ICANN New gTLD Application

Table of contents:

1. Mission of the OP3FT (the Applicant)
2. Frogans technology
3. Use of the .frogans gTLD
4. .frogans gTLD characteristics
5. The need for the .frogans gTLD
6. The .frogans gTLD registry back-end service provider
7. Financing the .frogans gTLD


1. Mission of the OP3FT (the Applicant)

The OP3FT (Organization for the Promotion, Protection and Progress of Frogans Technology) is a French non-profit organization established on March 17, 2012 for ensuring that, in the public interest, the development of Frogans technology, begun in 1999, be carried out irrevocably in a neutral, open and transparent manner and that its access be guaranteed to all Internet users.

As a non-profit organization acting in the public interest, the object of the OP3FT is to hold, promote, protect and progress Frogans technology in the form of an open standard for the Internet that is free to use by all.

More generally, the object of the OP3FT is to contribute to the development of a secure and stable Internet that is open to innovations, and to support and finance public-interest initiatives or innovative actions of a social, educational or cultural nature in the area of information and communication technologies.

The support and financing of these actions, programs and⁄or initiatives take place in compliance with the principles of openness and transparency that contribute to the spread of progress made in the area of information and communication technologies to the greatest number of people.

The works that the OP3FT must accomplish to fulfill its mission are detailed in its Bylaws which are available to the public.


2. Frogans technology

Frogans technology allows the implementation of a new application layer on the Internet, called the “Frogans layer,” alongside the other Internet application layers such as Email or the World Wide Web.

Frogans technology, which is secure and simple, allows for the publication of Frogans sites. A Frogans site is a set of Frogans pages, hyperlinked between each other, which is put on line on the Internet, or on an intranet, at a Frogans address. Frogans sites can be published by anyone.

Frogans technology uses standards that are open and free, such as the DNS, Unicode, the XML format, the HTTP protocol and the TLS (SSL) protocol.

The Frogans layer will be introduced on the Internet in the near future.


3. Use of the .frogans gTLD

The OP3FT will use the .frogans gTLD with the aim of ensuring the security, stability and reliability of the Frogans layer for the benefit of all Internet users.

The OP3FT will use .frogans gTLD domain names for addressing the computers that are dedicated to the functioning of the Frogans layer on the Internet.

These computers are to be used, for example:

- for the distribution of the free software application that allows all Internet users to navigate Frogans sites (Frogans Player);

- for the resolution of Frogans addresses from anywhere on the Internet.

This use of the .frogans gTLD is inspired by the use of the .net gTLD, as specified by Jon Postel in RFC 1591 (March 1994).

The OP3FT will not use .frogans gTLD domain names for naming Frogans sites. Indeed, the global addressing system of the Frogans layer (Frogans Network System) relies on Frogans addresses, whose specific and simple pattern (network*site) is not related to the Uniform Resource Identifier (URI) scheme.

The use of the .frogans gTLD adheres to ICANNʹs principles for keeping the Internet secure, stable and interoperable, while encouraging more innovation on the Internet.


4. .frogans gTLD characteristics

With respect to the OP3FTʹs mission and in consideration of the use of the .frogans gTLD specified above:

- The OP3FT will be the sole and unique holder of domain names in the .frogans gTLD registry;

- The OP3FT will not commercialize any domain names in the .frogans gTLD registry to third parties;

- The maximum number of domain names in the .frogans gTLD registry will be 1,000.

Additionally:

- For establishing a clear, reliable and perpetual framework of use for the Internet community, the character string “Frogans” is a worldwide trademark since 2000. It has been registered in 80 countries, including France, European countries, the USA, Canada, Australia and Latin American, Asian and African countries (classes 9, 16, 35, 38 and⁄or 42).

- In any event, the OP3FT cannot have any commercial activity. Thus, the .frogans gTLD shall never represent any commercial activity for the OP3FT.

- Should the maximum number of domain names in the .frogans gTLD registry (1,000) prove at some point to be insufficient for addressing all of the computers that are dedicated to the functioning of the Frogans layer, the OP3FT would likely reorganize the hierarchy of domain names with the introduction of third-level domain names so that the total number of domain names in the .frogans gTLD registry will not exceed 1,000.


5. The need for the .frogans gTLD

Two current members of the OP3FTʹs Board of Directors have participated or have been represented in all ICANN meetings from June 2008 to June 2011 in order to prepare the .frogans gTLD application (ICANN meetings in Paris, Cairo, Mexico City, Sydney, Seoul, Nairobi, Brussels, Cartagena, San Francisco and Singapore).

They also met with the head of the new gTLD program at ICANN, Kurt Pritz, in June 2010 in Brussels. They wanted to make sure that there would be no red flags concerning the envisioned use of the .frogans gTLD and that the application would be in line with ICANNʹs goals for the new gTLD program where it concerns the innovative use of the DNS and security.

The main concern in using a second-level domain (e.g. frogans.net), instead of using a gTLD for holding the computers dedicated to the functioning of the Frogans layer is that it would expose the entirety of Internet users who use the Frogans layer (publishers, hosting providers, end-users, etc.) to unmanageable security issues and threats, such as, for example, a nameserver change, unauthorized by the OP3FT, related to the delegation of that second-level domain.

The key benefit in using a gTLD instead of a second-level domain is that unauthorized delegation changes in the DNS root, which is under IANAʹs responsibility, are not possible.

Thus, the use of the .frogans gTLD is imperative for the OP3FT to guarantee the security, stability and reliability of the computers dedicated to the functioning of the Frogans layer for the benefit of all Internet users.


6. The .frogans gTLD registry back-end service provider

After having explored a number of back-end registry solutions for the operation of the .frogans gTLD, including the development of an in-house solution, the OP3FT selected AFNIC (Association Française pour le Nommage Internet en Coopération) as the registry back-end service provider for the .frogans gTLD.

AFNIC provides the OP3FT with assistance for the ICANN application and technical back-end registry services compliant with ICANNʹs requirements.


7. Financing the .frogans gTLD

The application for the .frogans gTLD and its ongoing operation is an investment, for the benefit of all Internet users, in the security, stability and reliability of the Frogans layer.

This investment conforms to the OP3FTʹs mission, which is to hold, promote, protect and progress Frogans technology.

The expenses for the ongoing operation of the .frogans gTLD are self-funded, representing a fixed sum of approximately 100,000 euros per year. This fixed sum represents 5.5% of the OP3FTʹs guaranteed minimum annual operating budget (1.8 million euros).

The annual operating budget of the OP3FT, a French “fonds de dotation,” is guaranteed for 10 years. It is provided through its endowment revenues according to the French Law on the Modernization of the Economy of August 4th, 2008, law number 2008-776.

Table of Contents :

1 - Receipt of data from registrars concerning registration of domain names and nameservers : Shared Registration System (SRS)
2 - Operation of the Registry zone servers
3 - Provision to registrars of status information relating to the zone servers for the TLD
3.1 - Standard DNS related status information
3.2 - Emergency DNS related status information
4 - Dissemination of TLD zone files.
4.1 - Incremental updates every 10 minutes
4.2 - Complete publication of the zone
4.3 - Propagation mechanism
4.4 - Zone File Access⁄Distribution
5 - Dissemination of contact or other information concerning domain name registrations (Whois service)
6 - Internationalized Domain Names
7 - DNS Security Extensions (DNSSEC)
7.1 - Registrar Services
7.2 - Signing Activity
8 - Other relevant services
8.1 - Security and Redundancy
8.2 - Consensus Policy Compliance

------------------------
1 - Receipt of data from registrars concerning registration of domain names and nameservers : Shared Registration System (SRS)

Operated by AFNIC, the .FROGANS TLD will adapt a domain shared registration system used in production by AFNIC to operate .fr zone and which has been fully functional for the past 15 years. This Extensible Provisioning Protocol (EPP) based Shared Registration System (SRS) has exhibited the ability to meet stringent SLAs as well as to scale from the operational management of an initial thousands of domain names to over 2 million in late 2011.

The SRS exists to interact with the Registrar’s systems, who are responsible for the provisioning of a registrant’s domain name with the .FROGANS TLD registry. Registrars interact with the registry through two primary mechanisms :
* Communication machine to machine via an EPP client (Registrar) to an EPP Server API (Registry).
* A Web Portal Interface that expresses the functionality of the EPP API. The Web Portal also provides access to user configuration and other back-office functions such as report and invoice retrieval.

