| Internet Engineering Task Force | S. Hollenbeck |
| Internet-Draft | Verisign Labs |
| Intended status: Best Current Practice | November 19, 2015 |
| Expires: May 22, 2016 |
Federated Authentication for the Registration Data Access Protocol (RDAP) using OpenID Connect
draft-hollenbeck-weirds-rdap-openid-03
The Registration Data Access Protocol (RDAP) provides "RESTful" web services to retrieve registration metadata from domain name and regional internet registries. RDAP allows a server to make access control decisions based on client identity, and as such it includes support for client identification features provided by the Hypertext Transfer Protocol (HTTP). Identification methods that require clients to obtain and manage credentials from every RDAP server operator present management challenges for both clients and servers, whereas a federated authentication system would make it easier to operate and use RDAP without the need to maintain server-specific client credentials. This document describes a federated authentication system for RDAP based on OpenID Connect.
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 May 22, 2016.
Copyright (c) 2015 IETF Trust and the persons identified as the document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.
The Registration Data Access Protocol (RDAP) provides "RESTful" web services to retrieve registration metadata from domain name and regional internet registries. RDAP allows a server to make access control decisions based on client identity, and as such it includes support for client identification features provided by the Hypertext Transfer Protocol (HTTP) [RFC7230].
RDAP is specified in multiple documents, including "HTTP Usage in the Registration Data Access Protocol (RDAP)" [RFC7480], "Security Services for the Registration Data Access Protocol (RDAP)" [RFC7481], "Registration Data Access Protocol Query Format" [RFC7482], and "JSON Responses for the Registration Data Access Protocol (RDAP)" [RFC7483]. RFC 7481 describes client identification and authentication services that can be used with RDAP, but it does not specify how any of these services can (or should) be used with RDAP.
The traditional "user name and password" authentication method does not scale well in the RDAP ecosystem. Assuming that all domain name and address registries will eventually provide RDAP service, it is impractical and inefficient for users to secure login credentials from the hundreds of different server operators. Authentication methods based on user names and passwords do not provide information that describes the user in sufficient detail (while protecting the personal privacy of the user) for server operators to make fine-grained access control decisions based on the user's identity. The authentication system used for RDAP needs to address all of these needs.
A basic level of RDAP service can be provided to users who possess an identifier issued by a recognized provider who is able to authenticate and validate the user. The identifiers issued by social media services, for example, can be used. Users who require higher levels of service (and who are willing to share more information about them self to gain access to that service) can secure identifiers from specialized providers who are or will be able to provide more detailed information about the user. Server operators can then make access control decisions based on the identification information provided by the user.
A federated authentication system would make it easier to operate and use RDAP by re-using existing identifiers to provide a basic level of access. It can also provide the ability to collect additional user identification information, and that information can be shared with the consent of the user. This document describes a federated authentication system for RDAP based on OpenID Connect [OIDC] that meets all of these needs.
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 [RFC2119].
RDAP itself does not include native security services. Instead, RDAP relies on features that are available in other protocol layers to provide needed security services including access control, authentication, authorization, availability, data confidentiality, data integrity, and identification. A description of each of these security services can be found in "Internet Security Glossary, Version 2" [RFC4949]. This document focuses on a federated authentication system for RDAP that provides services for authentication, authorization, and identification, allowing a server operator to make access control decisions. Section 3 of RFC 7481 [RFC7481] describes general considerations for RDAP access control, authentication, and authorization.
The traditional client-server authentication model requires clients to maintain distinct credentials for every RDAP server. This situation can become unwieldy as the number of RDAP servers increases. Federated authentication mechanisms allow clients to use one credential to access multiple RDAP servers and reduce client credential management complexity.
OpenID Connect 1.0 [OIDCC] is a decentralized, single sign-on (SSO) federated authentication system that allows users to access multiple web resources with one identifier instead of having to create multiple server-specific identifiers. Users acquire identifiers from Identity Providers, or IDPs. Relying Parties, or RPs, are applications (such as RDAP) that outsource their user authentication function to an IDP. OpenID Connect is built on top of the authorization framework provided by the OAuth 2.0 [RFC6749] protocol.
The OAuth authorization framework describes a method for users to access protected web resources without having to hand out their credentials. Instead, clients are issued access tokens by authorization servers with the permission of the resource owners. Using OpenID Connect and OAuth, multiple RDAP servers can form a federation and clients can access any server in the federation by providing one credential registered with any IDP in that federation. The OAuth authorization framework is designed for use with HTTP and thus can be used with RDAP.
This document uses the terms "client" and "server" defined by RDAP [RFC7480]. An RDAP client performs the role of an OpenID Connect Core [OIDCC] Entity or End-User. An RDAP server performs the role of an OpenID Connect Core Relying Party (RP). Additional terms from Section 1.2 of the OpenID Connect Core specification are incorporated by reference.
At a high level, RDAP authentication of a browser-based client using OpenID Connect requires completion of the following steps:
The RDAP server can then make identification, authorization, and access control decisions based on local policies, the ID Token received from the OP, and the received Claims. Note that OpenID Connect describes different process flows for other types of clients, such as script-based or command line clients.
End-Users MUST possess an identifier (an OpenID) issued by an OP to use OpenID Connect with RDAP. The OpenID Foundation maintains a list of OPs on its web site.
OpenID Connect requires RPs to register with OPs to use OpenID Connect services for an End-User. That process is REQUIRED and is described by the "OpenID Connect Dynamic Client Registration" protocol [OIDCR].
An RDAP server/RP needs to receive an identifier from an End-User that can be used to discover the End-User's OP. That process is REQUIRED and is documented in the "OpenID Connect Discovery" protocol [OIDCD].
