not allowed (e.g.,
      mailbox syntax incorrect)

   554 Transaction failed (Or, in the case of a connection-opening
      response, "No SMTP service here")

   555 MAIL FROM/RCPT TO parameters not recognized or not implemented

4.2.4. Reply Code 502

   Questions have been raised as to when reply code 502 (Command not
   implemented) SHOULD be returned in preference to other codes. 502
   SHOULD be used when the command is actually recognized by the SMTP
   server, but not implemented. If the command is not recognized, code
   500 SHOULD be returned. Extended SMTP systems MUST NOT list
   capabilities in response to EHLO for which they will return 502 (or
   500) replies.

4.2.5. Reply Codes after DATA and the Subsequent <CRLF>.<CRLF>

   When an SMTP server returns a positive completion status (2yz code)
   after the DATA command is completed with <CRLF>.<CRLF>, it accepts
   responsibility for:

   o delivering the message (if the recipient mailbox exists), or

   o if attempts to deliver the message fail due to transient
      conditions, retrying delivery some reasonable number of times at
      intervals as specified in Section 4.5.4.

   o if attempts to deliver the message fail due to permanent
      conditions, or if repeated attempts to deliver the message fail
      due to transient conditions, returning appropriate notification to
      the sender of the original message (using the address in the SMTP
      MAIL command).

   When an SMTP server returns a temporary error status (4yz) code after
   the DATA command is completed with <CRLF>.<CRLF>, it MUST NOT make a
   subsequent attempt to deliver that message. The SMTP client retains
   responsibility for the delivery of that message and may either return
   it to the user or requeue it for a subsequent attempt (see
   Section 4.5.4.1).





Klensin Standards Track [Page 53]
 
RFC 5321 SMTP October 2008


   The user who originated the message SHOULD be able to interpret the
   return of a transient failure status (by mail message or otherwise)
   as a non-delivery indication, just as a permanent failure would be
   interpreted. If the client SMTP successfully handles these
   conditions, the user will not receive such a reply.

   When an SMTP server returns a permanent error status (5yz) code after
   the DATA command is completed with <CRLF>.<CRLF>, it MUST NOT make
   any subsequent attempt to deliver the message. As with temporary
   error status codes, the SMTP client retains responsibility for the
   message, but SHOULD not again attempt delivery to the same server
   without user review of the message and response and appropriate
   intervention.

4.3. Sequencing of Commands and Replies

4.3.1. Sequencing Overview

   The communication between the sender and receiver is an alternating
   dialogue, controlled by the sender. As such, the sender issues a
   command and the receiver responds with a reply. Unless other
   arrangements are negotiated through service extensions, the sender
   MUST wait for this response before sending further commands. One
   important reply is the connection greeting. Normally, a receiver
   will send a 220 "Service ready" reply when the connection is
   completed. The sender SHOULD wait for this greeting message before
   sending any commands.

   Note: all the greeting-type replies have the official name (the
   fully-qualified primary domain name) of the server host as the first
   word following the reply code. Sometimes the host will have no
   meaningful name. See Section 4.1.3 for a discussion of alternatives
   in these situations.

   For example,

      220 ISIF.USC.EDU Service ready

   or

      220 mail.example.com SuperSMTP v 6.1.2 Service ready

   or

      220 [10.0.0.1] Clueless host service ready

   The table below lists alternative success and failure replies for
   each command. These SHOULD be strictly adhered to. A receiver MAY



Klensin Standards Track [Page 54]
 
RFC 5321 SMTP October 2008


   substitute text in the replies, but the meanings and actions implied
   by the code numbers and by the specific command reply sequence MUST
   be preserved.

4.3.2. Command-Reply Sequences

   Each command is listed with its usual possible replies. The prefixes
   used before the possible replies are "I" for intermediate, "S" for
   success, and "E" for error. Since some servers may generate other
   replies under special circumstances, and to allow for future
   extension, SMTP clients SHOULD, when possible, interpret only the
   first digit of the reply and MUST be prepared to deal with
   unrecognized reply codes by interpreting the first digit only.
   Unless extended using the mechanisms described in Section 2.2, SMTP
   servers MUST NOT transmit reply codes to an SMTP client that are
   other than three digits or that do not start in a digit between 2 and
   5 inclusive.

   These sequencing rules and, in principle, the codes themselves, can
   be extended or modified by SMTP extensions offered by the server and
   accepted (requested) by the client. However, if the target is more
   precise granularity in the codes, rather than codes for completely
   new purposes, the system described in RFC 3463 [25] SHOULD be used in
   preference to the invention of new codes.

   In addition to the codes listed below, any SMTP command can return
   any of the following codes if the corresponding unusual circumstances
   are encountered:

   500 For the "command line too long" case or if the command name was
      not recognized. Note that producing a "command not recognized"
      error in response to the required subset of these commands is a
      violation of this specification. Similarly, producing a "command
      too long" message for a command line shorter than 512 characters
      would violate the provisions of Section 4.5.3.1.4.

   501 Syntax error in command or arguments. In order to provide for
      future extensions, commands that are specified in this document as
      not accepting arguments (DATA, RSET, QUIT) SHOULD return a 501
      message if arguments are supplied in the absence of EHLO-
      advertised extensions.

   421 Service shutting down and closing transmission channel








Klensin Standards Track [Page 55]
 
RFC 5321 SMTP October 2008


   Specific sequences are:

      CONNECTION ESTABLISHMENT

         S: 220
         E: 554

      EHLO or HELO

         S: 250
         E: 504 (a conforming implementation could return this code only
         in fairly obscure cases), 550, 502 (permitted only with an old-
         style server that does not support EHLO)

      MAIL

         S: 250
         E: 552, 451, 452, 550, 553, 503, 455, 555

      RCPT

         S: 250, 251 (but see Section 3.4 for discussion of 251 and 551)
         E: 550, 551, 552, 553, 450, 451, 452, 503, 455, 555

      DATA

         I: 354 -> data -> S: 250

                           E: 552, 554, 451, 452

                           E: 450, 550 (rejections for policy reasons)

         E: 503, 554

      RSET

         S: 250

      VRFY

         S: 250, 251, 252
         E: 550, 551, 553, 502, 504

      EXPN

         S: 250, 252
         E: 550, 500, 502, 504




Klensin Standards Track [Page 56]
 
RFC 5321 SMTP October 2008


      HELP

         S: 211, 214
         E: 502, 504

      NOOP

         S: 250

      QUIT

         S: 221

4.4. Trace Information

   When an SMTP server receives a message for delivery or further
   processing, it MUST insert trace ("time stamp" or "Received")
   information at the beginning of the message content, as discussed in
   Section 4.1.1.4.

   This line MUST be structured as follows:

   o The FROM clause, which MUST be supplied in an SMTP environment,
      SHOULD contain both (1) the name of the source host as presented
      in the EHLO command and (2) an address literal containing the IP
      address of the source, determined from the TCP connection.

   o The ID clause MAY contain an "@" as suggested in RFC 822, but this
      is not required.

   o If the FOR clause appears, it MUST contain exactly one <path>
      entry, even when multiple RCPT commands have been given. Multiple
      <path>s raise some security issues and have been deprecated, see
      Section 7.2.

   An Internet mail program MUST NOT change or delete a Received: line
   that was previously added to the message header section. SMTP
   servers MUST prepend Received lines to messages; they MUST NOT change
   the order of existing lines or insert Received lines in any other
   location.

   As the Internet grows, comparability of Received header fields is
   important for detecting problems, especially slow relays. SMTP
   servers that create Received header fields SHOULD use explicit
   offsets in the dates (e.g., -0800), rather than time zone names of
   any type. Local time (with an offset) SHOULD be used rather than UT
   when feasible. This formulation allows slightly more information
   about local circumstances to be specified. If UT is needed, the



Klensin Standards Track [Page 57]
 
RFC 5321 SMTP October 2008


   receiver need merely do some simple arithmetic to convert the values.
   Use of UT loses information about the time zone-location of the
   server. If it is desired to supply a time zone name, it SHOULD be
   included in a comment.

   When the delivery SMTP server makes the "final delivery" of a
   message, it inserts a return-path line at the beginning of the mail
   data. This use of return-path is required; mail systems MUST support
   it. The return-path line preserves the information in the <reverse-
   path> from the MAIL command. Here, final delivery means the message
   has left the SMTP environment. Normally, this would mean it had been
   delivered to the destination user or an associated mail drop, but in
   some cases it may be further processed and transmitted by another
   mail system.

   It is possible for the mailbox in the return path to be different
   from the actual sender's mailbox, for example, if error responses are
   to be delivered to a special error handling mailbox rather than to
   the message sender. When mailing lists are involved, this
   arrangement is common and useful as a means of directing errors to
   the list maintainer rather than the message originator.

   The text above implies that the final mail data will begin with a
   return path line, followed by one or more time stamp lines. These
   lines will be followed by the rest of the mail data: first the
   balance of the mail header section and then the body (RFC 5322 [4]).

   It is sometimes difficult for an SMTP server to determine whether or
   not it is making final delivery since forwarding or other operations
   may occur after the message is accepted for delivery. Consequently,
   any further (forwarding, gateway, or relay) systems MAY remove the
   return path and rebuild the MAIL command as needed to ensure that
   exactly one such line appears in a delivered message.

   A message-originating SMTP system SHOULD NOT send a message that
   already contains a Return-path header field. SMTP servers performing
   a relay function MUST NOT inspect the message data, and especially
   not to the extent needed to determine if Return-path header fields
   are present. SMTP servers making final delivery MAY remove Return-
   path header fields before adding their own.

   The primary purpose of the Return-path is to designate the address to
   which messages indicating non-delivery or other mail system failures
   are to be sent. For this to be unambiguous, exactly one return path
   SHOULD be present when the message is delivered. Systems using RFC
   822 syntax with non-SMTP transports SHOULD designate an unambiguous
   address, associated with the transport envelope, to which error
   reports (e.g., non-delivery messages) should be sent.



