| Network Working Group | K. Moore |
| Internet-Draft | Network Heretics |
| Updates: 1939, 2595, 3464, 3501, 5068, | C. Newman |
| 6186, 6409 (if approved) | Oracle |
| Intended status: Standards Track | March 13, 2017 |
| Expires: September 14, 2017 |
Mail User Agent Strict Transport Security (MUA-STS)
draft-ietf-uta-email-deep-06
This specification defines a set of requirements and facilities designed to improve email confidentiality between a mail user agent (MUA) and a mail submission or mail access server. This provides mechanisms intended to increase use of already deployed Transport Layer Security (TLS) technology and provides a model for a mail user agent's confidentiality assurance. This enables mail service providers to advertise strict transport security (STS) policies that request MUAs increase confidentiality assurance.
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at http://datatracker.ietf.org/drafts/current/.
Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."
This Internet-Draft will expire on September 14, 2017.
Copyright (c) 2017 IETF Trust and the persons identified as the document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.
Software that provides email service via Internet Message Access Protocol (IMAP) [RFC3501], Post Office Protocol (POP) [RFC1939] and/or Simple Mail Transfer Protocol (SMTP) Submission [RFC6409] usually has Transport Layer Security (TLS) [RFC5246] support but often does not use it in a way that maximizes end-user confidentiality. This specification proposes changes to email software and deployments intended to increase the use of TLS and record when that use occurs. This adapts the strict transport security (STS) model described in [RFC6797] to cover mail user agents (MUAs).
In brief, this memo now recommends that:
This does not address use of TLS with SMTP for message relay (where Message Submission [RFC6409] does not apply). Improved use of TLS with SMTP for message relay requires a different approach. One approach to address that topic is described in [RFC7672].
The recommendations in this memo do not replace the functionality of, and are not intended as a substitute for, end-to-end encryption of electronic mail.
This draft is subject to change. Implementation of this proposal is not recommended at this time. Please discuss this proposal on the ietf-uta mailing list.
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 [RFC2119].
This specification expresses syntax using the Augmented Backus-Naur Form (ABNF) as described in [RFC5234], including the core rules in Appendix B and rules from [RFC5322].
In examples, "C:" and "S:" indicate lines sent by the client and server respectively. If a single "C:" or "S:" label applies to multiple lines, then the line breaks between those lines are for editorial clarity only and are not part of the actual protocol exchange.
A "mail account" refers to the network services an end user uses to read, submit and manage email communications on the Internet. This typically involves at least one mail access server (IMAP or POP) and at least one SMTP submission server. An end user uses a mail user agent (MUA) to access a mail account. (Most MUAs support the ability to access multiple mail accounts.) This document uses the term "confidentiality assurance level" to indicate the degree to which the network connections between an MUA and a mail account have confidentiality protection from both passive and active attackers on the network.
The configuration necessary for a mail account includes an email address, connection information, and authentication credentials for network services. MUAs compliant with this specification MUST also associate a minimum confidentiality assurance level with each mail account. If during a session with a network service, the requirements for the minimum confidentiality assurance level associated with that mail account are not met, the MUA MUST NOT continue the session with the network service. MUAs MUST support at least the ability to detect whether a session with a network service implements confidentiality assurance level 1 as described in the next section. Note that the minimum confidentiality assurance level associated with an account applies to all protocol interactions and all servers associated with the account.
MUAs SHOULD continuously indicate to the user the current confidentiality assurance level of any account currently in use when reading, submitting and managing mail (e.g., via a lock icon, background colors, or other indications similar to those commonly used in web browsers for a similar purpose) and SHOULD indicate the minimum confidentiality assurance level for each account whenever displaying a list of mail accounts. Note that the displayed confidentiality assurance level for a current session could be higher than the minimum confidentiality assurance level set at account configuration, but never lower. If multiple active connections are associated with an account or view, the indication of the current confidentiality assurance level associated with the account should reflect the level provided by the least confidential connection. It is therefore possible that at any given instant some services associated with a mail account meet the minimum confidentiality assurance level associated with the account, and other services do not. An MUA MAY continue to interact with those services for which the minimum confidentiality assurance level is met, while refusing to interact with those services for which the minimum confidentiality assurance level is not met. For example, if the IMAP service associated with a mail account meets the minimum confidentiality assurance level, but the Mail Submission service associated with that account does not, the MUA MAY continue to permit reading mail from that account but MUST NOT send mail until it can do so using a Submission service that meets the minimum confidentiality assurance level for that account.
Account configuration occurs when an MUA is first used to access a particular service, when a user wishes to access or submit mail through servers in addition to those specified or found during first use, or when a user explicitly requests to change account configuration parameters such as server names, user names, passwords, client certificates, etc. Account configuration can be entirely manual (entering server names explicitly) or partially automated via a mechanism such as DNS SRV records [RFC6186]. MUAs SHOULD require a minimum confidentiality assurance level of 1 as the default for newly configured accounts.
