Version: | 2.0 |
Date: | 2017-7-26 |
Editor: | Eve Maler, ForgeRock |
Authors: | Maciej Machulak, Self |
Justin Richer, Bespoke Engineering |
This specification defines a means for an UMA-enabled authorization server and resource server to be loosely coupled, or federated, in a resource owner context.
This technical specification is a Draft Recommendation produced by the User-Managed Access Work Group. See the Kantara Initiative Operating Procedures for more information.
Copyright © 2017 Kantara Initiative and the persons identified as the document authors. All rights reserved.
This document is subject to the Kantara IPR Policy - Option Patent & Copyright: Reciprocal Royalty Free with Opt-Out to Reasonable And Non discriminatory (RAND) (HTML version).
This specification extends and complements [UMAGrant] to loosely couple, or federate, the authorization process. This enables multiple resource servers operating in different domains to communicate with a single authorization server operating in yet another domain that acts on behalf of a resource owner. A service ecosystem can thus automate resource protection, and the resource owner can monitor and control authorization grant rules at a central service location over time. Further, authorization grants can increase and decrease at the level of individual resources and scopes.
Building on the example provided in the introduction in [UMAGrant], bank customer (resource owner) Alice has a bank account service (resource server), a cloud file system (different resource server hosted elsewhere), and a dedicated sharing management service (authorization server) hosted by the bank. She can manage access to her various protected resources by spouse Bob, accounting professional Charline, bank account aggregation company DecideAccount, and neighbor Erik (requesting parties), all using different client applications, each of them variously to view account data, get access to payment or withdrawal functions, and see or edit files for a community project. Alice can use the same sharing management service to monitor and control these different levels of access.
This specification, together with [UMAGrant], constitutes UMA 2.0. This specification is OPTIONAL to use with the UMA grant.
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 [RFC2119].
Unless otherwise noted, all parameter names and values are case sensitive. JSON [RFC7159] data structures defined in this specification MAY contain extension parameters that are not defined in this specification. Any entity receiving or retrieving a JSON data structure SHOULD ignore extension parameters it is unable to understand. Extension names that are unprotected from collisions are outside the scope of this specification.
The UMA grant defined in [UMAGrant] enhances the abstract protocol flow of OAuth. This specification enhances the UMA grant by defining formal communications between the UMA-enabled authorization server and resource server as they act on behalf of the resource owner, responding to authorization and resource requests, respectively, by a client that is acting on behalf of a requesting party.
A summary of UMA 2.0 communications, combining the UMA grant with federated authorization, is shown in Figure 1.
+------------------+ | resource | +------------manage (out of scope)----| owner | | +------------------+ | | | protection | | API access control | token (PAT) (out of scope) | | v v +------------+ +----------+------------------+ | | |protection| | | resource | | API | authorization | | server |<-----protect-------| (needs | server | | | | PAT) | | +------------+ +----------+------------------+ | protected | | UMA | | resource | | grant | |(needs RPT) | requesting | (PCT optional) | +------------+ party token +------------------+ ^ (RPT) ^ persisted ^ | | claims | | push token | | claim (PCT) | | tokens interact | +--------+ for +------------access--------------------| client | claims +--------+ gathering +---------------+ | requesting | | party | +---------------+
Figure 1: Federated Authorization Enhancements to UMA Grant Flow
This specification uses all of the terms and concepts in [UMAGrant]. This figure introduces the following new concepts:
This specification is designed for use with HTTP [RFC2616], and for interoperability and security in the context of loosely coupled services and applications operated by independent parties in independent domains. The use of UMA over any protocol other than HTTP is undefined. In such circumstances, it is RECOMMENDED to define profiles or extensions to achieve interoperability among independent implementations (see Section 4 of [UMAGrant]).
The authorization server MUST use TLS protection over its protection API endpoints, as governed by [BCP195], which discusses deployment and adoption characteristics of different TLS versions.
The authorization server MUST use OAuth and require a PAT to secure its protection API endpoints.
As defined in [UMAGrant], the resource owner -- the entity here authorizing PAT issuance -- MAY be an end-user (natural person) or a non-human entity treated as a person for limited legal purposes (legal person), such as a corporation. A PAT is unique to a resource owner, resource server used for resource management, and authorization server used for protection of those resources. The issuance of the PAT represents the authorization of the resource owner for the resource server to use the authorization server for protecting those resources.
