Network Working Group J. Hodges Internet-Draft PayPal Intended status: Standards Track C. Jackson Expires: July 31, 2012 Carnegie Mellon University A. Barth Google, Inc. January 28, 2012 HTTP Strict Transport Security (HSTS) draft-ietf-websec-strict-transport-sec-04 Abstract This specification defines a mechanism enabling Web sites to declare themselves accessible only via secure connections, and/or for users to be able to direct their user agent(s) to interact with given sites only over secure connections. This overall policy is referred to as HTTP Strict Transport Security (HSTS). The policy is declared by Web sites via the Strict-Transport-Security HTTP response header field, and/or by other means, such as user agent configuration, for example. Status of this Memo 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 July 31, 2012. Copyright Notice Copyright (c) 2012 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 Hodges, et al. Expires July 31, 2012 [Page 1] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 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. Table of Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.1. Organization of this specification . . . . . . . . . . . . 5 2. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.1. Use Cases . . . . . . . . . . . . . . . . . . . . . . . . 5 2.2. HTTP Strict Transport Security Policy Effects . . . . . . 6 2.3. Threat Model . . . . . . . . . . . . . . . . . . . . . . . 6 2.3.1. Threats Addressed . . . . . . . . . . . . . . . . . . 6 2.3.1.1. Passive Network Attackers . . . . . . . . . . . . 6 2.3.1.2. Active Network Attackers . . . . . . . . . . . . . 7 2.3.1.3. Web Site Development and Deployment Bugs . . . . . 7 2.3.2. Threats Not Addressed . . . . . . . . . . . . . . . . 8 2.3.2.1. Phishing . . . . . . . . . . . . . . . . . . . . . 8 2.3.2.2. Malware and Browser Vulnerabilities . . . . . . . 8 2.4. Requirements . . . . . . . . . . . . . . . . . . . . . . . 8 2.4.1. Overall Requirement . . . . . . . . . . . . . . . . . 8 2.4.1.1. Detailed Core Requirements . . . . . . . . . . . . 8 2.4.1.2. Detailed Ancillary Requirements . . . . . . . . . 9 3. Conformance Criteria . . . . . . . . . . . . . . . . . . . . . 10 3.1. Document Conventions . . . . . . . . . . . . . . . . . . . 10 4. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 10 5. HSTS Mechanism Overview . . . . . . . . . . . . . . . . . . . 12 6. Syntax . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 6.1. Strict-Transport-Security HTTP Response Header Field . . . 13 6.1.1. The max-age Directive . . . . . . . . . . . . . . . . 14 6.1.2. The includeSubDomains Directive . . . . . . . . . . . 14 6.2. Examples . . . . . . . . . . . . . . . . . . . . . . . . . 14 7. Server Processing Model . . . . . . . . . . . . . . . . . . . 15 7.1. HTTP-over-Secure-Transport Request Type . . . . . . . . . 15 7.2. HTTP Request Type . . . . . . . . . . . . . . . . . . . . 15 8. User Agent Processing Model . . . . . . . . . . . . . . . . . 16 8.1. Strict-Transport-Security Response Header Field Processing . . . . . . . . . . . . . . . . . . . . . . . . 16 8.1.1. Noting a HSTS Host . . . . . . . . . . . . . . . . . . 17 8.1.2. Known HSTS Host Domain Name Matching . . . . . . . . . 18 8.2. URI Loading and Port Mapping . . . . . . . . . . . . . . . 19 8.3. Errors in Secure Transport Establishment . . . . . . . . . 19 8.4. HTTP-Equiv Element Attribute . . . . . . . . . . . 19 8.5. Interstitially Missing Strict-Transport-Security Response Header Field . . . . . . . . . . . . . . . . . . 20 Hodges, et al. Expires July 31, 2012 [Page 2] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 9. Domain Name IDNA-Canonicalization . . . . . . . . . . . . . . 20 10. Server Implementation and Deployment Advice . . . . . . . . . 20 10.1. HSTS Policy expiration time considerations . . . . . . . . 21 10.2. Using HSTS in conjunction with self-signed public-key certificates . . . . . . . . . . . . . . . . . . . . . . . 21 10.3. Implications of includeSubDomains . . . . . . . . . . . . 22 11. User Agent Implementation Advice . . . . . . . . . . . . . . . 23 11.1. No User Recourse . . . . . . . . . . . . . . . . . . . . . 23 11.2. User-declared HSTS Policy . . . . . . . . . . . . . . . . 23 11.3. HSTS Pre-Loaded List . . . . . . . . . . . . . . . . . . . 24 11.4. Disallow Mixed Security Context . . . . . . . . . . . . . 24 11.5. HSTS Policy Deletion . . . . . . . . . . . . . . . . . . . 24 12. Constructing an Effective Request URI . . . . . . . . . . . . 24 12.1. ERU Fundamental Definitions . . . . . . . . . . . . . . . 25 12.2. Determining the Effective Requrest URI . . . . . . . . . . 25 12.2.1. Effective Requrest URI Examples . . . . . . . . . . . 26 13. Internationalized Domain Names for Applications (IDNA): Dependency and Migration . . . . . . . . . . . . . . . . . . . 26 14. Security Considerations . . . . . . . . . . . . . . . . . . . 26 14.1. Ramifications of HSTS Policy Establishment only over Error-free Secure Transport . . . . . . . . . . . . . . . 27 14.2. The Need for includeSubDomains . . . . . . . . . . . . . . 28 14.3. Denial of Service . . . . . . . . . . . . . . . . . . . . 28 14.4. Bootstrap MITM Vulnerability . . . . . . . . . . . . . . . 29 14.5. Network Time Attacks . . . . . . . . . . . . . . . . . . . 30 14.6. Bogus Root CA Certificate Phish plus DNS Cache Poisoning Attack . . . . . . . . . . . . . . . . . . . . . 30 14.7. Creative Manipulation of HSTS Policy Store . . . . . . . . 30 14.8. Internationalized Domain Names . . . . . . . . . . . . . . 30 15. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 31 16. References . . . . . . . . . . . . . . . . . . . . . . . . . . 31 16.1. Normative References . . . . . . . . . . . . . . . . . . . 31 16.2. Informative References . . . . . . . . . . . . . . . . . . 32 Appendix A. Design Decision Notes . . . . . . . . . . . . . . . . 35 Appendix B. Differences between HSTS Policy and Same-Origin Policy . . . . . . . . . . . . . . . . . . . . . . . 36 Appendix C. Acknowledgments . . . . . . . . . . . . . . . . . . . 37 Appendix D. Change Log . . . . . . . . . . . . . . . . . . . . . 37 D.1. For draft-ietf-websec-strict-transport-sec . . . . . . . . 37 D.2. For draft-hodges-strict-transport-sec . . . . . . . . . . 41 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 42 Hodges, et al. Expires July 31, 2012 [Page 3] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 1. Introduction [ Please discuss this draft on the WebSec@ietf.org mailing list [WEBSEC]. ] The HTTP protocol [RFC2616] may be used over various transports, typically the Transmission Control Protocol (TCP) [RFC0793]. However, TCP does not provide channel integrity protection, confidentiality, nor secure host identification. Thus the Secure Sockets Layer (SSL) protocol [I-D.ietf-tls-ssl-version3] and its successor Transport Layer Security (TLS) [RFC5246], were developed in order to provide channel-oriented security, and are typically layered between application protocols and TCP. [RFC2818] specifies how HTTP is layered onto TLS, and defines the Uniform Resource Identifier (URI) scheme of "https" (in practice however, HTTP user agents (UAs) typically offer their users choices among SSL2, SSL3, and TLS for secure transport). UAs employ various local security policies with respect to the characteristics of their interactions with web resources depending on (in part) whether they are communicating with a given web resource's host using HTTP or HTTP-over-a-Secure-Transport. For example, cookies ([RFC6265]) may be flagged as Secure. UAs are to send such Secure cookies to their addressed host only over a secure transport. This is in contrast to non-Secure cookies, which are returned to the host regardless of transport (although modulo other rules). UAs typically annunciate to their users any issues with secure connection establishment, such as being unable to validate a TLS server certificate trust chain, or if a TLS server certificate is expired, or if a TLS host's domain name appears incorrectly in the TLS server certificate (see section 3.1 of [RFC2818]). Often, UAs enable users to elect to continue to interact with a web resource's host in the face of such issues. This behavior is sometimes referred to as "click(ing) through" security [GoodDhamijaEtAl05] [SunshineEgelmanEtAl09], and thus can be described as "click-through insecurity". A key vulnerability enabled by click-through insecurity is the leaking of any cookies the web resource may be using to manage a user's session. The threat here is that an attacker could obtain the cookies and then interact with the legitimate web resource while impersonating the user. Jackson and Barth proposed an approach, in [ForceHTTPS], to enable web resources to declare that any interactions by UAs with the web resource must be conducted securely, and that any issues with establishing a secure transport session are to be treated as fatal Hodges, et al. Expires July 31, 2012 [Page 4] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 and without direct user recourse. The aim is to prevent click- through insecurity, and address other potential threats. This specification embodies and refines the approach proposed in [ForceHTTPS]. For example, rather than using a cookie to convey policy from a web resource's host to a UA, it defines an HTTP response header field for this purpose. Additionally, a web resource's host may declare its policy to apply to the entire domain name subtree rooted at its host name. This enables HSTS to protect so-called "domain cookies", which are applied to all subdomains of a given web resource's host name. This specification also incorporates notions from [JacksonBarth2008] in that policy is applied on an "entire-host" basis: it applies to all TCP ports of the issuing host. Note that the policy defined by this specification is distinctly different than the "same-origin policy" defined in "The Web Origin Concept" [RFC6454]. These differences are summarized below in Appendix B. 1.1. Organization of this specification This specification begins with an overview of the use cases, policy effects, threat models, and requirements for HSTS (in Section 2). Then, Section 3 defines conformance requirements. The HSTS mechanism itself is formally specified in Section 4 through Section 15. 