This document defines two initial confidentiality assurance levels, 1 and 0. It is expected that other levels may be defined in the future, as needed to thwart increasingly sophisticated and/or pervasive attacks.
A mail account has a confidentiality assurance level of 1 when the following conditions are met on all TCP server connections associated with an account. This includes connections to POP, IMAP and SMTP submission servers as well as any other associated protocols defined now or in the future. Examples of protocols associated with a mail account include managesieve [RFC5804] and MTQP [RFC3887].
The end user is part of the system that protects the user's confidentiality and security. As a result, it's critical not to present the end user with a simple action that reduces their confidentiality in response to certificate validation failure. An MUA which offers a user actions such as "connect anyway", "trust certificate for future connections" or "lower confidentiality assurance for this account" in response to certificate validation failure is not implementing a minimum confidentiality assurance of 1 as defined in this section and thus does not comply with this document. Examples of acceptable actions to offer would be "work offline", "try again later", and "open service provider status web page".
MUAs MAY support the ability to configure accounts with a minimum confidentiality assurance level of 0. At this level, the MUA MUST attempt to negotiate TLS, but MAY ignore server certificate validation failures. MUAs MAY support use of connections without TLS, or using TLS versions prior to TLS 1.1, for accounts with a minimum confidentiality assurance level of 0. Even for accounts with a minimum confidentiality assurance level of 0, MUAs SHOULD attempt TLS first if available, and MUST implement the ability to reconnect without TLS if TLS negotiation fails for reasons other than server certificate validity.
Note that if TLS is not used, or a version of TLS prior to TLS 1.1 is negotiated, or the TLS server certificate is not successfully validated as described in Section 3.1, the client MUST clearly indicate to the user that there is currently no assurance of confidentiality for the mail account or connection.
This specification is not intended to limit experimentation and innovation with respect to user confidentiality. As a result, an implementation MAY implement confidentiality assurance levels other than those defined in this document, as long as those levels are distinguished in user interfaces from those defined in this document, and the ordering associated with them reflects the actual expectation of confidentiality provided. However, implementation of levels below confidentiality assurance level 0, as described in the previous section, is discouraged. Implementers are also cautioned that end users may be confused by too many confidentiality assurance levels.
As stated above, higher confidentiality assurance levels may be standardized in the future. For example, a future confidentiality assurance levels might require multiple independent trust anchors for server certificate validation.
Previous standards for use of email protocols with TLS used the STARTTLS mechanism: [RFC2595], [RFC3207], and [RFC3501]. With STARTTLS, the client establishes a clear text application session and determines whether to issue a STARTTLS command based on server capabilities and client configuration. If the client issues a STARTTLS command, a TLS handshake follows that can upgrade the connection. While this mechanism has been deployed, an alternate mechanism where TLS is negotiated immediately at connection start on a separate port (referred to in this document as "Implicit TLS") has been deployed more successfully. To increase use of TLS, this specification recommends use of implicit TLS by new POP, IMAP and SMTP Submission software.
When a TCP connection is established for the "pop3s" service (default port 995), a TLS handshake begins immediately. Clients MUST implement the certificate validation mechanism described in [RFC7817]. Once the TLS session is established, POP3 [RFC1939] protocol messages are exchanged as TLS application data for the remainder of the TCP connection. After the server sends a +OK greeting, the server and client MUST enter AUTHORIZATION state, even if client credentials were supplied during the TLS handshake.
See Section 11.1.1 for additional information on client certificate authentication. See Section 13.3 for port registration information.
When a TCP connection is established for the "imaps" service (default port 993), a TLS handshake begins immediately. Clients MUST implement the certificate validation mechanism described in [RFC3501] and SHOULD implement the certificate validation mechanism described in [RFC7817]. Once the TLS session is established, IMAP [RFC3501] protocol messages are exchanged as TLS application data for the remainder of the TCP connection. If client credentials were provided during the TLS handshake that the server finds acceptable, the server MAY issue a PREAUTH greeting in which case both the server and client enter AUTHENTICATED state. If the server issues an OK greeting then both server and client enter NOT AUTHENTICATED state.
See Section 11.1.1 for additional information on client certificate authentication. See Section 13.4 for port registration information.
When a TCP connection is established for the "submissions" service (default port 465), a TLS handshake begins immediately. Clients MUST implement the certificate validation mechanism described in [RFC7817]. Once a TLS session is established, message submission protocol data [RFC6409] is exchanged as TLS application data for the remainder of the TCP connection. (Note: the "submissions" service name is defined in section 10.3 of this document, and follows the usual convention that the name of a service layered on top of Implicit TLS consists of the name of the service as used without TLS, with an "s" appended.)