Klensin Standards Track [Page 58]
 
RFC 5321 SMTP October 2008


   Historical note: Text in RFC 822 that appears to contradict the use
   of the Return-path header field (or the envelope reverse-path address
   from the MAIL command) as the destination for error messages is not
   applicable on the Internet. The reverse-path address (as copied into
   the Return-path) MUST be used as the target of any mail containing
   delivery error messages.

   In particular:
   o a gateway from SMTP -> elsewhere SHOULD insert a return-path
      header field, unless it is known that the "elsewhere" transport
      also uses Internet domain addresses and maintains the envelope
      sender address separately.

   o a gateway from elsewhere -> SMTP SHOULD delete any return-path
      header field present in the message, and either copy that
      information to the SMTP envelope or combine it with information
      present in the envelope of the other transport system to construct
      the reverse-path argument to the MAIL command in the SMTP
      envelope.

   The server must give special treatment to cases in which the
   processing following the end of mail data indication is only
   partially successful. This could happen if, after accepting several
   recipients and the mail data, the SMTP server finds that the mail
   data could be successfully delivered to some, but not all, of the
   recipients. In such cases, the response to the DATA command MUST be
   an OK reply. However, the SMTP server MUST compose and send an
   "undeliverable mail" notification message to the originator of the
   message.

   A single notification listing all of the failed recipients or
   separate notification messages MUST be sent for each failed
   recipient. For economy of processing by the sender, the former
   SHOULD be used when possible. Note that the key difference between
   handling aliases (Section 3.9.1) and forwarding (this subsection) is
   the change to the backward-pointing address in this case. All
   notification messages about undeliverable mail MUST be sent using the
   MAIL command (even if they result from processing the obsolete SEND,
   SOML, or SAML commands) and MUST use a null return path as discussed
   in Section 3.6.

   The time stamp line and the return path line are formally defined as
   follows (the definitions for "FWS" and "CFWS" appear in RFC 5322
   [4]):

   Return-path-line = "Return-Path:" FWS Reverse-path <CRLF>

   Time-stamp-line = "Received:" FWS Stamp <CRLF>



Klensin Standards Track [Page 59]
 
RFC 5321 SMTP October 2008


   Stamp = From-domain By-domain Opt-info [CFWS] ";"
                  FWS date-time
                  ; where "date-time" is as defined in RFC 5322 [4]
                  ; but the "obs-" forms, especially two-digit
                  ; years, are prohibited in SMTP and MUST NOT be used.

   From-domain = "FROM" FWS Extended-Domain

   By-domain = CFWS "BY" FWS Extended-Domain

   Extended-Domain = Domain /
                    ( Domain FWS "(" TCP-info ")" ) /
                    ( address-literal FWS "(" TCP-info ")" )

   TCP-info = address-literal / ( Domain FWS address-literal )
                  ; Information derived by server from TCP connection
                  ; not client EHLO.

   Opt-info = [Via] [With] [ID] [For]
                  [Additional-Registered-Clauses]

   Via = CFWS "VIA" FWS Link

   With = CFWS "WITH" FWS Protocol

   ID = CFWS "ID" FWS ( Atom / msg-id )
                  ; msg-id is defined in RFC 5322 [4]

   For = CFWS "FOR" FWS ( Path / Mailbox )

   Additional-Registered-Clauses = CFWS Atom FWS String
                                  ; Additional standard clauses may be
                                  added in this
                                  ; location by future standards and
                                  registration with
                                  ; IANA. SMTP servers SHOULD NOT use
                                  unregistered
                                  ; names. See Section 8.

   Link = "TCP" / Addtl-Link

   Addtl-Link = Atom
                  ; Additional standard names for links are
                  ; registered with the Internet Assigned Numbers
                  ; Authority (IANA). "Via" is primarily of value
                  ; with non-Internet transports. SMTP servers
                  ; SHOULD NOT use unregistered names.




Klensin Standards Track [Page 60]
 
RFC 5321 SMTP October 2008


   Protocol = "ESMTP" / "SMTP" / Attdl-Protocol

   Attdl-Protocol = Atom
                  ; Additional standard names for protocols are
                  ; registered with the Internet Assigned Numbers
                  ; Authority (IANA) in the "mail parameters"
                  ; registry [9]. SMTP servers SHOULD NOT
                  ; use unregistered names.

4.5. Additional Implementation Issues

4.5.1. Minimum Implementation

   In order to make SMTP workable, the following minimum implementation
   MUST be provided by all receivers. The following commands MUST be
   supported to conform to this specification:

      EHLO
      HELO
      MAIL
      RCPT
      DATA
      RSET
      NOOP
      QUIT
      VRFY

   Any system that includes an SMTP server supporting mail relaying or
   delivery MUST support the reserved mailbox "postmaster" as a case-
   insensitive local name. This postmaster address is not strictly
   necessary if the server always returns 554 on connection opening (as
   described in Section 3.1). The requirement to accept mail for
   postmaster implies that RCPT commands that specify a mailbox for
   postmaster at any of the domains for which the SMTP server provides
   mail service, as well as the special case of "RCPT TO:<Postmaster>"
   (with no domain specification), MUST be supported.

   SMTP systems are expected to make every reasonable effort to accept
   mail directed to Postmaster from any other system on the Internet.
   In extreme cases -- such as to contain a denial of service attack or
   other breach of security -- an SMTP server may block mail directed to
   Postmaster. However, such arrangements SHOULD be narrowly tailored
   so as to avoid blocking messages that are not part of such attacks.








Klensin Standards Track [Page 61]
 
RFC 5321 SMTP October 2008


4.5.2. Transparency

   Without some provision for data transparency, the character sequence
   "<CRLF>.<CRLF>" ends the mail text and cannot be sent by the user.
   In general, users are not aware of such "forbidden" sequences. To
   allow all user composed text to be transmitted transparently, the
   following procedures are used:

   o Before sending a line of mail text, the SMTP client checks the
      first character of the line. If it is a period, one additional
      period is inserted at the beginning of the line.

   o When a line of mail text is received by the SMTP server, it checks
      the line. If the line is composed of a single period, it is
      treated as the end of mail indicator. If the first character is a
      period and there are other characters on the line, the first
      character is deleted.

   The mail data may contain any of the 128 ASCII characters. All
   characters are to be delivered to the recipient's mailbox, including
   spaces, vertical and horizontal tabs, and other control characters.
   If the transmission channel provides an 8-bit byte (octet) data
   stream, the 7-bit ASCII codes are transmitted, right justified, in
   the octets, with the high-order bits cleared to zero. See
   Section 3.6 for special treatment of these conditions in SMTP systems
   serving a relay function.

   In some systems, it may be necessary to transform the data as it is
   received and stored. This may be necessary for hosts that use a
   different character set than ASCII as their local character set, that
   store data in records rather than strings, or which use special
   character sequences as delimiters inside mailboxes. If such
   transformations are necessary, they MUST be reversible, especially if
   they are applied to mail being relayed.

4.5.3. Sizes and Timeouts

4.5.3.1. Size Limits and Minimums

   There are several objects that have required minimum/maximum sizes.
   Every implementation MUST be able to receive objects of at least
   these sizes. Objects larger than these sizes SHOULD be avoided when
   possible. However, some Internet mail constructs such as encoded
   X.400 addresses (RFC 2156 [35]) will often require larger objects.
   Clients MAY attempt to transmit these, but MUST be prepared for a
   server to reject them if they cannot be handled by it. To the
   maximum extent possible, implementation techniques that impose no
   limits on the length of these objects should be used.



Klensin Standards Track [Page 62]
 
RFC 5321 SMTP October 2008


   Extensions to SMTP may involve the use of characters that occupy more
   than a single octet each. This section therefore specifies lengths
   in octets where absolute lengths, rather than character counts, are
   intended.

4.5.3.1.1. Local-part

   The maximum total length of a user name or other local-part is 64
   octets.

4.5.3.1.2. Domain

   The maximum total length of a domain name or number is 255 octets.

4.5.3.1.3. Path

   The maximum total length of a reverse-path or forward-path is 256
   octets (including the punctuation and element separators).

4.5.3.1.4. Command Line

   The maximum total length of a command line including the command word
   and the <CRLF> is 512 octets. SMTP extensions may be used to
   increase this limit.

4.5.3.1.5. Reply Line

   The maximum total length of a reply line including the reply code and
   the <CRLF> is 512 octets. More information may be conveyed through
   multiple-line replies.

4.5.3.1.6. Text Line

   The maximum total length of a text line including the <CRLF> is 1000
   octets (not counting the leading dot duplicated for transparency).
   This number may be increased by the use of SMTP Service Extensions.

4.5.3.1.7. Message Content

   The maximum total length of a message content (including any message
   header section as well as the message body) MUST BE at least 64K
   octets. Since the introduction of Internet Standards for multimedia
   mail (RFC 2045 [21]), message lengths on the Internet have grown
   dramatically, and message size restrictions should be avoided if at
   all possible. SMTP server systems that must impose restrictions
   SHOULD implement the "SIZE" service extension of RFC 1870 [10], and
   SMTP client systems that will send large messages SHOULD utilize it
   when possible.



Klensin Standards Track [Page 63]
 
RFC 5321 SMTP October 2008


4.5.3.1.8. Recipients Buffer

   The minimum total number of recipients that MUST be buffered is 100
   recipients. Rejection of messages (for excessive recipients) with
   fewer than 100 RCPT commands is a violation of this specification.
   The general principle that relaying SMTP server MUST NOT, and
   delivery SMTP servers SHOULD NOT, perform validation tests on message
   header fields suggests that messages SHOULD NOT be rejected based on
   the total number of recipients shown in header fields. A server that
   imposes a limit on the number of recipients MUST behave in an orderly
   fashion, such as rejecting additional addresses over its limit rather
   than silently discarding addresses previously accepted. A client
   that needs to deliver a message containing over 100 RCPT commands
   SHOULD be prepared to transmit in 100-recipient "chunks" if the
   server declines to accept more than 100 recipients in a single
   message.