2. Overview This section discusses the use cases, summarizes the HTTP Strict Transport Security (HSTS) policy, and continues with a discussion of the threat model, non-addressed threats, and derived requirements. 2.1. Use Cases The high-level use case is a combination of: o Web browser user wishes to interact with various web sites (some arbitrary, some known) in a secure fashion. o Web site deployer wishes to offer their site in an explicitly secure fashion for both their own, as well as their users', benefit. Hodges, et al. Expires July 31, 2012 [Page 5] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 2.2. HTTP Strict Transport Security Policy Effects The effects of the HTTP Strict Transport Security (HSTS) Policy, as applied by a UA in interactions with a web resource host wielding such policy (known as a HSTS Host), are summarized as follows: 1. All insecure ("http") connections to any TCP ports on a HSTS Host are redirected by the HSTS Host to be secure connections ("https"). 2. The UA terminates any secure transport connection attempts upon any and all secure transport errors or warnings, including those caused by a web application presenting self-signed certificates. 3. UAs transform insecure URI references to a HSTS Host into secure URI references before dereferencing them. 2.3. Threat Model HSTS is concerned with three threat classes: passive network attackers, active network attackers, and imperfect web developers. However, it is explicitly not a remedy for two other classes of threats: phishing and malware. Addressed and not addressed threats are briefly discussed below. Readers may wish refer to [ForceHTTPS] for details as well as relevant citations. 2.3.1. Threats Addressed 2.3.1.1. Passive Network Attackers When a user browses the web on a local wireless network (e.g., an 802.11-based wireless local area network) a nearby attacker can possibly eavesdrop on the user's unencrypted Internet Protocol-based connections, such as HTTP, regardless of whether or not the local wireless network itself is secured [BeckTews09]. Freely available wireless sniffing toolkits (e.g., [Aircrack-ng]) enable such passive eavesdropping attacks, even if the local wireless network is operating in a secure fashion. A passive network attacker using such tools can steal session identifiers/cookies and hijack the user's web session(s), by obtaining cookies containing authentication credentials [ForceHTTPS]. For example, there exist widely-available tools, such as Firesheep (a Firefox extension) [Firesheep], which enable their wielder to obtain other local users' session cookies for various web applications. To mitigate such threats, some Web sites support, but usually do not force, access using end-to-end secure transport -- e.g., signaled through URIs constructed with the "https" scheme [RFC2818]. This can Hodges, et al. Expires July 31, 2012 [Page 6] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 lead users to believe that accessing such services using secure transport protects them from passive network attackers. Unfortunately, this is often not the case in real-world deployments as session identifiers are often stored in non-Secure cookies to permit interoperability with versions of the service offered over insecure transport ("Secure cookies" are those cookies containing the "Secure" attribute [RFC6265]). For example, if the session identifier for a web site (an email service, say) is stored in a non- Secure cookie, it permits an attacker to hijack the user's session if the user's UA makes a single insecure HTTP request to the site. 2.3.1.2. Active Network Attackers A determined attacker can mount an active attack, either by impersonating a user's DNS server or, in a wireless network, by spoofing network frames or offering a similarly-named evil twin access point. If the user is behind a wireless home router, an attacker can attempt to reconfigure the router using default passwords and other vulnerabilities. Some sites, such as banks, rely on end-to-end secure transport to protect themselves and their users from such active attackers. Unfortunately, browsers allow their users to easily opt-out of these protections in order to be usable for sites that incorrectly deploy secure transport, for example by generating and self-signing their own certificates (without also distributing their CA certificate to their users' browsers). 2.3.1.3. Web Site Development and Deployment Bugs The security of an otherwise uniformly secure site (i.e. all of its content is materialized via "https" URIs), can be compromised completely by an active attacker exploiting a simple mistake, such as the loading of a cascading style sheet or a SWF movie over an insecure connection (both cascading style sheets and SWF movies can script the embedding page, to the surprise of many web developers, plus some browsers do not issue so-called "mixed content warnings" when SWF files are embedded via insecure connections). Even if the site's developers carefully scrutinize their login page for "mixed content", a single insecure embedding anywhere on the overall site compromises the security of their login page because an attacker can script (i.e., control) the login page by injecting script into another, insecurely loaded, site page. Note: "Mixed content" as used above (see also section 5.3 in [W3C.REC-wsc-ui-20100812]) refers to the notion termed "mixed security context" in this specification, and should not be confused with the same "mixed content" term used in the context of markup languages such as XML and HTML. Hodges, et al. Expires July 31, 2012 [Page 7] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 2.3.2. Threats Not Addressed 2.3.2.1. Phishing Phishing attacks occur when an attacker solicits authentication credentials from the user by hosting a fake site located on a different domain than the real site, perhaps driving traffic to the fake site by sending a link in an email message. Phishing attacks can be very effective because users find it difficult to distinguish the real site from a fake site. HSTS is not a defense against phishing per se; rather, it complements many existing phishing defenses by instructing the browser to protect session integrity and long-lived authentication tokens [ForceHTTPS]. 2.3.2.2. Malware and Browser Vulnerabilities Because HSTS is implemented as a browser security mechanism, it relies on the trustworthiness of the user's system to protect the session. Malicious code executing on the user's system can compromise a browser session, regardless of whether HSTS is used. 2.4. Requirements This section identifies and enumerates various requirements derived from the use cases and the threats discussed above, and lists the detailed core requirements HTTP Strict Transport Security addresses, as well as ancillary requirements that are not directly addressed. 2.4.1. Overall Requirement o Minimize the risks to web browser users and web site deployers that are derived from passive and active network attackers, web site development and deployment bugs, as well as insecure user actions. 2.4.1.1. Detailed Core Requirements These core requirements are derived from the overall requirement, and are addressed by this specification. 1. Web sites need to be able to declare to UAs that they should be interacted with using a strict security policy. 2. Web sites need to be able to instruct UAs that contact them insecurely to do so securely. 3. UAs need to persistently remember web sites that signal strict security policy enablement, for a web site declared time span. Hodges, et al. Expires July 31, 2012 [Page 8] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 Additionally, UAs need to cache the "freshest" strict security policy information, in order to allow web sites to update the information. 4. UAs need to re-write all insecure UA "http" URI loads to use the "https" secure scheme for those web sites for which secure policy is enabled. 5. Web site administrators need to be able to signal strict security policy application to subdomains of higher-level domains for which strict security policy is enabled, and UAs need to enforce such policy. For example, both example.com and foo.example.com could set policy for bar.foo.example.com. 6. UAs need to disallow security policy application to peer domains, and/or higher-level domains, by domains for which strict security policy is enabled. For example, neither bar.foo.example.com nor foo.example.com can set policy for example.com, nor can bar.foo.example.com set policy for foo.example.com. Also, foo.example.com cannot set policy for sibling.example.com. 7. UAs need to prevent users from clicking-through security warnings. Halting connection attempts in the face of secure transport exceptions is acceptable. Note: A means for uniformly securely meeting the first core requirement above is not specifically addressed by this specification (see Section 14.4 "Bootstrap MITM Vulnerability"). It may be addressed by a future revision of this specification or some other specification. Note also that there are means by which UA implementations may more fully meet the first core requirement, see Section 11 "User Agent Implementation Advice". 2.4.1.2. Detailed Ancillary Requirements These ancillary requirements are also derived from the overall requirement. They are not normatively addressed in this specification, but could be met by UA implementations at their implementor's discretion, although meeting these requirements may be complex. Hodges, et al. Expires July 31, 2012 [Page 9] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 1. Disallow "mixed security context" loads (see Section 2.3.1.3, above). 2. Facilitate user declaration of web sites for which strict security policy is enabled, regardless of whether the sites signal HSTS Policy. 3. Conformance Criteria This specification is written for hosts and user agents (UAs). 