Internet-Draft | OAuth DPoP | April 2020 |
Fett, et al. | Expires 3 October 2020 | [Page] |
This document describes a mechanism for sender-constraining OAuth 2.0 tokens via a proof-of-possession mechanism on the application level. This mechanism allows for the detection of replay attacks with access and refresh tokens.¶
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 https://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 3 October 2020.¶
Copyright (c) 2020 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 (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.¶
[RFC8705] describes methods to bind (sender-constrain) access tokens using mutual Transport Layer Security (TLS) authentication with X.509 certificates.¶
[I-D.ietf-oauth-token-binding] provides mechanisms to sender-constrain access tokens using HTTP token binding.¶
Due to a sub-par user experience of TLS client authentication in user agents and a lack of support for HTTP token binding, neither mechanism can be used if an OAuth client is a Single Page Application (SPA) running in a web browser.¶
This document outlines an application-level sender-constraining for access and refresh tokens that can be used in cases where neither mTLS nor OAuth Token Binding are available. It uses proof-of-possession based on a public/private key pair and application-level signing.¶
DPoP can be used with public clients and, in case of confidential clients, can be combined with any client authentication method.¶
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here.¶
This specification uses the terms "access token", "refresh token", "authorization server", "resource server", "authorization endpoint", "authorization request", "authorization response", "token endpoint", "grant type", "access token request", "access token response", and "client" defined by The OAuth 2.0 Authorization Framework [RFC6749].¶
Under the attacker model defined in [I-D.ietf-oauth-security-topics], the mechanism defined by this specification aims to prevent token replay at a different endpoint.¶
More precisely, if an adversary is able to get hold of an access token or refresh token because it set up a counterfeit authorization server or resource server, the adversary is not able to replay the respective token at another authorization or resource server.¶
The main data structure introduced by this specification is a DPoP proof JWT, described in detail below. A client uses a DPoP proof JWT to prove the possession of a private key belonging to a certain public key. Roughly speaking, a DPoP proof is a signature over some data of the HTTP request to which it is attached to and a timestamp.¶
Figure 1: Basic DPoP Flow¶
The basic steps of an OAuth flow with DPoP are shown in Figure 1:¶
(A) In the Token Request, the client sends an authorization code to the authorization server in order to obtain an access token (and potentially a refresh token). The client attaches a DPoP proof to the request in an HTTP header.¶
(B) The AS binds (sender-constrains) the access token to the
public key claimed by the client in the DPoP proof; that is, the access token cannot
be used without proving possession of the respective private key.
This is signaled to the client by using the token_type
value
DPoP
.¶
If a refresh token is issued to a public client, it is
sender-constrained in the same way. For confidential clients,
refresh tokens are bound to the client_id
, which is more
flexible than binding it to a particular public key.¶
(C) If the client wants to use the access token, it has to prove possession of the private key by, again, adding a header to the request that carries the DPoP proof. The resource server needs to receive information about the public key to which the access token is bound. This information is either encoded directly into the access token (for JWT structured access tokens), or provided at the token introspection endpoint of the authorization server (not shown).¶
(D) The resource server refuses to serve the request if the signature check fails or the data in the DPoP proof is wrong, e.g., the request URI does not match the URI claim in the DPoP proof JWT.¶
When a refresh token that is sender-constrained using DPoP is used by the client, the client has to provide a DPoP proof just as in the case of a resource access. The new access token will be bound to the same public key.¶
The mechanism presented herein is not a client authentication method.
In fact, a primary use case is public clients (single page
applications) that do not use client authentication. Nonetheless, DPoP
is designed such that it is compatible with private_key_jwt
and all
other client authentication methods.¶
DPoP does not directly ensure message integrity but relies on the TLS layer for that purpose. See Section 9 for details.¶
DPoP uses so-called DPoP proof JWTs for binding public keys and proving knowledge about private keys.¶
A DPoP proof is a JWT ([RFC7519]) that is signed (using JWS, [RFC7515]) using a private key chosen by the client (see below). The header of a DPoP JWT contains at least the following parameters:¶
typ
: type header, value dpop+jwt
(REQUIRED).¶
alg
: a digital signature algorithm identifier as per [RFC7518]
(REQUIRED). MUST NOT be none
or an identifier for a symmetric
algorithm (MAC).¶
jwk
: representing the public key chosen by the client, in JWK
format, as defined in [RFC7515] (REQUIRED)¶
The body of a DPoP proof contains at least the following claims:¶
jti
: Unique identifier for the DPoP proof JWT (REQUIRED).