SRS functionality includes standard functions and features such as :

* Domain Registration : The AFNIC SRS supports synchronous registrations (creations) of domain names through the EPP domain create command. It supports updates of associative status, DNS and DNSSEC delegation information and EPP contact objects with a domain and the deletion of existing domains. This allows Registrars to create domain registrations, modify them and ultimately delete them.

* Domain Renewal : The AFNIC SRS allows registrars to renew sponsored domains using the EPP renew command. The SRS automatically renews domain names upon expiry.

* Transfer : The AFNIC SRS supports the transfer of a given domain between two Registrars in a secure fashion by requiring two party confirmations and the exchange of a token (the EPP authinfo code) associated with the domain.

* Contact Objects : The AFNIC SRS supports the creation, update, association to domain objects, and deletion of EPP contact objects. This functionality supports the required information to supply contact data displayed in Registration Data Directory Services (RDDS) (Whois) systems.

* Hosts : A subordinate object of the domain object in an EPP based SRS, internal hosts are supported in the AFNIC SRS. These hosts cannot be removed when other 2nd level domains within the .FROGANS TLD zone are delegated to these nameservers. Delegation must be removed prior to the removal of the child hosts and a parent domain name to a given host in turn cannot be removed prior to the deletion of the related child host.

* Redemption Grace Period (RGP) & Restoring deleted domain name registrations : AFNIC SRS supports the RGP for the purpose of retrieving erroneously deleted domain names prior to being made available again for public registration.

Other features include :

* Additional EPP commands in order to manage and update both domain and contact objects in the registry which are EPP info, check, delete and update commands.

* An inline billing system which is synchronised with the SRS. Actions can be taken daily from simple alerts to concrete account blocking.

* Grace Periods and Refunds : the AFNIC SRS will support standard grace periods such as Add, Renew, Autorenew, Transfer and RGP grace periods. Refunds issued will reflect actual values deducted from registrar’s balance in consideration of any rebates issued conjunctively or separately for the relevant domain registration.

* The capacity to deal with reserved domain name registration. Reserved names are stored in a specific back office tool. Specific authorisations codes can be delivered out of band by support team to “unlock” creation of these reserved names. SRS uses standard EPP auth_info field in conformity with EPP RFCs to prevent or allow the registration of the domain name.

[see attached diagram Q23_1_authorisation_code_workflow.pdf]
Diagram : Reserved names unlock
Description : This diagram illustrates process to unlock registration of reserved names. An out of band email process is used to deliver a specific authorisation_code, that can be used in EPP or through the web interface to register the domain name.

SRS EPP functions are compatible with the following list of RFCs :
RFCs 5910, 5730, 5731, 5732, 5733 and 5734. Since AFNIC will implement the Registry Grace Period (RGP), it will comply with RFC 3915 and the successors of the aforementioned RFCs.

------------------------
2 - Operation of the Registry zone servers

The DNS resolution service is a core business of the Registry Operator. It is an essential function that must be provided with a very high level of service quality to satisfy queries concerning a zone 100% of the time with a response time as short as possible.

As the registry back-end service provider for the .FROGANS TLD, AFNIC has a set of sites, distributed internationally, to answer these queries. The high availability of responses is ensured by the number of servers that host the zone information; the response time is in turn linked to the geographical location of the servers (as near as possible to the exchange points and as a result to users).

To ensure a very high level of availability of information and a response time as short as possible to a DNS query for a given zone, AFNIC has chosen to deploy its own DNS architecture, operated by our teams, while also relying on a set of internationally recognized service providers in order to significantly increase the number of servers hosting the zone to be published.

The AFNIC DNS service is based on the standards of RFCs (RFCs 1034, 1035, 1982, 2181, 2182, 2671, 3226, 3596, 3597, 4343, and 5966 and any future successors), related to the Internet, and the DNS in particular.

In addition, special attention has been paid to the security component of the DNS servers and services in order to maintain a very high level of availability of the information, for example in the event of attacks or the denial of services. At present, a series of national and international servers are deployed as close as possible to the exchange points to ensure the DNS resolution service. To ensure a high level of availability, Anycast technology is applied to overcome the issues involved in the geographical location of sensitive servers. Through an effective pooling of DNS server resources, it ensures better resistance to denial of service attacks as the number of physical servers to attack is very high, and the geographical attraction of traffic by each server is very strong. Maintenance of the nodes is also improved since interventions on a given server have no effect on the visibility of the Anycast cloud for users.
As explained in the answer to Question 34 (Geographic Diversity), the registry also relies on two operators of Anycast clouds to expand the international coverage of the DNS nodes which must respond to queries for the domain extensions hosted on them. The two operators are Netnod Autonomica and PCH (Packet Clearing House) who are both known for their high quality services; in addition, Netnod Autonomica hosts one the root server i.root-servers.net.

------------------------
3 - Provision to registrars of status information relating to the zone servers for the TLD

Registrars interactions with the Registry Systems in two states in regards to the state of the TLD zone servers :
* an operational state where normal registry transactions and operational policies⁄practices result in a cause and effect in resolution of relevant domains AND
* an emergency state where resolution could be threatened by operational problems due to either internal or external factors to the DNS services.

------------------------
3.1 - Standard DNS related status information

The SRS supports related updates to domain objects that allow a Registrar to populate internal (glue record) and or external DNS hosts associated with the domain. External hosts result in the correct associated NS records being inserted into the current TLD zone file, this in turns results in DNS resolution being delegated to the identified external hosts. The SRS expresses this status to the Registrar as “Active” in both the EPP API and the SRS Web Portal. The registrar may suspend the NS records associated with the external hosts by applying an EPP client HOLD in the system, which will also be displayed as a status in the same manner. This holds true of the Registry when it applied “Server Hold”. Internal hosts follow the same behaviour with one exception, IP addresses must also be provided to the SRS by the registrar for Internal hosts, resulting in A records or⁄and AAAA records for IPv6 (also known as glue records) being added to the zone file.

------------------------
3.2 - Emergency DNS related status information

AFNIC registry services maintain emergency Network Operation Center (NOC) and Customer Service personnel on a 24⁄7⁄365 basis to address escalation and customer issue management. Part of these teams responsibility is to maintain contact lists for technical notification of regular or emergency situations including email lists, names and contact numbers. In the unlikely event that DNS resolution or DNS updates were or were expected to fall out of ICANN mandated SLAs, registrars will be contacted proactively by their email lists, status alerts will be posted to the Registry Operator’s Registrar Relations Web Portals and Customer Service personnel will be prepared to take and address calls on the current DNS status.

------------------------
4 - Dissemination of TLD zone files.

Publication of DNS resolution data to the TLD DNS nodes serving resolution :
One of the main challenges of DNS resolution is to provide updated information about the resources associated with a registered domain name. As soon as information is updated by a registrar on behalf of a customer, the latter expects the server to be accessible to its users as soon as possible.
For this reason, updates of DNS resolution data (publication) are entered into the AFNIC SRS, subsequently generated into incremented zone files, and are distributed to the authoritative DNS servers using the two following methods:
* Incremental updates every 10 minutes
and
* Complete publication of the zone.

------------------------
4.1 - Incremental updates every 10 minutes

The principle of publication by Dynamic Update (RFC 2136 and 2137) is designed to publish only the changes to the zone that have occurred since the last update. At the registry level, we have opted to propagate every 10 minutes the changes made during the last 10 minutes on all the zones managed. In this way, any changes made will naturally be published in the next 10 minutes.

------------------------
4.2 - Complete publication of the zone

In addition to the publication described above, the registry’s DNS operations team produces a complete publication of all the data for all the zones once a week by running a series of computer scripts which regenerates zonefile from database, through the same validation and integrity mechanisms as dynamic publication. This is used as a training for eventual recovery measures to be triggered.

------------------------
4.3 - Propagation mechanism

Whether during the publication by Dynamic Update or complete publication, the propagation mechanism is the same. The process involving the generation of the various zone files is triggered, without blocking any operation on the registration system.
These zone files are then transmitted in full to the authoritative server, via the AXFR protocol in conformance with RFC 5936. Once received and processed by the authoritative server, notifications are sent to secondary servers that will retrieve the changes in the different zones via the IXFR protocol in conformity with RFC 1995. The choice of an incremental (rather than complete) update of the zone files to the secondary servers during the dissemination process has been made to avoid sending large amounts of data to remote sites.

------------------------
4.4 - Zone File Access⁄Distribution

In compliance with Specification 4, Section 2, AFNIC registry services will offer a subscription service for qualifying applicants to download a stateful copy of the TLD zone file no more than once per 24 hours period. Distribution of the zone file will occur through the ICANN authorized Centralized Zone Data Access Provider.