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]. A conformant host is one that implements all the requirements listed in this specification that are applicable to hosts. A conformant user agent is one that implements all the requirements listed in this specification that are applicable to user agents. 3.1. Document Conventions Note: ..is a note to the reader. These are points that should be expressly kept in mind and/or considered. Warning: This is how a warning is shown. These are things that can have suboptimal downside risks if not heeded. 4. Terminology Terminology is defined in this section. ASCII case-insensitive comparison means comparing two strings exactly, codepoint for codepoint, except that the characters in the range U+0041 .. U+005A (i.e. LATIN CAPITAL LETTER A to LATIN CAPITAL LETTER Z) and the corresponding characters in the range U+0061 .. U+007A (i.e. LATIN SMALL LETTER A to LATIN SMALL LETTER Z) are considered to also match. See [Unicode6] for details. Hodges, et al. Expires July 31, 2012 [Page 10] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 codepoint is a colloquial contraction of Code Point, which is any value in the Unicode codespace; that is, the range of integers from 0 to 10FFFF(hex) [Unicode6]. domain name domain names, also referred to as DNS Names, are defined in [RFC1035] to be represented outside of the DNS protocol itself (and implementations thereof) as a series of labels separated by dots, e.g., "example.com" or "yet.another.example.org". In the context of this specification, domain names appear in that portion of a URI satisfying the reg- name production in "Appendix A. Collected ABNF for URI" in [RFC3986], and the host component from the Host HTTP header field production in section 14.23 of [RFC2616]. Note: The domain names appearing in actual URI instances and matching the aforementioned production components may or may not be a fully qualified domain name. domain name label is that portion of a domain name appearing "between the dots", i.e. consider "foo.example.com": "foo", "example", and "com" are all domain name labels. Effective Request URI is a URI, identifying the target resource, that can be inferred by an HTTP host for any given HTTP request it receives. Such inference is necessary because HTTP requests often do not contain a complete "absolute" URI identifying the target resource. See Section 12 "Constructing an Effective Request URI", below. HTTP Strict Transport Security is the overall name for the combined UA- and server-side security policy defined by this specification. HTTP Strict Transport Security Host is a HTTP host implementing the server aspects of the HSTS policy. This means that an HSTS Host returns the "Strict-Transport-Security" HTTP response header field in its HTTP response messages sent over secure transport. Hodges, et al. Expires July 31, 2012 [Page 11] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 HTTP Strict Transport Security Policy is the name of the combined overall UA- and server- side facets of the behavior specified in this specification. HSTS See HTTP Strict Transport Security. HSTS Host See HTTP Strict Transport Security Host. HSTS Policy See HTTP Strict Transport Security Policy. Known HSTS Host is a HSTS Host for which the UA has a HSTS Policy in effect. I.e., the UA has noted this host as a Known HSTS Host. Local policy is comprised of policy rules deployers specify and which are often manifested as "configuration settings". MITM is an acronym for man-in-the-middle. See "man-in- the-middle attack" in [RFC4949]. Request URI is the URI used to cause a UA to issue an HTTP request message. UA is a an acronym for user agent. For the purposes of this specification, a UA is an HTTP client application typically actively manipulated by a user [RFC2616] . unknown HSTS Host is a HSTS Host that the user agent in question has not yet noted. 5. HSTS Mechanism Overview This section provides an overview of the mechanism by which an HSTS Host conveys its HSTS Policy to UAs, and how UAs process the HSTS Policies received from HSTS Hosts. The mechanism details are specified in Section 6 through Section 15. An HSTS Host conveys its HSTS Policy to UAs, only over secure transport (e.g., TLS), via the Strict-Transport-Security HTTP response header field. Receipt of this header field signals to UAs to enforce the HSTS Policy for all subsequent secure transport connections made to the HSTS Host, for a specified time duration. Application of the HSTS Policy to subdomains of the HSTS Host name may optionally be specified. Hodges, et al. Expires July 31, 2012 [Page 12] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 HSTS Hosts manage their advertised HSTS Policies by sending Strict- Transport-Security HTTP response header fields to UAs with new values for policy time duration and application to subdomains. UAs cache the "freshest" HSTS Policy information on behalf of an HSTS Host. Specifying a zero time duration signals to the UA to delete the HSTS policy for that HSTS host. Section 6.2 presents examples of Strict-Transport-Security HTTP response header fields. 6. Syntax This section defines the syntax of the new header this specification introduces. It also provides a short description of the function the header. The Section 7 "Server Processing Model" section details how hosts are to use this header. Likewise, the Section 8 "User Agent Processing Model" section details how user agents are to use this header. 6.1. Strict-Transport-Security HTTP Response Header Field The Strict-Transport-Security HTTP response header field (STS header field) indicates to a UA that it MUST enforce the HSTS Policy in regards to the host emitting the response message containing this header field. The ABNF syntax for the STS header field is: Strict-Transport-Security = "Strict-Transport-Security" ":" directive *( ";" [ directive ] ) directive = token [ "=" ( token | quoted-string ) ] where: token = quoted-string = The two directives defined in this specification are described below. The overall requirements for directives are: o The order of appearance of directives is not significant. Hodges, et al. Expires July 31, 2012 [Page 13] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 o All directives MUST appear only once in an STS header field. o Directive names are case-insensitive. o UAs MUST ignore any STS header fields containing directives that do not conform to their ABNF definition. Additional directives extending the the semantic functionality of the STS header field may be defined in other specifications (which "update" this specification), using the STS directive extension point. 6.1.1. The max-age Directive The REQUIRED max-age directive specifies the number of seconds, after the reception of the STS header field, during which the UA regards the host, from whom the message was received, as a Known HSTS Host (see also Section 8.1.1 "Noting a HSTS Host", below). The delta- seconds production is specified in [RFC2616]. The syntax of the max-age directive is defined as: max-age = "max-age" "=" delta-seconds delta-seconds = <1*DIGIT, defined in [RFC2616], Section 3.3.2> Note: A max-age value of zero signals the UA to cease regarding the host as a Known HSTS Host. 6.1.2. The includeSubDomains Directive The OPTIONAL includeSubDomains directive is a flag which, if present, signals to the UA that the HSTS Policy applies to this HSTS Host as well as any subdomains of the host's domain name. The syntax of the includeSubDomains directive is defined as: includeSubDomains = "includeSubDomains" 6.2. Examples The below HSTS header field stipulates that the HSTS policy is to remain in effect for one year (there are approximately 31 536 000 seconds in a year), and the policy applies only to the domain of the HSTS Host issuing it: Strict-Transport-Security: max-age=31536000 Hodges, et al. Expires July 31, 2012 [Page 14] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 The below HSTS header field stipulates that the HSTS policy is to remain in effect for approximately six months and the policy applies only to the domain of the issuing HSTS Host and all of its subdomains: Strict-Transport-Security: max-age=15768000 ; includeSubDomains 7. Server Processing Model This section describes the processing model that HSTS Hosts implement. The model is comprised of two facets: the first being the processing rules for HTTP request messages received over a secure transport (e.g., TLS [RFC5246], SSL [I-D.ietf-tls-ssl-version3], or perhaps others, the second being the processing rules for HTTP request messages received over non-secure transports, i.e. over TCP/IP [RFC0793]. 7.1. HTTP-over-Secure-Transport Request Type When replying to an HTTP request that was conveyed over a secure transport, a HSTS Host SHOULD include in its response message a STS header field that MUST satisfy the grammar specified above in Section 6.1 "Strict-Transport-Security HTTP Response Header Field". If a STS header field is included, the HSTS Host MUST include only one such header field. Note: Including the STS header field is stipulated as a "SHOULD" in order to accommodate various server- and network-side caches and load-balancing configurations where it may be difficult to uniformly emit STS header fields on behalf of a given HSTS Host. Establishing a given host as a Known HSTS Host, in the context of a given UA, MAY be accomplished over the HTTP protocol by correctly returning, per this specification, at least one valid STS header field to the UA. Other mechanisms, such as a client-side pre-loaded Known HSTS Host list MAY also be used. E.g., see Section 11 "User Agent Implementation Advice". 7.2. HTTP Request Type If a HSTS Host receives a HTTP request message over a non-secure transport, it SHOULD send a HTTP response message containing a Status-Code of 301 and a Location header field value containing either the HTTP request's original Effective Request URI (see Section 12 "Constructing an Effective Request URI", below) altered as necessary to have a URI scheme of "https", or a URI generated Hodges, et al. Expires July 31, 2012 [Page 15] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 according to local policy (which SHOULD employ a URI scheme of "https"). Note: The above behavior is a "SHOULD" rather than a "MUST" because: * There are risks in server-side non-secure-to-secure redirects [owaspTLSGuide]. * Site deployment characteristics -- e.g., a site that incorporates third-party components may not behave correctly when doing server-side non-secure-to-secure redirects in the case of being accessed over non-secure transport, but does behave correctly when accessed uniformly over secure transport. The latter is the case given a HSTS-capable UA that has already noted the site as a Known HSTS Host (by whatever means, e.g., prior interaction or UA configuration). A HSTS Host MUST NOT include the STS header field in HTTP responses conveyed over non-secure transport. 