Different grant types for PAT issuance might be appropriate for different types of resource owners; for example, the client credentials grant is useful in the case of an organization acting as a resource owner, whereas an interactive grant type is typically more appropriate for capturing the approval of an end-user resource owner. Where an identity token is desired in addition to an access token, it is RECOMMENDED to use [OIDCCore] in addition.
Federation of authorization for the UMA grant delivers a conceptual separation of responsibility and authority:
The separation of authorization decision making and authorization enforcement is similar to the architectural separation often used in enterprises between policy decision points and policy enforcement points. However, the resource server MAY apply additional authorization controls beyond those imposed by the authorization server. For example, even if an RPT provides sufficient permissions for a particular case, the resource server can choose to bar access to certain additional requesting parties.
Practical control of access among loosely coupled parties typically requires more than just messaging protocols. It is outside the scope of this specification to define more than the technical contract between UMA-conforming entities. Laws may govern authorization-granting relationships. It is RECOMMENDED for the resource owner, authorization server, and resource server to establish agreements about which parties are responsible for establishing and maintaining authorization grant rules and other authorization rules on a legal or contractual level, and parties operating entities claiming to be UMA-conforming should provide documentation of rights and obligations between and among them. See Section 4 of [UMAGrant] for more information.
The setting of policy conditions is outside the scope of this specification.
Except for PAT issuance, the resource owner-authorization server and resource owner-resource server interfaces are outside the scope of this specification. A variety of flows and user interfaces for policy condition setting involving user agents for both of these servers are possible, each with different privacy consequences for end-user resource owners. Some elements of the protection API enable the building of user interfaces for policy condition setting (see Section 3).
A variety of authorization, security, and time-to-live policies could be managed on a per-resource owner basis or a per-authorization server basis, as the entities see fit. Validity periods of PATs, RPTs, refresh tokens, permissions, caching periods for various responses, and even OAuth client credentials are all subject to management. Different time-to-live strategies might be suitable for different resources and scopes. These options are all outside the scope of this specification; where deployment ecosystems using independent implementations seek a consistent high level of security, profiling is RECOMMENDED (see Section 4 of [UMAGrant]).
Note: The resource server generally requires access to the protection API when an end-user resource owner is not available ("offline" access). Thus, the authorization server needs to manage the PAT in a way that ensures this outcome.
The protection API defines the following endpoints:
Use of these endpoints assumes that the resource server has acquired OAuth client credentials from the authorization server by static or dynamic means, and has a valid PAT.
The authorization server MUST declare its endpoints in the discovery document (see Section 2).
Examples in this specification show the use of bearer-style PATs in [RFC6750] format.
A permission is (requested or granted) authorized access to a particular resource with some number of scopes bound to that resource. The concept of permissions is used in authorization assessment, results calculation, and RPT issuance in [UMAGrant]. This concept takes on greater significance in relation to the protection API.
The resource server's resource registration operations at the authorization server result in a set of resource owner-specific resource identifiers. When the client makes a resource request that is unaccompanied by an access token or its resource request fails, the resource server is responsible for interpreting that request and mapping it to a choice of authorization server, resource owner, resource identifier(s), and set of scopes for each identifier, in order to request one or more permissions -- resource identifiers and a set of scopes -- and obtain a permission ticket on the client's behalf. Finally, when the client has made a resource request accompanied by an RPT and token introspection is in use, the returned token introspection object reveals the structure of permissions, potentially including expiration of individual permissions.
This specification makes use of the authorization server discovery document structure and endpoint defined in [UMAGrant]. The resource server uses this discovery document to discover the endpoints it needs.
In addition to the metadata defined in that specification and [OAuthMeta], this specification defines the following metadata for inclusion in the discovery document:
Following are additional requirements related to metadata:
The authorization server SHOULD document any profiled or extended features it supports explicitly, ideally by supplying the URI identifying each UMA profile and extension as an uma_profiles_supported metadata array value (defined in [UMAGrant]), and by using extension metadata to indicate specific usage details as necessary.