------------------------
5 - Dissemination of contact or other information concerning domain name registrations (Whois service)

The AFNIC RDDS (Whois) service is in direct connection with the database of the Shared Registration System and offers access to the public administrative and technical data of the TLD. Contact data associated with registrations in the SRS is accessible both on port 43 (following specifications of RFC 3912) and through web access.

Data that can be accessed through the RDDS include:
* contact data : holder, administrative, technical, billing
* domain data : domain name, status
* host data : name servers, IP addresses
* ephemeris : creation, expiration dates
* registrar data
These data elements are fully compliant to the mapping of RFCs 5730 to 5734 and an example of standard port 43 output is given at the end of answer to Question 26 (WHOIS).

Both web and port 43 RDDS offer natively compliance with privacy law with a “restricted diffusion” flag. This option is activated through EPP (see Question 25 (EPP)) while creating or updating a contact and automatically understood by the Whois server to anonymize the data. The choice to activate restricted diffusion is made in compliance with the policy and the local rules of the TLD.

This service is accessible both in IPv4 and IPv6. The AFNIC RDDS service access is rate limited to ensure performance in the event of extreme query volumes generated in the cases of distributed denial of service (DDOS) and⁄or RDDS data-mining activities.

------------------------
6 - Internationalized Domain Names

Based on AFNIC’s Back-end registry’s operation experience, the .FROGANS TLD will allow registration of IDN domain names in full compliance with RFCs 5890 to 5893 and based on the character set described in detail in our answer to Question 44 (IDN). IDN will be implemented following the policies presented in detail in our answer to Question 44 (IDN). For the purpose of clarity, a brief summary of this information is presented below.

The list of characters includes the French language as well as several other regional languages in use in France : Occitan, Breton, Frankish, Reunion Creole, Catalan, Corsican and Guadeloupe Creole. The list consists of some of the characters of the Latin1 standard (ISO-8859-1) and the Latin9 standard (ISO-8859-15), respectively in Unicode Latin-1 Supplement and Latin Extended-A blocks.

------------------------
7 - DNS Security Extensions (DNSSEC).

AFNIC registry services fully support DNSSEC and will sign the .FROGANS TLD zone from initiation into the root servers.

------------------------
7.1 - Registrar Services

Operations are available for registrars through EPP with the SecDNS EPP extension version 1.1 exclusively (as defined in RFC 5910) or through registrars extranet (with a web form). Among the two interfaces defined in the RFC 5910, AFNIC chose the “dsData” interface : domain names keys are solely under registrars management and are not exchanged, only the keys hashes (DS records) are sent by the registrars to the registry back-end service provider. Each domain name can be associated to 6 distinct key materials at most.

Zonecheck : A complementary monitoring and validation service.
AFNIC notes that “Zonecheck” is a DNS monitoring and validation service that is outside standard registry services and could be offered by third parties other than a Registry Operator. In respect of DNSSEC monitoring, each change of DS data related to a domain name is verified by the AFNIC ZoneCheck tool, out of band of standard EPP registry functions. Registrar are notified via email of detected errors. This helps Registrars ensure the DNSSEC validation will operate correctly, for example by avoiding the “Security Lameness” scenario outlined in section 4.4.3 of RFC 4641.

Registrar transfer by default removes DS data from the zonefile. This is done to cover cases when a current signed domain names goes from a DNSSEC enabled registrar to another registrar that is not yet prepared to handle DNSSEC materials (the registrar can also be the DNS hoster or not, but in both cases DS data of the domain name has to flow from the registrar to the registry, hence the registrar must have the technical capabilities to do so).

------------------------
7.2 - Signing Activity

Each public-facing DNS server operated by AFNIC or through its anycast providers is fully DNSSEC enabled through RFC 4033, 4034, and 4035 by virtue of using standard open source software (BIND & NSD) that are developed according to these RFCs.

Each zone uses a standard Key Signing Key (KSK)⁄Zone Signing Key (ZSK) split (as defined in RFC 4641, section 3.1), which enables longer KSKs and frequent re-signing of zone content to deter DNSSEC-related brute force attacks and to make sure that keys rollovers are part of registry staff operational habits. All keys are created using RSA algorithms, as defined in RFC 4641 section 3.4 : KSKs are 2048 bits long (as recommended for “high value domains” in section 3.5 of RFC 4641), and ZSKs are 1024 bits. Algorithm SHA-256 (as defined in RFC 4509) is used for DS generations. Signatures of zone resources records are done using SHA-2 and more specifically RSA⁄SHA-256 as defined by RFC 5702.

Each zone has its set of dedicated KSKs and ZSKs: one of each is active at all time, while a second of each is ready to be used at next rollover. A third ZSK may be kept in the zone after being inactive (not used any more for signing) to ease transitions and make sure DNS caches can still use it to verify old resource records signatures. Following recommendations in section 4.1.1 “Time considerations” of RFC 4641, with a zone maximum TTL being 2 days and a zone minimum TTL of 1.5 hour, ZSK rollovers are done each 2 months, KSK rollovers are done each 2 years. Their expirations are monitored. Rollovers are operated according to the “Pre-Publish Key Rollover” procedure detailed in section 4.2.1.1 of RFC 4641.

1 year worth of key materials is generated in advance. Encrypted backup of keys is made on Hardware Security Module (HSM) cards (Storage Master Key), which are securely stored physically.

------------------------
8 - Other relevant services

------------------------
8.1 - Security and Redundancy

AFNIC maintains primary and secondary datacenter locations as well as redundant key personal operating locations. High availability of AFNIC Registry infrastructure is provided through the implementation of either load‐balancing, or fail‐over capacity in various layers of the architecture. It also enables fast scalability through expertise in virtualization technologies. AFNIC’s infrastructure is globally virtualized apart from services requiring very high performance rate and⁄or specific access to dedicated CPU for demanding computation such as DNSSEC zone signing or databases.
AFNIC maintains robust secure policies, protocols and third party testing and certification of security measures and practises. Systems involved in the AFNIC registry services used standard multi-factor authentication, high encryption transmission of data and are kept current with industry advancement in security technologies and best practices in prevention of data breaches. Registry systems follow standard EPP practices including required passphrases associated with each domain object and the use of those passphrases to successfully negotiate and verify domain transfers. Registrars are networked source restricted (2 IP addresses authorized by registrar) for SRS access in addition to the use of digital certificates and contact to Customer Service is restricted to registered Registrar personnel only (identified by personal passphrases⁄credentials listed on file).

------------------------
8.2 - Consensus Policy Compliance

AFNIC registry services will fully comply with Specification 1 of the Application Guidebook, below is a list of current consensus policies that have cause and effect on the systems of a registry operator. This list will be updated from time to time as per the ICANN process and the AFNIC registry services will be adjusted to maintain and support full compliance.

* Uniform Domain Name Dispute Resolution Policy (adopted by ICANN Board 26 August 1999; form of implementation documents approved 24 October 1999).
* Inter-Registrar Transfer Policy (effective on 12 November 2004, adopted by ICANN Board 25 April 2003; implementation documents issued 13 July 2004).
* Registry Services Evaluation Policy (effective on 15 August 2006, adopted by ICANN Board 8 November 2005; implementation documents posted 25 July 2006)
* AGP Limits Policy (effective on 1 April 2009, adopted by ICANN Board on 26 June 2008; implementation documents posted 17 December 2008)

Table of Contents

1 - Global description
2 - Shared Registration System (SRS) architecture
3 - SRS architecture diagram
4 - Detailed infrastructure
5 - Rate limitation
6 - Interconnectivity and synchronization with other systems
7 - Performance and scalability
8 - Resources
8.1 - Initial implementation
8.2 - On-going maintenance

------------------------
1 - Global description

As one of the critical registry functions, the SRS is part of the core of AFNIC back-end registry solution as deployed to fit the needs of the .FROGANS TLD.
It both provides services for registrars and generates the data used for DNS publication and resolution service. In that aspect, it is responsible for most of the SLA’s to be respected. The following description will provide full and detailed description of the architecture of the SRS both from an application and from an infrastructure point of view.
This architecture is the same as the one used in production by AFNIC to operate .fr zone and has been fully functional for the last 15 years, with the ability to meet stringent SLAs as well as to scale from the management of a few thousands domain names in operations to over 2 million in late 2011.

