Simple Web Discovery (SWD)
draft-jones-simple-web-discovery-01

Abstract

Simple Web Discovery (SWD) defines a HTTPS GET based mechanism to discover the location of a given type of service for a given principal starting only with a domain name.

Requirements Language

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].

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 January 12, 2012.

Copyright Notice

Copyright (c) 2011 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.

Table of Contents

• 1. Introduction
• 2. Simple Web Discovery Request
• 3. Simple Web Discovery Responses
• 3.1. Response Containing One or More Locations
• 3.2. Redirecting All Simple Web Discovery Requests
• 3.3. 401 Unauthorized Response
• 3.4. Other HTTP 1.1 Responses
• 4. IANA Considerations
• 5. Security Considerations
• 6. References
• 6.1. Normative References
• 6.2. Informative References
• Appendix A. Document History
• Authors' Addresses

1. Introduction

Simple Web Discovery (SWD) defines a HTTPS GET based mechanism to discover the location of a given type of service for a given principal starting only with a domain name. SWD requests use the x-www-form-urlencoded format to specify a URI for the principal and another URI for the type of service being sought. If the request is successful then the response, by default, is a JSON object containing an array of URIs that point to where the principal has instances of services of the requested type.

For example, let us say that a requester wants to discover where Joe keeps his calendar. The requester could take Joe's e-mail address, joe@example.com and take from it its domain to create a HTTPS GET request of the following form:

GET /.well-known/simple-web-discovery?principal=mailto:joe@example.com&service=urn:adatum.com:calendar HTTP/1.1
Host: example.com

HTTP/1.1 200 O.K.
Content-Type: application/json

{
 "locations": ["http://calendars.proseware.com/calendars/joseph"]
}
        

Note: The request-URI is left un-encoded in the above example for the sake of readability in the above example.

2. Simple Web Discovery Request

Domains that support SWD requests MUST make available a SWD server for their domain at the path .well-known/simple-web-discovery. The syntax and semantics of ".well-known" are defined in RFC 5785 [RFC5785]. "simple-web-discovery" MUST point to a SWD server compliant with this specification.

SWD servers MUST support receiving SWD requests via TLS 1.2 as defined in RFC 5246 [RFC5246] and MAY support other transport layer security mechanisms of equivalent security. SWD servers MUST reject SWD requests sent over plain HTTP or any other transport that does not provide both privacy and validation of the server's identity.

A SWD server is queried using a HTTPS GET request with the previously specified path along with a query segment containing a x-www-form-urlencoded form as defined in HTML 4.01 [W3C.REC-html401-19991224]. The form MUST contain two name/value pairs that MUST appear exactly once, principal and service. Both name/value pairs MUST have values that are set to URIs (as defined in RFC 3986 [RFC3986]). If any of the previous requirements are not met in a SWD request, then the request MUST be rejected with a 400 Bad Request.

The SWD request form MAY contain additional name/value pairs but if those name/value pairs are not recognized by the SWD server then the SWD server MUST ignore them for processing purposes.

The principal query component is a URI that identifies an entity. The service query component is a URI that identifies a service type. The semantics of the SWD query is "Please return the location(s) of instances of the specified service type associated with the specified principal". The definition of URIs used to identify principals and services are outside the scope of this specification.

3. Simple Web Discovery Responses

3.1. Response Containing One or More Locations

Unless another content-type is negotiated, a 200 O.K. response to a SWD request that contains the information requested MUST return content of type application/json as defined in RFC 4627 [RFC4627]. The JSON response MUST contain a JSON object that contains a member pair whose name is the string locations and whose value is an array of strings that are each a URI pointing to a location where the desired service type belonging to the specified principal can be found. There are no semantics associated with the order in which the URIs are listed in the array.

The JSON object MAY contain other members but a receiver of the object MAY ignore any member pairs whose name it does not recognize.

3.2. Redirecting All Simple Web Discovery Requests

SWD requests by definition start off by being issued to the .well-known/simple-web-discovery location. But locating a SWD server at a root location can prove inconvenient. To enable service level redirection a SWD server MAY return a 200 O.k. to an HTTPS request with a content type of application/json (or whatever other content type has been negotiated) that contains a JSON object that contains a member pair whose name is the string SWD_service_redirect whose value is a JSON object with a member pair whose name is location and whose value is a string that encodes a URI. Optionally the JSON object value of SWD_service_redirect MAY also contain a member whose name is expires and whose value is a JSON number that encodes an integer.

