American Registry for Internet Numbers

3635 Concorde Parkway Chantilly VA United States 20151 andy@arin.net http://www.arin.net

Asia Pacific Network Information Centre

6 Cordelia Street South Brisbane Australia QLD 4101 bje@apnic.net http://www.apnic.net

China Internet Network Information Center

4 South 4th Street, Zhongguancun, Haidian District China 100190 Beijing +86 10 5881 3147 nkong@cnnic.cn

This document is one of a collection that together describes the Registration Data Access Protocol (RDAP). It describes how RDAP is transported using the Hypertext Transfer Protocol (HTTP). RDAP is a successor protocol to the very old WHOIS protocol. The purpose of this document is to clarify the use of standard HTTP mechanisms for this application.

This document describes the usage of the Hypertext Transfer Protocol (HTTP) for the Registration Data Access Protocol (RDAP). The goal of this document is to tie together usage patterns of HTTP into a common profile applicable to the various types of directory services serving registration data using practices informed by the Representational State Transfer (REST) architectural style. By giving the various directory services common behavior, a single client is better able to retrieve data from directory services adhering to this behavior. Registration data expected to be presented by this service is Internet resource registration data -- registration of domain names and Internet number resources. This data is typically provided by WHOIS services, but the WHOIS protocol is insufficient to modern registration data service requirements. A replacement protocol is expected to retain the simple transactional nature of WHOIS, while providing a specification for queries and responses, redirection to authoritative sources, support for Internationalized Domain Names (IDNs) , and support for localized registration data such as addresses and organization or person names. In designing these common usage patterns, this document introduces considerations for a simple use of HTTP. Where complexity may reside, it is the goal of this document to place it upon the server and to keep the client as simple as possible. A client implementation should be possible using common operating system scripting tools (e.g., bash and wget). This is the basic usage pattern for this protocol: A client determines an appropriate server to query along with the appropriate base Uniform Resource Locator (URL) to use in such queries. describes one method to determine the server and the base URL. See for more information. A client issues an HTTP (or HTTPS) query using GET . As an example, a query URL for the network registration 192.0.2.0 might be http://example.com/rdap/ip/192.0.2.0 details the various queries used in RDAP. If the receiving server has the information for the query, it examines the Accept header field of the query and returns a 200 response with a response entity appropriate for the requested format. details a response in JavaScript Object Notation (JSON). If the receiving server does not have the information for the query but does have knowledge of where the information can be found, it will return a redirection response (3xx) with the Location header field containing an HTTP(S) URL pointing to the information or another server known to have knowledge of the location of the information. The client is expected to requery using that HTTP URL. If the receiving server does not have the information being requested and does not have knowledge of where the information can be found, it returns a 404 response. If the receiving server will not answer a request for policy reasons, it will return an error response (4xx) indicating the reason for giving no answer. It is not the intent of this document to redefine the meaning and semantics of HTTP. The purpose of this document is to clarify the use of standard HTTP mechanisms for this application.

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 . As is noted in "Security and Stability Advisory Committee (SSAC) Report on WHOIS Terminology and Structure", the term "WHOIS" is overloaded, often referring to a protocol, a service, and data. In accordance with , this document describes the base behavior for an RDAP. describes a protocol profile of RDAP for Domain Name Registries (DNRs), the Domain Name Registration Data Access Protocol (DNRD-AP). In this document, an RDAP client is an HTTP user agent performing an RDAP query, and an RDAP server is an HTTP server providing an RDAP response. RDAP query and response formats are described in and , while this document describes how RDAP clients and servers use HTTP to exchange queries and responses. describes security considerations for RDAP.

There are a few design criteria this document attempts to meet. First, each query is meant to require only one path of execution to obtain an answer. A response may contain an answer, no answer, or a redirect, and clients are not expected to fork multiple paths of execution to make a query. Second, the semantics of the request/response allow for future and/or non-standard response formats. In this document, only a JSON response media type is noted, with the response contents to be described separately (see ). This document only describes how RDAP is transported using HTTP with this format. Third, this protocol is intended to be able to make use of the range of mechanisms available for use with HTTP. HTTP offers a number of mechanisms not described further in this document. Operators are able to make use of these mechanisms according to their local policy, including cache control, authorization, compression, and redirection. HTTP also benefits from widespread investment in scalability, reliability, and performance, as well as widespread programmer understanding of client behaviors for web services styled after REST, reducing the cost to deploy Registration Data Directory Services and clients. This protocol is forward compatible with HTTP 2.0.

Clients use the GET method to retrieve a response body and use the HEAD method to determine existence of data on the server. Clients SHOULD use either the HTTP GET or HEAD methods (see ). Servers are under no obligation to support other HTTP methods; therefore, clients using other methods will likely not interoperate properly. Clients and servers MUST support HTTPS to support security services.

To indicate to servers that an RDAP response is desired, clients include an Accept header field with an RDAP-specific JSON media type, the generic JSON media type, or both. Servers receiving an RDAP request return an entity with a Content-Type header containing the RDAP-specific JSON media type. This specification does not define the responses a server returns to a request with any other media types in the Accept header field, or with no Accept header field. One possibility would be to return a response in a media type suitable for rendering in a web browser.