4.5.3.1.9. Treatment When Limits Exceeded

   Errors due to exceeding these limits may be reported by using the
   reply codes. Some examples of reply codes are:

      500 Line too long.

   or

      501 Path too long

   or

      452 Too many recipients (see below)

   or

      552 Too much mail data.

4.5.3.1.10. Too Many Recipients Code

   RFC 821 [1] incorrectly listed the error where an SMTP server
   exhausts its implementation limit on the number of RCPT commands
   ("too many recipients") as having reply code 552. The correct reply
   code for this condition is 452. Clients SHOULD treat a 552 code in
   this case as a temporary, rather than permanent, failure so the logic
   below works.

   When a conforming SMTP server encounters this condition, it has at
   least 100 successful RCPT commands in its recipients buffer. If the
   server is able to accept the message, then at least these 100



Klensin Standards Track [Page 64]
 
RFC 5321 SMTP October 2008


   addresses will be removed from the SMTP client's queue. When the
   client attempts retransmission of those addresses that received 452
   responses, at least 100 of these will be able to fit in the SMTP
   server's recipients buffer. Each retransmission attempt that is able
   to deliver anything will be able to dispose of at least 100 of these
   recipients.

   If an SMTP server has an implementation limit on the number of RCPT
   commands and this limit is exhausted, it MUST use a response code of
   452 (but the client SHOULD also be prepared for a 552, as noted
   above). If the server has a configured site-policy limitation on the
   number of RCPT commands, it MAY instead use a 5yz response code. In
   particular, if the intent is to prohibit messages with more than a
   site-specified number of recipients, rather than merely limit the
   number of recipients in a given mail transaction, it would be
   reasonable to return a 503 response to any DATA command received
   subsequent to the 452 (or 552) code or to simply return the 503 after
   DATA without returning any previous negative response.

4.5.3.2. Timeouts

   An SMTP client MUST provide a timeout mechanism. It MUST use per-
   command timeouts rather than somehow trying to time the entire mail
   transaction. Timeouts SHOULD be easily reconfigurable, preferably
   without recompiling the SMTP code. To implement this, a timer is set
   for each SMTP command and for each buffer of the data transfer. The
   latter means that the overall timeout is inherently proportional to
   the size of the message.

   Based on extensive experience with busy mail-relay hosts, the minimum
   per-command timeout values SHOULD be as follows:

4.5.3.2.1. Initial 220 Message: 5 Minutes

   An SMTP client process needs to distinguish between a failed TCP
   connection and a delay in receiving the initial 220 greeting message.
   Many SMTP servers accept a TCP connection but delay delivery of the
   220 message until their system load permits more mail to be
   processed.

4.5.3.2.2. MAIL Command: 5 Minutes

4.5.3.2.3. RCPT Command: 5 Minutes

   A longer timeout is required if processing of mailing lists and
   aliases is not deferred until after the message was accepted.





Klensin Standards Track [Page 65]
 
RFC 5321 SMTP October 2008


4.5.3.2.4. DATA Initiation: 2 Minutes

   This is while awaiting the "354 Start Input" reply to a DATA command.

4.5.3.2.5. Data Block: 3 Minutes

   This is while awaiting the completion of each TCP SEND call
   transmitting a chunk of data.

4.5.3.2.6. DATA Termination: 10 Minutes.

   This is while awaiting the "250 OK" reply. When the receiver gets
   the final period terminating the message data, it typically performs
   processing to deliver the message to a user mailbox. A spurious
   timeout at this point would be very wasteful and would typically
   result in delivery of multiple copies of the message, since it has
   been successfully sent and the server has accepted responsibility for
   delivery. See Section 6.1 for additional discussion.

4.5.3.2.7. Server Timeout: 5 Minutes.

   An SMTP server SHOULD have a timeout of at least 5 minutes while it
   is awaiting the next command from the sender.

4.5.4. Retry Strategies

   The common structure of a host SMTP implementation includes user
   mailboxes, one or more areas for queuing messages in transit, and one
   or more daemon processes for sending and receiving mail. The exact
   structure will vary depending on the needs of the users on the host
   and the number and size of mailing lists supported by the host. We
   describe several optimizations that have proved helpful, particularly
   for mailers supporting high traffic levels.

   Any queuing strategy MUST include timeouts on all activities on a
   per-command basis. A queuing strategy MUST NOT send error messages
   in response to error messages under any circumstances.

4.5.4.1. Sending Strategy

   The general model for an SMTP client is one or more processes that
   periodically attempt to transmit outgoing mail. In a typical system,
   the program that composes a message has some method for requesting
   immediate attention for a new piece of outgoing mail, while mail that
   cannot be transmitted immediately MUST be queued and periodically
   retried by the sender. A mail queue entry will include not only the
   message itself but also the envelope information.




Klensin Standards Track [Page 66]
 
RFC 5321 SMTP October 2008


   The sender MUST delay retrying a particular destination after one
   attempt has failed. In general, the retry interval SHOULD be at
   least 30 minutes; however, more sophisticated and variable strategies
   will be beneficial when the SMTP client can determine the reason for
   non-delivery.

   Retries continue until the message is transmitted or the sender gives
   up; the give-up time generally needs to be at least 4-5 days. It MAY
   be appropriate to set a shorter maximum number of retries for non-
   delivery notifications and equivalent error messages than for
   standard messages. The parameters to the retry algorithm MUST be
   configurable.

   A client SHOULD keep a list of hosts it cannot reach and
   corresponding connection timeouts, rather than just retrying queued
   mail items.

   Experience suggests that failures are typically transient (the target
   system or its connection has crashed), favoring a policy of two
   connection attempts in the first hour the message is in the queue,
   and then backing off to one every two or three hours.

   The SMTP client can shorten the queuing delay in cooperation with the
   SMTP server. For example, if mail is received from a particular
   address, it is likely that mail queued for that host can now be sent.
   Application of this principle may, in many cases, eliminate the
   requirement for an explicit "send queues now" function such as ETRN,
   RFC 1985 [36].

   The strategy may be further modified as a result of multiple
   addresses per host (see below) to optimize delivery time versus
   resource usage.

   An SMTP client may have a large queue of messages for each
   unavailable destination host. If all of these messages were retried
   in every retry cycle, there would be excessive Internet overhead and
   the sending system would be blocked for a long period. Note that an
   SMTP client can generally determine that a delivery attempt has
   failed only after a timeout of several minutes, and even a one-minute
   timeout per connection will result in a very large delay if retries
   are repeated for dozens, or even hundreds, of queued messages to the
   same host.

   At the same time, SMTP clients SHOULD use great care in caching
   negative responses from servers. In an extreme case, if EHLO is
   issued multiple times during the same SMTP connection, different
   answers may be returned by the server. More significantly, 5yz
   responses to the MAIL command MUST NOT be cached.



Klensin Standards Track [Page 67]
 
RFC 5321 SMTP October 2008


   When a mail message is to be delivered to multiple recipients, and
   the SMTP server to which a copy of the message is to be sent is the
   same for multiple recipients, then only one copy of the message
   SHOULD be transmitted. That is, the SMTP client SHOULD use the
   command sequence: MAIL, RCPT, RCPT, ..., RCPT, DATA instead of the
   sequence: MAIL, RCPT, DATA, ..., MAIL, RCPT, DATA. However, if there
   are very many addresses, a limit on the number of RCPT commands per
   MAIL command MAY be imposed. This efficiency feature SHOULD be
   implemented.

   Similarly, to achieve timely delivery, the SMTP client MAY support
   multiple concurrent outgoing mail transactions. However, some limit
   may be appropriate to protect the host from devoting all its
   resources to mail.

4.5.4.2. Receiving Strategy

   The SMTP server SHOULD attempt to keep a pending listen on the SMTP
   port (specified by IANA as port 25) at all times. This requires the
   support of multiple incoming TCP connections for SMTP. Some limit
   MAY be imposed, but servers that cannot handle more than one SMTP
   transaction at a time are not in conformance with the intent of this
   specification.

   As discussed above, when the SMTP server receives mail from a
   particular host address, it could activate its own SMTP queuing
   mechanisms to retry any mail pending for that host address.

4.5.5. Messages with a Null Reverse-Path

   There are several types of notification messages that are required by
   existing and proposed Standards to be sent with a null reverse-path,
   namely non-delivery notifications as discussed in Section 3.7, other
   kinds of Delivery Status Notifications (DSNs, RFC 3461 [32]), and
   Message Disposition Notifications (MDNs, RFC 3798 [37]). All of
   these kinds of messages are notifications about a previous message,
   and they are sent to the reverse-path of the previous mail message.
   (If the delivery of such a notification message fails, that usually
   indicates a problem with the mail system of the host to which the
   notification message is addressed. For this reason, at some hosts
   the MTA is set up to forward such failed notification messages to
   someone who is able to fix problems with the mail system, e.g., via
   the postmaster alias.)

   All other types of messages (i.e., any message which is not required
   by a Standards-Track RFC to have a null reverse-path) SHOULD be sent
   with a valid, non-null reverse-path.




Klensin Standards Track [Page 68]
 
RFC 5321 SMTP October 2008


   Implementers of automated email processors should be careful to make
   sure that the various kinds of messages with a null reverse-path are
   handled correctly. In particular, such systems SHOULD NOT reply to
   messages with a null reverse-path, and they SHOULD NOT add a non-null
   reverse-path, or change a null reverse-path to a non-null one, to
   such messages when forwarding.