------------------------
2 - Shared Registration System (SRS) architecture

AFNIC SRS is based on a three-layer architecture : front-end, business logic, middleware.
These three layers are supported by the data layer which is described in detail in Question 33 (Database Capabilities).

= Front end : Extensible Provisioning Protocol (EPP) and extranet =

The automated front-end of the SRS is EPP.
The EPP interface and implementation complies with RFCs 3735 and 5730-5734. It is itself described in detail in Question 25 (EPP).
An extranet web interface also offers the same functions as the EPP interface.
Both theses interfaces are supported by the same middleware layer.

= Business logic : flexible policies =

The Business logic enables configurability in order to allow for the adjustment of registry systems to the chosen registry policies. Various policy-related parameters such as delay for redemption, access rate-limiting and penalties can be configured in this layer.
The Business logic also incorporates a scheduler which provides for semi-automated processes with human validation in order to address specific policy needs which cannot or should not be fully automated.

= Middleware : a guaranty for evolution and scalability =

The Middleware layer guarantees a consistent and registry oriented access for all the TLD data. All registry applications operate through this layer in order to centralize object management rules. It enables access through different programming languages (Java, php and Perl in AFNIC solution) with same rules and ease of switching from one language to another in case of application refactoring or migration.

= Data =

The Data layer is the structured data repository for domain, contact, operations, historization of transactions, as well as registrars and contracts data. It provides all the necessary resilient mechanisms to ensure 100% uptime and full recovery and backup.
It also provides a complete toolbox for the fine tuning of the various applications. This layer is described in more details in Question 33 (Database capacities).

------------------------
3 - SRS architecture diagram

[see attached diagram Q24_3_SRS_architecture_diagram.pdf]
Diagram : SRS architecture diagram
Description : This diagram shows global interaction between Internet, DMZ (Demilitarized Zone) and private network zones. Topology of network and servers is illustrated including dedicated IP address scheme and network flows.

This diagram does not shows additional sandbox and preproduction services. These services are offered respectively for registrars and back-end developer team to stabilize developments before production delivery. They are fully iso-functional to the SRS description above.

= SRS logical diagram =

Our robust infrastructure shows dual Internet Service Provider (ISP) connectivity both in IPv4 and IPv6 (Jaguar and RENATER), redundant firewall and switching infrastructure. This part of the architecture is mutualised for all TLDs hosted.

The networking architecture dedicates LAN for administration, backup and production.

Servers are hosted on different network zones : database for database, private for servers not visible on the internet and public for external servers visible on the DMZ. Dedicated zones are also set up for monitoring servers, administration servers or desktop and backup servers.
Each server is load balanced and the service is not impacted by the loss of one server, the capacity of each server being sized to be able to host the whole traffic.

Servers hosting the .FROGANS TLD are shared with up to an estimated number of 20 TLDs of comparable scale and use case.

= SRS physical diagram =

The IP scheme used is the following :

2001:67c:2218:1::4:0⁄64 for IPv6 Internet homing
192.134.4.0⁄24 for Ipv4 Internet homing

Production LAN
192.134.4.0⁄24 for public network IP range
10.1.50.0⁄24, 10.1.30.0⁄24 for private network IP ranges distributed on the zones described above.

Backup LAN
172.x.y.0⁄24 : x is different on each network zone. y is fixed to the value of the associated production LAN in the same zone (for example Private zone production LAN being 10.1.”50”.0⁄24, Private zone backup LAN is 172.16.”50”.0⁄24)

Administration LAN
172.z.y.0⁄24 : z is the value of x+1, x being the digit chosen for the corresponding Backup LAN in the same zone. y is fixed to the value of the associated production LAN in the same zone (for example Private zone production LAN being 10.1.”50”.0⁄24, Private zone administration LAN is 172.17.”50”.0⁄24).

Hot standby of the production database is automatically taken into account by the SRS Oracle Transparent Network Substrate configuration. Therefore if the database are migrated in hot standby due to failure of part of the system, the SRS access is automatically swapped to the new base.

------------------------
4 - Detailed infrastructure

The SRS modules play a central role in the back-end registry infrastructure. This is highlighted in terms of capacity expenditures (CAPEX) by the fact that SRS modules account for approximately 30% of the global CAPEX of the solution.

In the following description “server” will refer to either a physical or a virtual server.
Due to very fast growth of performance in storage and processors technologies, the infrastructure described below could be replaced by more powerful one available at the time of the set up for the same cost.

At the applicative and system level, AFNIC’s SRS systems are shared with up to an estimated number of 20 TLDs of comparable scale and use case.

AFNIC has invested in very efficient VMWare Vsphere virtualization infrastructure. It provides a flexible approach to recovery both through quick activation of a new fresh server in case of local failure (cold standby) and through global failover to a mirrored infrastructure on another site.
This comes in addition to natural redundancy provided by the load balanced servers.

Nevertheless, internal protocols and best practices for server virtualization have shown that very high I⁄O-intensive (Input⁄Output) application servers are not good clients for virtualization. The SRS is therefore hosted on virtualized infrastructure to the exception of the database, which presents very high rate of I⁄O, and is hosted on a dedicated physical infrastructure.

The whole SRS service is located in the primary datacenter used by AFNIC in production, the secondary datacenter serves as failover capacity.

The Front end is hosted on two load balanced virtual servers and two load balanced reverse proxies ensuring authentication of registrars.

The Business logic is hosted on two load balanced dedicated virtual servers. Scalability of these servers is ensured by quick resizing offered by virtualization technology if needed.

The Middleware is hosted on two load balanced dedicated virtual servers. It can be extended to any amount of servers needed to ensure performance commensurate with the amount of traffic expected. The dual use of Apache HAproxy and of a centralized lock mechanism ensure good queuing of each request in the system despite heavy load and parallelized middleware data access.

Scalability of all these servers are ensured by quick resizing offered by virtualization technology if needed.

All databases are based on Oracle technologies. The main database is replicated logically on two sites. Full local recovery processes are in place in case of loss of integrity through the Oracle redolog functions which provides full recovery by replay of historized logged requests.

The whole SRS service is located in the primary Tier 3 datacenter used by AFNIC in production, the secondary datacenter serves as failover capacity. Continuity mechanisms at a datacenter level are described in Questions 34 (Geographic Diversity), 39 (Registry Continuity) and 41 (Failover testing).

The detailed list of infrastructures involved can be described as follows :

This infrastructure is designed to host up to an estimated number of 20 TLDs of comparable scale and use case.

= Virtual servers =

EPP proxy : 2 servers
* Processor : 1 bi-core CPU
* Main memory : 8 GB of RAM
* Operating system : RedHat RHEL 6
* Disk space : 500 GB

EPP service : 2 servers
* Processor : 1 quad-core CPU
* Main memory : 16 GB of RAM
* Operating system : RedHat RHEL 6
* Disk space : 1 TB

Business logic : 2 servers
* Processor : 1 bi-core CPU
* Main memory : 16 GB of RAM
* Operating system : RedHat RHEL 6
* Disk space : 500 GB

Data Gateway : 2 servers
* Processor : 1 quad-core CPU
* Main memory : 16 GB of RAM
* Operating system : RedHat RHEL 6
* Disk space : 1 TB

= Data storage : see Question 33 (Database Capabilities) =

= Physical server =

Rate limiting database : 1 server
* Processor : 1 bi-core CPU
* Main memory : 8 GB of RAM
* Operating system : RedHat RHEL 6
* Disk space : 500 GB

Back up servers, backup libraries, Web whois server : mutualized with the global registry service provider infrastructure

= Additionnal infrastructure =

Failover infrastructure : 6 servers
* 1 bi-core CPU, 8 GB of RAM, RedHat RHEL 6, 500 GB

Sandbox infrastructure : 6 servers
* 1 bi-core CPU, 8 GB of RAM, RedHat RHEL 6, 500 GB

Preproduction infrastructure : 1 server
* 1 quad-core CPU, 16 GB of RAM, RedHat RHEL 6, 1 TB

------------------------
5 - Rate limitation

To ensure resiliency of the SRS a rate limitation and penalty mechanisms are in place.
Rate limitation and penalties are directly implemented on the front end server.