Servers MUST ignore unknown query parameters. Use of unknown query parameters for cache busting is described in .

This section describes the various types of responses a server may send to a client. While no standard HTTP response code is forbidden in usage, this section defines the minimal set of response codes in common use by servers that a client will need to understand. While some clients may be constructed with simple tooling that does not account for all of these response codes, a more robust client accounting for these codes will likely provide a better user experience. It is expected that usage of response codes and types for this application not defined here will be described in subsequent documents.

If a server has the information requested by the client and wishes to respond to the client with the information according to its policies, it returns that answer in the body of a 200 (OK) response (see ).

If a server wishes to inform a client that the answer to a given query can be found elsewhere, it returns either a 301 (Moved Permanently) response code to indicate a permanent move or a 302 (Found), 303 (See Other), or 307 (Temporary Redirect) response code to indicate a non-permanent redirection, and it includes an HTTP(S) URL in the Location header field (see ). The client is expected to issue a subsequent request to satisfy the original query using the given URL without any processing of the URL. In other words, the server is to hand back a complete URL, and the client should not have to transform the URL to follow it. Servers are under no obligation to return a URL conformant to . For this application, such an example of a permanent move might be a Top-Level Domain (TLD) operator informing a client the information being sought can be found with another TLD operator (i.e., a query for the domain bar in foo.example is found at http://foo.example/domain/bar). For example, if the client uses http://serv1.example.com/weirds/domain/example.com the server redirecting to https://serv2.example.net/weirds2/ would set the Location: field to the value https://serv2.example.net/weirds2/domain/example.com

If a server wishes to respond that it has an empty result set (that is, no data appropriately satisfying the query), it returns a 404 (Not Found) response code. Optionally, it MAY include additional information regarding the negative answer in the HTTP entity body. If a server wishes to inform the client that information about the query is available, but cannot include the information in the response to the client for policy reasons, the server MUST respond with an appropriate response code out of HTTP's 4xx range. A client MAY retry the query if that is appropriate for the respective response code.

If a server receives a query that it cannot interpret as an RDAP query, it returns a 400 (Bad Request) response code. Optionally, it MAY include additional information regarding this negative answer in the HTTP entity body.

Some servers apply rate limits to deter address scraping and other abuses. When a server declines to answer a query due to rate limits, it returns a 429 (Too Many Requests) response code as described in . A client that receives a 429 response SHOULD decrease its query rate and honor the Retry-After header field if one is present. Servers may place stricter limits upon clients that do not honor the Retry-After header. Optionally, the server MAY include additional information regarding the rate limiting in the HTTP entity body. Note that this is not a defense against denial-of-service (DoS) attacks, since a malicious client could ignore the code and continue to send queries at a high rate. A server might use another response code if it did not wish to reveal to a client that rate limiting is the reason for the denial of a reply.

When responding to queries, it is RECOMMENDED that servers use the Access-Control-Allow-Origin header field, as specified by . A value of "*" is suitable when RDAP is used for public resources. This header (often called the CORS header) helps in-browser web applications by lifting the "same-origin" restriction (i.e., a browser may load RDAP client code from one web server but query others for RDAP data). By default, browsers do not send cookies when cross origin requests are allowed. Setting the Access-Control-Allow-Credentials header field to "true" will send cookies. Use of the Access&nbhy;Control&nbhy;Allow&nbhy;Credentials header field is NOT RECOMMENDED.

For extensibility purposes, this document defines an IANA registry for prefixes used in JSON data serialization and URI path segments (see ). Prefixes and identifiers SHOULD only consist of the alphabetic US-ASCII characters A through Z in both uppercase and lowercase, the numerical digits 0 through 9, and the underscore character, and they SHOULD NOT begin with an underscore character, numerical digit, or the characters "xml". The following describes the production of JSON names in ABNF.

name = ALPHA *( ALPHA / DIGIT / "_" )

This restriction is a union of the Ruby programming language identifier syntax and the XML element name syntax and has two purposes. First, client implementers using modern programming languages such as Ruby or Java can use libraries that automatically promote JSON names to first-order object attributes or members. Second, a clean mapping between JSON and XML is easy to accomplish using these rules.

This document does not pose strong security requirements to the RDAP protocol. However, it does not restrict against the use of security mechanisms offered by the HTTP protocol. It does require that RDAP clients and servers MUST support HTTPS. This document makes recommendations for server implementations against DoS () and interoperability with existing security mechanisms in HTTP clients (). Additional security considerations to the RDAP protocol are covered in .