5. Address Resolution and Mail Handling

5.1. Locating the Target Host

   Once an SMTP client lexically identifies a domain to which mail will
   be delivered for processing (as described in Sections 2.3.5 and 3.6),
   a DNS lookup MUST be performed to resolve the domain name (RFC 1035
   [2]). The names are expected to be fully-qualified domain names
   (FQDNs): mechanisms for inferring FQDNs from partial names or local
   aliases are outside of this specification. Due to a history of
   problems, SMTP servers used for initial submission of messages SHOULD
   NOT make such inferences (Message Submission Servers [18] have
   somewhat more flexibility) and intermediate (relay) SMTP servers MUST
   NOT make them.

   The lookup first attempts to locate an MX record associated with the
   name. If a CNAME record is found, the resulting name is processed as
   if it were the initial name. If a non-existent domain error is
   returned, this situation MUST be reported as an error. If a
   temporary error is returned, the message MUST be queued and retried
   later (see Section 4.5.4.1). If an empty list of MXs is returned,
   the address is treated as if it was associated with an implicit MX
   RR, with a preference of 0, pointing to that host. If MX records are
   present, but none of them are usable, or the implicit MX is unusable,
   this situation MUST be reported as an error.

   If one or more MX RRs are found for a given name, SMTP systems MUST
   NOT utilize any address RRs associated with that name unless they are
   located using the MX RRs; the "implicit MX" rule above applies only
   if there are no MX records present. If MX records are present, but
   none of them are usable, this situation MUST be reported as an error.

   When a domain name associated with an MX RR is looked up and the
   associated data field obtained, the data field of that response MUST
   contain a domain name. That domain name, when queried, MUST return
   at least one address record (e.g., A or AAAA RR) that gives the IP
   address of the SMTP server to which the message should be directed.
   Any other response, specifically including a value that will return a
   CNAME record when queried, lies outside the scope of this Standard.
   The prohibition on labels in the data that resolve to CNAMEs is
   discussed in more detail in RFC 2181, Section 10.3 [38].



Klensin Standards Track [Page 69]
 
RFC 5321 SMTP October 2008


   When the lookup succeeds, the mapping can result in a list of
   alternative delivery addresses rather than a single address, because
   of multiple MX records, multihoming, or both. To provide reliable
   mail transmission, the SMTP client MUST be able to try (and retry)
   each of the relevant addresses in this list in order, until a
   delivery attempt succeeds. However, there MAY also be a configurable
   limit on the number of alternate addresses that can be tried. In any
   case, the SMTP client SHOULD try at least two addresses.

   Two types of information are used to rank the host addresses:
   multiple MX records, and multihomed hosts.

   MX records contain a preference indication that MUST be used in
   sorting if more than one such record appears (see below). Lower
   numbers are more preferred than higher ones. If there are multiple
   destinations with the same preference and there is no clear reason to
   favor one (e.g., by recognition of an easily reached address), then
   the sender-SMTP MUST randomize them to spread the load across
   multiple mail exchangers for a specific organization.

   The destination host (perhaps taken from the preferred MX record) may
   be multihomed, in which case the domain name resolver will return a
   list of alternative IP addresses. It is the responsibility of the
   domain name resolver interface to have ordered this list by
   decreasing preference if necessary, and the SMTP sender MUST try them
   in the order presented.

   Although the capability to try multiple alternative addresses is
   required, specific installations may want to limit or disable the use
   of alternative addresses. The question of whether a sender should
   attempt retries using the different addresses of a multihomed host
   has been controversial. The main argument for using the multiple
   addresses is that it maximizes the probability of timely delivery,
   and indeed sometimes the probability of any delivery; the counter-
   argument is that it may result in unnecessary resource use. Note
   that resource use is also strongly determined by the sending strategy
   discussed in Section 4.5.4.1.

   If an SMTP server receives a message with a destination for which it
   is a designated Mail eXchanger, it MAY relay the message (potentially
   after having rewritten the MAIL FROM and/or RCPT TO addresses), make
   final delivery of the message, or hand it off using some mechanism
   outside the SMTP-provided transport environment. Of course, neither
   of the latter require that the list of MX records be examined
   further.

   If it determines that it should relay the message without rewriting
   the address, it MUST sort the MX records to determine candidates for



Klensin Standards Track [Page 70]
 
RFC 5321 SMTP October 2008


   delivery. The records are first ordered by preference, with the
   lowest-numbered records being most preferred. The relay host MUST
   then inspect the list for any of the names or addresses by which it
   might be known in mail transactions. If a matching record is found,
   all records at that preference level and higher-numbered ones MUST be
   discarded from consideration. If there are no records left at that
   point, it is an error condition, and the message MUST be returned as
   undeliverable. If records do remain, they SHOULD be tried, best
   preference first, as described above.

5.2. IPv6 and MX Records

   In the contemporary Internet, SMTP clients and servers may be hosted
   on IPv4 systems, IPv6 systems, or dual-stack systems that are
   compatible with either version of the Internet Protocol. The host
   domains to which MX records point may, consequently, contain "A RR"s
   (IPv4), "AAAA RR"s (IPv6), or any combination of them. While RFC
   3974 [39] discusses some operational experience in mixed
   environments, it was not comprehensive enough to justify
   standardization, and some of its recommendations appear to be
   inconsistent with this specification. The appropriate actions to be
   taken either will depend on local circumstances, such as performance
   of the relevant networks and any conversions that might be necessary,
   or will be obvious (e.g., an IPv6-only client need not attempt to
   look up A RRs or attempt to reach IPv4-only servers). Designers of
   SMTP implementations that might run in IPv6 or dual-stack
   environments should study the procedures above, especially the
   comments about multihomed hosts, and, preferably, provide mechanisms
   to facilitate operational tuning and mail interoperability between
   IPv4 and IPv6 systems while considering local circumstances.

6. Problem Detection and Handling

6.1. Reliable Delivery and Replies by Email

   When the receiver-SMTP accepts a piece of mail (by sending a "250 OK"
   message in response to DATA), it is accepting responsibility for
   delivering or relaying the message. It must take this responsibility
   seriously. It MUST NOT lose the message for frivolous reasons, such
   as because the host later crashes or because of a predictable
   resource shortage. Some reasons that are not considered frivolous
   are discussed in the next subsection and in Section 7.8.

   If there is a delivery failure after acceptance of a message, the
   receiver-SMTP MUST formulate and mail a notification message. This
   notification MUST be sent using a null ("<>") reverse-path in the
   envelope. The recipient of this notification MUST be the address
   from the envelope return path (or the Return-Path: line). However,



Klensin Standards Track [Page 71]
 
RFC 5321 SMTP October 2008


   if this address is null ("<>"), the receiver-SMTP MUST NOT send a
   notification. Obviously, nothing in this section can or should
   prohibit local decisions (i.e., as part of the same system
   environment as the receiver-SMTP) to log or otherwise transmit
   information about null address events locally if that is desired. If
   the address is an explicit source route, it MUST be stripped down to
   its final hop.

   For example, suppose that an error notification must be sent for a
   message that arrived with:

      MAIL FROM:<@a,@b:user@d>

   The notification message MUST be sent using:

      RCPT TO:<user@d>

   Some delivery failures after the message is accepted by SMTP will be
   unavoidable. For example, it may be impossible for the receiving
   SMTP server to validate all the delivery addresses in RCPT command(s)
   due to a "soft" domain system error, because the target is a mailing
   list (see earlier discussion of RCPT), or because the server is
   acting as a relay and has no immediate access to the delivering
   system.

   To avoid receiving duplicate messages as the result of timeouts, a
   receiver-SMTP MUST seek to minimize the time required to respond to
   the final <CRLF>.<CRLF> end of data indicator. See RFC 1047 [40] for
   a discussion of this problem.

6.2. Unwanted, Unsolicited, and "Attack" Messages

   Utility and predictability of the Internet mail system requires that
   messages that can be delivered should be delivered, regardless of any
   syntax or other faults associated with those messages and regardless
   of their content. If they cannot be delivered, and cannot be
   rejected by the SMTP server during the SMTP transaction, they should
   be "bounced" (returned with non-delivery notification messages) as
   described above. In today's world, in which many SMTP server
   operators have discovered that the quantity of undesirable bulk email
   vastly exceeds the quantity of desired mail and in which accepting a
   message may trigger additional undesirable traffic by providing
   verification of the address, those principles may not be practical.

   As discussed in Section 7.8 and Section 7.9 below, dropping mail
   without notification of the sender is permitted in practice.
   However, it is extremely dangerous and violates a long tradition and
   community expectations that mail is either delivered or returned. If



Klensin Standards Track [Page 72]
 
RFC 5321 SMTP October 2008


   silent message-dropping is misused, it could easily undermine
   confidence in the reliability of the Internet's mail systems. So
   silent dropping of messages should be considered only in those cases
   where there is very high confidence that the messages are seriously
   fraudulent or otherwise inappropriate.

   To stretch the principle of delivery if possible even further, it may
   be a rational policy to not deliver mail that has an invalid return
   address, although the history of the network is that users are
   typically better served by delivering any message that can be
   delivered. Reliably determining that a return address is invalid can
   be a difficult and time-consuming process, especially if the putative
   sending system is not directly accessible or does not fully and
   accurately support VRFY and, even if a "drop messages with invalid
   return addresses" policy is adopted, it SHOULD be applied only when
   there is near-certainty that the return addresses are, in fact,
   invalid.

   Conversely, if a message is rejected because it is found to contain
   hostile content (a decision that is outside the scope of an SMTP
   server as defined in this document), rejection ("bounce") messages
   SHOULD NOT be sent unless the receiving site is confident that those
   messages will be usefully delivered. The preference and default in
   these cases is to avoid sending non-delivery messages when the
   incoming message is determined to contain hostile content.