The API available at the resource registration endpoint enables the resource server to put resources under the protection of an authorization server on behalf of the resource owner and manage them over time. Protection of a resource at the authorization server begins on successful registration and ends on successful deregistration.
The resource server uses a RESTful API at the authorization server's resource registration endpoint to create, read, update, and delete resource descriptions, along with retrieving lists of such descriptions. The descriptions consist of JSON documents that are maintained as web resources at the authorization server. (Note carefully the similar but distinct senses in which the word "resource" is used in this section.)
Figure 2 illustrates the resource registration API operations.
authorization resource resource server server owner | | | |*PROTECTION API: | | |*Resource registration | | |endpoint/API | | | | | |*Create resource (POST)| | |<----------------------| | |*201 Created with | | |resource ID | | |---------------------->| | | | | |Set policy conditions (anytime | |before deletion/deregistration) | |<- - - - - - - - - - - - - - - - - - - -| | | | |*Read (GET) with | | |resource ID | | |<----------------------| | |*200 OK with resource | | |representation | | |---------------------->| | |*Update (PUT) with | | |resource ID | | |<----------------------| | |*200 OK with resource | | |ID | | |---------------------->| | |*List (GET) | | |<----------------------| | |*200 OK with list of | | |resource IDs | | |---------------------->| | |*Delete (DELETE) with | | |resource ID | | |<----------------------| | |*200 OK or 204 No | | |Content | | |---------------------->| |
Figure 2: Resource Registration Endpoint and API
The resource server MAY protect any subset of the resource owner's resources using different authorization servers or other means entirely, or to protect some resources and not others. Additionally, the choice of protection regimes MAY be made explicitly by the resource owner or implicitly by the resource server. Any such partitioning by the resource server or owner is outside the scope of this specification.
The resource server MAY register a single resource for protection that, from its perspective, has multiple parts, or has dynamic elements such as the capacity for querying or filtering, or otherwise has internal complexity. The resource server alone is responsible for maintaining any required mappings between internal representations and the resource identifiers and scopes known to the authorization server.
Note: The resource server is responsible for managing the process and timing of registering resources, maintaining the registration of resources, and deregistering resources at the authorization server. Motivations for updating a resource might include, for example, new scopes added at a new API version or resource owner actions at a service provider that result in new resource description text. See [UMA-Impl] for a discussion of initial resource registration timing options.
A resource description is a JSON document that describes the characteristics of a resource sufficiently for an authorization server to protect it. A resource description has the following parameters:
For example, this description characterizes a resource (a photo album) that can potentially be viewed or printed; the scope URI points to a scope description as defined in Section 3.1.1:
{ "resource_scopes":[ "view", "http://photoz.example.com/dev/scopes/print" ], "description":"Collection of digital photographs", "icon_uri":"http://www.example.com/icons/flower.png", "name":"Photo Album", "type":"http://www.example.com/rsrcs/photoalbum" }
A scope description is a JSON document that describes the characteristics of a scope sufficiently for an authorization server to protect the resource with this scope.
While a scope URI appearing in a resource description (see Section 3.1) MAY resolve to a scope description document, and thus scope description documents are possible to standardize and reference publicly, the authorization server is not expected to resolve scope description details at resource registration time or at any other run-time requirement. The resource server and authorization server are presumed to have negotiated any required interpretation of scope handling out of band.
A scope description has the following parameters:
For example, this scope description characterizes a scope that involves viewing (as opposed to, say, creating or editing in some fashion):
{ "description":"View, inspect, and zoom in on photos", "icon_uri":"http://www.example.com/icons/reading-glasses", "name":"View" }
The authorization server MUST support the following five registration options and MUST require a valid PAT for access to them; any other operations are undefined by this specification. Here, rreguri stands for the resource registration endpoint and _id stands for the authorization server-assigned identifier for the web resource corresponding to the resource at the time it was created, included within the URL returned in the Location header. Each operation is defined in its own section below.
Within the JSON body of a successful response, the authorization server includes common parameters, possibly in addition to method-specific parameters, as follows:
If the request to the resource registration endpoint is incorrect, then the authorization server instead responds as follows (see Section 6 for information about error messages):
Adds a new resource description to the authorization server using the POST method. If the request is successful, the resource is thereby registered and the authorization server MUST respond with an HTTP 201 status message that includes a Location header and an _id parameter.
