httpstate Working GroupA. Barth
Internet-DraftU.C. Berkeley
Obsoletes: 2109 (if approved)March 7, 2010
Intended status: Standards Track
Expires: September 8, 2010

HTTP State Management Mechanism


This document defines the HTTP Cookie and Set-Cookie headers. These headers can be used by HTTP servers to store state on HTTP user agents, letting the servers maintain a stateful session over the mostly stateless HTTP protocol. The cookie protocol has many historical infelicities that degrade its security and privacy.

Status of this Memo

This Internet-Draft is submitted to IETF in full conformance with the provisions of BCP 78 and BCP 79.

Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet-Drafts.

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”.

The list of current Internet-Drafts can be accessed at

The list of Internet-Draft Shadow Directories can be accessed at

This Internet-Draft will expire on September 8, 2010.

Copyright Notice

Copyright (c) 2010 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 ( 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 BSD License.

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

1. Introduction

This document defines the HTTP Cookie and Set-Cookie header. Using the Set-Cookie header, an HTTP server can store name/value pairs and associated metadata (called cookies) at the user agent. When the user agent makes subsequent requests to the server, the user agent uses the metadata to determine whether to return the name/value pairs in the Cookie header.

Although simple on its surface, the cookie protocol has a number of complexities. For example, the server indicates a scope for each cookie when sending them to the user agent. The scope indicates the maximum amount of time the user agent should retain the cookie, to which servers the user agent should return the cookie, and for which protocols the cookie is applicable.

For historical reasons, the cookie protocol contains a number of security and privacy infelicities. For example, a server can indicate that a given cookie is intended for "secure" connections, but the Secure attribute provides only confidentiality (not integrity) from active network attackers. Similarly, cookies for a given host are shared across all the ports on that host, even though the usual "same-origin policy" used by web browsers isolates content retrieved from different ports.

2. General Nonsense

2.1. Conformance Criteria

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

Requirements phrased in the imperative as part of algorithms (such as "strip any leading space characters" or "return false and abort these steps") are to be interpreted with the meaning of the key word ("MUST", "SHOULD", "MAY", etc) used in introducing the algorithm.

2.2. Syntax Notation

This specification uses the Augmented Backus-Naur Form (ABNF) notation of [RFC5234].

The following core rules are included by reference, as defined in [RFC5234], Appendix B.1: ALPHA (letters), CR (carriage return), CRLF (CR LF), CTL (controls), DIGIT (decimal 0-9), DQUOTE (double quote), HEXDIG (hexadecimal 0-9/A-F/a-f), LF (line feed), OCTET (any 8-bit sequence of data), SP (space), HTAB (horizontal tab), VCHAR (any visible [USASCII] character), and WSP (whitespace).

2.3. Terminology

The terms user agent, client, server, proxy, and origin server have the same meaning as in the HTTP/1.1 specification ([RFC2616]).

The terms request-host and request-URI refer to the values the user agent would send to the server as, respectively, the host (but not port) and abs_path portions of the absoluteURI (http_URL) of the HTTP Request-Line.

3. Overview

We outline here a way for an origin server to send state information to a user agent, and for the user agent to return the state information to the origin server.

To initiate a session, the origin server includes a Set-Cookie header in an HTTP response. (Note that "session" here does not refer to a persistent network connection but to a logical session created from HTTP requests and responses. The presence or absence of a persistent connection should have no effect on the use of cookie-derived sessions).

The user agent returns a Cookie request header to the origin server if it chooses to continue a session. The Cookie header contains a number of cookies the user agent received in previous Set-Cookie headers. The origin server MAY ignore the Cookie header or use the header to determine the current state of the session. The origin server MAY send the user agent a Set-Cookie response header with the same or different information, or it MAY send no Set-Cookie header at all.

Servers MAY return a Set-Cookie response header with any response. User agents SHOULD send a Cookie request header, subject to other rules detailed below, with every request.

An origin server MAY include multiple Set-Cookie header fields in a single response. Note that an intervening gateway MUST NOT fold multiple Set-Cookie header fields into a single header field.

3.1. Examples

Using the cookie protocol, a server can send the user agent a short string in an HTTP response that the user agent will return in future HTTP requests. For example, the server can send the user agent a "session identifier" named SID with the value 31d4d96e407aad42. The user agent then returns the session identifier in subsequent requests.

== Server -> User Agent ==
Set-Cookie: SID=31d4d96e407aad42

== User Agent -> Server ==
Cookie: SID=31d4d96e407aad42

The server can alter the default scope of the cookie using the Path and Domain attributes. For example, the server can instruct the user agent to return the cookie to every path and every subdomain of

== Server -> User Agent ==
Set-Cookie: SID=31d4d96e407aad42; Path=/;

== User Agent -> Server ==
Cookie: SID=31d4d96e407aad42

The server can store multiple cookies in the user agent. For example, the server can store a session identifier as well as the user's preferred language by returning two Set-Cookie response headers. Notice that the server uses the Secure and HttpOnly attributes to provide additional security protections for the more-sensitive session identifier.