The STARTTLS mechanism on port 587 is relatively widely deployed due to the situation with port 465 (discussed in Section 13.5). This differs from IMAP and POP services where implicit TLS is more widely deployed on servers than STARTTLS. It is desirable to migrate core protocols used by MUA software to implicit TLS over time for consistency as well as the additional reasons discussed in Appendix A. However, to maximize use of encryption for submission it is desirable to support both mechanisms for Message Submission over TLS for a transition period of several years. As a result, clients and servers SHOULD implement both STARTTLS on port 587 and implicit TLS on port 465 for this transition period. Note that there is no significant difference between the security properties of STARTTLS on port 587 and implicit TLS on port 465 if the implementations are correct and both client and server are configured to require successful negotiation of TLS prior to message submission (as required in Section 11.1).
Note that the submissions port provides access to a Mail Submission Agent (MSA) as defined in [RFC6409] so requirements and recommendations for MSAs in that document apply to the submissions port, including the requirement to implement SMTP AUTH [RFC4954].
See Section 11.1.1 for additional information on client certificate authentication. See Section 13.5 for port registration information.
When a client or server wishes to close the connection, it SHOULD initiate the exchange of TLS close alerts before TCP connection termination. The client MAY, after sending a TLS close alert, gracefully close the TCP connection without waiting for a TLS response from the server.
Once an improved email security mechanism is deployed and ready for general use, it is desirable to continue using it for all future email service. For example, TLS is widely deployed in email software, but use of TLS is often not required. At the time this is written, deployed mail user agents (MUAs) [RFC5598] usually make a determination if TLS is available when an account is first configured and may require use of TLS with that account if and only if it was initially available. If the service provider makes TLS available after initial client configuration, many MUAs will not notice the change.
Alternatively, a security feature may be purely opportunistic and thus subject to downgrade attacks. For example, at the time this was written, most TLS stacks that support TLS 1.2 will use an older TLS version if the peer does not support TLS 1.2 and many do so without alerting the user of the reduced security. Thus a variety of active attacks could cause the loss of TLS 1.2 benefits. Only if client policy is upgraded to require TLS 1.2 can the client prevent all downgrade attacks. However, this sort of security policy upgrade will be ignored by most users unless it is automated.
This section describes a mechanism, called "security directives", which is designed to permit an MUA to recognize when a service provider has committed to provide certain server security features, and that it's safe for the client to change its configuration for that account to require that such features be present in future sessions with that server. Once the client has changed the configuration for a mail service to require specific server security features, those features are said to be "latched".
Note that security directives are a separate mechanism from minimum confidentiality assurance levels. A connection between a client and a service MUST meet the requirements of both the minimum confidentiality assurance level associated with the account, and the conditions of any security directives established for that service. Otherwise the client MUST abandon the connection. When an MUA implements both minimum confidentiality assurance levels and security directives, then both the end-user and the service provider independently have the ability to improve the end-user's confidentiality.
directive = directive-name [ "=" directive-value ] directive-name = token directive-value = token token = <As defined in RFC 7230>
A security directive has the following formal syntax: [RFC6797] as revised in [RFC7230]; but simplified for use by protocols other than HTTP.
Servers supporting this extension MUST advertise an STS policy. This includes a list of security directives the server administrator has explicitly configured as recommended for use by clients (the list MAY be empty). When a server advertises a security directive associated with a security facility, it is making a commitment to support that facility (or a revised version of that facility) indefinitely and recommending that the client save that directive with the account configuration and require that security facility for future connections to that server.
sts-policy = [directive *(";" [SP] directive)]
Server STS policy may also include a "sts-url" directive with a value containing an https Uniform Resource Locator (URL) [RFC2818] that the client can save and subsequently resolve for the user in the event of a security connection problem. Server STS policy has the following formal syntax: Section 9.
The IANA Considerations Section 13 defines a registry so that more directives can be defined in the future. Three initial directives are defined for use by MUAs in Section 13.2: tls-version, sts-url, and tls-cert.
Before a client can consider storing any security directives, it MUST verify that the connection to the server uses TLS, the server has been authenticated, and any requirements for any previously saved security directives are met. Then the client performs the following steps for each security directive in the STS policy:
Once a security directive is saved, all subsequent connections to that host require any associated security feature. For this confidentiality protection to work as desired clients MUST NOT offer a click-through-to-connect action when unable to achieve connection security matching the saved security directives.
Suppose a server advertises the "tls-version" directive name with value "1.1". A client that successfully negotiates either TLS 1.1 or TLS 1.2 SHOULD save this directive. The server may subsequently change the value to "1.2". When a client with "1.1" saved value connects and negotiates TLS 1.2, it will upgrade the saved directive value to "1.2". However, a client that only supports TLS 1.1 will continue to require use of TLS 1.1 and work with that server as long as it permits TLS 1.1. This way individual clients can require the newer/stronger protocol (e.g., TLS 1.2), while older clients can continue to communicate securely (albeit potentially less so) using the older protocol.