Form of a create request, with a PAT in the header:
POST /rreg/ HTTP/1.1 Content-Type: application/json Authorization: Bearer MHg3OUZEQkZBMjcx ... { "resource_scopes":[ "read-public", "post-updates", "read-private", "http://www.example.com/scopes/all" ], "icon_uri":"http://www.example.com/icons/sharesocial.png", "name":"Tweedl Social Service", "type":"http://www.example.com/rsrcs/socialstream/140-compatible" }
Form of a successful response, also containing an optional user_access_policy_uri parameter:
HTTP/1.1 201 Created Content-Type: application/json Location: /rreg/KX3A-39WE ... { "_id":"KX3A-39WE", "user_access_policy_uri":"http://as.example.com/rs/222/resource/KX3A-39WE/policy" }
Reads a previously registered resource description using the GET method. If the request is successful, the authorization server MUST respond with an HTTP 200 status message that includes a body containing the referenced resource description, along with an _id parameter.
Form of a read request, with a PAT in the header:
GET /rreg/KX3A-39WE HTTP/1.1 Authorization: Bearer MHg3OUZEQkZBMjcx ...
Form of a successful response, containing all the parameters that were registered as part of the description:
HTTP/1.1 200 OK Content-Type: application/json ... { "_id":"KX3A-39WE", "resource_scopes":[ "read-public", "post-updates", "read-private", "http://www.example.com/scopes/all" ], "icon_uri":"http://www.example.com/icons/sharesocial.png", "name":"Tweedl Social Service", "type":"http://www.example.com/rsrcs/socialstream/140-compatible" }
If the referenced resource does not exist, the authorization server MUST produce an error response with an error parameter value of not_found, as defined in Section 3.2.
Updates a previously registered resource description, by means of a complete replacement of the previous resource description, using the PUT method. If the request is successful, the authorization server MUST respond with an HTTP 200 status message that includes an _id parameter.
Form of an update request adding a description parameter to a resource description that previously had none, with a PAT in the header:
PUT /rreg/9UQU-DUWW HTTP/1.1 Content-Type: application/json Authorization: Bearer 204c69636b6c69 ... { "resource_scopes":[ "http://photoz.example.com/dev/scopes/view", "public-read" ], "description":"Collection of digital photographs", "icon_uri":"http://www.example.com/icons/sky.png", "name":"Photo Album", "type":"http://www.example.com/rsrcs/photoalbum" }
Form of a successful response, not containing the optional user_access_policy_uri parameter:
HTTP/1.1 200 OK ... { "_id":"9UQU-DUWW" }
Deletes a previously registered resource description using the DELETE method. If the request is successful, the resource is thereby deregistered and the authorization server MUST respond with an HTTP 200 or 204 status message.
Form of a delete request, with a PAT in the header:
DELETE /rreg/9UQU-DUWW Authorization: Bearer 204c69636b6c69 ...
Form of a successful response:
HTTP/1.1 204 No content ...
Lists all previously registered resource identifiers for this resource owner using the GET method. The authorization server MUST return the list in the form of a JSON array of _id string values.
The resource server can use this method as a first step in checking whether its understanding of protected resources is in full synchronization with the authorization server's understanding.
Form of a list request, with a PAT in the header:
GET /rreg/ HTTP/1.1 Authorization: Bearer 204c69636b6c69 ...
Form of a successful response:
HTTP/1.1 200 OK ... [ "KX3A-39WE", "9UQU-DUWW" ]
The permission endpoint defines a means for the resource server to request one or more permissions (resource identifiers and corresponding scopes) with the authorization server on the client's behalf, and to receive a permission ticket in return, in order to respond as indicated in Section 3.2 of [UMAGrant]. The resource server uses this endpoint on the following occasions:
The use of the permission endpoint is illustrated in Figure 3.
authorization resource client server server | | | |Request resource (no or insufficient | |access token) | | |--------------------------------------->| | | | | |*PROTECTION API: | | |*Permission endpoint | | | | | |*Request permissions | | |(POST) | | |<--------------------| | |*201 Created with | | |permission ticket | | |-------------------->| | | | |401 response with permission ticket, | |authz server location | |<---------------------------------------|
Figure 3: Permission Endpoint
The PAT provided in the API request enables the authorization server to map the resource server's request to the appropriate resource owner. It is possible to request permissions for access to the resources of only one resource owner, protected by only one authorization server, at a time.