6.3. Loop Detection

   Simple counting of the number of "Received:" header fields in a
   message has proven to be an effective, although rarely optimal,
   method of detecting loops in mail systems. SMTP servers using this
   technique SHOULD use a large rejection threshold, normally at least
   100 Received entries. Whatever mechanisms are used, servers MUST
   contain provisions for detecting and stopping trivial loops.

6.4. Compensating for Irregularities

   Unfortunately, variations, creative interpretations, and outright
   violations of Internet mail protocols do occur; some would suggest
   that they occur quite frequently. The debate as to whether a well-
   behaved SMTP receiver or relay should reject a malformed message,
   attempt to pass it on unchanged, or attempt to repair it to increase
   the odds of successful delivery (or subsequent reply) began almost
   with the dawn of structured network mail and shows no signs of
   abating. Advocates of rejection claim that attempted repairs are
   rarely completely adequate and that rejection of bad messages is the
   only way to get the offending software repaired. Advocates of
   "repair" or "deliver no matter what" argue that users prefer that



Klensin Standards Track [Page 73]
 
RFC 5321 SMTP October 2008


   mail go through it if at all possible and that there are significant
   market pressures in that direction. In practice, these market
   pressures may be more important to particular vendors than strict
   conformance to the standards, regardless of the preference of the
   actual developers.

   The problems associated with ill-formed messages were exacerbated by
   the introduction of the split-UA mail reading protocols (Post Office
   Protocol (POP) version 2 [15], Post Office Protocol (POP) version 3
   [16], IMAP version 2 [41], and PCMAIL [42]). These protocols
   encouraged the use of SMTP as a posting (message submission)
   protocol, and SMTP servers as relay systems for these client hosts
   (which are often only intermittently connected to the Internet).
   Historically, many of those client machines lacked some of the
   mechanisms and information assumed by SMTP (and indeed, by the mail
   format protocol, RFC 822 [28]). Some could not keep adequate track
   of time; others had no concept of time zones; still others could not
   identify their own names or addresses; and, of course, none could
   satisfy the assumptions that underlay RFC 822's conception of
   authenticated addresses.

   In response to these weak SMTP clients, many SMTP systems now
   complete messages that are delivered to them in incomplete or
   incorrect form. This strategy is generally considered appropriate
   when the server can identify or authenticate the client, and there
   are prior agreements between them. By contrast, there is at best
   great concern about fixes applied by a relay or delivery SMTP server
   that has little or no knowledge of the user or client machine. Many
   of these issues are addressed by using a separate protocol, such as
   that defined in RFC 4409 [18], for message submission, rather than
   using originating SMTP servers for that purpose.

   The following changes to a message being processed MAY be applied
   when necessary by an originating SMTP server, or one used as the
   target of SMTP as an initial posting (message submission) protocol:

   o Addition of a message-id field when none appears

   o Addition of a date, time, or time zone when none appears

   o Correction of addresses to proper FQDN format

   The less information the server has about the client, the less likely
   these changes are to be correct and the more caution and conservatism
   should be applied when considering whether or not to perform fixes
   and how. These changes MUST NOT be applied by an SMTP server that
   provides an intermediate relay function.




Klensin Standards Track [Page 74]
 
RFC 5321 SMTP October 2008


   In all cases, properly operating clients supplying correct
   information are preferred to corrections by the SMTP server. In all
   cases, documentation SHOULD be provided in trace header fields and/or
   header field comments for actions performed by the servers.

7. Security Considerations

7.1. Mail Security and Spoofing

   SMTP mail is inherently insecure in that it is feasible for even
   fairly casual users to negotiate directly with receiving and relaying
   SMTP servers and create messages that will trick a naive recipient
   into believing that they came from somewhere else. Constructing such
   a message so that the "spoofed" behavior cannot be detected by an
   expert is somewhat more difficult, but not sufficiently so as to be a
   deterrent to someone who is determined and knowledgeable.
   Consequently, as knowledge of Internet mail increases, so does the
   knowledge that SMTP mail inherently cannot be authenticated, or
   integrity checks provided, at the transport level. Real mail
   security lies only in end-to-end methods involving the message
   bodies, such as those that use digital signatures (see RFC 1847 [43]
   and, e.g., Pretty Good Privacy (PGP) in RFC 4880 [44] or Secure/
   Multipurpose Internet Mail Extensions (S/MIME) in RFC 3851 [45]).

   Various protocol extensions and configuration options that provide
   authentication at the transport level (e.g., from an SMTP client to
   an SMTP server) improve somewhat on the traditional situation
   described above. However, in general, they only authenticate one
   server to another rather than a chain of relays and servers, much
   less authenticating users or user machines. Consequently, unless
   they are accompanied by careful handoffs of responsibility in a
   carefully designed trust environment, they remain inherently weaker
   than end-to-end mechanisms that use digitally signed messages rather
   than depending on the integrity of the transport system.

   Efforts to make it more difficult for users to set envelope return
   path and header "From" fields to point to valid addresses other than
   their own are largely misguided: they frustrate legitimate
   applications in which mail is sent by one user on behalf of another,
   in which error (or normal) replies should be directed to a special
   address, or in which a single message is sent to multiple recipients
   on different hosts. (Systems that provide convenient ways for users
   to alter these header fields on a per-message basis should attempt to
   establish a primary and permanent mailbox address for the user so
   that Sender header fields within the message data can be generated
   sensibly.)





Klensin Standards Track [Page 75]
 
RFC 5321 SMTP October 2008


   This specification does not further address the authentication issues
   associated with SMTP other than to advocate that useful functionality
   not be disabled in the hope of providing some small margin of
   protection against a user who is trying to fake mail.

7.2. "Blind" Copies

   Addresses that do not appear in the message header section may appear
   in the RCPT commands to an SMTP server for a number of reasons. The
   two most common involve the use of a mailing address as a "list
   exploder" (a single address that resolves into multiple addresses)
   and the appearance of "blind copies". Especially when more than one
   RCPT command is present, and in order to avoid defeating some of the
   purpose of these mechanisms, SMTP clients and servers SHOULD NOT copy
   the full set of RCPT command arguments into the header section,
   either as part of trace header fields or as informational or private-
   extension header fields. Since this rule is often violated in
   practice, and cannot be enforced, sending SMTP systems that are aware
   of "bcc" use MAY find it helpful to send each blind copy as a
   separate message transaction containing only a single RCPT command.

   There is no inherent relationship between either "reverse" (from
   MAIL, SAML, etc., commands) or "forward" (RCPT) addresses in the SMTP
   transaction ("envelope") and the addresses in the header section.
   Receiving systems SHOULD NOT attempt to deduce such relationships and
   use them to alter the header section of the message for delivery.
   The popular "Apparently-to" header field is a violation of this
   principle as well as a common source of unintended information
   disclosure and SHOULD NOT be used.

7.3. VRFY, EXPN, and Security

   As discussed in Section 3.5, individual sites may want to disable
   either or both of VRFY or EXPN for security reasons (see below). As
   a corollary to the above, implementations that permit this MUST NOT
   appear to have verified addresses that are not, in fact, verified.
   If a site disables these commands for security reasons, the SMTP
   server MUST return a 252 response, rather than a code that could be
   confused with successful or unsuccessful verification.

   Returning a 250 reply code with the address listed in the VRFY
   command after having checked it only for syntax violates this rule.
   Of course, an implementation that "supports" VRFY by always returning
   550 whether or not the address is valid is equally not in
   conformance.

   On the public Internet, the contents of mailing lists have become
   popular as an address information source for so-called "spammers."



Klensin Standards Track [Page 76]
 
RFC 5321 SMTP October 2008


   The use of EXPN to "harvest" addresses has increased as list
   administrators have installed protections against inappropriate uses
   of the lists themselves. However, VRFY and EXPN are still useful for
   authenticated users and within an administrative domain. For
   example, VRFY and EXPN are useful for performing internal audits of
   how email gets routed to check and to make sure no one is
   automatically forwarding sensitive mail outside the organization.
   Sites implementing SMTP authentication may choose to make VRFY and
   EXPN available only to authenticated requestors. Implementations
   SHOULD still provide support for EXPN, but sites SHOULD carefully
   evaluate the tradeoffs.

   Whether disabling VRFY provides any real marginal security depends on
   a series of other conditions. In many cases, RCPT commands can be
   used to obtain the same information about address validity. On the
   other hand, especially in situations where determination of address
   validity for RCPT commands is deferred until after the DATA command
   is received, RCPT may return no information at all, while VRFY is
   expected to make a serious attempt to determine validity before
   generating a response code (see discussion above).

7.4. Mail Rerouting Based on the 251 and 551 Response Codes

   Before a client uses the 251 or 551 reply codes from a RCPT command
   to automatically update its future behavior (e.g., updating the
   user's address book), it should be certain of the server's
   authenticity. If it does not, it may be subject to a man in the
   middle attack.

7.5. Information Disclosure in Announcements

   There has been an ongoing debate about the tradeoffs between the
   debugging advantages of announcing server type and version (and,
   sometimes, even server domain name) in the greeting response or in
   response to the HELP command and the disadvantages of exposing
   information that might be useful in a potential hostile attack. The
   utility of the debugging information is beyond doubt. Those who
   argue for making it available point out that it is far better to
   actually secure an SMTP server rather than hope that trying to
   conceal known vulnerabilities by hiding the server's precise identity
   will provide more protection. Sites are encouraged to evaluate the
   tradeoff with that issue in mind; implementations SHOULD minimally
   provide for making type and version information available in some way
   to other network hosts.