Access is rate limited through token-bucket algorithms with rate-limiting IP data stored on a dedicated database.
Penalties are applied as follow :
* Any command that follows a login command is immediately executed but the next one is only taken into account 2 seconds later. The following commands are not penalized (unless they do not follow one of the limitation rules).
* For the same domain name, the domain:check commands will not be able to follow themselves more than 2 times every 4 seconds. Beyond this rate, a 2 second penalty will be applied on the following domain:check commands (for the same domain name). For instance, it is possible to have a domain:check follow a domain:create command that already followed a first domain:check on a same domain name without any penalty.
* On the other hand, a customer making several domain:check commands on a same domain name will need to respect a 4 second delay between the first and the third call if he wishes not to be penalized.
* Any domain:create command on an already existing domain name induce an additional 2 seconds in the answer time of this command.
* Any domain:info command on a domain name that is not in your portfolio and for which you do not indicate the auth_info induce an additional 1 second in the answer time of this command.

------------------------
6 - Interconnectivity and synchronization with other systems

= Whois (RDDS) =

The whois service will be described in detail in question 27. It is hosted on two servers directly connected to the main production database through read only API. Data updated by the SRS are immediately visible in the Whois with no further synchronisation needed. Rate limitation is applied on RDDS service to avoid any load on the database due to Whois direct access. Hot standby of the production database is automatically taken into account by the Whois Oracle Transparent Network Substrate configuration. Therefore if SRS and database are migrated in hot standby due to failure of part of the system, the Whois service is automatically swapped to the new architecture.

= Back office⁄billing⁄Escrow =

Back-office, escrow and billing system is hosted on mutualized server. It operates directly on production data through the middleware layer to ensure integrity of data. These can be considered as fully synchronous applications. Hot standby of the production database is automatically taken into account by the Middleware layer Transparent Network Substrate configuration. Therefore if SRS and database are migrated in hot standby due to failure of part of the system, the back office and billing service are automatically swapped to the new architecture.

= Monitoring =

Monitoring is operated through probes and agents scanning systems with a 5 minutes period. The monitoring system gets snmp data from all servers described in the SRS architecture and also from dedicated Oracle monitoring agent for the database. A specific prove for EPP simulating a full domain creation is also activated, still with the 5 minutes period.

= Dispute resolution =

Any operation on domain names triggered in the context of a dispute resolution is made through a back-office tool (see Back office)

= DNS publication =

DNS publication relies on a specific table of the production database hosted on the same oracle instance. These data are directly generated by the SRS system. Dynamic Update batches are generated at each operation. The use of theses batches for DNS Dynamic update or of the whole data for full zonefile generation are made directly from these production data. No further synchronization is needed. The detail of frequency and workflow for dns publication is described in Question 35 (DNS) and Question 32 (Architecture). Hot standby of the production database is automatically taken into account by the DNS publication Transparent Network Substrate configuration. Therefore if SRS and database are migrated in hot standby due to failure of part of the system, the dns publication is automatically swapped to the new architecture.

------------------------
7 - Performance and scalability

The Registry’s SRS offers high level production SLAs and derives from the branch of systems that have evolved over the last 15 years to successfully operate a set of french ccTLDs.

The Registry’s SRS is used to operate .fr, .re, .yt, .pm, .tf, .wf TLDs. It is used by more than 800 registrars in parallel managing more than 2 millions domain names.

AFNIC’s SRS is designed to meet ICANN’s Service-level requirements as specified in Specification 10 (SLA Matrix) attached to the Registry Agreement.

Actual and current average performance of AFNIC’s SRS is :
* SRS availability : 99,4%
* SRS session-command RTT : 400ms for 99,4% of requests
* SRS query command RTT : 500ms
* SRS transform command RTT : 1,4 s on availability period
* SRS max downtime : 2 hours⁄month

As described in Question 31 (Technical Overview) in relation to each of the phases of the TLD’s operations, the following transaction loads are expected on the SRS :
* launch phase : up to 200 requests⁄hour
* routine ongoing operations : up to 1,000 requests⁄hour

The system is designed to handle up to 50,000 domain names and up to 2 requests per second.

The targeted TLD size being approximately 1,000 domain names after 3 years of operations and the expected peak transaction rate being 1,000 requests⁄hour this ensures that enough capacity is available to handle the launch phase, unexpected demand peaks, as well as rapid scalability needs.

Capacity planning indicators are set up to anticipate exceptional growth of the TLD.
Technologies used enables quick upgrade on all fields :
* Servers : virtual resizing to add CPUs or disk space if resource is available on the production ESX servers. If not, 2 spare additional ESX servers can be brought live if additional performance is needed.
* Database : database capacity has been greatly oversized to avoid need of replacement of this physical highly capable server. Precise capacity planning will ensure that sufficient delay will be available to acquire new server if needed. A threshold of 40% of CPU use or total storage capacity triggers alert for acquisition.

------------------------
8 - Resources

Four categories of profiles are needed to run the Registry’s Technical Operations : Registry Operations Specialists (I), Registry Systems Administrators (II), Registry Software Developer (III) and Registry Expert Engineers (IV). These categories, skillset and global availability of resources have been detailed in Question 31 (Technical Overview of Proposed Registry) including specific resources set and organisation to provide 24⁄7 coverage and maintenance capacity.
Specific workload for SRS management is detailed below.

------------------------
8.1 - Initial implementation

The set up is operated on the pre-installed virtualization infrastructure. It implies actions by system, database and network administrators to create the virtual servers and install the applicative packages.

Then, developers, assisted by a team of experts and senior staff members apply proper configuration for the given TLD. Specific policy rules are configured and tested.

The initial implementation effort is estimated as follows :

Database Administrator 0.03 man.day
Network Administrator 0.03 man.day
System Administrator 0.03 man.day
Software Developer 0.10 man.day
Database Engineer 0.10 man.day
Software Engineer 0.20 man.day
DNS Expert Engineer 0.10 man.day

------------------------
8.2 - On-going maintenance

On-going maintenance on the SRS includes integration of new policy rules, evolution of technology, bug fixing, infrastructure evolution, failover testing.

Although all the defined technical profiles are needed for such on-going maintenance operations, on a regular basis, it is mainly a workload handled by monitoring and development teams for alert management and new functional developments, respectively.

The on-going maintenance effort per year is estimated as follows, on a yearly basis :

Operations Specialist 0.40 man.day
Database Administrator 0.10 man.day
Network Administrator 0.10 man.day
System Administrator 0.10 man.day
Software Developer 0.20 man.day
Database Engineer 0.05 man.day
Network Engineer 0.05 man.day
System Engineer 0.05 man.day
Software Engineer 0.05 man.day

Table of Contents

1 - Global description
2 - Description of commands
2.1 - Introduction
2.2 - Global commands
2.2.1 - session management commands ‘greeting’, ‘hello’, ‘login’, ‘logout’
2.2.2 - poll command ‘poll’
2.3 - domain commands
2.3.1 - query commands ‘check’, ‘info’
2.3.2 - transform commands
2.4 - contact command
2.5 - Return Codes
3 - Compliance to RFCs
3.1 - Delivery process
3.2 - XML validation
3.3 - Cross checking
4 - Specific extensions
4.1 - Specific extension : DNSSEC
4.2 - Specific extension : IDN
4.3 - Specific extension : Sunrise period
4.3.1 - New objects
4.3.2 - command extensions
4.3.2.1 - EPP Query Commands
4.3.2.2 - EPP Transform Commands
4.3.2.2.1 - EPP ʹcreateʹ Command
4.3.2.2.2 - EPP ʹupdateʹ Command
4.3.2.2.3 - EPP ʹdeleteʹ Command
5 - Resources
5.1 - Initial implementation
5.2 - On-going maintenance

------------------------
1 - Global description

The main service of the Shared Registration System (SRS) for its registrars is the Extensible Provisioning Protocol (EPP) interface. The interface has been developed and is maintained in full compliance with the relevant standards RFCs 5730-5732 and with RFCs 5910 and 3735 for the standard registration interface. Contacts are handled as described in RFC 5733. Transport is guaranteed according to RFC 5734. In addition, AFNIC’s EPP implementation is also compliant with RFCs 4034, 5730 and 5731 for DNSSEC support and with RFCs 5890 and 5891 for Internationalized Domain Name (IDN) support.

The EPP service is available through IPv4 and IPv6, based on a SSL certificate authentication.
No specific extension is used.

Note : Throughout the document we write the XML markups describing the EPP requests between the two characters ʹ and ʹ.

