--- 1/draft-ietf-regext-rdap-reverse-search-01.txt 2019-10-08 04:13:56.703347301 -0700 +++ 2/draft-ietf-regext-rdap-reverse-search-02.txt 2019-10-08 04:13:56.727347915 -0700 @@ -1,18 +1,18 @@ Registration Protocols Extensions M. Loffredo Internet-Draft M. Martinelli Intended status: Standards Track IIT-CNR/Registro.it -Expires: October 13, 2019 April 11, 2019 +Expires: April 10, 2020 October 8, 2019 Registration Data Access Protocol (RDAP) Reverse search capabilities - draft-ietf-regext-rdap-reverse-search-01 + draft-ietf-regext-rdap-reverse-search-02 Abstract The Registration Data Access Protocol (RDAP) does not include query capabilities to find the list of domains related to a set of entities matching a given search pattern. Even if such capabilities, commonly referred as reverse search, respond to some needs not yet readily fulfilled by the current Whois protocol, they have raised concerns from two perspectives: server processing impact and data privacy. Anyway, the impact of the reverse queries on RDAP servers processing @@ -31,21 +31,21 @@ 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 October 13, 2019. + This Internet-Draft will expire on April 10, 2020. Copyright Notice Copyright (c) 2019 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 @@ -71,69 +71,69 @@ 9. References . . . . . . . . . . . . . . . . . . . . . . . . . 8 9.1. Normative References . . . . . . . . . . . . . . . . . . 8 9.2. Informative References . . . . . . . . . . . . . . . . . 9 Appendix A. Change Log . . . . . . . . . . . . . . . . . . . . . 10 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 10 1. Introduction Reverse Whois is a service provided by many web applications that allow users to find domain names owned by an individual or a company - starting from the owner details, such as name and email. Even if it - has been considered useful for some legal purposes (e.g. uncovering - trademark infringements, detecting cybercrime cases), its - availability as a standardised Whois capability has been objected for + starting from the owner's details, such as name and email. Even if + it has been considered useful for some legal purposes (e.g. + uncovering trademark infringements, detecting cybercrime cases), its + availability as a standardized Whois capability has been objected for two main reasons, which now don't seem to conflict with an RDAP implementation. The first objection has been caused by the potential risks of privacy violation. However, TLDs community is considering a new generation of Registration Directory Services ([ICANN-RDS1],[ICANN-RDS2]), which provide access to sensitive data under some permissible purposes and according to adequate policies to enforce the requestor accreditation, authentication, authorization, and terms and conditions of data use. It is well known that such security policies are not implemented in Whois ([RFC3912]), while they are in RDAP ([RFC7481]). Therefore, RDAP permits a reverse search implementation complying with privacy protection principles. Another objection to the implementation of a reverse search capability has been connected with its impact on server processing. Since RDAP supports search queries, the impact of both standard and reverse searches is equivalent and can be mitigated by servers - adopting ad hoc strategies. Furthermore, reverse search is almost - always performed by specifying an entity role (e.g. registrant, - technical contact) and this can contribute to restricting the result - set. + adopting ad hoc strategies. Furthermore, the reverse search is + almost always performed by specifying an entity role (e.g. + registrant, technical contact) and this can contribute to restricting + the result set. Reverse searches, such as finding the list of domain names associated with contacts, nameservers or DNSSEC keys, may be useful to registrars as well. Usually, registries adopt out-of-band mechanisms to provide results to registrars asking for reverse searches on their - domains. Possible reasons of such requests are: + domains. Possible reasons for such requests are: o the loss of synchronization between the registrar database and the registry database; - o the need of such data to perform massive EPP ([RFC5730]) updates + o the need for such data to perform massive EPP ([RFC5730]) updates (e.g. changing the contacts of a set of domains, etc.). Currently, RDAP does not provide any way for a client to search for the collection of domains associated with an entity ([RFC7482]). A query (lookup or search) on domains can return the array of entities related to a domain with different roles (registrant, registrar, administrative, technical, reseller, etc.), but the reverse operation is not allowed. Only reverse searches to find the collection of domains related to a nameserver (ldhName or ip) can be requested. - Since entities can be in relation with all RDAP objects ([RFC7483]), - the availability of a reverse search can be common to all RDAP query - paths. + Since entities can be in relationship with all RDAP objects + ([RFC7483]), the availability of a reverse search can be common to + all RDAP query paths. The protocol described in this specification aims to extend the RDAP query capabilities to enable reverse search based on the domains- entities relationship (the classic Reverse Whois scenario). The extension is implemented by adding new path segments (i.e. search paths) and using a RESTful web service ([REST]). The service is implemented using the Hypertext Transfer Protocol (HTTP) ([RFC7230]) and the conventions described in RFC 7480 ([RFC7480]). 