8. User Agent Processing Model This section describes the HTTP Strict Transport Security processing model for UAs. There are several facets to the model, enumerated by the following subsections. This processing model assumes that the UA implements IDNA2008 [RFC5890], or possibly IDNA2003 [RFC3490], as noted in Section 13 "Internationalized Domain Names for Applications (IDNA): Dependency and Migration". It also assumes that all domain names manipulated in this specification's context are already IDNA-canonicalized as outlined in Section 9 "Domain Name IDNA-Canonicalization" prior to the processing specified in this section. The above assumptions mean that this processing model also specifically assumes that appropriate IDNA and Unicode validations and character list testing have occurred on the domain names, in conjunction with their IDNA-canonicalization, prior to the processing specified in this section. See the IDNA-specific security considerations in Section 14.8 "Internationalized Domain Names" for rationale and further details. 8.1. Strict-Transport-Security Response Header Field Processing If an HTTP response, received over a secure transport, includes a STS header field, conforming to the grammar specified in Section 6.1 "Strict-Transport-Security HTTP Response Header Field" (above), and Hodges, et al. Expires July 31, 2012 [Page 16] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 there are no underlying secure transport errors or warnings (see Section 8.3, below), the UA MUST either: o Note the host as a Known HSTS Host if it is not already so noted (see Section 8.1.1 "Noting a HSTS Host", below), or, o Update its cached information for the Known HSTS Host if the max- age and/or includeSubDomains header field value tokens are conveying information different than that already maintained by the UA. Note: The max-age value is essentially a "time to live" value relative to the reception time of the STS header field. If the max-age header field value token has a value of zero, the UA MUST remove its cached HSTS Policy information if the HSTS Host is known, or, MUST NOT note this HSTS Host if it is not yet known. If a UA receives more than one STS header field in a HTTP response message over secure transport, then the UA MUST process only the first such header field. Otherwise: o If an HTTP response is received over insecure transport, the UA MUST ignore any present STS header field(s). o The UA MUST ignore any STS header fields not conforming to the grammar specified in Section 6.1 "Strict-Transport-Security HTTP Response Header Field" (above). 8.1.1. Noting a HSTS Host If the substring matching the host production from the Request-URI, that the host responded to, syntactically matches the IP-literal or IPv4address productions from section 3.2.2 of [RFC3986], then the UA MUST NOT note this host as a Known HSTS Host. Otherwise, if the substring does not congruently match a presently known HSTS Host, per the matching procedure specified in Section 8.1.2 "Known HSTS Host Domain Name Matching" below, then the UA MUST note this host as a Known HSTS Host, caching the HSTS Host's domain name and noting along with it the expiry time of this information, as effectively stipulated per the given max-age value, as well as whether the includeSubDomains flag is asserted or not. Hodges, et al. Expires July 31, 2012 [Page 17] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 8.1.2. Known HSTS Host Domain Name Matching A UA determines whether a domain name represents a Known HSTS Host by looking for a match between the query Domain Name and the UA's set of Known HSTS Hosts. 1. Compare the query domain name string with the Domain Names of the UA's set of Known HSTS Hosts. For each Known HSTS Host's domain name, the comparison is done with the query domain name label-by- label using an ASCII case-insensitive comparison beginning with the rightmost label, and continuing right-to-left, and ignoring separator characters (see clause 3.1(4) of [RFC3986]. * If a label-for-label match between an entire Known HSTS Host's domain name and a right-hand portion of the query domain name is found, then the Known HSTS Host's domain name is a superdomain match for the query domain name. For example: Query Domain Name: bar.foo.example.com Superdomain matched Known HSTS Host DN: foo.example.com At this point, the query domain name is ascertained to effectively represent a Known HSTS Host. There may also be additional matches further down the domain name label tree, up to and including a congruent match. * If a label-for-label match between a Known HSTS Host's domain name and the query domain name is found, i.e. there are no further labels to compare, then the query domain name congruently matches this Known HSTS Host. For example: Query Domain Name: foo.example.com Congruently matched Known HSTS Host DN: foo.example.com The query domain name is ascertained to represent a Known HSTS Host. However, if there are also superdomain matches, the one highest in the tree asserts the HSTS Policy for this Known HSTS Host. Hodges, et al. Expires July 31, 2012 [Page 18] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 * Otherwise, if no matches are found, the query domain name does not represent a Known HSTS Host. 8.2. URI Loading and Port Mapping Whenever the UA prepares to "load", also known as "dereference", any URI where the host component of the authority component of the URI [RFC3986] matches that of a Known HSTS Host (either as a congruent match or as a superdomain match where the superdomain Known HSTS Host has includeSubDomains asserted), then before proceeding with the load: If the URI's scheme is "http", then the UA MUST replace the URI scheme with "https", and, if the URI contains an explicit port component [RFC3986] of "80", then the UA MUST convert the port component to be "443", or, if the URI contains an explicit port component that is not equal to "80", the port component value MUST be preserved, otherwise, if the URI does not contain an explicit port component, the UA MUST NOT add one. Otherwise, if the URI's scheme is "https", then the UA MUST NOT modify the URI before dereferencing it. Note that the implication of the above steps is that the HSTS policy applies to all TCP ports on a host advertising the HSTS policy. 8.3. Errors in Secure Transport Establishment When connecting to a Known HSTS Host, the UA MUST terminate the connection (see also Section 11 "User Agent Implementation Advice", below) if there are any errors (e.g., certificate errors), whether "warning" or "fatal" or any other error level, with the underlying secure transport. This includes any issues with certificate revocation checking whether via the Certificate Revocation List (CRL) [RFC5280], or via the Online Certificate Status Protocol (OCSP) [RFC2560]. 8.4. HTTP-Equiv Element Attribute UAs MUST NOT heed http-equiv="Strict-Transport-Security" attribute settings on elements in received content. Hodges, et al. Expires July 31, 2012 [Page 19] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 8.5. Interstitially Missing Strict-Transport-Security Response Header Field If a UA receives HTTP responses from a Known HSTS Host over a secure channel, but they are missing the STS header field, the UA MUST continue to treat the host as a Known HSTS Host until the max-age value for the knowledge that Known HSTS Host is reached. Note that the max age could be infinite for a given Known HSTS Host. For example, if the Known HSTS Host is part of a pre-configured list that is implemented such that the list entries never "age out". 9. Domain Name IDNA-Canonicalization An IDNA-canonicalized domain name is the string generated by the following algorithm, whose input must be a valid Unicode-encoded (in NFC form [Unicode6]) string-serialized domain name: 1. Convert the domain name to a sequence of individual domain name label strings. 2. When implementing IDNA2008, convert each label that is not a Non- Reserved LDH (NR-LDH) label, to an A-label. See Section 2.3.2 of [RFC5890] for definitions of the former and latter, refer to Sections 5.3 through 5.5 of [RFC5891] for the conversion algorithm and requisite input validation and character list testing procedures. Otherwise, when implementing IDNA2003, convert each label using the "ToASCII" conversion in Section 4 of [RFC3490] (see also the definition of "equivalence of labels" in Section 2 of the latter specification). 3. Concatenate the resulting labels, separating each label from the next with (".") a %x2E character. See also Section 13 "Internationalized Domain Names for Applications (IDNA): Dependency and Migration" and Section 14.8 "Internationalized Domain Names" of this specification for further details and considerations. 10. Server Implementation and Deployment Advice This section is non-normative. Hodges, et al. Expires July 31, 2012 [Page 20] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 10.1. HSTS Policy expiration time considerations Server implementations and deploying web sites need to consider whether they are setting an expiry time that is a constant value into the future, e.g., by constantly sending the same max-age value to UAs. For example, a max-age value of 778000 is 90 days: Strict-Transport-Security: max-age=778000 Note that each receipt of this header by a UA will require the UA to update its notion of when it must delete its knowledge of this Known HSTS Host. The specifics of how this is accomplished is out of the scope of this specification. Or, whether they are setting an expiry time that is a fixed point in time, e.g., by sending max-age values that represent the remaining time until the expiry time. A consideration here is whether a deployer wishes to have the signaled HSTS Policy expiry time match that for the web site's domain certificate. Additionally, server implementers should consider employing a default max-age value of zero in their deployment configuration systems. This will require deployers to wilfully set max-age in order to have UAs enforce the HSTS Policy for their host, and protects them from inadvertently enabling HSTS with some arbitrary non-zero duration. 