Klensin Standards Track [Page 77]
 
RFC 5321 SMTP October 2008


7.6. Information Disclosure in Trace Fields

   In some circumstances, such as when mail originates from within a LAN
   whose hosts are not directly on the public Internet, trace
   ("Received") header fields produced in conformance with this
   specification may disclose host names and similar information that
   would not normally be available. This ordinarily does not pose a
   problem, but sites with special concerns about name disclosure should
   be aware of it. Also, the optional FOR clause should be supplied
   with caution or not at all when multiple recipients are involved lest
   it inadvertently disclose the identities of "blind copy" recipients
   to others.

7.7. Information Disclosure in Message Forwarding

   As discussed in Section 3.4, use of the 251 or 551 reply codes to
   identify the replacement address associated with a mailbox may
   inadvertently disclose sensitive information. Sites that are
   concerned about those issues should ensure that they select and
   configure servers appropriately.

7.8. Resistance to Attacks

   In recent years, there has been an increase of attacks on SMTP
   servers, either in conjunction with attempts to discover addresses
   for sending unsolicited messages or simply to make the servers
   inaccessible to others (i.e., as an application-level denial of
   service attack). While the means of doing so are beyond the scope of
   this Standard, rational operational behavior requires that servers be
   permitted to detect such attacks and take action to defend
   themselves. For example, if a server determines that a large number
   of RCPT TO commands are being sent, most or all with invalid
   addresses, as part of such an attack, it would be reasonable for the
   server to close the connection after generating an appropriate number
   of 5yz (normally 550) replies.

7.9. Scope of Operation of SMTP Servers

   It is a well-established principle that an SMTP server may refuse to
   accept mail for any operational or technical reason that makes sense
   to the site providing the server. However, cooperation among sites
   and installations makes the Internet possible. If sites take
   excessive advantage of the right to reject traffic, the ubiquity of
   email availability (one of the strengths of the Internet) will be
   threatened; considerable care should be taken and balance maintained
   if a site decides to be selective about the traffic it will accept
   and process.




Klensin Standards Track [Page 78]
 
RFC 5321 SMTP October 2008


   In recent years, use of the relay function through arbitrary sites
   has been used as part of hostile efforts to hide the actual origins
   of mail. Some sites have decided to limit the use of the relay
   function to known or identifiable sources, and implementations SHOULD
   provide the capability to perform this type of filtering. When mail
   is rejected for these or other policy reasons, a 550 code SHOULD be
   used in response to EHLO (or HELO), MAIL, or RCPT as appropriate.

8. IANA Considerations

   IANA maintains three registries in support of this specification, all
   of which were created for RFC 2821 or earlier. This document expands
   the third one as specified below. The registry references listed are
   as of the time of publication; IANA does not guarantee the locations
   associated with the URLs. The registries are as follows:

   o The first, "Simple Mail Transfer Protocol (SMTP) Service
      Extensions" [46], consists of SMTP service extensions with the
      associated keywords, and, as needed, parameters and verbs. As
      specified in Section 2.2.2, no entry may be made in this registry
      that starts in an "X". Entries may be made only for service
      extensions (and associated keywords, parameters, or verbs) that
      are defined in Standards-Track or Experimental RFCs specifically
      approved by the IESG for this purpose.

   o The second registry, "Address Literal Tags" [47], consists of
      "tags" that identify forms of domain literals other than those for
      IPv4 addresses (specified in RFC 821 and in this document). The
      initial entry in that registry is for IPv6 addresses (specified in
      this document). Additional literal types require standardization
      before being used; none are anticipated at this time.

   o The third, "Mail Transmission Types" [46], established by RFC 821
      and renewed by this specification, is a registry of link and
      protocol identifiers to be used with the "via" and "with"
      subclauses of the time stamp ("Received:" header field) described
      in Section 4.4. Link and protocol identifiers in addition to
      those specified in this document may be registered only by
      standardization or by way of an RFC-documented, IESG-approved,
      Experimental protocol extension. This name space is for
      identification and not limited in size: the IESG is encouraged to
      approve on the basis of clear documentation and a distinct method
      rather than preferences about the properties of the method itself.

      An additional subsection has been added to the "VIA link types"
      and "WITH protocol types" subsections of this registry to contain
      registrations of "Additional-registered-clauses" as described
      above. The registry will contain clause names, a description, a



Klensin Standards Track [Page 79]
 
RFC 5321 SMTP October 2008


      summary of the syntax of the associated String, and a reference.
      As new clauses are defined, they may, in principle, specify
      creation of their own registries if the Strings consist of
      reserved terms or keywords rather than less restricted strings.
      As with link and protocol identifiers, additional clauses may be
      registered only by standardization or by way of an RFC-documented,
      IESG-approved, Experimental protocol extension. The additional
      clause name space is for identification and is not limited in
      size: the IESG is encouraged to approve on the basis of clear
      documentation, actual use or strong signs that the clause will be
      used, and a distinct requirement rather than preferences about the
      properties of the clause itself.

   In addition, if additional trace header fields (i.e., in addition to
   Return-path and Received) are ever created, those trace fields MUST
   be added to the IANA registry established by BCP 90 (RFC 3864) [11]
   for use with RFC 5322 [4].

9. Acknowledgments

   Many people contributed to the development of RFC 2821. That
   document should be consulted for those acknowledgments. For the
   present document, the editor and the community owe thanks to Dawn
   Mann and Tony Hansen who assisted in the very painful process of
   editing and converting the internal format of the document from one
   system to another.

   Neither this document nor RFC 2821 would have been possible without
   the many contribution and insights of the late Jon Postel. Those
   contributions of course include the original specification of SMTP in
   RFC 821. A considerable quantity of text from RFC 821 still appears
   in this document as do several of Jon's original examples that have
   been updated only as needed to reflect other changes in the
   specification.

   Many people made comments or suggestions on the mailing list or in
   notes to the author. Important corrections or clarifications were
   suggested by several people, including Matti Aarnio, Glenn Anderson,
   Derek J. Balling, Alex van den Bogaerdt, Stephane Bortzmeyer, Vint
   Cerf, Jutta Degener, Steve Dorner, Lisa Dusseault, Frank Ellerman,
   Ned Freed, Randy Gellens, Sabahattin Gucukoglu, Philip Guenther, Arnt
   Gulbrandsen, Eric Hall, Richard O. Hammer, Tony Hansen, Peter J.
   Holzer, Kari Hurtta, Bryon Roche Kain, Valdis Kletnieks, Mathias
   Koerber, John Leslie, Bruce Lilly, Jeff Macdonald, Mark E. Mallett,
   Mark Martinec, S. Moonesamy, Lyndon Nerenberg, Chris Newman, Douglas
   Otis, Pete Resnick, Robert A. Rosenberg, Vince Sabio, Hector Santos,
   David F. Skoll, Paul Smith, and Brett Watson.




Klensin Standards Track [Page 80]
 
RFC 5321 SMTP October 2008


   The efforts of the Area Directors -- Lisa Dusseault, Ted Hardie, and
   Chris Newman -- to get this effort restarted and keep it moving, and
   of an ad hoc committee with the same purpose, are gratefully
   acknowledged. The members of that committee were (in alphabetical
   order) Dave Crocker, Cyrus Daboo, Tony Finch, Ned Freed, Randall
   Gellens, Tony Hansen, the author, and Alexey Melnikov. Tony Hansen
   also acted as ad hoc chair on the mailing list reviewing this
   document; without his efforts, sense of balance and fairness, and
   patience, it clearly would not have been possible.

10. References

10.1. Normative References

   [1] Postel, J., "Simple Mail Transfer Protocol", STD 10, RFC 821,
         August 1982.

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

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

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

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

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

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

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

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

   [9] Newman, C., "ESMTP and LMTP Transmission Types Registration",
         RFC 3848, July 2004.

   [10] Klensin, J., Freed, N., and K. Moore, "SMTP Service Extension
         for Message Size Declaration", STD 10, RFC 1870, November 1995.




Klensin Standards Track [Page 81]
 
RFC 5321 SMTP October 2008


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

10.2. Informative References

   [12] Partridge, C., "Mail routing and the domain system", RFC 974,
         January 1986.

   [13] Klensin, J., Freed, N., Rose, M., Stefferud, E., and D.
         Crocker, "SMTP Service Extensions", STD 10, RFC 1869,
         November 1995.

   [14] Klensin, J., "Simple Mail Transfer Protocol", RFC 2821,
         April 2001.

   [15] Butler, M., Postel, J., Chase, D., Goldberger, J., and J.
         Reynolds, "Post Office Protocol: Version 2", RFC 937,
         February 1985.

   [16] Myers, J. and M. Rose, "Post Office Protocol - Version 3",
         STD 53, RFC 1939, May 1996.

   [17] Crispin, M., "INTERNET MESSAGE ACCESS PROTOCOL - VERSION
         4rev1", RFC 3501, March 2003.

   [18] Gellens, R. and J. Klensin, "Message Submission for Mail",
         RFC 4409, April 2006.

   [19] Freed, N., "SMTP Service Extension for Command Pipelining",
         STD 60, RFC 2920, September 2000.

   [20] Vaudreuil, G., "SMTP Service Extensions for Transmission of
         Large and Binary MIME Messages", RFC 3030, December 2000.

   [21] Freed, N. and N. Borenstein, "Multipurpose Internet Mail
         Extensions (MIME) Part One: Format of Internet Message Bodies",
         RFC 2045, November 1996.

   [22] Klensin, J., Freed, N., Rose, M., Stefferud, E., and D.
         Crocker, "SMTP Service Extension for 8bit-MIMEtransport",
         RFC 1652, July 1994.

   [23] Moore, K., "MIME (Multipurpose Internet Mail Extensions) Part
         Three: Message Header Extensions for Non-ASCII Text", RFC 2047,
         November 1996.





Klensin Standards Track [Page 82]
 
RFC 5321 SMTP October 2008


   [24] Freed, N. and K. Moore, "MIME Parameter Value and Encoded Word
         Extensions: Character Sets, Languages, and Continuations",
         RFC 2231, November 1997.