For contact management, the registry service provider uses a dedicated “Repository Identifier” for each TLD, this Repository identifier being declared to IANA prior to the launch of the TLD. It is also used as a post-extension to contact nic-handles, each contact for a given TLD being then identified by a unique code XX1234-REPID. An example of this declaration can be found for .fr extension (2008-05-10) at IANA epp repository identifier’s page :

[...]
NORID, #x004E #x004F #x0052 #x0049 #x0044 UNINETT Norid AS 2007-12-10 info&norid.no
FRNIC, #x0046 #x0052 #x004e #x0049 #x0043 AFNIC 2008-05-29 tld-tech&afnic.fr
CIRA, #x0043 #x0049 #x0052 #x0041 Canadian Internet Registration Authority 2009-07-22 info&cira.ca
[...]

------------------------
2 - Description of commands

------------------------
2.1 - Introduction

The EPP interface, based on a double system of real-time answer by the server and asynchronous notifications, implements all standard operations : ‘domain:create’ (1 to 10 years), ‘domain:info’, ‘domain:check’, ‘domain:transfer’, ‘domain:update’, ‘domain:renew’. Similar commands are available concerning contact objects.
The registry’s EPP server implement name servers management as domain name attributes in conformity with RFC 5732.

[see attached diagram Q25_2.1_EPP_xsd_main_schema.pdf]
Diagram : EPP xsd main schema
Description : Registry service provider SRS EPP interface is based on standard xsd schema as defined in RFC 5730.

In the following description of the commands, an example of client command and server answer has been added only for the create command as an example. All other commands work in the same way in full compliance with descriptions and schema of RFCs 5730-5734 and same examples can be found in the RFCs text.

------------------------
2.2 - Global commands

------------------------
2.2.1 - session management commands ‘greeting’, ‘hello’, ‘login’, ‘logout’

As all of these commands are basic and totally compliant with the IETF’s STD69 (RFCs 5730 to 5734), they have not be described again here.

Focus points are :
* Enforcing a limit of 2 simultaneous connection per registrar (checked at login), ensuring equitable access for all registrars.
* List of namespaces announced in ʹgreetingʹ is strictly checked in registrar ʹloginʹ command.
* ʹhelloʹ can be used by registrars as a keepalive command, otherwise inactive sessions are closed by server after 20 minutes.

------------------------
2.2.2 - poll command ʹpollʹ

For some operation on objects, notifications are added in a queue that can be read by using the ʹpollʹ command. The use of the ʹpollʹ command will retrieve the oldest message in the queue. The number of messages awaiting in the queue is indicated at each command answer with the ʹmsgQʹ element. To delete a message from the queue, the ʹpollʹ command should be used with the message number as indicated in RFC 5730.

------------------------
2.3 - domain commands

------------------------
2.3.1 - query commands ʹcheckʹ, ʹinfoʹ

ʹcheckʹ command allows the client to check if a domain object is available.
ʹinfoʹ command allows the client to retrieve information on any objects (domain names or contacts) that are indicated in the command. Registrars can only use this command for objects they already manage in their portfolio. This command can also be used for domain names outside the registrar’s portfolio if the ʹauth_infoʹ code that protects the domain is given as well.

------------------------
2.3.2 - transform commands

In compliance with RFCs 5730 (commands presentation), 5731 (domain objects), 5732 (contact objects) and 5910 (DNSSEC specifications) AFNIC’s Registry solution use the following commands that allow for objects updates :

= ʹcreateʹ =

The EPP protocol (RFC 5730) allows domain name creation (RFC 5731). The registry service provider allows two types of creations: direct domain creations (with auth_info freely determined by the registrar) and domain names creation “with authorization code” (the correct auth_info value must be sent for the creation to succeed)

Both are standard domain:create command as defined in the RFCs.

[see attached diagram Q25_2.3.2_EPP_create_command_example.pdf]
Diagram : EPP client create command and server answer example
Description : This is a standard EPP client create command following RFC 5731. Parameters sent in the following example are domain name, period of registration, registrant identifier, administrative, technical and billing identifier, and auth_info password followed by standard EPP server create command answer compliant with RFC 5731. Parameters sent in the answer are result code, message, creation and expiry date, and client and server transaction ID.

Creation “with authorization code” enables the registry service provider to manage protected names or names under specific registration conditions. An authorization code is associated to three items (the registrar, the domain name and the holder nic-handle ) and is delivered outside the automated process through a manual process defined by a specific policy rule. The registry-generated authorization code must be present in the ʹdomain:authInfoʹ item of the creation request. No registrar-computed value is permitted.
In every case, domain creation proceeds through standard EPP command.

[see attached diagram Q25_2.3.2_SRS_authorisation_code.pdf]
Diagram : SRS authorisation code
Description : The EPP auth_info field that can usually be freely filled in by the registrar has a specific use for registration of reserved names : an authorisation_code is delivered through an out of band process and must be used in the create command for the answer to be successful.

= ʹupdateʹ =

The registry offers EPP ʹdomain:updateʹ command to :
* update the administrative, technical, registrant contacts of a domain name
* update the DNS and DNSsec configuration of a domain name
* update the status of a domain name or its auth_info

This command is also used to add or delete signed delegations (DS records), through a ʹsecDNS:updateʹ extension if DNSSEC operations are wanted and if the secDNS extension was chosen by the client at login.

When requested the status of domain name is changed to “pendingUpdate”.

= ʹdeleteʹ =

The whole deletion process (including redemption grace period and pending delete) of a domain name comes with a restoration mechanism (restore). This mechanism, based on RFC 3915, is applied to the deletion operation only.

The status of the domain name is switched to ʺpendingDeleteʺ for the total duration of the ʺredemption grace periodʺ and as long as the domain is not restored or totally deleted.

= ʹtransferʹ =

The registry offers standard EPP ʹdomain:transferʹ command to allow a change of registrar to the registrant.

A transfer can be initiated only by an incoming registrar and using the auth_info that the registrant has given him. This standard mechanism acts as a security and associates the triggering of transfer to the acceptance of the owner of the domain.
The transfer operation can be triggered only if the domain is not protected by a clientTransferProhibited lock.

The transfer implementation follows RFC 5730 section 2.9.3.4 and its lifecycle follow the inter registrar transfer policy as revised by the ICANN in 2008.

------------------------
2.4 - contact command

Postal addresses are managed as indicated in RFC 5731 with the following specific rules : only the type “loc” for postal addresses is accepted and only one element of type ʹcontact:postalInfoʹ can be indicated for the contact .

ʹdiscloseʹ parameters is implemented and enables to activate restricted publication in the RDDS.
The choice to activate restricted diffusion is made in compliance with the policy and the local rules of the TLD towards privacy law.

------------------------
2.5 - Return Codes

Some operations under normal working conditions of the SRS will answer with a 1000 return code. Otherwise, two different levels of return codes have been chosen according to the two different types of problems that can happen on the SRS :
* minor problems answer with Return code 1001 : Minor problems do not affect requests reception. This code indicates the command was taken into account but that its complete execution is delayed. The final result will be known later on and will be sent in a message placed in the notification queue of the concerned registrar(s).
* blocking problems answer with Return code 2400 “command failed” : no operations that transform a domain name can be taken into account.

------------------------
3 - Compliance to RFCs

The system has been launched compliant with RFCs. Mechanisms are in place to ensure that ongoing maintenance and new functional delivery stay compliant with RFCs.

------------------------
3.1 - Delivery process

The SRS evolutions are developed on the development environment.
The development process implies strict coding rules and use of shared best practices. Pair programming is standard practice. Unit test are developed prior to function development to ensure resiliency of the produced code.

Delivery process take place in four steps :
* 1st step : XML validation and RFC compliance is checked through automated tools. A 100% compliance signal must be received to be able to proceed to second step.
* 2nd step : delivery to the pre-production environment. The development is delivered on the preproduction environment. This environment is available for internal testing team. They proceed through a standard Operational Test which goes through a full lifecycle of a domain name. Specific tests are made on new functions in any.
* 3rd step : delivery to the sandbox environment. This sandbox environment is opened for registrar where they have two accounts to validate their clients before production activation.
* 4th step : the new release is delivered in production.

------------------------
3.2 - XML validation

EPP RFC compliance is reached through three mechanisms :
* a batch of unitary tests on each operation, each answer of the server being validated through the XSD schema.
* XML validation through perl XML::LibXML::Schema library
* fuzzy testing, by sending garbage input and checking error return codes.

------------------------
3.3 - Cross checking

EPP cross checking partnership is established with .AT Registry operator to validate in sandbox environment prior to delivery in production through mutual agreement.