Once the OP is known, an RP MUST form an Authentication Request and send it to the OP as described in Section 3 of the OpenID Connect Core protocol [OIDCC]. The authentication path followed (authorization, implicit, or hybrid) will depend on the Authentication Request response_type set by the RP. The remainder of the processing steps described here assume that the Authorization Flow is being used by setting "response_type=code" in the Authentication Request.
The benefits of using the Authorization Flow for authenticating a human user are described in Section 3.1 of the OpenID Connect Core protocol. The Implicit Flow is more commonly used by clients implemented in a web browser using a scripting language; it is described in Section 3.2 of the OpenID Connect Core protocol. The Hybrid Flow (described in Section 3.3 of the OpenID Connect Core protocol) combines elements of the Authorization and Implicit Flows by returning some tokens from the Authorization Endpoint and others from the Token Endpoint.
TBD: which of the optional request parameters should be specified?
The OP receives the Authentication Request and attempts to validate it as described in Section 3.1.2.2 of the OpenID Connect Core protocol [OIDCC]. If the request is valid, the OP attempts to authenticate the End-User as described in Section 3.1.2.3 of the OpenID Connect Core protocol [OIDCC]. The OP returns an error response if the request is not valid or if any error is encountered.
After the End-User is authenticated, the OP MUST obtain authorization information from the End-User before releasing information to the RDAP Server/RP. This process is described in Section 3.1.2.4 of the OpenID Connect Core protocol [OIDCC].
After the End-User is authenticated, the OP will send a response to the RP that describes the result of the authorization process in the form of an Authorization Grant. The RP MUST validate the response. This process is described in Sections 3.1.2.5 - 3.1.2.7 of the OpenID Connect Core protocol [OIDCC].
The RP sends a Token Request using the Authorization Grant to a Token Endpoint to obtain a Token Response containing an Access Token, ID Token, and an OPTIONAL Refresh Token. The RP MUST validate the Token Response. This process is described in Sections 3.1.3 - 3.1.3.8 of the OpenID Connect Core protocol [OIDCC].
The set of Claims can be retrieved by sending a request to a UserInfo Endpoint using the Access Token. The Claims MAY be returned in the ID Token. The process of retrieving Claims from a UserInfo Endpoint is described in Sections 5.3 - 5.3.4 of the OpenID Connect Core protocol [OIDCC].
OpenID Connect specified a set of standard Claims in Section 5.1. Additional Claims for RDAP are described in Section 3.1.4.1.
OpenID Connect claims are pieces of information used to make assertions about an entity. Section 5 of the OpenID Connect Core protocol [OIDCC] describes a set of standard claims that can be used to identify a person. Section 5.1.2 notes that additional claims MAY be used, and it describes a method to create them.
There are communities of RDAP users and operators who wish to make and validate claims about a user's "need to know" when it comes to requesting access to a resource. For example, a law enforcement agent or a trademark attorney may wish to be able to assert that they have a legal right to access a protected resource, and a server operator will need to be able to receive and validate that claim. These needs can be met by defining and using an additional "purpose" claim.
The "purpose" claim identifies the purpose for which access to a protected resource is being requested. The processing of this claim is subject the server acceptance of the purpose and successful authentication of the End-User. The "purpose" value is a case-sensitive string containing a StringOrURI value as specified in Section 2 of the JSON Web Token (JWT) specification ([I-D.ietf-oauth-json-web-token]). Use of this claim is OPTIONAL.
Client authentication is requested by adding a query component to an RDAP request URI using the syntax described in Section 3.4 of RFC 3986 [RFC3986]. The query used to request client authentication is represented as a "key=value" pair using a key value of "authenticate" and a value component of "true", e.g. "authenticate=true". An example:
https://example.com/rdap/domain/example.com?authenticate=true
Unrecognized query parameters MAY be ignored. An RDAP request that does not include a "authenticate=true" query component MUST be processed as an unauthenticated query.
The flow described in Section 3.1.3 requires a client to interact with a server using a web browser. This will not work well in situations where the client is automated or using a command-line client such as curl or wget. This is a known issue with OpenID, and is typically addressed using a two-step process:
This topic needs additional research and discussion. Specific examples will be provided when an approach can be confirmed.
This document describes a proposal that is almost certainly incomplete. For the time being this section will serve as a place to capture unanswered questions, topics for future discussion, and anything else that might deserve additional text in the future.
Recursive or proxy RDAP servers: how might federated authentication work in a model where a subset of RDAP servers act as proxies to other RDAP servers? is it possible to cache user credentials in such a way that authentication process latency can be reduced?
Additional claims: are there any other claims that need to be defined and registered?
Implementations: does it make sense to add text describing existing implementations that can be used for experimentation?
Non-browser clients: how can they best be accommodated?
IANA is requested to register the following value in the JSON Web Token Claims Registry:
Security considerations for RDAP can be found in RFC 7481 [RFC7481]. Security considerations for OpenID Connect Core [OIDCC] and OAuth [RFC6749] can be found in their reference specifications. OpenID Connect defines optional mechanisms for robust signing and encryption that can be used to provide data integrity and data confidentiality services as needed.
Having completed the client identification, authorization, and validation process, an RDAP server can make access control decisions based on a comparison of client-provided information and local policy. For example, a client who provides an email address (and nothing more) might be entitled to receive a subset of the information that would be available to a client who provides an email address, a full name, and a stated purpose. Development of these access control policies is beyond the scope of this document.
The author would like to acknowledge the following individuals for their contributions to this document: Rhys Smith and Jaromir Talir.
| [RFC4949] | Shirey, R., "Internet Security Glossary, Version 2", FYI 36, RFC 4949, DOI 10.17487/RFC4949, August 2007. |