When a security directive has been saved for connections from a client to a server and the facility identified by that directive is no longer available, this results in a connection failure. An MUA SHOULD inform the user of a potential threat to their confidentiality and offer to resolve a previously-recorded sts-url https URL if one is available. MUAs are discouraged from offering a lightweight option to reset or ignore directives as this defeats the benefit they provide to end users.
tls-cipher-clause = CFWS "tls" FWS tls-cipher
tls-cipher = tls-cipher-suite-name / tls-cipher-suite-hex
tls-cipher-name = ALPHA *(ALPHA / DIGIT / "_")
; as registered in IANA cipher suite registry
tls-cipher-hex = "0x" 4HEXDIG
The ESMTPS transmission type [RFC3848] provides trace information that can indicate TLS was used when transferring mail. However, TLS usage by itself is not a guarantee of confidentiality or security. The TLS cipher suite provides additional information about the level of security made available for a connection. This defines a new SMTP "tls" Received header additional-registered-clause that is used to record the TLS cipher suite that was negotiated for the connection. The value included in this additional clause SHOULD be the registered cipher suite name (e.g., TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) included in the TLS cipher suite registry. In the event the implementation does not know the name of the cipher suite (a situation that should be remedied promptly), a four-digit hexadecimal cipher suite identifier MAY be used. The ABNF for the field follows:
This memo defines optional mechanisms for use by MUAs to communicate saved STS policy to servers and for servers to advertise policy. One purpose of such mechanisms is to permit servers to determine which and how many clients have saved security directives, and thus, to permit operators to be aware of potential impact to their users should support for such facilities be changed. For IMAP, the existing ID command is extended to provide this capability. For SMTP Submission, a new CLIENT command is defined. No similar mechanism is defined for POP in this version of the memo to keep POP simpler, but one may be added in the future if deemed necessary.
In addition, for each of IMAP, POP, and SMTP, a new STS capability is defined so the client can access the server's STS policy.
When an IMAP server advertises the STS capability, that indicates the IMAP server implements IMAP4 ID [RFC2971] with additional field values defined here. This is grouped with the ID command because that is the existing IMAP mechanism for clients to report data for server logging, and provides a way for the server to report the STS policy.
IMAP clients SHOULD use the IMAP ID command to report policy failures and determine the server STS policy. Clients MAY use the ID command to report other security directive information. IMAP servers MUST implement the ID command at least to report STS policy to clients.
<client connected to port 993 and negotiated TLS successfully>
S: * OK [CAPABILITY IMAP4rev1 STS ID AUTH=PLAIN
AUTH=SCRAM-SHA-1] hello
C: a001 ID ("name" "Demo Mail" "version" "1.5" "saved"
"tls-version=1.1; tls-cert"
"directives" "tls-version=1.2")
S: * ID ("name" "Demo Server" "version" "1.7" "sts-policy"
"tls-version=1.1; tls-cert;
sts-url=https://www.example.com/security-support.html")
S: a001 OK ID completed
Example 1
This example shows a client that successfully negotiated TLS version 1.1 or later and verified the server's certificate as required by IMAP. Even if the client successfully validates the server certificate, it will not require tls-version 1.2 in the future as the server does not advertise that version as policy. The client has not yet saved an STS URL, but if the client successfully validated the server certificate, it will save the provided URL.
<client connected to port 993 and negotiated TLS successfully>
S: * OK [CAPABILITY IMAP4rev1 DEEP ID AUTH=PLAIN
AUTH=SCRAM-SHA-1] hello
C: a001 ID ("name" "Demo Mail" "version" "1.5" "policy-failure"
"tls-cert=pkix")
S: * ID ("name" "Demo Server" "version" "1.7" "sts-policy"
"tls-version=1.1;
sts-url=<https://www.example.com/security-support.html>")
S: a001 OK ID completed
C: a002 LOGOUT
Example 2
This example shows a client that negotiated TLS, but was unable to verify the server's certificate using PKIX. The policy-failure informs the server of this problem, at which point the client can disconnect. If the client had previously saved the sts-url security directive from this server, it could offer to resolve that URI. However, the sts-policy in this exchange is ignored due to the failure to meet the conditions of the tls-version security directive.
<IMAP Proxy connected over private network on port 143, there is
a client connected to the proxy on port 993 that negotiated TLS>
S: * OK [CAPABILITY IMAP4rev1 DEEP ID AUTH=PLAIN
AUTH=SCRAM-SHA-1] hello
C: a001 ID ("name" "Demo Mail" "version" "1.5" "saved"
"tls-version=1.1; tls-cert=pkix"
"tls" "TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256")
S: * ID ("name" "Demo Server" "version" "1.7" "sts-policy"
"tls-version=1.1; tls-cert=pkix;
sts-url=https://www.example.com/support.html")
S: a001 OK ID completed
Example 3
This example shows the connection from an IMAP proxy to a back-end server. The client connected to the proxy and sent the ID command shown in example 1, and the proxy has added the "tls" item to the ID command so the back-end server can log the cipher suite that was used on the connection from the client.