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

   [26] Hansen, T. and J. Klensin, "A Registry for SMTP Enhanced Mail
         System Status Codes", BCP 138, RFC 5248, June 2008.

   [27] Freed, N., "Behavior of and Requirements for Internet
         Firewalls", RFC 2979, October 2000.

   [28] Crocker, D., "Standard for the format of ARPA Internet text
         messages", STD 11, RFC 822, August 1982.

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

   [30] Fenton, J., "Analysis of Threats Motivating DomainKeys
         Identified Mail (DKIM)", RFC 4686, September 2006.

   [31] Allman, E., Callas, J., Delany, M., Libbey, M., Fenton, J., and
         M. Thomas, "DomainKeys Identified Mail (DKIM) Signatures",
         RFC 4871, May 2007.

   [32] Moore, K., "Simple Mail Transfer Protocol (SMTP) Service
         Extension for Delivery Status Notifications (DSNs)", RFC 3461,
         January 2003.

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

   [34] Postel, J. and J. Reynolds, "File Transfer Protocol", STD 9,
         RFC 959, October 1985.

   [35] Kille, S., "MIXER (Mime Internet X.400 Enhanced Relay): Mapping
         between X.400 and RFC 822/MIME", RFC 2156, January 1998.

   [36] De Winter, J., "SMTP Service Extension for Remote Message Queue
         Starting", RFC 1985, August 1996.

   [37] Hansen, T. and G. Vaudreuil, "Message Disposition
         Notification", RFC 3798, May 2004.

   [38] Elz, R. and R. Bush, "Clarifications to the DNS Specification",
         RFC 2181, July 1997.



Klensin Standards Track [Page 83]
 
RFC 5321 SMTP October 2008


   [39] Nakamura, M. and J. Hagino, "SMTP Operational Experience in
         Mixed IPv4/v6 Environments", RFC 3974, January 2005.

   [40] Partridge, C., "Duplicate messages and SMTP", RFC 1047,
         February 1988.

   [41] Crispin, M., "Interactive Mail Access Protocol: Version 2",
         RFC 1176, August 1990.

   [42] Lambert, M., "PCMAIL: A distributed mail system for personal
         computers", RFC 1056, June 1988.

   [43] Galvin, J., Murphy, S., Crocker, S., and N. Freed, "Security
         Multiparts for MIME: Multipart/Signed and Multipart/Encrypted",
         RFC 1847, October 1995.

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

   [45] Ramsdell, B., "Secure/Multipurpose Internet Mail Extensions
         (S/MIME) Version 3.1 Message Specification", RFC 3851,
         July 2004.

   [46] Internet Assigned Number Authority (IANA), "IANA Mail
         Parameters", 2007,
         <http://www.iana.org/assignments/mail-parameters>.

   [47] Internet Assigned Number Authority (IANA), "Address Literal
         Tags", 2007,
         <http://www.iana.org/assignments/address-literal-tags>.





















Klensin Standards Track [Page 84]
 
RFC 5321 SMTP October 2008


Appendix A. TCP Transport Service

   The TCP connection supports the transmission of 8-bit bytes. The
   SMTP data is 7-bit ASCII characters. Each character is transmitted
   as an 8-bit byte with the high-order bit cleared to zero. Service
   extensions may modify this rule to permit transmission of full 8-bit
   data bytes as part of the message body, or, if specifically designed
   to do so, in SMTP commands or responses.

Appendix B. Generating SMTP Commands from RFC 822 Header Fields

   Some systems use an RFC 822 header section (only) in a mail
   submission protocol, or otherwise generate SMTP commands from RFC 822
   header fields when such a message is handed to an MTA from a UA.
   While the MTA-UA protocol is a private matter, not covered by any
   Internet Standard, there are problems with this approach. For
   example, there have been repeated problems with proper handling of
   "bcc" copies and redistribution lists when information that
   conceptually belongs to the mail envelope is not separated early in
   processing from header field information (and kept separate).

   It is recommended that the UA provide its initial ("submission
   client") MTA with an envelope separate from the message itself.
   However, if the envelope is not supplied, SMTP commands SHOULD be
   generated as follows:

   1. Each recipient address from a TO, CC, or BCC header field SHOULD
       be copied to a RCPT command (generating multiple message copies
       if that is required for queuing or delivery). This includes any
       addresses listed in a RFC 822 "group". Any BCC header fields
       SHOULD then be removed from the header section. Once this
       process is completed, the remaining header fields SHOULD be
       checked to verify that at least one TO, CC, or BCC header field
       remains. If none do, then a BCC header field with no additional
       information SHOULD be inserted as specified in [4].

   2. The return address in the MAIL command SHOULD, if possible, be
       derived from the system's identity for the submitting (local)
       user, and the "From:" header field otherwise. If there is a
       system identity available, it SHOULD also be copied to the Sender
       header field if it is different from the address in the From
       header field. (Any Sender header field that was already there
       SHOULD be removed.) Systems may provide a way for submitters to
       override the envelope return address, but may want to restrict
       its use to privileged users. This will not prevent mail forgery,
       but may lessen its incidence; see Section 7.1.





Klensin Standards Track [Page 85]
 
RFC 5321 SMTP October 2008


   When an MTA is being used in this way, it bears responsibility for
   ensuring that the message being transmitted is valid. The mechanisms
   for checking that validity, and for handling (or returning) messages
   that are not valid at the time of arrival, are part of the MUA-MTA
   interface and not covered by this specification.

   A submission protocol based on Standard RFC 822 information alone
   MUST NOT be used to gateway a message from a foreign (non-SMTP) mail
   system into an SMTP environment. Additional information to construct
   an envelope must come from some source in the other environment,
   whether supplemental header fields or the foreign system's envelope.

   Attempts to gateway messages using only their header "To" and "Cc"
   fields have repeatedly caused mail loops and other behavior adverse
   to the proper functioning of the Internet mail environment. These
   problems have been especially common when the message originates from
   an Internet mailing list and is distributed into the foreign
   environment using envelope information. When these messages are then
   processed by a header-section-only remailer, loops back to the
   Internet environment (and the mailing list) are almost inevitable.

Appendix C. Source Routes

   Historically, the <reverse-path> was a reverse source routing list of
   hosts and a source mailbox. The first host in the <reverse-path> was
   historically the host sending the MAIL command; today, source routes
   SHOULD NOT appear in the reverse-path. Similarly, the <forward-path>
   may be a source routing lists of hosts and a destination mailbox.
   However, in general, the <forward-path> SHOULD contain only a mailbox
   and domain name, relying on the domain name system to supply routing
   information if required. The use of source routes is deprecated (see
   Appendix F.2); while servers MUST be prepared to receive and handle
   them as discussed in Section 3.3 and Appendix F.2, clients SHOULD NOT
   transmit them and this section is included in the current
   specification only to provide context. It has been modified somewhat
   from the material in RFC 821 to prevent server actions that might
   confuse clients or subsequent servers that do not expect a full
   source route implementation.

   For relay purposes, the forward-path may be a source route of the
   form "@ONE,@TWO:JOE@THREE", where ONE, TWO, and THREE MUST be fully-
   qualified domain names. This form is used to emphasize the
   distinction between an address and a route. The mailbox (here, JOE@
   THREE) is an absolute address, and the route is information about how
   to get there. The two concepts should not be confused.

   If source routes are used, RFC 821 and the text below should be
   consulted for the mechanisms for constructing and updating the



Klensin Standards Track [Page 86]
 
RFC 5321 SMTP October 2008


   forward-path. A server that is reached by means of a source route
   (e.g., its domain name appears first in the list in the forward-path)
   MUST remove its domain name from any forward-paths in which that
   domain name appears before forwarding the message and MAY remove all
   other source routing information. The reverse-path SHOULD NOT be
   updated by servers conforming to this specification.

   Notice that the forward-path and reverse-path appear in the SMTP
   commands and replies, but not necessarily in the message. That is,
   there is no need for these paths and especially this syntax to appear
   in the "To:" , "From:", "CC:", etc. fields of the message header
   section. Conversely, SMTP servers MUST NOT derive final message
   routing information from message header fields.

   When the list of hosts is present despite the recommendations above,
   it is a "reverse" source route and indicates that the mail was
   relayed through each host on the list (the first host in the list was
   the most recent relay). This list is used as a source route to
   return non-delivery notices to the sender. If, contrary to the
   recommendations here, a relay host adds itself to the beginning of
   the list, it MUST use its name as known in the transport environment
   to which it is relaying the mail rather than that of the transport
   environment from which the mail came (if they are different). Note
   that a situation could easily arise in which some relay hosts add
   their names to the reverse source route and others do not, generating
   discontinuities in the routing list. This is another reason why
   servers needing to return a message SHOULD ignore the source route
   entirely and simply use the domain as specified in the Mailbox.

Appendix D. Scenarios

   This section presents complete scenarios of several types of SMTP
   sessions. In the examples, "C:" indicates what is said by the SMTP
   client, and "S:" indicates what is said by the SMTP server.

















Klensin Standards Track [Page 87]
 
RFC 5321 SMTP October 2008


D.1. A Typical SMTP Transaction Scenario

   This SMTP example shows mail sent by Smith at host bar.com, and to
   Jones, Green, and Brown at host foo.com. Here we assume that host
   bar.com contacts host foo.com directly. The mail is accepted for
   Jones and Brown. Green does not have a mailbox at host foo.com.