The value MUST be assigned such that there is a negligible
probability that the same value will be assigned to any
other DPoP proof used in the same context during the time window of validity.
Such uniqueness can be accomplished by encoding (base64url or any other
suitable encoding) at least 96 bits of
pseudorandom data or by using a version 4 UUID string according to [RFC4122].
The jti
SHOULD be used by the server for replay
detection and prevention, see Section 9.1.¶
htm
: The HTTP method for the request to which the JWT is
attached, as defined in [RFC7231] (REQUIRED).¶
htu
: The HTTP URI used for the request, without query and
fragment parts (REQUIRED).¶
iat
: Time at which the JWT was created (REQUIRED).¶
Figure 2 shows the JSON header and payload of a DPoP proof JWT.¶
Figure 2: Example JWT content for DPoP
proof header.¶
Note: To keep DPoP simple to implement, only the HTTP method and URI are signed in DPoP proofs. Nonetheless, DPoP proofs can be extended to contain other information of the HTTP request (see also Section 9.4).¶
To check if a string that was received as part of an HTTP Request is a valid DPoP proof, the receiving server MUST ensure that¶
the string value is a well-formed JWT,¶
all required claims are contained in the JWT,¶
the typ
field in the header has the value dpop+jwt
,¶
the algorithm in the header of the JWT indicates an asymmetric digital
signature algorithm, is not none
, is supported by the
application, and is deemed secure,¶
that the JWT is signed using the public key contained in the jwk
header of the JWT,¶
the htm
claim matches the HTTP method value of the HTTP
request in which the JWT was received (case-insensitive),¶
the htu
claims matches the HTTP URI value for the HTTP
request in which the JWT was received, ignoring any query and
fragment parts,¶
the token was issued within an acceptable timeframe (see Section 9.1), and¶
that, within a reasonable consideration of accuracy and resource utilization,
a JWT with the same jti
value has not been received
previously (see Section 9.1).¶
Servers SHOULD employ Syntax-Based Normalization and Scheme-Based
Normalization in accordance with Section 6.2.2. and Section 6.2.3. of
[RFC3986] before comparing the htu
claim.¶
To bind a token to a public key in the token request, the client MUST
provide a valid DPoP proof JWT in a DPoP
header. The HTTPS request shown
in Figure 3 illustrates the protocol for this (with extra line breaks
for display purposes only).¶
Figure 3: Token Request for a DPoP sender-constrained token.¶
The HTTP header DPoP
MUST contain a valid DPoP proof.¶
The authorization server, after checking the validity of the DPoP proof,
MUST associate the access token issued at the token endpoint with the
public key. It then sets token_type
to DPoP
in the token
response.¶
A client typically cannot know whether a certain AS supports DPoP. It
therefore SHOULD use the value of the token_type
parameter returned
from the AS to determine support for DPoP: If the token type returned
is Bearer
or another value, the AS does not support DPoP. If it is
DPoP
, DPoP is supported. Only then, the client needs to send
the DPoP
header in subsequent requests and use the token type
DPoP
in the Authorization
header as described below.¶
If a refresh token is issued to a public client at the token endpoint and a valid DPoP proof is presented, the refresh token MUST be bound to the public key contained in the header of the DPoP proof JWT.¶
If a DPoP-bound refresh token is to be used at the token endpoint by a public client, the AS MUST ensure that the DPoP proof contains the same public key as the one the refresh token is bound to. The access token issued MUST be bound to the public key contained in the DPoP proof.¶
To make use of an access token that is token-bound to a public key
using DPoP, a client MUST prove the possession of the corresponding
private key by providing a DPoP proof in the DPoP
request header.¶
The DPoP-bound access token must be sent in the Authorization
header
with the prefix DPoP
.¶
If a resource server detects that an access token that is to be used
for resource access is bound to a public key using DPoP (via the
methods described in Section 7) it MUST check that a header
DPoP
was received in the HTTP request, and check the header's
contents according to the rules in Section 4.2.¶
The resource server MUST NOT grant access to the resource unless all checks are successful.¶
Figure 4: Protected Resource Request with a DPoP sender-constrained access token.¶
It MUST be ensured that resource servers can reliably identify whether a token is bound using DPoP and learn the public key to which the token is bound.¶
Access tokens that are represented as JSON Web Tokens (JWT) [RFC7519]
MUST contain information about the DPoP public key (in JWK format) in
the member jkt
of the cnf
claim, as shown in Figure 5.¶
The value in jkt
MUST be the base64url encoding [RFC7515] of