POP servers supporting this specification MUST implement the POP3 extension mechanism [RFC2449]. POP servers MUST advertise the DEEP capability with an argument indicating the server's DEEP status. (Note: DEEP is an ancronym for the original name of this specification, before the terms were changed to align better with those used in HSTS.)
<client connected to port 995 and negotiated TLS successfully>
S: +OK POP server ready
C: CAPA
S: +OK Capability list follows
S: TOP
S: SASL PLAIN SCRAM-SHA-1
S: RESP-CODES
S: PIPELINING
S: UIDL
S: STS tls-version=1.2
sts-url=<https://www.example.com/security-support.html>
S: .
Example 4
After verifying the TLS server certificate and issuing CAPA, the client can save any or all of the STS policy. If the client connects to this same server later and has a security failure, the client can direct the user's browser to the previously-saved URL where the service provider can provide advice to the end user.
SMTP Submission servers supporting this specification MUST implement the MSTS SMTP extension. The name of this extension is MSTS. The EHLO keyword value is MSTS and the sts-policy ABNF is the syntax of the EHLO keyword parameters. This does not add parameters to the MAIL FROM or RCPT TO commands. This also adds a CLIENT command to SMTP which is used to report client information to the server. The formal syntax for the command follows:
deep-cmd = "CLIENT" 1*(SP deep-parameter)
deep-parameter = name / version / policy-fail
/ directives / tls / future-extension
name = "name" SP esmtp-value
version = "version" SP esmtp-value
saved = "saved" SP directive-list
policy-fail = "policy-fail" SP directive-list
directive-list = DQUOTE [directive
*(";" [SP] directive)] DQUOTE
directives = "directives" SP directive-list
tls = "tls" SP tls-cipher
future-extension = Atom SP String
Atom = <as defined in RFC 5321>
String = <as defined in RFC 5321>
<client connected to port 465 and negotiated TLS successfully>
S: 220 example.com Demo SMTP Submission Server
C: EHLO client.example.com
S: 250-example.com
S: 250-8BITMIME
S: 250-PIPELINING
S: 250-DSN
S: 250-AUTH PLAIN LOGIN
S: 250-MSTS tls-version=1.2; tls-cert;
sts-url=<https://www.example.com/status.html>
S: 250-BURL imap
S: 250 SIZE 0
C: CLIENT name demo_submit version 1.5 saved "tls-version=1.1;
tls-cert=pkix+dane" directives "tls-version=1.2"
S: 250 OK
Example 5
The CLIENT command parameters listed here have the same meaning as the parameters used in the IMAP STS extension (Section 9.1). The server responds to the CLIENT command with a "250" if the command has correct syntax and a "501" if the command has incorrect syntax.
This section updates [RFC6186] by changing the preference rules and adding a new SRV service label _submissions._tcp to refer to Message Submission with implicit TLS.
User-configurable MUAs SHOULD support use of [RFC6186] for account setup. However, when using configuration information obtained by this method, MUAs SHOULD default to a minimum confidentiality assurance level of 1, unless the user has explicitly requested reduced confidentiality. This will have the effect of causing the MUA to ignore advertised configurations that do not support TLS, even when those advertised configurations have a higher priority than other advertised configurations.
When using [RFC6186] configuration information, Mail User Agents SHOULD NOT automatically establish new configurations that do not require TLS for all servers, unless there are no advertised configurations using TLS. If such a configuration is chosen, prior to attempting to authenticate to the server or use the server for message submission, the MUA SHOULD warn the user that traffic to that server will not be encrypted and that it will therefore likely be intercepted by unauthorized parties. The specific wording is to be determined by the implementation, but it should adequately capture the sense of risk given the widespread incidence of mass surveillance of email traffic.
When establishing a new configuration for connecting to an IMAP, POP, or SMTP Submission server, an MUA SHOULD NOT blindly trust SRV records unless they are signed by DNSSEC and have a valid signature. Instead, the MUA SHOULD warn the user that the DNS-advertised mechanism for connecting to the server is not authenticated, and request the user to manually verify the connection details by reference to his or her mail service provider's documentation.
Similarly, an MUA MUST NOT consult SRV records to determine which servers to use on every connection attempt, unless those SRV records are signed by DNSSEC and have a valid signature. However, an MUA MAY consult SRV records from time to time to determine if an MSP's server configuration has changed, and alert the user if it appears that this has happened. This can also serve as a means to encourage users to upgrade their configurations to require TLS if and when their MSPs support it.
During account setup, the MUA will identify servers that provide account services such as mail access and mail submission (the previous section describes one way to do this). The certificates for these servers are verified using the rules described in [RFC7817] and PKIX [RFC5280]. In the event the certificate does not validate due to an expired certificate, lack of appropriate chain of trust or lack of identifier match, the MUA MAY create a persistent binding between that certificate and the saved host name for the server. This is called certificate pinning. Certificate pinning is only appropriate during account setup and MUST NOT be offered in response to a failed certificate validation for an existing account. An MUA that allows certificate pinning MUST NOT allow a certificate pinned for one account to validate connections for other accounts.