      S: 220 foo.com Simple Mail Transfer Service Ready
      C: EHLO bar.com
      S: 250-foo.com greets bar.com
      S: 250-8BITMIME
      S: 250-SIZE
      S: 250-DSN
      S: 250 HELP
      C: MAIL FROM:<Smith@bar.com>
      S: 250 OK
      C: RCPT TO:<Jones@foo.com>
      S: 250 OK
      C: RCPT TO:<Green@foo.com>
      S: 550 No such user here
      C: RCPT TO:<Brown@foo.com>
      S: 250 OK
      C: DATA
      S: 354 Start mail input; end with <CRLF>.<CRLF>
      C: Blah blah blah...
      C: ...etc. etc. etc.
      C: .
      S: 250 OK
      C: QUIT
      S: 221 foo.com Service closing transmission channel





















Klensin Standards Track [Page 88]
 
RFC 5321 SMTP October 2008


D.2. Aborted SMTP Transaction Scenario

      S: 220 foo.com Simple Mail Transfer Service Ready
      C: EHLO bar.com
      S: 250-foo.com greets bar.com
      S: 250-8BITMIME
      S: 250-SIZE
      S: 250-DSN
      S: 250 HELP
      C: MAIL FROM:<Smith@bar.com>
      S: 250 OK
      C: RCPT TO:<Jones@foo.com>
      S: 250 OK
      C: RCPT TO:<Green@foo.com>
      S: 550 No such user here
      C: RSET
      S: 250 OK
      C: QUIT
      S: 221 foo.com Service closing transmission channel
































Klensin Standards Track [Page 89]
 
RFC 5321 SMTP October 2008


D.3. Relayed Mail Scenario

   Step 1 -- Source Host to Relay Host

   The source host performs a DNS lookup on XYZ.COM (the destination
   address) and finds DNS MX records specifying xyz.com as the best
   preference and foo.com as a lower preference. It attempts to open a
   connection to xyz.com and fails. It then opens a connection to
   foo.com, with the following dialogue:

      S: 220 foo.com Simple Mail Transfer Service Ready
      C: EHLO bar.com
      S: 250-foo.com greets bar.com
      S: 250-8BITMIME
      S: 250-SIZE
      S: 250-DSN
      S: 250 HELP
      C: MAIL FROM:<JQP@bar.com>
      S: 250 OK
      C: RCPT TO:<Jones@XYZ.COM>
      S: 250 OK
      C: DATA
      S: 354 Start mail input; end with <CRLF>.<CRLF>
      C: Date: Thu, 21 May 1998 05:33:29 -0700
      C: From: John Q. Public <JQP@bar.com>
      C: Subject: The Next Meeting of the Board
      C: To: Jones@xyz.com
      C:
      C: Bill:
      C: The next meeting of the board of directors will be
      C: on Tuesday.
      C: John.
      C: .
      S: 250 OK
      C: QUIT
      S: 221 foo.com Service closing transmission channel















Klensin Standards Track [Page 90]
 
RFC 5321 SMTP October 2008


   Step 2 -- Relay Host to Destination Host

   foo.com, having received the message, now does a DNS lookup on
   xyz.com. It finds the same set of MX records, but cannot use the one
   that points to itself (or to any other host as a worse preference).
   It tries to open a connection to xyz.com itself and succeeds. Then
   we have:

           S: 220 xyz.com Simple Mail Transfer Service Ready
           C: EHLO foo.com
           S: 250 xyz.com is on the air
           C: MAIL FROM:<JQP@bar.com>
           S: 250 OK
           C: RCPT TO:<Jones@XYZ.COM>
           S: 250 OK
           C: DATA
           S: 354 Start mail input; end with <CRLF>.<CRLF>
           C: Received: from bar.com by foo.com ; Thu, 21 May 1998
           C: 05:33:29 -0700
           C: Date: Thu, 21 May 1998 05:33:22 -0700
           C: From: John Q. Public <JQP@bar.com>
           C: Subject: The Next Meeting of the Board
           C: To: Jones@xyz.com
           C:
           C: Bill:
           C: The next meeting of the board of directors will be
           C: on Tuesday.
           C: John.
           C: .
           S: 250 OK
           C: QUIT
           S: 221 foo.com Service closing transmission channel



















Klensin Standards Track [Page 91]
 
RFC 5321 SMTP October 2008


D.4. Verifying and Sending Scenario

      S: 220 foo.com Simple Mail Transfer Service Ready
      C: EHLO bar.com
      S: 250-foo.com greets bar.com
      S: 250-8BITMIME
      S: 250-SIZE
      S: 250-DSN
      S: 250-VRFY
      S: 250 HELP
      C: VRFY Crispin
      S: 250 Mark Crispin <Admin.MRC@foo.com>
      C: MAIL FROM:<EAK@bar.com>
      S: 250 OK
      C: RCPT TO:<Admin.MRC@foo.com>
      S: 250 OK
      C: DATA
      S: 354 Start mail input; end with <CRLF>.<CRLF>
      C: Blah blah blah...
      C: ...etc. etc. etc.
      C: .
      S: 250 OK
      C: QUIT
      S: 221 foo.com Service closing transmission channel

Appendix E. Other Gateway Issues

   In general, gateways between the Internet and other mail systems
   SHOULD attempt to preserve any layering semantics across the
   boundaries between the two mail systems involved. Gateway-
   translation approaches that attempt to take shortcuts by mapping
   (such as mapping envelope information from one system to the message
   header section or body of another) have generally proven to be
   inadequate in important ways. Systems translating between
   environments that do not support both envelopes and a header section
   and Internet mail must be written with the understanding that some
   information loss is almost inevitable.














Klensin Standards Track [Page 92]
 
RFC 5321 SMTP October 2008


Appendix F. Deprecated Features of RFC 821

   A few features of RFC 821 have proven to be problematic and SHOULD
   NOT be used in Internet mail.

F.1. TURN

   This command, described in RFC 821, raises important security issues
   since, in the absence of strong authentication of the host requesting
   that the client and server switch roles, it can easily be used to
   divert mail from its correct destination. Its use is deprecated;
   SMTP systems SHOULD NOT use it unless the server can authenticate the
   client.

F.2. Source Routing

   RFC 821 utilized the concept of explicit source routing to get mail
   from one host to another via a series of relays. The requirement to
   utilize source routes in regular mail traffic was eliminated by the
   introduction of the domain name system "MX" record and the last
   significant justification for them was eliminated by the
   introduction, in RFC 1123, of a clear requirement that addresses
   following an "@" must all be fully-qualified domain names.
   Consequently, the only remaining justifications for the use of source
   routes are support for very old SMTP clients or MUAs and in mail
   system debugging. They can, however, still be useful in the latter
   circumstance and for routing mail around serious, but temporary,
   problems such as problems with the relevant DNS records.

   SMTP servers MUST continue to accept source route syntax as specified
   in the main body of this document and in RFC 1123. They MAY, if
   necessary, ignore the routes and utilize only the target domain in
   the address. If they do utilize the source route, the message MUST
   be sent to the first domain shown in the address. In particular, a
   server MUST NOT guess at shortcuts within the source route.

   Clients SHOULD NOT utilize explicit source routing except under
   unusual circumstances, such as debugging or potentially relaying
   around firewall or mail system configuration errors.

F.3. HELO

   As discussed in Sections 3.1 and 4.1.1, EHLO SHOULD be used rather
   than HELO when the server will accept the former. Servers MUST
   continue to accept and process HELO in order to support older
   clients.





Klensin Standards Track [Page 93]
 
RFC 5321 SMTP October 2008


F.4. #-literals

   RFC 821 provided for specifying an Internet address as a decimal
   integer host number prefixed by a pound sign, "#". In practice, that
   form has been obsolete since the introduction of TCP/IP. It is
   deprecated and MUST NOT be used.

F.5. Dates and Years

   When dates are inserted into messages by SMTP clients or servers
   (e.g., in trace header fields), four-digit years MUST BE used. Two-
   digit years are deprecated; three-digit years were never permitted in
   the Internet mail system.

F.6. Sending versus Mailing

   In addition to specifying a mechanism for delivering messages to
   user's mailboxes, RFC 821 provided additional, optional, commands to
   deliver messages directly to the user's terminal screen. These
   commands (SEND, SAML, SOML) were rarely implemented, and changes in
   workstation technology and the introduction of other protocols may
   have rendered them obsolete even where they are implemented.

   Clients SHOULD NOT provide SEND, SAML, or SOML as services. Servers
   MAY implement them. If they are implemented by servers, the
   implementation model specified in RFC 821 MUST be used and the
   command names MUST be published in the response to the EHLO command.

Author's Address

   John C. Klensin
   1770 Massachusetts Ave, Suite 322
   Cambridge, MA 02140
   USA

   EMail: john+smtp@jck.com















Klensin Standards Track [Page 94]
 
RFC 5321 SMTP October 2008


Full Copyright Statement

   Copyright (C) The IETF Trust (2008).

   This document is subject to the rights, licenses and restrictions
   contained in BCP 78, and except as set forth therein, the authors
   retain all their rights.

   This document and the information contained herein are provided on an
   "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS
   OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND
   THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS
   OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF
   THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED
   WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.

Intellectual Property

   The IETF takes no position regarding the validity or scope of any
   Intellectual Property Rights or other rights that might be claimed to
   pertain to the implementation or use of the technology described in
   this document or the extent to which any license under such rights
   might or might not be available; nor does it represent that it has
   made any independent effort to identify any such rights. Information
   on the procedures with respect to rights in RFC documents can be
   found in BCP 78 and BCP 79.

   Copies of IPR disclosures made to the IETF Secretariat and any
   assurances of licenses to be made available, or the result of an
   attempt made to obtain a general license or permission for the use of
   such proprietary rights by implementers or users of this
   specification can be obtained from the IETF on-line IPR repository at
   http://www.ietf.org/ipr.

   The IETF invites any interested party to bring to its attention any
   copyrights, patents or patent applications, or other proprietary
   rights that may cover technology that may be required to implement
   this standard. Please address the information to the IETF at
   ietf-ipr@ietf.org.












Klensin Standards Track [Page 95]