WHOIS

Elizabeth Feinler and her team (who had created the Resource Directory for ARPANET) were responsible for creating the first WHOIS directory in the early 1970s.[3] Feinler set up a server in Stanford's Network Information Center (NIC) which acted as a directory which could retrieve relevant information about people or entities.[4] She and the team created domains, with Feinler's suggestion that domains be divided into categories based on the physical address of the computer.[5]

The process of registration was established in RFC 920. WHOIS was standardized in the early 1980s to look up domains, people and other resources related to domain and number registrations. As all registration was done by one organization at that time, one centralized server was used for WHOIS queries. This made looking up such information very easy.

At the time of the emergence of the internet from the ARPANET, the only organization that handled all domain registrations was the Defense Advanced Research Projects Agency (DARPA) of the United States government (created during 1958.[6]). Responsibility of domain registration remained with DARPA as the ARPANET became the Internet during the 1980s. UUNET began offering domain registration service; however they simply handled the paperwork which they forwarded to the DARPA Network Information Center (NIC). Then the National Science Foundation directed that management of Internet domain registration would be handled by commercial, third-party entities. InterNIC was formed in 1993 under contract with the NSF, consisting of Network Solutions, Inc., General Atomics and AT&T. The General Atomics contract was canceled after several years due to performance issues.

20th century WHOIS servers were highly permissive and would allow wild-card searches. A WHOIS query of a person's last name would yield all individuals with that name. A query with a given keyword returned all registered domains containing that keyword. A query for a given administrative contact returned all domains the administrator was associated with. Since the advent of the commercialized Internet, multiple registrars and unethical spammers, such permissive searching is no longer available.

On December 1, 1999, management of the top-level domains (TLDs) com, net, and org was assigned to ICANN. At the time, these TLDs were converted to a thin WHOIS model. Existing WHOIS clients stopped working at that time. A month later, it had self-detecting Common Gateway Interface support so that the same program could operate a web-based WHOIS lookup, and an external TLD table to support multiple WHOIS servers based on the TLD of the request. This eventually became the model of the modern WHOIS client.

By 2005, there were many more generic top-level domains than there had been in the early 1980s. There are also many more country-code top-level domains. This has led to a complex network of domain name registrars and registrar associations, especially as the management of Internet infrastructure has become more internationalized. As such, performing a WHOIS query on a domain requires knowing the correct, authoritative WHOIS server to use. Tools to do WHOIS proxy searches have become common.

The WHOIS protocol had its origin in the ARPANET NICNAME protocol and was based on the NAME/FINGER Protocol, described in RFC 742 (1977). The NICNAME/WHOIS protocol was first described in RFC 812 in 1982 by Ken Harrenstien and Vic White of the Network Information Center at SRI International.

WHOIS was originally implemented on the Network Control Program (NCP) but found its major use when the TCP/IP suite was standardized across the ARPANET and later the Internet.

The protocol specification is the following (original quote):[15]

Connect to the service host
   TCP: service port 43 decimal
   NCP: ICP to socket 43 decimal, establishing two 8-bit connections
Send a single "command line", ending with <CRLF>.
Receive information in response to the command line.  The
server closes its connections as soon as the output is
finished.

The command line server query is normally a single name specification. i.e. the name of a resource. However, servers accept a query, consisting of only the question mark (?) to return a description of acceptable command line formats. Substitution or wild-card formats also exist, e.g., appending a full-stop (period) to the query name returns all entries beginning with the query name.

On the modern Internet, WHOIS services are typically communicated using the Transmission Control Protocol (TCP). Servers listen to requests on the well-known port number 43. Clients are simple applications that establish a communications channel to the server, transmit a text record with the name of the resource to be queried and await the response in form of a sequence of text records found in the database. This simplicity of the protocol also permits an application, and a command line interface user, to query a WHOIS server using the Telnet protocol.

WHOIS lookups were traditionally performed with a command line interface application, but now many alternative web-based tools exist.

A WHOIS database consists of a set of text records for each resource. These text records consists of various items of information about the resource itself, and any associated information of assignees, registrants, administrative information, such as creation and expiration dates.

Two data models exist for storing resource information in a WHOIS database, the thick and the thin model.