A pinned certificate is subject to a man-in-the-middle attack at account setup time, and lacks a mechanism to revoke or securely refresh the certificate. Therefore use of a pinned certificate does not meet the requirement for a minimum confidentiality assurance level of 1, and an MUA MUST NOT indicate a confidentiality assurance level of 1 for an account or connection using a pinned certificate. Additional advice on certificate pinning is present in [RFC6125].
This section details requirements for implementations of electronic mail protocol clients and servers. A requirement for a client or server implementation to support a particular feature is not the same thing as a requirement that a client or server running a conforming implementation be configured to use that feature. Requirements for Mail Service Providers (MSPs) are distinct from requirements for protocol implementations, and are listed in a separate section.
These requirements apply to MUAs as well as POP, IMAP and SMTP Submission servers.
MUAs and mail servers MAY implement client certificate authentication on the implicit TLS port. Servers MUST NOT request a client certificate during the TLS handshake unless the server is configured to accept some client certificates as sufficient for authentication and the server has the ability to determine a mail server authorization identity matching such certificates. How to make this determination is presently implementation specific. Clients MUST NOT provide a client certificate during the TLS handshake unless the server requests one and the client has determined the certificate can be safely used with that specific server, OR the client has been explicitly configured by the user to use that particular certificate with that server. How to make this determination is presently implementation specific. If the server accepts the client's certificate as sufficient for authorization, it MUST enable the SASL EXTERNAL [RFC4422] mechanism. An IMAPS server MAY issue a PREAUTH greeting instead of enabling SASL EXTERNAL. A client supporting client certificate authentication with implicit TLS MUST implement the SASL EXTERNAL [RFC4422] mechanism using the appropriate authentication command (AUTH for POP3 [RFC5034], AUTH for SMTP Submission [RFC4954], AUTHENTICATE for IMAP [RFC3501]).
These requirements apply to servers that implement POP, IMAP or SMTP Submission.
This section describes requirements on Mail User Agents (MUAs) using IMAP, POP, and/or Submission protocols. Note: Requirements pertaining to use of Submission servers are also applicable when using SMTP servers (e.g., port 25) for mail submission.
MUAs which are not configurable to use user-specified servers MUST implement TLS or similarly other strong encryption mechanism when communicating with their mail servers. This generally applies to MUAs that are pre-configured to operate with one or more specific services, whether or not supplied by the vendor of those services.
MUAs using protocols other than IMAP, POP, and Submission to communicate with mail servers, MUST implement TLS or other similarly robust encryption mechanism in conjunction with those protocols.
There are multiple ways to connect an Anti-Virus and/or Anti-Spam (AVAS) service to a mail server. Some mechanisms, such as the de-facto milter protocol, are out of scope for this specification. However, some services use an SMTP relay proxy that intercepts mail at the application layer to perform a scan and proxy or forward to another MTA. Deploying AVAS services in this way can cause many problems [RFC2979] including direct interference with this specification, and other forms of confidentiality or security reduction. An AVAS product or service is considered compliant with this specification if all IMAP, POP and SMTP-related software (including proxies) it includes are compliant with this specification, and each of these services advertise and support all security directives that the actual end-servers advertise.
Note that end-to-end email encryption prevents AVAS software and services from using email content as part of a spam or virus assessment. Furthermore, while a minimum confidentiality assurance level of 1 or better can prevent a man-in-the-middle from introducing spam or virus content between the MUA and Submission server, it does not prevent other forms of client or account compromise. Use of AVAS services for submitted email therefore remains necessary.
This section details requirements for providers of IMAP, POP, and/or SMTP submission services, for providers who claim to conform to this specification.
Mail Service Providers MUST use server implementations that conform to this specification.
This document updates the advice in [RFC5068] by making Implicit TLS on port 465 the preferred submission port.
Mail Service Providers that accept mail submissions from end-users using the Internet Protocol MUST provide one or more SMTP Submission services, separate from the SMTP MTA services used to process incoming mail. Those submission services MUST be configured to support Implicit TLS on port 465 and SHOULD support STARTTLS if port 587 is used.
MSPs MAY also support submission of messages via one or more designated SMTP servers to facilitate compatibility with legacy MUAs.
Discussion: SMTP servers used to accept incoming mail or to relay mail are expected to accept mail in cleartext. This is incompatible with the purpose of this memo which is to encourage encryption of traffic between mail servers. There is no such requirement for mail submission servers to accept mail in cleartext or without authentication. For other reasons, use of separate SMTP submission servers has been best practice for many years.
MSPs MUST maintain valid server certificates for all servers. See [RFC7817] for the recommendations and requirements necessary to achieve this.