IANA has created a new category in the protocol registries labeled "Registration Data Access Protocol (RDAP)", and within that category, has established a URL-referenceable, stand-alone registry labeled "RDAP Extensions". The purpose of this registry is to ensure uniqueness of extension identifiers. The extension identifier is used as a prefix in JSON names and as a prefix of path segments in RDAP URLs. The production rule for these identifiers is specified in . In accordance with , the IANA policy for assigning new values, shall be Specification Required: values and their meanings must be documented in an RFC or in some other permanent and readily available reference, in sufficient detail that interoperability between independent implementations is possible. The following is a template for an RDAP extension registration: Extension identifier: the identifier of the extension Registry operator: the name of the registry operator Published specification: RFC number, bibliographical reference, or URL to a permanent and readily available specification Person & email address to contact for further information: The names and email addresses of individuals to contact regarding this registry entry Intended usage: brief reasons for this registry entry (as defined by ). The following is an example of a registration in the RDAP extension registry: Extension identifier: lunarNic Registry operator: The Registry of the Moon, LLC Published specification: http://www.example/moon_apis/rdap Person & email address to contact for further information: Professor Bernardo de la Paz <berny@moon.example> Intended usage: COMMON

Clients can use Internationalized Resource Identifiers (IRIs) for internal use as they see fit but MUST transform them to URIs for interaction with RDAP servers. RDAP servers MUST use URIs in all responses, and again clients can transform these URIs to IRIs for internal use as they see fit.

Under most scenarios, clients requesting data will not signal that the data be returned in a particular language or script. On the other hand, when servers return data and have knowledge that the data is in a language or script, the data SHOULD be annotated with language identifiers whenever they are available, thus allowing clients to process and display the data accordingly. provides such a mechanism.

Given the description of the use of language identifiers in , unless otherwise specified, servers SHOULD ignore the HTTP Accept-Language header field when formulating HTTP entity responses, so that clients do not conflate the Accept-Language header with the 'lang' values in the entity body. However, servers MAY return language identifiers in the Content-Language header field so as to inform clients of the intended language of HTTP layer messages.

LACNIC

To demonstrate typical behavior of an RDAP client and server, the following is an example of an exchange, including a redirect. The data in the response has been elided for brevity, as the data format is not described in this document. The media type used here is described in .

An example of an RDAP client and server exchange: Client: <TCP connect to rdap.example.com port 80> GET /rdap/ip/203.0.113.0/24 HTTP/1.1 Host: rdap.example.com Accept: application/rdap+json rdap.example.com: HTTP/1.1 301 Moved Permanently Location: http://rdap-ip.example.com/rdap/ip/203.0.113.0/24 Content-Length: 0 Content-Type: application/rdap+json <TCP disconnect> Client: <TCP connect to rdap-ip.example.com port 80> GET /rdap/ip/203.0.113.0/24 HTTP/1.1 Host: rdap-ip.example.com Accept: application/rdap+json rdap-ip.example.com: HTTP/1.1 200 OK Content-Type: application/rdap+json Content-Length: 9001 { ... } <TCP disconnect>

Some HTTP cache infrastructures do not adhere to caching standards adequately and could cache responses longer than is intended by the server. To overcome these issues, clients can use an ad hoc and improbably used query parameter with a random value of their choosing. As instructs servers to ignore unknown parameters, this is compatible with the RDAP definition.

An example of using an unknown query parameter to bust caches: http://example.com/ip/192.0.2.0?__fuhgetaboutit=xyz123

Use of an unknown parameter to overcome misbehaving caches is not part of any specification and is offered here for informational purposes.

The traditional deployment model of WHOIS does not provide a mechanism for determining the authoritative source for information. Some approaches have been implemented in the past, most notably the Joint WHOIS initiative. However, among other shortcomings, Joint WHOIS is implemented using proxies and server-side referrals. These issues are solved in RDAP using HTTP redirects and bootstrapping. Bootstrapping is discussed in . In constrained environments, the processes outlined in may not be viable, and there may be the need for servers acting as a "redirector". Redirector servers issue HTTP redirects to clients using a redirection table informed by . diagrams a client using a redirector for bootstrapping.

| | | | <---------- HTTP 301 --------| | ('Try ARIN RDAP') | | | | | Q: 23.1.1.1? -------------------------------> | | <---------- HTTP 200 --------------------- | (JSON response is returned) | | | . ]]>

In some cases, particularly sub-delegations made between Regional Internet Registries (RIRs) known as "ERX space" and transfers of networks, multiple HTTP redirects will be issued. shows such a scenario.

| | | | | | <-- HTTP 301 --- | | | ('Try LACNIC') | | | | | | | | | Q: 23.1.1.1? -----------------> | | | | <---------- HTTP 301 --------| | ('Try ARIN RDAP') | | | | | Q: 23.1.1.1? -------------------------------> | | <---------- HTTP 200 --------------------- | (JSON response is returned) | | | . ]]>

John Levine provided text to tighten up the Accept header field usage and the text for the section on 429 responses. Marc Blanchet provided some clarifying text regarding the use of URLs with redirects, as well as very useful feedback during Working Group Last Call (WGLC). Normative language reviews were provided by Murray S. Kucherawy, Andrew Sullivan, Tom Harrison, Ed Lewis, and Alexander Mayrhofer. Jean-Phillipe Dionne provided text for the Security Considerations section. The concept of the redirector server informatively discussed in was documented by Carlos M. Martinez and Gerardo Rada of LACNIC and Linlin Zhou of CNNIC and subsequently incorporated into this document. This document is the work product of the IETF's WEIRDS working group, of which Olaf Kolkman and Murray Kucherawy were chairs.