the JWK SHA-256 Thumbprint (according to [RFC7638]) of the public
key to which the access token is bound.¶
Figure 5: Example access token body with cnf
claim.¶
When access token introspection is used, the same cnf
claim as above
MUST be contained in the introspection response.¶
Resource servers MUST ensure that the fingerprint of the public key in
the DPoP proof JWT equals the value in the jkt
claim in the access
token or introspection response.¶
We would like to thank David Waite, Filip Skokan, Mike Engan, and Justin Richer for their valuable input and feedback.¶
This document resulted from discussions at the 4th OAuth Security Workshop in Stuttgart, Germany. We thank the organizers of this workshop (Ralf Kusters, Guido Schmitz).¶
In DPoP, the prevention of token replay at a different endpoint (see Section 2) is achieved through the binding of the DPoP proof to a certain URI and HTTP method. DPoP does not, however, achieve the same level of protection as TLS-based methods such as OAuth Mutual TLS [RFC8705] or OAuth Token Binding [I-D.ietf-oauth-token-binding] (see also Section 9.1 and Section 9.4). TLS-based mechanisms can leverage a tight integration between the TLS layer and the application layer to achieve a very high level of message integrity and replay protection. Therefore, it is RECOMMENDED to prefer TLS-based methods over DPoP if such methods are suitable for the scenario at hand.¶
If an adversary is able to get hold of a DPoP proof JWT, the adversary
could replay that token at the same endpoint (the HTTP endpoint
and method are enforced via the respective claims in the JWTs). To
prevent this, servers MUST only accept DPoP proofs for a limited time
window after their iat
time, preferably only for a relatively brief period.
Servers SHOULD store the jti
value of each DPoP proof for the time window in
which the respective DPoP proof JWT would be accepted and decline HTTP requests
for which the jti
value has been seen before. In order to guard against
memory exhaustion attacks a server SHOULD reject DPoP proof JWTs with unnecessarily
large jti
values or store only a hash thereof.¶
Note: To accommodate for clock offsets, the server MAY accept DPoP
proofs that carry an iat
time in the near future (e.g., up to a few
seconds in the future).¶
Servers accepting signed DPoP proof JWTs MUST check the typ
field in the
headers of the JWTs to ensure that adversaries cannot use JWTs created
for other purposes in the DPoP headers.¶
Implementers MUST ensure that only asymmetric digital signature algorithms that
are deemed secure can be used for signing DPoP proofs. In particular,
the algorithm none
MUST NOT be allowed.¶
DPoP does not ensure the integrity of the payload or headers of requests. The signature of DPoP proofs only contains the HTTP URI and method, but not, for example, the message body or other request headers.¶
This is an intentional design decision to keep DPoP simple to use, but as described, makes DPoP potentially susceptible to replay attacks where an attacker is able to modify message contents and headers. In many setups, the message integrity and confidentiality provided by TLS is sufficient to provide a good level of protection.¶
Implementers that have stronger requirements on the integrity of messages are encouraged to either use TLS-based mechanisms or signed requests. TLS-based mechanisms are in particular OAuth Mutual TLS [RFC8705] and OAuth Token Binding [I-D.ietf-oauth-token-binding].¶
Note: While signatures on (parts of) requests are out of the scope of this specification, signatures or information to be signed can be added into DPoP proofs.¶
This specification registers the following access token type in the OAuth Access Token Types registry defined in [RFC6749].¶
This specification registers the dpop+jwt
type value in the IANA
JSON Web Signature and Encryption Type Values registry [RFC7515]:¶
[[ To be removed from the final specification ]]¶
-00 [[ Working Group Draft ]]¶
Working group draft¶
-04¶
-03¶
rework the text around uniqueness requirements on the jti claim in the DPoP proof JWT¶
make tokens a bit smaller by using htm
, htu
, and jkt
rather than http_method
, http_uri
, and jkt#S256
respectively¶
more explicit recommendation to use mTLS if that is available¶
added David Waite as co-author¶
editorial updates¶
-02¶
added normalization rules for URIs¶
removed distinction between proof and binding¶
"jwk" header again used instead of "cnf" claim in DPoP proof¶
renamed "Bearer-DPoP" token type to "DPoP"¶
removed ability for key rotation¶
added security considerations on request integrity¶
explicit advice on extending DPoP proofs to sign other parts of the HTTP messages¶
only use the jkt#S256 in ATs¶
iat instead of exp in DPoP proof JWTs¶
updated guidance on token_type evaluation¶
-01¶
fixed inconsistencies¶
moved binding and proof messages to headers instead of parameters¶
extracted and unified definition of DPoP JWTs¶
improved description¶
-00¶
first draft¶