10.2. Using HSTS in conjunction with self-signed public-key certificates If a web site/organization/enterprise is generating their own secure transport public-key certificates for web sites, and that organization's root certification authority (CA) certificate is not typically embedded by default in browser CA certificate stores, and if HSTS Policy is enabled on a site identifying itself using a self- signed certificate, then secure connections to that site will fail, per the HSTS design. This is to protect against various active attacks, as discussed above. However, if said organization strongly wishes to employ self-signed certificates, and their own CA in concert with HSTS, they can do so by deploying their root CA certificate to their users' browsers. They can also, in addition or instead, distribute to their users' browsers the end-entity certificate(s) for specific hosts. There are various ways in which this can be accomplished (details are out of Hodges, et al. Expires July 31, 2012 [Page 21] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 scope for this specification). Once their root CA certificate is installed in the browsers, they may employ HSTS Policy on their site(s). Note: Interactively distributing root CA certificates to users, e.g., via email, and having the users install them, is arguably training the users to be susceptible to a possible form of phishing attack, see Section 14.6 "Bogus Root CA Certificate Phish plus DNS Cache Poisoning Attack". 10.3. Implications of includeSubDomains The includeSubDomains directive has some practical implications -- for example, if a HSTS Host offers HTTP-based services on various ports or at various subdomains of its host domain name, then they will all have to be available over secure transport in order to work properly. For example, certification authorities often offer their CRL distribution and OCSP services over plain HTTP, and sometimes at a subdomain of a publicly-available web application which may be secured by TLS/SSL. For example, is a publicly-available web application for "Example CA", a certification authority. Customers use this web application to register their public keys and obtain certificates. Example CA generates certificates for customers containing as the value for the "CRL Distribution Points" and "Authority Information Access:OCSP" certificate fields. If example-ca.com were to issue an HSTS Policy with the includeSubDomains directive, then HTTP-based user agents implementing HSTS, and that have interacted with the example-ca.com web application, would fail to retrieve CRLs and fail to check OCSP for certificates because these services are offered over plain HTTP. In this case, Example CA can either: o not use the includeSubDomains directive, or, o ensure HTTP-based services offered at subdomains of example-ca.com are uniformly offered over TLS/SSL, or, o offer plain HTTP-based services at a different domain name, e.g., example-ca-services.net. Hodges, et al. Expires July 31, 2012 [Page 22] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 11. User Agent Implementation Advice This section is non-normative. In order to provide users and web sites more effective protection, as well as controls for managing their UA's caching of HSTS Policy, UA implementors should consider including features such as: 11.1. No User Recourse Failing secure connection establishment on any warnings or errors (per Section 8.3 "Errors in Secure Transport Establishment"), should be done with "no user recourse". This means that the user should not be presented with an explanatory dialog giving her the option to proceed. Rather, it should be treated similarly to a server error where there is nothing further the user can do with respect to interacting with the target web application, other than wait and re- try. Essentially, "any warnings or errors" means anything that would cause the UA implementation to annunciate to the user that something is not entirely correct with the connection establishment. Not doing this, i.e., allowing user recourse such as "clicking- through warning/error dialogs", is a recipe for a Man-in-the-Middle attack. If a web application advertises HSTS, then it is opting into this scheme, whereby all certificate errors or warnings cause a connection termination, with no chance to "fool" the user into making the wrong decision and compromising themselves. 11.2. User-declared HSTS Policy Ability for users to explicitly declare a given Domain Name as representing a HSTS Host, thus seeding it as a Known HSTS Host before any actual interaction with it. This would help protect against the Section 14.4 "Bootstrap MITM Vulnerability". Note: Such a feature is difficult to get right on a per-site basis -- see the discussion of "rewrite rules" in section 5.5 of [ForceHTTPS]. For example, arbitrary web sites may not materialize all their URIs using the "https" scheme, and thus could "break" if a UA were to attempt to access the site exclusively using such URIs. Also note that this feature would complement, but is independent of the following described facility. Hodges, et al. Expires July 31, 2012 [Page 23] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 11.3. HSTS Pre-Loaded List Facility whereby web site administrators can have UAs pre-configured with HSTS Policy for their site(s) by the UA vendor(s) -- a so-called "pre-loaded list" -- in a manner similar to how root CA certificates are embedded in browsers "at the factory". This would help protect against the Section 14.4 "Bootstrap MITM Vulnerability". Note: Such a facility would complement a "User-declared HSTS Policy" feature. 11.4. Disallow Mixed Security Context Disallowing "mixed security context" (also known as "mixed content") loads (see section 5.3 "Mixed Content" in [W3C.REC-wsc-ui-20100812]). Note: In order to provide behavioral uniformity across UA implementations, the notion of mixed security context will require further standardization work, e.g., to more clearly define the term(s) and to define specific behaviors with respect to it. 11.5. HSTS Policy Deletion Ability to delete UA's cached HSTS Policy on a per HSTS Host basis. Note: Adding such a feature should be done very carefully in both the user interface and security senses. Deleting a cache entry for a Known HSTS Host should be a very deliberate and well-considered act -- it shouldn't be something users get used to just "clicking through" in order to get work done. Also, it shouldn't be possible for an attacker to inject script into the UA that silently and programmatically removes entries from the UA's cache of Known HSTS Hosts. 12. Constructing an Effective Request URI This section specifies how an HSTS Host must construct the Effective Request URI for a received HTTP request. HTTP requests often do not carry an absoluteURI for the target resource; instead, the URI needs to be inferred from the Request-URI, Host header field, and connection context ([RFC2616], Sections 3.2.1 and 5.1.2). The result of this process is called the "effective request URI (ERU)". The "target resource" is the resource identified by the effective request URI. Hodges, et al. Expires July 31, 2012 [Page 24] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 12.1. ERU Fundamental Definitions The first line of an HTTP request message, Request-Line, is specified by the following ABNF from [RFC2616], section 5.1: Request-Line = Method SP Request-URI SP HTTP-Version CRLF The Request-URI, within the Request-Line, is specified by the following ABNF from [RFC2616], section 5.1.2: Request-URI = "*" | absoluteURI | abs_path | authority The Host request header field is specified by the following ABNF from [RFC2616], section 14.23: Host = "Host" ":" host [ ":" port ] 12.2. Determining the Effective Requrest URI If the Request-URI is an absoluteURI, then the effective request URI is the Request-URI. If the Request-URI uses the abs_path form or the asterisk form, and the Host header field is present, then the effective request URI is constructed by concatenating: o the scheme name: "http" if the request was received over an insecure TCP connection, or "https" when received over a TLS/ SSL-secured TCP connection, and, o the octet sequence "://", and, o the host, and the port (if present), from the Host header field, and o the Request-URI obtained from the Request-Line, unless the Request-URI is just the asterisk "*". If the Request-URI uses the abs_path form or the asterisk form, and the Host header field is not present, then the effective request URI is undefined. Otherwise, when Request-URI uses the authority form, the effective request URI is undefined. Effective request URIs are compared using the rules described in [RFC2616] Section 3.2.3, except that empty path components MUST NOT be treated as equivalent to an absolute path of "/". Hodges, et al. Expires July 31, 2012 [Page 25] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 12.2.1. Effective Requrest URI Examples Example 1: the effective request URI for the message GET /pub/WWW/TheProject.html HTTP/1.1 Host: www.example.org:8080 (received over an insecure TCP connection) is "http", plus "://", plus the authority component "www.example.org:8080", plus the request-target "/pub/WWW/TheProject.html". Thus it is: "http://www.example.org:8080/pub/WWW/TheProject.html". Example 2: the effective request URI for the message OPTIONS * HTTP/1.1 Host: www.example.org (received over an SSL/TLS secured TCP connection) is "https", plus "://", plus the authority component "www.example.org". Thus it is: "https://www.example.org". 