1.1. Conventions Used in This Document @@ -149,37 +149,40 @@ Syntax: domains?entityHandle= Syntax: domains?entityFn= Syntax: domains?entityEmail= Syntax: domains?entityAddr= The reverse search pattern is a JSON ([RFC8259]) object including two - members: "value" and "role". The "value" member represents the - search pattern to be applied to the corresponding entity field and - can be a JSON type primitive or object. The "role" member is a - string whose possible values are those detailed in Section 10.2.4 of - RFC 7483 ([RFC7483]). The former is REQUIRED while the latter is - OPTIONAL to allow RDAP servers to provide reverse search capabilities - without specifying any role. + members: + + "value" (REQUIRED): represents the search pattern to be applied to + the corresponding entity field and can be a JSON type primitive or + object; + + "role" (OPTIONAL): is a string whose possible values are those + detailed in Section 10.2.4 of RFC 7483 ([RFC7483]). When it is + missing, the reverse search is performed without considering the + entity role. The search patterns corresponding to the "value" in the first two cases (Figure 1) are the same as specified in paragraph Section 3.2.3 of RFC 7482 ([RFC7482]). domains?entityHandle={"value":"CID-40*","role":"registrant"} domains?entityFn={"value":"Bobby*","role":"registrant"} - Figure 1: Examples of RDAP queries to find all domains related to a + Figure 1: Examples of RDAP queries to find all domains related to any registrant whose handle matches "CID-40*" and whose formatted name matches "Bobby*" The last two reverse searches are considered by gTLD stakeholders very useful to improve RDS searchability ([ICANN-RDS1], [ICANN-RA]). Searches for domains by related entity email are specified using this form: domains?entityEmail={"value":"XXXX","role":"ZZZZ"} @@ -193,47 +195,47 @@ domains?entityAddr={"value":YYYY,"role":"ZZZZ"} where YYYY is a JSON object containing the information described in Section 2.4 of RFC 5733 ([RFC5733]), respectively: "street", "city", "sp", "pc" and "cc" (Figure 2). All the members of the postal address object are OPTIONAL but at least one is REQUIRED. The constraints on the members are implicitly joined by AND. domains?entityAddr={"value":{"cc":"CA","city":"Sydney"},"role":"registrant"} - Figure 2: Example of a RDAP query to find all domains related to a + Figure 2: Example of an RDAP query to find all domains related to any registrant whose postal address contains the country code equals to "CA" and the city equals to "Sydney" 3. Implementation Considerations The implementation of the proposed extension is technically feasible. The search paths "handle" and "fn" are used as standard paths to search for entities. With regards to the last two reverse searches, both email and postal address information are usually required by the registries but, while the former is usually mapped onto a DBMS indexed field, the latter is mapped onto a combination of non-indexed - fields. As a consequence while the former should not significantly + fields. As a consequence, while the former should not significantly decrease the performance, the latter might have an impact on server processing. Anyway, this impact is evaluated to be the same as other query capabilities already presented in RDAP (e.g. wildcard prefixed search pattern) so the risks to generate huge result sets are the same as those related to other standard searches and can be mitigated - by adopting the same policies (e.g. restricting search + by adopting the same policies (e.g. restricting the search functionalities, limiting the rate of search requests according to the user profile, truncating and paging the results, returning partial responses). 3.1. JSON in URLs Many web services, including RDAP, rely on the HTTP GET method to - take advantage from some of its features: + take advantage of some of its features: o GET requests can be cached; o GET requests remain in the browser history; o GET requests can be bookmarked. Sometimes, it happens that such advantages should be combined with the requirement to pass objects and arrays in the query string. JSON is the best candidate as data interchange format, but it contains some characters that are forbidden from appearing in a URL. Anyway, escaping the invalid characters is not an issue because, on the @@ -298,24 +301,24 @@ 5. Privacy Considerations The use of the capability described in this document MUST be compliant with the rules about privacy protection each RDAP provider is subject to. Sensitive registration data MUST be protected and accessible for permissible purposes only. Therefore, RDAP servers MUST provide reverse search only to those requestors who are authorized according to a lawful basis. Some potential users of this capability include registrars searching for their own domains and operators in the exercise of an official authority or performing a - specific task in the public interest that is set out in law. Another - scenario consists of permitting reverse searches, which take into - account only those entities that have previously given the explicit - consent for publishing and processing their personal data. + specific task in the public interest that is set out in a law. + Another scenario consists of permitting reverse searches, which take + into account only those entities that have previously given the + explicit consent for publishing and processing their personal data. 6. Security Considerations Security services required to provide controlled access to the operations specified in this document are described in RFC 7481 ([RFC7481]). 7. IANA Considerations This document has no actions for IANA. @@ -387,29 +390,29 @@ Internet Corporation For Assigned Names and Numbers, "Registry Agreement", July 2017, . [ICANN-RDS1] Internet Corporation For Assigned Names and Numbers, "Final Report from the Expert Working Group on gTLD Directory Services: A Next-Generation Registration Directory Service (RDS)", June 2014, - . + . [ICANN-RDS2] Internet Corporation For Assigned Names and Numbers, "Final Issue Report on a Next-Generation gTLD RDS to Replace WHOIS", October 2015, - . + . [JSURL] github.com, "JSURL", 2016, . [REST] Fielding, R., "Architectural Styles and the Design of Network-based Software Architectures", 2000, . [RFC4648] Josefsson, S., "The Base16, Base32, and Base64 Data @@ -425,20 +428,21 @@ . [URLON] github.com, "URL Object Notation", 2017, . Appendix A. Change Log 00: Initial working group version ported from draft-loffredo-regext- rdap-reverse-search-04 01: Updated "Privacy Considerations" section. + 02: Revised the text. Authors' Addresses Mario Loffredo IIT-CNR/Registro.it Via Moruzzi,1 Pisa 56124 IT Email: mario.loffredo@iit.cnr.it