If a protocol server provides service for more than one mail domain, it MAY use a separate IP address for each domain and/or a server certificate that advertises multiple domains. This will generally be necessary unless and until it is acceptable to impose the constraint that the server and all clients support the Server Name Indication extension to TLS [RFC6066]. For more discussion of this problem, see section 5.1 of [RFC7817].
This section discusses not only the DNS records that are recommended, but also implications of DNS records for server configuration and TLS server certificates.
It is recommended that MSPs advertise MX records for handling of inbound mail (instead of relying entirely on A or AAAA records), and that those MX records be signed using DNSSEC. This is mentioned here only for completeness, as handling of inbound mail is out of scope for this document.
MSPs SHOULD advertise SRV records to aid MUAs in determination of proper configuration of servers, per the instructions in [RFC6186].
MSPs SHOULD advertise servers that support Implicit TLS in preference to those which support cleartext and/or STARTTLS operation.
All DNS records advertised by an MSP as a means of aiding clients in communicating with the MSP's servers, SHOULD be signed using DNSSEC.
MSPs SHOULD advertise TLSA records to provide an additional trust anchor for public keys used in TLS server certificates. However, TLSA records MUST NOT be advertised unless they are signed using DNSSEC.
MSPs SHOULD regularly and frequently monitor their various servers to make sure that: TLS server certificates remain valid and are not about to expire, TLSA records match the public keys advertised in server certificates, are signed using DNSSEC, server configurations are consistent with SRV advertisements, and DNSSEC signatures are valid and verifiable. Failure to detect expired certificates and DNS configuration errors in a timely fashion can result in significant loss of service for an MSP's users and a significant support burden for the MSP.
MSPs SHOULD advertise STS policies that include at least tls11, tls-cert and sts-url, with the latter having an associated https URL that can be used to inform clients of service outages or problems impacting client confidentiality. Note that advertising tls-cert is a commitment to maintain and renew server certificates. A MSP MAY also specifically indicate a commitment to support PKIX validation, DANE validation, or both, using tls-cert=pkix, tls-cert=dane, or tls-cert=pkix+dane, respectively.
New servers and services SHOULD be configured to require TLS unless it's necessary to support legacy clients or existing client configurations.
When an MSP changes the Internet Facing Servers providing mail access and mail submission services, including SMTP-based spam/virus filters, it is generally necessary to support the same and/or a newer version of TLS and the same security directives that were previously advertised.
IANA shall create (has created) the registry "STS Security Directives". This registry is a single table and will use an expert review process [RFC5226]. Each registration will contain the following fields:
The expert reviewer will verify the directive name follows the ABNF, and that the value and description fields are clear, unambiguous, do not overlap existing deployed technology, do not create security problems and appropriately considers interoperability issues. Email security directives intended for LIMITED USE have a lower review bar (interoperability and overlap issues are less of a concern). The reviewer may approve a registration, reject for a stated reason or recommend the proposal have standards track review due to importance or difficult subtleties.
Standards-track registrations may be updated if the relevant standards are updated as a consequence of that action. Non-standards-track entries may be updated by the listed change controller. The entry's name and submitter may not be changed. In exceptional cases, any aspect of any registered entity may be updated at the direction of the IESG (for example, to correct a conflict).
This document defines three initial security directives for the registry as follows, and registers the two additional directives specified in [RFC6797].
Service Name: pop3s Transport Protocol: TCP Assignee: IETF <iesg@ietf.org> Contact: IESG <iesg@ietf.org> Description: POP3 over TLS protocol Reference: RFC XXXX (this document once published) Port Number: 995
IANA is asked to update the registration of the TCP well-known port 995 using the following template ([RFC6335]):
Service Name: imaps Transport Protocol: TCP Assignee: IETF <iesg@ietf.org> Contact: IESG <iesg@ietf.org> Description: IMAP over TLS protocol Reference: RFC XXXX (this document once published) Port Number: 993
IANA is asked to update the registration of the TCP well-known port 993 using the following template ([RFC6335]):
Service Name: submissions Transport Protocol: TCP Assignee: IETF <iesg@ietf.org> Contact: IESG <iesg@ietf.org> Description: Message Submission over TLS protocol Reference: RFC XXXX (this document once published) Port Number: 465
IANA is asked to assign an alternate usage of port 465 in addition to the current assignment using the following template ([RFC6335]):
This is a one time procedural exception to the rules in RFC 6335. This requires explicit IESG approval and does not set a precedent. Historically, port 465 was briefly registered as the "smtps" port. This registration made no sense as the SMTP transport MX infrastructure has no way to specify a port so port 25 is always used. As a result, the registration was revoked and was subsequently reassigned to a different service. In hindsight, the "smtps" registration should have been renamed or reserved rather than revoked. Unfortunately, some widely deployed mail software interpreted "smtps" as "submissions" [RFC6409] and used that port for email submission by default when an end-user requests security during account setup. If a new port is assigned for the submissions service, email software will either continue with unregistered use of port 465 (leaving the port registry inaccurate relative to de-facto practice and wasting a well-known port), or confusion between the de-facto and registered ports will cause harmful interoperability problems that will deter use of TLS for message submission. The authors believe both of these outcomes are less desirable than a wart in the registry documenting real-world usage of a port for two purposes. Although STARTTLS-on-port-587 has deployed, it has not replaced deployed use of implicit TLS submission on port 465.