A SWD compliant client MUST support the SWD_service_redirect response.

The JSON objects MAY contain other members but a receiver of the objects MAY ignore any pairs whose name it does not recognize.

The location member identifies the URI that the caller MUST redirect all SWD requests for that domain to until the expires time is met. SWD requests for the redirected domain MUST be constructed by taking the URI returned in the location and using it as the base URI to which the SWD form arguments are then added as query parameters. The location URI MUST NOT include a query component.

GET /.well-known/simple-web-discovery?principal=mailto:joe@example.com&service=urn:adatum.com:calendar HTTP/1.1
Host: example.com

HTTP/1.1 200 O.K.
Content-Type: application/json

{
 "SWD_service_redirect":
  {
   "location": "https://swd.proseware.com/swd_server",
   "expires": 1300752001
  }
}

GET /swd_server?principal=mailto:joe@example.com&service=urn:adatum.com:calendar HTTP/1.1
Host: swd.proseware.com

HTTP/1.1 200 O.K.
Content-Type: application/json

{
 "locations": ["http://calendars.proseware.com/calendars/joseph"]
}
          

Note: The request-URIs are left un-encoded in the above example for the sake of readability in the above example.

The location URI MUST be an HTTPS URL.

The optional expires member identifies the point in time at which the caller MUST NOT redirect its SWD requests for that domain to the previously obtained location and MUST instead return to the .well-known/simple-web-discovery location. The value of the expires member MUST encode the number of seconds from 1970-01-01T0:0:0Z as measured in UTC until the desired date/time. See RFC 3339 [RFC3339] for details regarding date/times in general and UTC in particular. If the expires value is in the past or if the value is more than one hour in the future then the response MUST be treated as if it didn't contain an expires value.

If the expires value is omitted or if its value is incorrect then the expires value MUST be treated as having a value of exactly one hour into the future.

If a JSON response is received that contains both a member pair with the name SWD_service_redirect and a member pair with the name locations as children of the object root then the SWD_service_redirect member pair MUST be ignored.

3.3. 401 Unauthorized Response

A SWD server MAY respond to a request with a 401 Unauthorized Response, as described in RFC 2616 [RFC2616], Section 10. Per the RFC, the request MAY be repeated with a suitable Authorization header field. Authorization information may be communicated in this manner, including a JSON Web Token [JWT].

3.4. Other HTTP 1.1 Responses

A SWD server MAY return other HTTP 1.1 responses, including 404 Not Found, 400 Bad Request, and 403 Forbidden. SWD implementations MUST correctly handle these responses.

4. IANA Considerations

Per RFC 5785 [RFC5785], the following registration template is offered:

URI suffix

simple-web-discovery

Change controller

IETF

Specification document

This RFC

5. Security Considerations

SWD responses can contain confidential information. Therefore a, general approach is used to require TLS in all cases. But TLS can only provide for privacy and server validation, it cannot validate that the requester is authorized to see the results of a query. The exact mechanism used to determine if the requester is authorized to see the result of the query is outside the scope of this specification.

Because SWD responses can contain confidential information, the requestor may need authorization to receive them. Standard HTTP authorization mechanisms MAY be employed to request authorized access, including the use of an HTTP Authorization header field in requests, which in turn, may contain a JSON Web Token [JWT], among other authorization data formats.

The ability to redirect an entire SWD server as defined in this document is an obvious attack point. This is another reason why we have mandated TLS, so as to be sure that the redirect can only be received over a secure connection. We have also put in the upper limit of 60 minutes for a redirect so as to provide a path for regaining control over queries should a successful attack be launched to return false redirects.

The SWD_service_redirect capability may cause unanticipated failures in cases where a requestor may have permissions to discover content at the original SWD endpoint but not the one redirected to, or vice-versa.

6. References

6.1. Normative References

6.2. Informative References

Appendix A. Document History

-01

• Refresh draft before expiration of -00. No normative changes.

-00

• Initial version.

Authors' Addresses