------------------------
4 - Specific extensions

------------------------
4.1 - Specific extension : DNSSEC

The EPP server provides the secDNS-1-1 extension as described in RFC 5910. Implementation specifications are as follows :
* The server only supports “the DS data interface” (ʹsecDNS:dsDataʹ); section 4.1 of RFC 5910, without information on the associated key (the ʹsecDNS:keyDataʹ element is not included); if information on the key is indicated the server will answer with a 2102 error code.
* DNSSEC elements are only accepted during an update operation request. If included during a create operation the server will answer with a 2103 error code.
* Each domain name can have up to 6 associated DS records : the number of elements ʹsecDNS:dsDataʹ present in the ʹsecDNS:addʹ section during an update operation is therefore limited in order to have the domain name’s final status with no more than 6 DS records.
* The maxSigLife attribute is not supported, its presence inside a client request will generate a 2102 error code.
* The urgent attribute is not supported, its presence inside a client request will generate a 2102 error code.

[see attached diagram Q25_4.1_EPP_xsd_dnssec_extension_schema.pdf]
Diagram : EPP xsd dnssec extension schema
Description : Registry service provider DNSsec EPP secDNS-1-1 extension is based on standard xsd schema as defined in RFC 5910.

------------------------
4.2 - Specific extension : IDN

No specific IDN extension has been used. The script used for the TLD is declared in the greetings and no further indication is needed in the following transaction. Usage is in full compliance with RFCs 5890, 5891, 5892, 5893, and 5894. This may be a pending situation : if a standard IDN extension was to be produced in the months to come it would be added to the EPP schema in order to deal more precisely with each specific language management policies.

------------------------
4.3 - Specific extension : Sunrise period

Sunrise period is managed through a specific EPP extension. The sunrise registration workflow is described in Question 29 (Right Protection Mechanism).

The extension used is described below but will follow work in progress at the IETF initiated by Cloud Registry (draft-tan-epp-launchphase-01.txt). The xsd schema has been designed by AFNIC’s partner CORE and is fully in accordance with the draft. It could be modified before the launch if the IETF draft was to be accepted as an RFC with modifications.

AFNIC Registry extension is fully compatible with extension mechanism described in RFC 5730. It offers trademark holders a specific mapping to provide information related to trademarks. It also enables query function to keep the sunrise process transparent to everybody.

For illustration and further information purposes, please refer to the Q25_4.3_EPP_xsd_sunrise_extension_schema.pdf file attached (EPP XSD sunrise extension schema) which describes the registry back-end services provider’s EPP extension XSD schema used to deal with sunrise period. This schema is designed based on the work in progress at IETF, as initiated by Cloud Registry (draft-tan-epp-launchphase-01.txt). This extension is fully compatible with extension mechanism described in RFC 5730.

------------------------
4.3.1 - New objects

application : to deal with multiple demands on same domain name. The server creates an application object corresponding to the request and assigns an identifier for the application and returns it to the client. This mapping defines an ʹlp:applicationIDʹ element which is used to specify an ID to this object.

phase : optionnal element ʹlp:phaseʹ to be used in case of multiple sunrise phases.

status : status of each application in link with internal state of the process of the application. The ʹlp:statusʹ values that can be used in order to process the applications are pending, invalid, validated, allocated, rejected. These statuses have to be mapped with the sunrise workflow described in Question 29 (Right Protection Mechanism).

claim : claim object contains the details needed to applicantʹs prior right to the domain name.
The ʹlp:claimʹ element has the boolean ʺpreValidatedʺ attribute, which indicates whether a third party validation agency has already validated the claim in case of inter connection with the IP clearing house.

Several child elements of the ʹlp:claimʹ element are defined :
ʹlp:pvrcʹ, the Pre-Validation Result Code, is a string issued by a third-party validation agent. ʹlp:claimIssuerʹ contains the ID of a contact object (as described in RFC 5733) identifying the contact information of the authority which issued the right (for example, a trade mark office or company registration bureau).
ʹlp:claimNameʹ identifies the text string in which the applicant is claiming a prior right. ʹlp:claimNumberʹ contains the registration number of the right (i.e. trademark number or company registration number).
ʹlp:claimTypeʹ indicates the type of claim being made (e.g. trademark, symbol, combined mark,
company name).
ʹlp:claimEntitlementʹ indicates the applicantʹs entitlement to the claim (i.e. owner or licensee). ʹlp:claimRegDateʹ contains the date of registration of the claim.
ʹlp:claimExDateʹ contains the date of expiration of the claim.
ʹlp:claimCountryʹ indicates the country in which the claim is valid.
ʹlp:claimRegionʹ indicates the name of a city, state, province or other geographic region in which the claim is valid. This may be a two-character code from WIPO standard ST.3.

------------------------
4.3.2 - command extensions

------------------------
4.3.2.1 - EPP Query Commands

ʹinfoʹ command is the only extended query command.

In order to indicate that the query is meant for an application object, an ʹlp:infoʹ element is sent along with the regular ʹinfoʹ domain command.

The ʹlp:infoʹ element contains the following child elements :
ʹlp:applicationIDʹ, the application identifier for which the client wishes to query, and ʹlp:phaseʹ (optional), the phase the application is associated with.
If the query was successful, the server replies with an ʹlp:infDataʹ element along with the regular EPP ʹresDataʹ.
The ʹlp:infData contains the following child elements :
* ʹlp:applicationIDʹ the application identifier of the returned application.
* ʹlp:phaseʹ (optional) the phase during which the application was submitted or is associated with.
* ʹlp:statusʹ (optional) status of the application.
* ʹlp:claimʹ (optional) one or more ʹlp:claimʹ elements.
If present, the ʹlp:claimʹ elements may contain the child elements as described above in the claim object description.

------------------------
4.3.2.2 - EPP Transform Commands

There are three extended EPP transform commands : ʹcreateʹ, ʹdeleteʹ and ʹrenewʹ

------------------------
4.3.2.2.1 - EPP ʹcreateʹ Command

The EPP ʹcreateʹ command is used to create an application. Additional information is required to submit a domain name application during a launch phase :
* ʹlp:phaseʹ (optional), the phase the application should be associated with
* ʹlp:claimʹ (optional) elements to substantiate the prior rights of the applicant.

When such a ʹcreateʹ command has been processed successfully, the EPP ʹextensionʹ element in the response contains a child ʹlp:creDataʹ element that identifies the registry launchphase namespace and the location of the registry launchphase schema. The ʹlp:creDataʹ element contains a child ʹlp:applicationIDʹ element, which informs the registrar about the application ID the server has assigned.

------------------------
4.3.2.2.2 - EPP ʹupdateʹ Command

This extension defines additional elements to extend the EPP ʹupdateʹ command to be used in conjunction with the domain name mapping.
Registry policies permitting, clients may update an application object by submitting an EPP ʹupdateʹ command along with an ʹlp:updateʹ element to indicate the application object to be updated.
The ʹlp:updateʹ element contains the following child elements:
* ʹlp:applicationIDʹ the application identifier for which the client wishes to update.
* ʹlp:phaseʹ (optional) the phase during which the application was submitted or is associated with.

------------------------
4.3.2.2.3 - EPP ʹdeleteʹ Command

Registry policies permitting, clients may withdraw an application by submitting an EPP ʹdeleteʹ command along with an ʹlp:deleteʹ element to indicate the application object to be deleted. The ʹlp:deleteʹ element contains the following child elements:
* ʹlp:applicationIDʹ the application identifier for which the client wishes to delete.
* ʹlp:phaseʹ (optional) the phase during which the application was submitted or is associated with.

------------------------
5 - Resources

Four categories of profiles are needed to run the Registry’s Technical Operations : Registry Operations Specialists (I), Registry Systems Administrators (II), Registry Software Developer (III) and Registry Expert Engineers (IV). These categories, skill set and global availability of resources have been detailed in Question 31 (Technical Overview of Proposed Registry) including specific resources set and organisation to provide 24⁄7 coverage and maintenance capacity.
Specific workload for EPP management is detailed below.

------------------------
5.1 - Initial implementation

Then, developers, assisted by a senior staff member expert in internet technologies and RFCs apply proper configuration for the given TLD. Compliance is strictly tested.

The initial implementation effort is estimated as follows :

Database Administrator 0.03 man.day
Network Administrator 0.03 man.day
System Administrator 0.03 man.day
Software Developer 0.10 man.day
Software Engineer 0.20 man.day

------------------------
5.2 - On-going maintenance

On-going maintenance on the SRS includes integration of new policy rules, evolution of technology, bug fixing, infrastructure evolution, failover testing.