== Server -> User Agent ==
Set-Cookie: SID=31d4d96e407aad42; Path=/; Secure, HttpOnly
Set-Cookie: lang=en-US; Path=/;

== User Agent -> Server ==
Cookie: SID=31d4d96e407aad42; lang=en-US

If the server wishes the user agent to persist the cookie over multiple sessions, the server can specify a expiration date in the Expires attribute. Note that the user agent might might delete the cookie before the expiration date if the user agent's cookie store exceeds its quota or if the user manually deletes the server's cookie.

== Server -> User Agent ==
Set-Cookie: lang=en-US; Expires=Wed, 09 Jun 2021 10:18:14 GMT

== User Agent -> Server ==
Cookie: lang=en-US

Finally, to remove a cookie, the server returns a Set-Cookie header with an expiration date in the past. The server will be successful in removing the cookie only if the Path and the Domain attribute in the Set-Cookie header match the values used when the cookie was created.

== Server -> User Agent ==
Set-Cookie: lang=; Expires=Sun, 06 Nov 1994 08:49:37 GMT

== User Agent -> Server ==
(No Cookie header)

4. A Well-Behaved Profile

This section describes the syntax and semantics of a well-behaved profile of the protocol. Servers SHOULD use the profile described in this section, both to maximize interoperability with existing user agents and because a future version of the cookie protocol could remove support for some of the most esoteric aspects of the protocol. User agents, however, MUST implement the full protocol to ensure interoperability with servers making use of the full protocol.

6. Implementation Considerations

6.1. Limits

Practical user agent implementations have limits on the number and size of cookies that they can store. General-use user agents SHOULD provide each of the following minimum capabilities:

  • At least 4096 bytes per cookie (as measured by the sum of the length of the cookie's name, value, and attributes).
  • At least 50 cookies per domain.
  • At least 3000 cookies total.

Servers SHOULD use as few and as small cookies as possible to avoid reaching these implementation limits and to avoid network latency due to the Cookie header being included in every request.

Servers should gracefully degrade if the user agent fails to return one or more cookies in the Cookie header because the user agent might evict any cookie at any time on orders from the user.

6.2. Application Programmer Interfaces

One reason the cookie protocol uses such an esoteric syntax is because many platforms (both in servers and user agents) provide string-based application programmer interfaces (APIs), requiring application-layer programmers to generate and parse the syntax used by the cookie protocol.

Instead of providing string-based APIs to the cookie protocols, implementations would be well-served by providing more semantic APIs. It is beyond the scope of this document to recommend specific API designs, but there are clear benefits to accepting a abstract "Date" object instead of a serialized date string.

7. Security Considerations

7.1. Overview

The cookie protocol has a number of security and privacy pitfalls.

In particular, cookies encourage developers to rely on ambient authority for authentication, often creating vulnerabilities such as cross-site request forgery. When storing session identifiers in cookies, developers often create session fixation vulnerabilities.

Transport-layer encryption, such as that employed in HTTPS, is insufficient to prevent a network attacker from obtaining or altering a victim's cookies because the cookie protocol itself has various vulnerabilities (see "Weak Confidentiality" and "Weak Integrity", below). In addition, by default, the cookie protocol does not provide confidentiality or integrity from network attackers, even when used in conjunction with HTTPS.

7.2. Ambient Authority

A server that uses cookies to authenticate users can suffer security vulnerabilities because some user agents let remote parties issue HTTP requests from the user agent (e.g., via HTTP redirects and HTML forms). When issuing those requests, user agent attaches cookies even if the entity does not know the contents of the cookies, possibly letting the remote entity exercise authority at an unwary server.

Although this security concern goes by a number of names (e.g., cross-site request forgery, confused deputy), the issue stems from cookies being a form of ambient authority. Cookies encourage server operators to separate designation (in the form of URLs) from authorization (in the form of cookies). Consequently, the user agent might supply the authorization for a resource designated by the attacker, possibly causing the server or its clients to undertake actions designated by the attacker as though they were authorized by the user.

Instead of using cookies for authorization, server operators might wish to consider entangling designation and authorization by treating URLs as capabilities. Instead of storing secrets in cookies, this approach stores secrets in URLs, requiring the remote entity to supply the secret itself. Although this approach is not a panacea, judicious use of these principles can lead to more robust security.

7.3. Clear Text

Unless sent over a secure channel (such as TLS), the information in the Set-Cookie and Cookie headers is transmitted in the clear.

  1. All sensitive information conveyed in these headers is exposed to an eavesdropper.
  2. A malicious intermediary could alter the headers as they travel in either direction, with unpredictable results.
  3. A malicious client could alter the Cookie header before transmission, with unpredictable results.

Servers SHOULD encrypt and sign the contents of cookies when transmitting them to the user agent (even when sending the cookies over a secure channel). However, encrypting and signing cookie contents does not prevent an attacker from transplanting a cookie from one user agent to another or from replaying the cookie at a later time.