In its response, the authorization server returns a permission ticket for the resource server to give to the client that represents the same permissions that the resource server requested.
The process of choosing what permissions to request from the authorization server may require interpretation and mapping of the client's resource request. The resource server SHOULD request a set of permissions with scopes that is reasonable for the client's resource request. The resource server MAY request multiple permissions, and any permission MAY have zero scopes associated with it. Requesting multiple permissions might be appropriate, for example, in cases where the resource server expects the requesting party to need access to several related resources if they need access to any one of the resources. Requesting a permission with no scopes might be appropriate, for example, in cases where an access attempt involves an API call that is ambiguous without further context (role-based scopes such as user and admin could have this ambiguous quality, and an explicit client request for a particular scope at the token endpoint later can clarify the desired access). The resource server SHOULD document its intended pattern of permission requests in order to assist the client in pre-registering for and requesting appropriate scopes at the authorization server. See [UMA-Impl] for a discussion of permission request patterns.
Note: In order for the resource server to know which authorization server to approach for the permission ticket and on which resource owner's behalf (enabling a choice of permission endpoint and PAT), it needs to derive the necessary information using cues provided by the structure of the API where the resource request was made, rather than by an access token. Commonly, this information can be passed through the URI, headers, or body of the client's request. Alternatively, the entire interface could be dedicated to the use of a single resource owner and protected by a single authorization server.
The resource server uses the POST method at the permission endpoint. The body of the HTTP request message contains a JSON object (for a request containing a single resource identifier) or array of objects (for a request containing more than one resource identifier) providing the request, using a format derived from the resource description format specified in Section 3.1, as follows. The object used in both forms has the following parameters:
Example of an HTTP request for a single permission at the authorization server's permission endpoint, with a PAT in the header:
POST /perm HTTP/1.1 Content-Type: application/json Host: as.example.com Authorization: Bearer 204c69636b6c69 ... { "resource_id":"112210f47de98100", "resource_scopes":[ "view", "http://photoz.example.com/dev/actions/print" ] }
Example of an HTTP request for multiple permissions at the authorization server's permission endpoint, with a PAT in the header:
POST /perm HTTP/1.1 Content-Type: application/json Host: as.example.com Authorization: Bearer 204c69636b6c69 ... [ { "resource_id":"7b727369647d", "resource_scopes":[ "view", "crop", "lightbox" ] }, { "resource_id":"7b72736964327d", "resource_scopes":[ "view", "layout", "print" ] }, { "resource_id":"7b72736964337d", "resource_scopes":[ "http://www.example.com/scopes/all" ] } ]
If the authorization server is successful in creating a permission ticket in response to the resource server's request, it responds with an HTTP 201 (Created) status code and includes the ticket parameter in the JSON-formatted body. Regardless of whether the request contained one or multiple permissions, only a single permission ticket is returned.
For example:
HTTP/1.1 201 Created Content-Type: application/json ... { "ticket":"016f84e8-f9b9-11e0-bd6f-0021cc6004de" }
If the resource server's permission registration request is authenticated properly but fails due to other reasons, the authorization server responds with an HTTP 400 (Bad Request) status code and includes one of the following error codes (see Section 6 for more information about error codes and responses):
When the client makes a resource request accompanied by an RPT, the resource server needs to determine whether the RPT is active and, if so, its associated permissions. Depending on the nature of the RPT and operative caching parameters, the resource server MAY take any of the following actions as appropriate to determine the RPT's status:
The use of the token introspection endpoint is illustrated in Figure 4.
authorization resource client server server | | | |Resource request with RPT | |----------------------------------------->| | | | | |*PROTECTION API: | | |*Introspection endpoint| | | | | |*Request to introspect | | |token (POST) | | |<----------------------| | |*Response with token | | |introspection object | | |---------------------->| | | | |Protected resource | |<-----------------------------------------|
Figure 4: Token Introspection Endpoint
The authorization server MAY support both UMA-extended and non-UMA introspection requests and responses.