13. Internationalized Domain Names for Applications (IDNA): Dependency and Migration Textual domain names on the modern Internet may contain one or more "internationalized" domain name labels. Such domain names are referred to as "internationalized domain names" (IDNs). The specification suites defining IDNs and the protocols for their use are named "Internationalized Domain Names for Applications (IDNA)". At this time, there are two such specification suites: IDNA2008 [RFC5890] and its predecessor IDNA2003 [RFC3490]. IDNA2008 obsoletes IDNA2003, but there are differences between the two specifications, and thus there can be differences in processing (e.g., converting) domain name labels that have been registered under one from those registered under the other. There will be a transition period of some time during which IDNA2003-based domain name labels will exist in the wild. User agents SHOULD implement IDNA2008 [RFC5890] and MAY implement [RFC5895] (see also Section 7 of [RFC5894]) or [UTS46] in order to facilitate their IDNA transition. If a user agent does not implement IDNA2008, the user agent MUST implement IDNA2003. 14. Security Considerations Hodges, et al. Expires July 31, 2012 [Page 26] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 14.1. Ramifications of HSTS Policy Establishment only over Error-free Secure Transport The User Agent Processing Model defined in Section 8, stipulates that a host is initially noted as a Known HSTS Host, or that updates are made to a Known HSTS Host's cached information, only if the UA receives the STS header field over a secure transport connection having no underlying secure transport errors or warnings. The rationale behind this is that if there is a man-in-the-middle (MITM) -- whether a legitimately deployed proxy or an illegitimate entity -- it could cause various mischief (see also Appendix A "Design Decision Notes", item 3, as well as Section 14.4 "Bootstrap MITM Vulnerability" ), for example: o Unauthorized notation of the host as a Known HSTS Host, potentially leading to a denial of service situation if the host does not uniformly offer its services over secure transport (see also Section 14.3 "Denial of Service". o Resetting the time-to-live for the host's designation as a Known HSTS Host by manipulating the max-age header field parameter value that is returned to the UA. If max-age is returned as zero, this will cause the host to no longer be regarded as an Known HSTS Host by the UA, leading to either insecure connections to the host or possibly denial-of-service if the host delivers its services only over secure transport. However, this means that if a UA is "behind" a proxy -- within a corporate intranet, for example -- and interacts with an unknown HSTS Host beyond the proxy, the user will possibly be presented with the legacy secure connection error dialogs. And even if the risk is accepted and the user clicks-through, the host will not be noted as a HSTS Host. Thus as long as the UA is behind such a proxy the user will be vulnerable, and possibly be presented with the legacy secure connection error dialogs for as yet unknown HSTS Hosts. But once the UA successfully connects to an unknown HSTS Host over error-free secure transport, the host will be noted as a Known HSTS Host. This will result in the failure of subsequent connection attempts from behind interfering proxies. The above discussion relates to the recommendation in Section 11 "User Agent Implementation Advice" that the secure connection be terminated with "no user recourse" whenever there are warnings and errors and the host is a Known HSTS Host. Such a posture protects users from "clicking through" security warnings and putting themselves at risk. Hodges, et al. Expires July 31, 2012 [Page 27] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 14.2. The Need for includeSubDomains Without the includeSubDomains directive, a web application would not be able to adequately protect so-called "domain cookies" (even if these cookies have their "Secure" flag set and thus are conveyed only on secure channels). These are cookies the web application expects UAs to return to any and all subdomains of the web application. For example, suppose example.com represents the top-level DNS name for a web application. Further suppose that this cookie is set for the entire example.com domain, i.e. it is a "domain cookie", and it has its Secure flag set. Suppose example.com is a Known HSTS Host for this UA, but the includeSubDomains flag is not set. Now, if an attacker causes the UA to request a subdomain name that is unlikely to already exist in the web application, such as "https://uxdhbpahpdsf.example.com/", but the attacker has established somewhere and registered in the DNS, then: 1. The UA is unlikely to already have an HSTS policy established for "uxdhbpahpdsf.example.com", and, 2. The HTTP request sent to uxdhbpahpdsf.example.com will include the Secure-flagged domain cookie. 3. If "uxdhbpahpdsf.example.com" returns a certificate during TLS establishment, and the user clicks through any warning that might be annunciated (it is possible, but not certain, that one may obtain a requisite certificate for such a domain name such that a warning may or may not appear), then the attacker can obtain the Secure-flagged domain cookie that's ostensibly being protected. Without the "includeSubDomains" directive, HSTS is unable to protect such Secure-flagged domain cookies. 14.3. Denial of Service HSTS could be used to mount certain forms of Denial-of- Service (DoS) attacks against web sites. A DoS attack is an attack in which one or more network entities target a victim entity and attempt to prevent the victim from doing useful work. This section discusses such scenarios in terms of HSTS, though this list is not exhaustive. See also [RFC4732] for a discussion of overall Internet DoS considerations. o Web applications available over HTTP There is an opportunity for perpetrating DoS attacks with web Hodges, et al. Expires July 31, 2012 [Page 28] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 applications that are -- or critical portions of them are -- available only over HTTP without secure transport, if attackers can cause UAs to set HSTS Policy for such web applications' host(s). This is because once the HSTS Policy is set for a web application's host in a UA, the UA will only use secure transport to communicate with the host. If the host is not using secure transport, or is not for critical portions of its web application, then the web application will be rendered unusable for the UA's user. An HSTS Policy can be set for a victim host in various ways: * If the web application has a HTTP response splitting vulnerability [CWE-113] (which can be abused in order to facilitate "HTTP Header Injection"). * If an attacker can spoof a redirect from an insecure victim site, e.g., to , where the latter is attacker-controlled and has an apparently valid certificate, then the attacker can set an HSTS Policy for example.com, and also for all subdomains of example.com. * If an attacker can convince users to manually configure HSTS Policy for a victim host, assuming their UAs offer this capability (see Section 11 "User Agent Implementation Advice"). Or if such UA configuration is scriptable, and the attacker can cause UAs to execute his script. o Inadvertent use of includeSubDomains The includeSubDomains directive instructs UAs to automatically regard all subdomains of the given HSTS Host as Known HSTS Hosts. If any such subdomains do not support properly configured secure transport, then they will be rendered unreachable from such UAs. 14.4. Bootstrap MITM Vulnerability The bootstrap MITM (Man-In-The-Middle) vulnerability is a vulnerability users and HSTS Hosts encounter in the situation where the user manually enters, or follows a link, to an unknown HSTS Host using a "http" URI rather than a "https" URI. Because the UA uses an insecure channel in the initial attempt to interact with the specified server, such an initial interaction is vulnerable to various attacks [ForceHTTPS] . Hodges, et al. Expires July 31, 2012 [Page 29] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 Note: There are various features/facilities that UA implementations may employ in order to mitigate this vulnerability. Please see Section 11 "User Agent Implementation Advice". 14.5. Network Time Attacks Active network attacks can subvert network time protocols (such as NTP) - making HSTS less effective against clients that trust NTP or lack a real time clock. Network time attacks are beyond the scope of this specification. Note that modern operating systems use NTP by default. See also section 2.10 of [RFC4732]. 14.6. Bogus Root CA Certificate Phish plus DNS Cache Poisoning Attack If an attacker can convince users of, say, https://bank.example.com (which is protected by HSTS Policy), to install their own version of a root CA certificate purporting to be bank.example.com's CA, e.g., via a phishing email message with a link to such a certificate. Then, if they can perform an attack on the users' DNS, (e.g., via cache poisoning) and turn on HSTS Policy for their fake bank.example.com site, then they have themselves some new users. 14.7. Creative Manipulation of HSTS Policy Store Since an HSTS Host may select its own host name and subdomains thereof, and this information is cached in the HSTS Policy store of conforming UAs, it is possible for those who control a HSTS Host(s) to encode information into domain names they control and cause such UAs to cache this information as a matter of course in the process of noting the HSTS Host. This information can be retrieved by other hosts through clever loaded page construction causing the UA to send queries to (variations of) the encoded domain names. Such queries can reveal whether the UA had prior visited the original HSTS Host (and subdomains). Such a technique could potentially be abused as yet another form of "web tracking" [WebTracking]. 