In addition to encrypting and signing the contents of every cookie, servers that require a higher level of security SHOULD use the cookie protocol only over a secure channel. When using the cookie protocol over a secure channel, servers SHOULD set the Secure attribute in every cookie. If a server does not set the Secure attribute, the protection provided by the secure channel will be largely moot.

7.4. Session Identifiers

Instead of storing session information directly in a cookie (where it might be exposed to or replayed by an attacker), servers commonly store a nonce (or "session identifier") in a cookie. When the server receives an HTTP request with a nonce, the server can look up state information associated with the cookie using the nonce as a key.

Using session identifier cookies limits the damage an attacker can cause if the attacker learns the contents of a cookie because the nonce is useful only for interacting with the server (unlike non-nonce cookie content, which might itself be sensitive). Furthermore, using a single nonce prevents an attacker from "splicing" together cookie content from two interactions with the server, which could cause the server to behave unexpectedly.

Using session identifiers is not without risk. For example, the server SHOULD take care to avoid "session fixation" vulnerabilities. A session fixation attack proceeds in three steps. First, the attacker transplants a session identifier from his or her user agent to the victim's user agent. Second, the victim uses that session identifier to interact with the server, possibly imbuing the session identifier with the user's credentials or confidential information. Third, the attacker uses the session identifier to interact with server directly, possibly obtaining the user's authority or confidential information.

7.5. Weak Confidentiality

Cookies do not provide isolation by port. If a cookie is readable by a service running on one port, the cookie is also readable by a service running on another port of the same server. If a cookie is writable by a service on one port, the cookie is also writable by a service running on another port of the same server. For this reason, servers SHOULD NOT both run mutually distrusting services on different ports of the same host and use cookies to store security-sensitive information.

Cookies do not provide isolation by scheme. Although most commonly used with the http and https schemes, the cookies for a given host might also available to other schemes, such as ftp and gopher. Although this lack of isolation by scheme is most apparent in via non-HTTP APIs that permit access to cookies (e.g., HTML's document.cookie API), the lack of isolation by scheme is actually present in the cookie protocol itself (e.g., consider retrieving a URI with the gopher scheme via HTTP).

Cookies do not always provide isolation by path. Although the network-level protocol does not send cookie stored for one path to another, some user agents expose cookies via non-HTTP APIs, such as HTML's document.cookie API. Because some of these user agents (e.g., web browsers) do not isolate resources received from different paths, a resource retrieved from one path might be able to access cookies stored for another path.

7.6. Weak Integrity

Cookies do not provide integrity guarantees for sibling domains (and their subdomains). For example, consider and The server can set a cookie with a Domain attribute of "" (possibly overwriting an existing "" cookie set by, and the user agent will include that cookie in HTTP requests to In the worst case, will be unable to distinguish this cookie from a cookie it set itself. The server might be able to leverage this ability to mount an attack against

Even though the cookie protocol supports the Path attribute, the Path attribute does not provide any integrity protection because the user agent will accept an arbitrary Path attribute in a Set-Cookie header. For example, an HTTP response to a request for can set a cookie with a Path attribute of "/qux". Consequently, servers SHOULD NOT both run mutually distrusting services on different paths of the same host and use cookies store security sensitive information.

An active network attacker can also inject cookies into the Cookie header sent to by impersonating a response from and injecting a Set-Cookie header. The HTTPS server at will be unable to distinguish these cookies from cookies that it set itself in an HTTPS response. An active network attacker might be able to leverage this ability to mount an attack against even if uses HTTPS exclusively.

Servers can partially mitigate these attacks by encrypting and signing the contents of their cookies. However, using cryptography does not mitigate the issue completely because an attacker can replay a cookie he or she received from the authentic server in the user's session, with unpredictable results.

Finally, an attacker might be able to force the user agent to delete cookies by storing large number of cookies. Once the user agent reaches its storage limit, the user agent will be forced to evict some cookies. Servers SHOULD NOT rely upon user agents retaining cookies.

7.7. Reliance on DNS

The cookie protocol relies upon the Domain Name System (DNS) for security. If the DNS is partially or fully compromised, the cookie protocol might fail to provide the security properties required by applications.

8. References

8.2. Informative References

Kristol, D. and L. Montulli, “HTTP State Management Mechanism”, RFC 2109, February 1997.

Appendix A. Acknowledgements

This document borrows heavily from RFC 2109 [RFC2109]. We are indebted to David M. Kristol and Lou Montulli for their efforts to specify the cookie protocol. David M. Kristol, in particular, provided invaluable advice on navigating the IETF process. We would also like to thank Thomas Broyer, Tyler Close, Bil Corry, corvid, Roy T. Fielding, Blake Frantz, Eran Hammer-Lahav, Jeff Hodges, Achim Hoffmann, Georg Koppen, Dean McNamee, Mark Miller, Yngve N. Pettersen, Julian Reschke, Mark Seaborn, Maciej Stachowiak, Daniel Stenberg, David Wagner, Dan Winship, and Dan Witte for their valuable feedback on this document.

Author's Address

Adam Barth
University of California, Berkeley