WHOIS services are mainly run by registrars and registries; for example the Public Interest Registry (PIR) maintains the .ORG registry and associated WHOIS service.[23]

Regional Internet registries

Regional Internet registries

WHOIS servers operated by regional Internet registries (RIR) can be queried directly to determine the Internet service provider responsible for a particular resource.

The records of each of these registries are cross-referenced, so that a query to ARIN for a record which belongs to RIPE will return a placeholder pointing to the RIPE WHOIS server. This lets the WHOIS user making the query know that the detailed information resides on the RIPE server. In addition to the RIRs servers, commercial services exist, such as the Routing Assets Database used by some large networks (e.g., large Internet providers that acquired other ISPs in several RIR areas).

Normally the contact information of the resources assignee is returned. However, some registrars offer private registration, in which case the contact information of the registrar is shown instead.

Some registry operators are wholesalers, meaning that they typically provide domain name services to a large number of retail registrars, who in turn offer them to consumers. For private registration, only the identity of the wholesale registrar may be returned. In this case, the identity of the individual as well as the retail registrar may be hidden.

Below is an example of WHOIS data returned for an individual resource holder. This is the result of a WHOIS query of example.com:

whois example.com
[Querying whois.verisign-grs.com]
[Redirected to whois.iana.org]
[Querying whois.iana.org]
[whois.iana.org]
% IANA WHOIS server
% for more information on IANA, visit http://www.iana.org
% This query returned 1 object
domain:       EXAMPLE.COM
organisation: Internet Assigned Numbers Authority
created:      1992-01-01
source:       IANA

Referral Whois (RWhois) is an extension of the original Whois protocol and service. RWhois extends the concepts of Whois in a scalable, hierarchical fashion, potentially creating a system with a tree-like architecture. Queries are deterministically routed to servers based on hierarchical labels, reducing a query to the primary repository of information.[24]

Lookups of IP address allocations are often limited to the larger Classless Inter-Domain Routing (CIDR) blocks (e.g., /24, /22, /16), because usually only the regional Internet registries (RIRs) and domain registrars run RWhois or Whois servers, although RWhois is intended to be run by even smaller local Internet registries, to provide more granular information about IP address assignment.

RWhois is intended to replace Whois, providing an organized hierarchy of referral services where one could connect to any RWhois server, request a look-up and be automatically re-directed to the correct server(s). However, while the technical functionality is in place, adoption of the RWhois standard has been weak.

RWhois services are typically communicated using the Transmission Control Protocol (TCP). Servers listen to requests on the well-known port number 4321.

Rwhois was first specified in RFC 1714 in 1994 by Network Solutions,[24] but the specification was superseded in 1997 by RFC 2167.[25]

The referral features of RWhois are different than the feature of a Whois server to refer responses to another server, which RWhois also implements.

One criticism of WHOIS is the lack of full access to the data.[26][27] Few parties have realtime access to the complete databases.

Others cite the competing goal of domain privacy as a criticism, although this problem is strongly mitigated by domain privacy services. Currently, the Internet Corporation for Assigned Names and Numbers (ICANN) broadly requires that the mailing address, phone number and e-mail address of those owning or administrating a domain name to be made publicly available through the "WHOIS" directories. The registrant's (domain owner's) contact details, such as address and telephone number, are easily accessible to anyone who queries a WHOIS server. However, that policy enables spammers, direct marketers, identity thieves or other attackers to loot the directory for personal information about these people. Although ICANN has been exploring changing WHOIS to enable greater privacy, there is a lack of consensus among major stakeholders as to what type of change should be made.[28] Some domain registrars offer private registrations (also known as domain privacy), by which the contact information of the registrar is shown instead of the customer's. With the offer of private registration from many registrars, some of the risk has been mitigated.[29]

Studies have shown that spammers can and do harvest plain-text email addresses from WHOIS servers.[30] For this reason, some WHOIS servers and websites offering WHOIS queries have implemented rate-limiting systems, such as web-based CAPTCHA and limited amounts of search queries per user IP address.[29]