Note: In order for the resource server to know which authorization server, PAT (representing a resource owner), and endpoint to use in making the token introspection API call, it may need to interpret the client's resource request.
Example of the resource server's request to the authorization server for introspection of an RPT, with a PAT in the header:
POST /introspect HTTP/1.1 Host: as.example.com Authorization: Bearer 204c69636b6c69 ... token=sbjsbhs(/SSJHBSUSSJHVhjsgvhsgvshgsv }
Because an RPT is an access token, if the resource server chooses to supply a token type hint, it would use a token_type_hint of access_token.
The authorization server's response to the resource server MUST use [RFC7662], responding with a JSON object with the structure dictated by that specification, extended as follows.
If the introspection object's active parameter has a Boolean value of true, then the object MUST NOT contain a scope parameter, and MUST contain an extension parameter named permissions that contains an array of objects, each one (representing a single permission) containing these parameters:
Example of a response containing the introspection object with the permissions parameter containing a single permission:
HTTP/1.1 200 OK Content-Type: application/json Cache-Control: no-store ... { "active":true, "exp":1256953732, "iat":1256912345, "permissions":[ { "resource_id":"112210f47de98100", "resource_scopes":[ "view", "http://photoz.example.com/dev/actions/print" ], "exp":1256953732 } ] }
If a request is successfully authenticated, but is invalid for another reason, the authorization server produces an error response by supplying a JSON-encoded object with the following members in the body of the HTTP response:
HTTP/1.1 400 Bad Request Content-Type: application/json Cache-Control: no-store ... { "error": "invalid_resource_id", "error_description": "Permission request failed with bad resource ID.", "error_uri": "https://as.example.com/uma_errors/invalid_resource_id" }
This specification inherits the security considerations of [UMAGrant] and has the following additional security considerations.
In the context of federated authorization, more parties may be operating and using UMA software entities, and thus may need to establish agreements about the parties' rights and responsibilities on a legal or contractual level, as discussed in Section 5.4 of [UMAGrant].
The protection API is secured by means of OAuth (through the use of the PAT). Therefore, it is susceptible to OAuth threats.
This specification inherits the privacy considerations of [UMAGrant] and has the following additional privacy considerations.
In the context of federated authorization, more parties may be operating and using UMA software entities, and thus may need to establish agreements about mutual rights, responsibilities, and common interpretations of UMA constructs for consistent and expected software behavior, as discussed in Section 6.3 of [UMAGrant].
The authorization server comes to be in possession of resource details that may reveal information about the resource owner, which the authorization server's trust relationship with the resource server is assumed to accommodate. However, the client is a less-trusted party -- in fact, entirely untrustworthy until permissions are associated with its RPT. The more information about a resource that is registered, the more risk of privacy compromise there is through a less-trusted authorization server. For example, if resource owner Alice introduces her electronic health record resource server to a authorization server in the cloud, the authorization server may come to learn a great deal of detail about Alice's health information just so that she can control access by others to that information.
This document makes the following requests of IANA.
This specification registers OAuth 2.0 authorization server metadata defined in Section 2, as required by Section 7.1 of [OAuthMeta].
This specification registers the name defined in Section 5.1.1, as required by Section 3.1 of [RFC7662].
The following people made significant text contributions to the specification:
Additional contributors to this specification include the Kantara UMA Work Group participants, a list of whom can be found at [UMAnitarians].
[UMA-Impl] | Maler, E., “UMA Implementer's Guide”, 2017, <https://kantarainitiative.org/confluence/display/uma/UMA+Implementer%27s+Guide>. |
[UMAnitarians] | Maler, E., “UMA Participant Roster”, 2017, <https://kantarainitiative.org/confluence/display/uma/Participant+Roster>. |
Eve Maler
(editor)
ForgeRock
EMail: eve.maler@forgerock.com
Maciej Machulak
Self
EMail: maciej.machulak@gmail.com
Justin Richer
Bespoke Engineering
EMail: justin@bspk.io