14.8. Internationalized Domain Names Internet security relies in part on the DNS and the domain names it hosts. Domain names are used by users to identify and connect to Internet hosts and other network resources. For example, Internet security is compromised if a user entering an internationalized domain name (IDN) is connected to different hosts based on different interpretations of the IDN. The processing models specified in this specification assume that the Hodges, et al. Expires July 31, 2012 [Page 30] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 domain names they manipulate are IDNA-canonicalized, and that the canonicalization process correctly performed all appropriate IDNA and Unicode validations and character list testing per the requisite specifications (e.g., as noted in Section 9 "Domain Name IDNA- Canonicalization"). These steps are necessary in order to avoid various potentially compromising situations. In brief, some examples of issues that could stem from lack of careful and consistent Unicode and IDNA validations are things such as unexpected processing exceptions, truncation errors, and buffer overflows, as well as false-positive and/or false-negative domain name matching results. Any of the foregoing issues could possibly be leveraged by attackers in various ways. Additionally, IDNA2008 [RFC5890] differs from IDNA2003 [RFC3490] in terms of disallowed characters and character mapping conventions. This situation can also lead to false-positive and/or false-negative domain name matching results, resulting in, for example, users possibly communicating with unintended hosts, or not being able to reach intended hosts. For details, refer to the Security Considerations sections of [RFC5890], [RFC5891], and [RFC3490], as well as the specifications they normatively reference. Additionally, [RFC5894] provides detailed background and rationale for IDNA2008 in particular, as well as IDNA and its issues in general, and should be consulted in conjunction with the former specifications. 15. IANA Considerations Below is the Internet Assigned Numbers Authority (IANA) Provisional Message Header Field registration information per [RFC3864]. Header field name: Strict-Transport-Security Applicable protocol: HTTP Status: provisional Author/Change controller: TBD Specification document(s): this one 16. References 16.1. Normative References [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. Hodges, et al. Expires July 31, 2012 [Page 31] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 [RFC2616] Fielding, R., Gettys, J., Mogul, J., Frystyk, H., Masinter, L., Leach, P., and T. Berners-Lee, "Hypertext Transfer Protocol -- HTTP/1.1", RFC 2616, June 1999. [RFC3864] Klyne, G., Nottingham, M., and J. Mogul, "Registration Procedures for Message Header Fields", BCP 90, RFC 3864, September 2004. [RFC3986] Berners-Lee, T., Fielding, R., and L. Masinter, "Uniform Resource Identifier (URI): Generic Syntax", STD 66, RFC 3986, January 2005. 16.2. Informative References [Aircrack-ng] d'Otreppe, T., "Aircrack-ng", Accessed: 11-Jul-2010, . [BeckTews09] Beck, M. and E. Tews, "Practical Attacks Against WEP and WPA", Second ACM Conference on Wireless Network Security Zurich, Switzerland, 2009, . [CWE-113] "CWE-113: Improper Neutralization of CRLF Sequences in HTTP Headers ('HTTP Response Splitting')", Common Weakness Enumeration , The Mitre Corporation , . [Firesheep] Various, "Firesheep", Wikipedia Online, on-going, . [ForceHTTPS] Jackson, C. and A. Barth, "ForceHTTPS: Protecting High- Security Web Sites from Network Attacks", In Proceedings of the 17th International World Wide Web Conference (WWW2008) , 2008, . [GoodDhamijaEtAl05] Good, N., Dhamija, R., Grossklags, J., Thaw, D., Aronowitz, S., Mulligan, D., and J. Konstan, "Stopping Spyware at the Gate: A User Study of Privacy, Notice and Spyware", In Proceedings of Symposium On Usable Privacy Hodges, et al. Expires July 31, 2012 [Page 32] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 and Security (SOUPS) Pittsburgh, PA, USA, July 2005, . [I-D.ietf-tls-ssl-version3] Freier, A., Karlton, P., and P. Kocher, "The SSL Protocol Version 3.0", draft-ietf-tls-ssl-version3-00 (work in progress), November 1996, . [JacksonBarth2008] Jackson, C. and A. Barth, "Beware of Finer-Grained Origins", Web 2.0 Security and Privacy Oakland, CA, USA, 2008, . [RFC0793] Postel, J., "Transmission Control Protocol", STD 7, RFC 793, September 1981. [RFC1035] Mockapetris, P., "Domain names - implementation and specification", STD 13, RFC 1035, November 1987. [RFC2560] Myers, M., Ankney, R., Malpani, A., Galperin, S., and C. Adams, "X.509 Internet Public Key Infrastructure Online Certificate Status Protocol - OCSP", RFC 2560, June 1999. [RFC2818] Rescorla, E., "HTTP Over TLS", RFC 2818, May 2000. [RFC3454] Hoffman, P. and M. Blanchet, "Preparation of Internationalized Strings ("stringprep")", RFC 3454, December 2002. [RFC3490] Faltstrom, P., Hoffman, P., and A. Costello, "Internationalizing Domain Names in Applications (IDNA)", RFC 3490, March 2003. TODO: explanation of including this obsoleted reference [RFC3492] Costello, A., "Punycode: A Bootstring encoding of Unicode for Internationalized Domain Names in Applications (IDNA)", RFC 3492, March 2003. TODO: explanation of including this obsoleted reference [RFC4732] Handley, M., Rescorla, E., and IAB, "Internet Denial-of- Service Considerations", RFC 4732, December 2006. Hodges, et al. Expires July 31, 2012 [Page 33] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 [RFC4949] Shirey, R., "Internet Security Glossary, Version 2", RFC 4949, August 2007. [RFC5246] Dierks, T. and E. Rescorla, "The Transport Layer Security (TLS) Protocol Version 1.2", RFC 5246, August 2008. [RFC5280] Cooper, D., Santesson, S., Farrell, S., Boeyen, S., Housley, R., and W. Polk, "Internet X.509 Public Key Infrastructure Certificate and Certificate Revocation List (CRL) Profile", RFC 5280, May 2008. [RFC5890] Klensin, J., "Internationalized Domain Names for Applications (IDNA): Definitions and Document Framework", RFC 5890, August 2010. [RFC5891] Klensin, J., "Internationalized Domain Names in Applications (IDNA): Protocol", RFC 5891, August 2010. [RFC5894] Klensin, J., "Internationalized Domain Names for Applications (IDNA): Background, Explanation, and Rationale", RFC 5894, August 2010. [RFC5895] Resnick, P. and P. Hoffman, "Mapping Characters for Internationalized Domain Names in Applications (IDNA) 2008", RFC 5895, September 2010. [RFC6265] Barth, A., "HTTP State Management Mechanism", RFC 6265, April 2011. [RFC6454] Barth, A., "The Web Origin Concept", RFC 6454, December 2011. [SunshineEgelmanEtAl09] Sunshine, J., Egelman, S., Almuhimedi, H., Atri, N., and L. Cranor, "Crying Wolf: An Empirical Study of SSL Warning Effectiveness", In Proceedings of 18th USENIX Security Symposium Montreal, Canada, Augus 2009, . [UTS46] Davis, M. and M. Suignard, "Unicode IDNA Compatibility Processing", Unicode Technical Standards # 46, 2010, . [Unicode6] The Unicode Consortium, "The Unicode Standard, Version 6.0 - Core Specification", Unicode 6.0.0, Mountain View, CA, The Unicode Consortium ISBN 978-1-936213-01-6, 2011, Hodges, et al. Expires July 31, 2012 [Page 34] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 . [W3C.REC-wsc-ui-20100812] Saldhana, A. and T. Roessler, "Web Security Context: User Interface Guidelines", World Wide Web Consortium Recommendation REC-wsc-ui-20100812, August 2010, . [WEBSEC] "WebSec -- HTTP Application Security Minus Authentication and Transport", . Mailing list for IETF WebSec Working Group. [RFCEditor: please remove this reference upon publication as an RFC.] [WebTracking] Schmucker, N., "Web Tracking", SNET2 Seminar Paper Summer Term, 2011, . [owaspTLSGuide] Coates, M., Wichers, d., Boberski, M., and T. Reguly, "Transport Layer Protection Cheat Sheet", Accessed: 11- Jul-2010, . Appendix A. Design Decision Notes This appendix documents various design decisions. 1. Cookies aren't appropriate for HSTS Policy expression as they are potentially mutable (while stored in the UA), therefore an HTTP header field is employed. 2. We chose to not attempt to specify how "mixed security context loads" (aka "mixed content loads") are handled due to UA implementation considerations as well as classification difficulties. 3. A HSTS Host may update UA notions of HSTS Policy via new HSTS header field parameter values. We chose to have UAs honor the "freshest" information received from a server because there is the chance of a web site sending out an errornous HSTS Policy, such as a multi-year max-age value, and/or an incorrect includeSubDomains flag. If the HSTS Host couldn't correct such errors over protocol, it would require some form of annunciation to users and manual intervention on the users' part, which could Hodges, et al. Expires July 31, 2012 [Page 35] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 be a non-trivial problem for both web application providers and their users. 