This document adds the STS capability to the IMAP capabilities registry. This is described in Section 9.1.
This document adds the STS capability to the POP3 capabilities registry.
This document adds the MSTS EHLO Keyword to the SMTP Service Extension registry. This is described in Section 9.3.
This document adds the following entry to the "Additional-registered-clauses" sub-registry of the "MAIL Parameters" registry, created by [RFC5321]:
This entire document is about security considerations. In general, this is targeted to improve mail confidentiality and to mitigate threats external to the email system such as network-level snooping or interception; this is not intended to mitigate active attackers who have compromised service provider systems.
It could be argued that sharing the name and version of the client software with the server has privacy implications. Although providing this information is not required, it is encouraged so that mail service providers can more effectively inform end-users running old clients that they need to upgrade to protect their security, or know which clients to use in a test deployment prior to upgrading a server to have higher security requirements.
This section is not normative.
The first version of this was written independently from draft-moore-email-tls-00.txt; subsequent versions merge ideas from both drafts.
One author of this document was also the author of RFC 2595 that became the standard for TLS usage with POP and IMAP, and the other author was perhaps the first to propose that idea. In hindsight both authors now believe that that approach was a mistake. At this point the authors believe that while anything that makes it easier to deploy TLS is good, the desirable end state is that these protocols always use TLS, leaving no need for a separate port for cleartext operation except to support legacy clients while they continue to be used. The separate port model for TLS is inherently simpler to implement, debug and deploy. It also enables a "generic TLS load-balancer" that accepts secure client connections for arbitrary foo-over-TLS protocols and forwards them to a server that may or may not support TLS. Such load-balancers cause many problems because they violate the end-to-end principle and the server loses the ability to log security-relevant information about the client unless the protocol is designed to forward that information (as this specification does for the cipher suite). However, they can result in TLS deployment where it would not otherwise happen which is a sufficiently important goal that it overrides the problems.
Although STARTTLS appears only slightly more complex than separate-port TLS, we again learned the lesson that complexity is the enemy of security in the form of the STARTTLS command injection vulnerability (CERT vulnerability ID #555316). Although there's nothing inherently wrong with STARTTLS, the fact it resulted in a common implementation error (made independently by multiple implementers) suggests it is a less secure architecture than Implicit TLS.
Section 7 of RFC 2595 critiques the separate-port approach to TLS. The first bullet was a correct critique. There are proposals in the http community to address that, and use of SRV records as described in RFC 6186 resolves that critique for email. The second bullet is correct as well, but not very important because useful deployment of security layers other than TLS in email is small enough to be effectively irrelevant. The third bullet is incorrect because it misses the desirable option of "use and latch-on TLS if available". The fourth bullet may be correct, but is not a problem yet with current port consumption rates. The fundamental error was prioritizing a perceived better design based on a mostly valid critique over real-world deployability. But getting security and confidentiality facilities actually deployed is so important it should trump design purity considerations.
Port 465 is presently used for two purposes: for submissions by a large number of clients and service providers and for the "urd" protocol by one vendor. Actually documenting this current state is controversial as discussed in the IANA considerations section. However, there is no good alternative. Registering a new port for submissions when port 465 is widely used for that purpose already will just create interoperability problems. Registering a port that's only used if advertised by an SRV record (RFC 6186) would not create interoperability problems but would require all client and server deployments and software to change significantly which is contrary to the goal of promoting more TLS use. Encouraging use of STARTTLS on port 587 would not create interoperability problems, but is unlikely to have impact on current undocumented use of port 465 and makes the guidance in this document less consistent. The remaining option is to document the current state of the world and support future use of port 465 for submission as this increases consistency and ease-of-deployment for TLS email submission.
Changes since draft-ietf-uta-email-deep-05:
Changes since draft-ietf-uta-email-deep-04:
Changes since draft-ietf-uta-email-deep-03:
Changes since draft-ietf-uta-email-deep-02:
Changes since draft-ietf-uta-email-deep-01:
Changes since draft-ietf-uta-email-deep-00:
Changes since draft-newman-email-deep-02:
Changes since -01:
Changes since -00:
Thanks to Ned Freed for discussion of the initial latch concepts in this document. Thanks to Alexey Melnikov for draft-melnikov-pop3-over-tls-02, which was the basis of the POP3 implicit TLS text. Thanks to Russ Housley, Alexey Melnikov and Dan Newman for review feedback. Thanks to Paul Hoffman for interesting feedback in initial conversations about this idea.