The WHOIS requirements conflict with the General Data Protection Regulation (GDPR), effective in the European Union 25 May 2018, which places strict regulations on the processing and publication of personally identifiable information. ICANN stated in November 2017 that it would not reprimand "noncompliance with contractual obligations related to the handling of registration data" if registrars provide alternative solutions for compliance with its rules, until the WHOIS requirements are updated to take GDPR into account.[29][31]

The WHOIS protocol was not written with an international audience in mind. A WHOIS server and/or client cannot determine the text encoding in effect for the query or the database content. Many servers were originally using US-ASCII and Internationalization concerns were not taken into consideration until much later.[32] This might impact the usability or usefulness of the WHOIS protocol in countries outside the USA.[1] In the case of internationalized domain names it is the responsibility of the client application to perform the translation of the domain name between its native language script and the DNS name in punycode.

In cases where the registrant's (Domain Owner) identity is public, anyone can easily confirm the status of a domain via WHOIS.

In the case of private registrations, ascertaining registration information may be more difficult. If a registrant, who acquired a domain name, wants to verify the registrar has completed the registration process, three steps may be required:

1. Perform a WHOIS and confirm that the resource is at least registered with ICANN,
2. Determine the name of the wholesale registrar, and
3. Contact the wholesaler and obtain the name of the retail registrar.

This provides some confidence that the retailer actually registered the name. But if the registrar goes out of business, as with the failure of RegisterFly in 2007, the rightful domain holder with privacy-protected registrations may have difficulty regaining the administration of their domain name.[17] Registrants using "private registration" can attempt to protect themselves by using a registrar that places customer data in escrow with a third party.

ICANN requires that every registrant of a domain name be given the opportunity to correct any inaccurate contact data associated with their domain. For this reason, registrars are required to periodically send the holder the contact information on record for verification, but they do not provide any guarantee about the accuracy of information if the registrant provided inaccurate information.

WHOIS has generated policy issues in the United States federal government. As noted above, WHOIS creates a privacy issue which is also tied to free speech and anonymity. However, WHOIS is an important tool for law enforcement officers investigating violations like spam and phishing to track down the holders of domain names. As a result, law enforcement agencies have sought to make WHOIS records both open and verified:[33]

• The Federal Trade Commission has testified about how inaccurate WHOIS records thwart their investigations.[34]
• Congressional hearings have been conducted about the importance of WHOIS in 2001, 2002 and 2006.[35]
• The Fraudulent Online Identity Sanctions Act[36] "make it a violation of trademark and copyright law if a person knowingly provided, or caused to be provided, materially false contact information in making, maintaining, or renewing the registration of a domain name used in connection with the violation,"[37] where the latter "violation" refers to a prior violation of trademark or copyright law. The act does not make the submission of false WHOIS data illegal in itself, only if used to shield oneself from prosecution for crimes committed using that domain name.

The Expert Working Group (EWG) of the Internet Corporation for Assigned Names and Numbers (ICANN) recommended on 24 June 2013 that WHOIS should be scrapped. It recommends that WHOIS be replaced with a system that keeps information secret from most Internet users, and only discloses information for "permissible purposes".[38] ICANN's list of permissible purposes includes domain-name research, domain-name sale and purchase, regulatory enforcement, personal data protection, legal actions, and abuse mitigation.[39] Although WHOIS has been a key tool of journalists in determining who was disseminating certain information on the Internet,[40] the use of WHOIS by the free press is not included in ICANN's proposed list of permissible purposes.

The EWG collected public input on the initial report until 13 September 2013. Its final report was issued on 6 June 2014, without meaningful changes to the recommendations.[41] As of March 2015, ICANN is in the "process of re-inventing WHOIS," working on "ICANN WHOIS Beta."[42][43]

• RFC 812 – NICNAME/WHOIS (1982, obsolete)
• RFC 954 – NICNAME/WHOIS (1985, obsolete)
• RFC 3912 – WHOIS protocol specification (2004, current)

• Domain name registry
• Regional Internet registry
• Routing Assets Database
• Routing Policy Specification Language
• Shared Whois Project
• Registration Data Access Protocol

• The Internet Corporation for Assigned Names and Numbers
• The Internet Assigned Numbers Authority
• XML Whois Server list
• ICANN Whois Inaccuracy Complaint Form
• Whois status codes