4. HSTS Hosts are identified only via domain names -- explicit IP address identification of all forms is excluded. This is for simplification and also is in recognition of various issues with using direct IP address identification in concert with PKI-based security. Appendix B. Differences between HSTS Policy and Same-Origin Policy HSTS Policy has the following primary characteristics: HSTS Policy stipulates requirements for the security characteristics of UA-to-host connection establishment, on a per- host basis. Hosts explicitly declare HSTS Policy to UAs. Conformant UAs are obliged to implement hosts' declared HSTS Policies. HSTS Policy is conveyed over protocol from the host to the UA. The UA maintains a cache of Known HSTS Hosts. UAs apply HSTS Policy whenever making a HTTP connection to a Known HSTS Host, regardless of host port number. I.e., it applies to all ports on a Known HSTS Host. Hosts are unable to affect this aspect of HSTS Policy. Hosts may optionally declare that their HSTS Policy applies to all subdomains of their host domain name. In contrast, the Same-Origin Policy (SOP) [RFC6454] has the following primary characteristics: An origin is the scheme, host, and port of a URI identifying a resource. A UA may dereference a URI, thus loading a representation of the resource the URI identifies. UAs label resource representations with their origins, which are derived from their URIs. The SOP refers to a collection of principles, implemented within UAs, governing the isolation of and communication between resource representations within the UA, as well as resource representations' access to network resources. Hodges, et al. Expires July 31, 2012 [Page 36] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 In summary, although both HSTS Policy and SOP are enforced by by UAs, HSTS Policy is optionally declared by hosts and is not origin-based, while the SOP applies to all resource representations loaded from all hosts by conformant UAs. Appendix C. Acknowledgments The authors thank Devdatta Akhawe, Michael Barrett, Paul Hoffman, Yoav Nir, Laksh Raghavan, Marsh Ray, Julian Reschke, Tom Ritter, Peter Saint-Andre, Sid Stamm, Maciej Stachowiak, Andy Steingrubl, Brandon Sterne, Martin Thomson, Daniel Veditz, as well as all the websec working group participants and others for their review and contributions. Thanks to Julian Reschke for his elegant re-writing of the effective request URI text, which he did when incorporating the ERU notion into the HTTPbis work. Subsequently, the ERU text in this spec was lifted from Julian's work in [I-D.draft-ietf-httpbis-p1-messaging-17] and adapted to the [RFC2616] ABNF. Appendix D. Change Log [RFCEditor: please remove this section upon publication as an RFC.] Changes are grouped by spec revision listed in reverse issuance order. D.1. For draft-ietf-websec-strict-transport-sec Changes from -03 to -04: 1. Clarified that max-age=0 will cause UA to forget a known HSTS host, and more generally clarified that the "freshest" info from the HSTS host is cached, and thus HSTS hosts are able to alter the cached max-age in UAs. This addresses issue ticket #13. 2. Updated section on "Constructing an Effective Request URI" to remove remaining reference to RFC3986 and reference RFC2616 instead. Further addresses issue ticket #14. 3. Addresses further ABNF issues noted in comment:1 of issue ticket #27. Hodges, et al. Expires July 31, 2012 [Page 37] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 4. Reworked the introduction to clarify the denotation of "HSTS policy" and added the new Appendix B summarizing the primary characteristics of HSTS Policy and Same-Origin Policy, and identifying their differences. Added ref to [RFC4732]. This addresses issue ticket #28. 5. Reworked language in Section 2.3.1.3. wrt "mixed content", more clearly explain such vulnerability, disambiguate "mixed content" in web security context from its usage in markup language context. This addresses issue ticket #29. 6. Expanded Denial of Service discussion in Security Considerations. Added refs to [RFC4732] and [CWE-113]. This addresses issue ticket #30. 7. Mentioned in prose the case-insensitivity of directive names. This addresses issue ticket #31. 8. Added Section 10.3 "Implications of includeSubDomains". This addresses issue ticket #32. 9. Further refines text and ABNF definitions of STS header field directives. Retains use of quoted-string in directive grammar. This addresses issue ticket #33. 10. Added Section 14.7 "Creative Manipulation of HSTS Policy Store", including reference to [WebTracking]. This addresses issue ticket #34. 11. Added Section 14.1 "Ramifications of HSTS Policy Establishment only over Error-free Secure Transport" and made some accompanying editorial fixes in some other sections. This addresses issue ticket #35. 12. Refined references. Cleaned out un-used ones, updated to latest RFCs for others, consigned many to Informational. This addresses issue ticket #36. Hodges, et al. Expires July 31, 2012 [Page 38] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 13. Fixed-up some inaccuracies in the "Changes from -02 to -03" section. Changes from -02 to -03: 1. Updated section on "Constructing an Effective Request URI" to remove references to RFC3986. Addresses issue ticket #14. 2. Reference RFC5890 for IDNA, retaining subordinate refs to RFC3490. Updated IDNA-specific language, e.g. domain name canonicalization and IDNA dependencies. Addresses issue ticket #26 . 3. Completely re-wrote the STS header ABNF to be fully based on RFC2616, rather than a hybrid of RFC2616 and httpbis. Addresses issue ticket #27 . Changes from -01 to -02: 1. Updated Section 8.2 "URI Loading and Port Mapping" fairly thoroughly in terms of refining the presentation of the steps, and to ensure the various aspects of port mapping are clear. Nominally fixes issue ticket #1 2. Removed dependencies on [I-D.draft-ietf-httpbis-p1-messaging-15]. Thus updated STS ABNF in Section 6.1 "Strict-Transport-Security HTTP Response Header Field" by lifting some productions entirely from [I-D.draft-ietf-httpbis-p1-messaging-15] and leveraging [RFC2616]. Addresses issue ticket #2 . 3. Updated Effective Request URI section and definition to use language from [I-D.draft-ietf-httpbis-p1-messaging-15] and ABNF from [RFC2616]. Fixes issue ticket #3 . 4. Added explicit mention that the HSTS policy applies to all TCP ports of a host advertising the HSTS policy. Nominally fixes issue ticket #4 5. Clarified the need for the "includeSubDomains" directive, e.g. to protect Secure-flagged domain cookies. In Hodges, et al. Expires July 31, 2012 [Page 39] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 Section 14.2 "The Need for includeSubDomains". Nominally fixes issue ticket #5 6. Cited Firesheep as real-live threat in Section 2.3.1.1 "Passive Network Attackers". Nominally fixes issue ticket #6 . 7. Added text to Section 11 "User Agent Implementation Advice" justifying connection termination due to tls warnings/errors. Nominally fixes issue ticket #7 . 8. Added new subsection Section 8.5 "Interstitially Missing Strict-Transport-Security Response Header Field". Nominally fixes issue ticket #8 . 9. Added text to Section 8.3 "Errors in Secure Transport Establishment" explicitly note revocation check failures as errors causing connection termination. Added references to [RFC5280] and [RFC2560]. Nominally fixes issue ticket #9 . 10. Added a sentence, noting that distributing specific end- entity certificates to browsers will also work for self- signed/private-CA cases, to Section 10 "Server Implementation and Deployment Advice" Nominally fixes issue ticket #10 . 11. Moved "with no user recourse" language from Section 8.3 "Errors in Secure Transport Establishment" to Section 11 "User Agent Implementation Advice". This nominally fixes issue ticket #11 . 12. Removed any and all dependencies on [I-D.draft-ietf-httpbis-p1-messaging-15], instead depending on [RFC2616] only. Fixes issue ticket #12 . 13. Removed the inline "XXX1" issue because no one had commented on it and it seems reasonable to suggest as a SHOULD that web apps should redirect incoming insecure connections to secure connections. 14. Removed the inline "XXX2" issue because it was simply for raising consciousness about having some means for Hodges, et al. Expires July 31, 2012 [Page 40] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 distributing secure web application metadata. 15. Removed "TODO1" because description prose for "max-age" in the Note following the ABNF in Section 6 seems to be fine. 16. Decided for "TODO2" that "the first STS header field wins". TODO2 had read: "Decide UA behavior in face of encountering multiple HSTS headers in a message. Use first header? Last?". Removed TODO2. 17. Added Section 1.1 "Organization of this specification" for readers' convenience. 18. Moved design decision notes to be a proper appendix Appendix A. Changes from -00 to -01: 1. Changed the "URI Loading" section to be "URI Loading and Port Mapping". 2. [HASMAT] reference changed to [WEBSEC]. 3. Changed "server" -> "host" where applicable, notably when discussing "HSTS Hosts". Left as "server" when discussing e.g. "http server"s. 4. Fixed minor editorial nits. Changes from draft-hodges-strict-transport-sec-02 to draft-ietf-websec-strict-transport-sec-00: 1. Altered spec metadata (e.g. filename, date) in order to submit as a WebSec working group Internet-Draft. D.2. For draft-hodges-strict-transport-sec Changes from -01 to -02: 1. updated abstract such that means for expressing HSTS Policy other than via HSTS header field is noted. 2. Changed spec title to "HTTP Strict Transport Security (HSTS)" from "Strict Transport Security". Updated use of "STS" acronym throughout spec to HSTS (except for when specifically discussing syntax of Strict-Transport-Security HTTP Response Header field), updated "Terminology" appropriately. Hodges, et al. Expires July 31, 2012 [Page 41] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 3. Updated the discussion of "Passive Network Attackers" to be more precise and offered references. 4. Removed para on nomative/non-normative from "Conformance Criteria" pending polishing said section to IETF RFC norms. 5. Added examples subsection to "Syntax" section. 6. Added OWS to maxAge production in Strict-Transport-Security ABNF. 7. Cleaned up explanation in the "Note:" in the "HTTP-over- Secure-Transport Request Type" section, folded 3d para into "Note:", added conformance clauses to the latter. 8. Added exaplanatory "Note:" and reference to "HTTP Request Type" section. Added "XXX1" issue. 9. Added conformance clause to "URI Loading". 10. Moved "Notes for STS Server implementors:" from "UA Implementation dvice " to "HSTS Policy expiration time considerations:" in "Server Implementation Advice", and also noted another option. 11. Added cautionary "Note:" to "Ability to delete UA's cached HSTS Policy on a per HSTS Server basis". 12. Added some informative references. 13. Various minor editorial fixes. Changes from -00 to -01: 1. Added reference to HASMAT mailing list and request that this spec be discussed there. Authors' Addresses Jeff Hodges PayPal 2211 North First Street San Jose, California 95131 US Email: Jeff.Hodges@PayPal.com Hodges, et al. Expires July 31, 2012 [Page 42] Internet-Draft HTTP Strict Transport Security (HSTS) January 2012 Collin Jackson Carnegie Mellon University Email: collin.jackson@sv.cmu.edu Adam Barth Google, Inc. Email: ietf@adambarth.com URI: http://www.adambarth.com/ Hodges, et al. Expires July 31, 2012 [Page 43]