The on-going maintenance effort per year is estimated as follows, on a yearly basis :

Operations Specialist 0.20 man.day
System Administrator 0.10 man.day
Software Developer 0.15 man.day
Software Engineer 0.10 man.day

Table of Contents

1 - General description
2 - Data access
2.1 Typology of accessible data
2.2 Profiles for data access control
3 - RDDS architecture
4 - RDDS infrastructure
5 - Rate limitation
6 - Reverse lookups
7 - Interconnectivity and synchronization with other systems
8 - Performance and scalability
9 - ICANN Bulk access compliance
10 - RFC compliance
11 - Resources
11.1 - Initial implementation
11.2 - On-going maintenance

------------------------
1 - General description

Registration Data Directory Service (RDDS) is one of the five vital functions of the Registry.
It is in direct connection with the database of the Shared Registration System and offers access to the public administrative and technical data of the registry.
The registry back-end solution implements data access through various interfaces that will be described below as well as their data access policies.

The main focus will be made on Whois on port 43 following RFC 3912 which is the main point of access.
The web Whois offers similar functionalities, is based on the same architecture and will be presented through screenshots.

The following description will provide full and detailed description of the architecture of the RDDS both from an application and from an infrastructure point of view.
This architecture is the same as the one used in production by AFNIC for .FR zone and has been fully functional for the last 15 years, with the ability to meet stringent SLAs as well as to scale from the management of a few thousands domain names in operations to over 2 million in late 2011.

------------------------
2 - Data access

When considering the data access services, we must address :
* the typology of accessible data
* access control : who can access what kind of data
* performance : guarantee of availability and performance for requesting data

Potential limitations to the systems will also be described.
To be able to maintain a good access to everybody (registrar, holders, outside world), our back-end solution provides multiple access with consistent role and access policies.

------------------------
2.1 Typology of accessible data

Data that can be accessed through the RDDS are mainly :
* contact data : holder, administrative, technical, billing
* domain data : domain name, status
* host data : name servers, IP addresses
* ephemeris : creation, expiration dates
* registrar data

These data are all described in the RFCs and fully compliant to the mapping of RFCs 5730 to 5734 and an example of standard port 43 output is given at the end of the present answer.

------------------------
2.2 Profiles for data access control

= Whois for registrars =

The main registrar access tool is our RDDS service accessible both on port 43 following specifications of RFC 3912 and through web access.
Both web and port 43 RDDS offer natively compliance with privacy law with a “restricted disclosure” flag if needed by the TLD. This option is activated through Extensible Provisioning Protocol (EPP) standard ʹdiscloseʹ parameters while creating or updating a contact and automatically understood by the whois server to anonymize the data.
This service is accessible both in IPv4 and IPv6.
RDDS access for registrar is rate limited to ensure performance. (see performance)

= Public whois =

RDDS access is also available on port 43 to everybody through a rate limited access to ensure performance. (see performance)

= Legal requirements =

AFNIC back end solution implements by default French privacy laws with opt-out holder personal data privacy.
This option can be deactivated if necessary to be compliant with the policy of the TLD.

------------------------
3 - RDDS architecture

= RDDS architecture =

RDDS is running on two load balanced front virtual servers directly connected to two databases : the production database for data access, and a rate-limiting service database which applies rate-limiting policies and store IP involved. This server implements token bucket algorithm to flatten traffic on the server.

The two front servers are load balanced using classical round robin implementation.

The network infrastructure is the same as described in the global architecture (referred to below) and no specific dedicated switch or router is to be considered as the rate limiting tool is an applicative one. A global description of the network infrastructure (switch and routers involved) can be found in answers to Question 32 (Architecture).

[see attached diagram Q26_3_RDDS_architecture_diagram.pdf]
Diagram : RDDS architecture diagram
Description : This diagram shows global interaction between Internet, DMZ and private network zones. Topology of network and servers is illustrated including dedicated IP address scheme and network flows.

= RDDS logical diagram =

Our robust infrastructure shows dual Internet Service Provider (ISP) connectivity both in Ipv4 and Ipv6 (Jaguar and RENATER), redundant firewall and switching infrastructure. This part of the architecture is mutualized for all TLDs hosted.

The networking architecture dedicates LAN for administration, backup and production.

Servers are hosted on different network zones : database for database, private for servers not visible on the internet and public for external servers visible on the DMZ. Dedicated zones are also set up for monitoring servers, administration servers or desktop and backup servers.
RDDS servers are directly on the public zone.
Each server is load balanced and the service is not impacted by the loss of one server, the capacity of each server being sized to be able to host the whole traffic.

Servers hosting the .FROGANS TLD are shared with up to an estimated number of 20 TLDs of comparable scale and use case.

= RDDS physical diagram =

The IP scheme used is the following :

2001:67c:2218:1::4:0⁄64 for IPv6 Internet homing
192.134.4.0⁄24 for Ipv4 Internet homing

* Production LAN
192.134.4.0⁄24 for public network IP range
10.1.50.0⁄24, 10.1.30.0⁄24 for private network IP ranges distributed on the zones described above.

* Backup LAN
172.x.y.0⁄24 : x is a different on each network zone. y is fixed to the value of the associated production LAN in the same zone (for example Private zone production LAN being 10.1.”50”.0⁄24, Private zone backup LAN is 172.16.”50”.0⁄24)

* Administration LAN
172.z.y.0⁄24 : z is the value of x+1, x being the digit chosen for the corresponding Backup LAN in the same zone. y is fixed to the value of the associated production LAN in the same zone (for example Private zone production LAN being 10.1.”50”.0⁄24, Private zone administration LAN is 172.17.”50”.0⁄24)

Hot standby of the production database is automatically taken into account by the RDDS Oracle Transparent Network Substrate configuration. Therefore if the database are migrated in hot standby due to failure of part of the system, the Registration Data Directory Services (RDDS) access is automatically swapped to the new base.

------------------------
4 - RDDS infrastructure

At the applicative and system level, AFNIC’s SRS systems are shared with up to an estimated number of 20 TLDs of comparable scale and use case.

The RDDS is therefore hosted on virtualized infrastructure on the public zone (Demilitarized Zone - MZ) to the exception of the database, which presents very high rate of I⁄O (Input⁄Output), and is hosted on a dedicated physical infrastructure on the private zone.

The rate limiting database is hosted on one physical dedicated physical server. This server represents no failure point as a failure of the rate limiting system doesn’t affect the service (a standard uniform limitation is then applied instead of intelligent rate limiting).
The main database is the production database also used by the SRS and other registry vital functions and is described more in detail in Question 33 (Database Capabilities).

Databases are based on Oracle technologies. The main database is replicated logically on two sites. Full local recovery processes are in place in case of loss of integrity through the Oracle redolog functions which provides full recovery by replay of historized logged requests.

The whole RDDS service is located in the primary Tier 3 datacenter used by AFNIC in production, the
secondary datacenter serves as failover capacity. Continuity mechanisms at a datacenter level are described in Questions 34 (Geographic Diversity), 39 (Registry Continuity) and 41 (Failover testing).

The detailed list of infrastructures involved can be described as follows :

This infrastructure is designed to host up to an estimated number of 20 TLDs of comparable scale and use case.

= Virtual servers =

RDDS server : 2 servers
* Processor: 1 bi-core CPU
* Main memory: 16 GB of RAM
* Operating system: RedHat RHEL 6
* Disk space: 500 GB

= Data storage : see Question 33 (Database Capabilities) =

= Physical server =

Rate limiting database : 1 server
* Processor: 1 bi-core CPU
* Main memory: 8 GB of RAM
* Operating system: RedHat RHEL 6
* Disk space: 500 GB

Back up servers, backup libraries, Web whois server : mutualized with the global registry service provider infrastructure

= Additionnal infrastructure =

Failover, sandbox, preproduction infrastructure : 3 server
* 1 bi-core CPU, 16 GB of RAM, RedHat RHEL 6, 500 GB

------------------------
5 - Rate limitation

To ensure resiliency of the RDDS a rate limitation mechanism is in place.
RDDS is largely used by various public users and registrars, some of them for domain name drop catching. Potentiality of heavy load on this service is very high.
Therefore a rate limitation is applied with threshold calculated from the level of activity expected in order not to penalize normal use of the service. A double level mechanism enables different threshold for identified IP (white list) from registrar and for the public access.

Rate limitation is directly implemented on the front end server.

Access is rate limited through token-bucket algorithms with rate-limiting IP data stored on a dedicated database.
Penalties are applied as follow :
* any IP : 7,200 request ⁄ 24 hour ⁄ IP.
* white listed IP for registrars : 86,400 requests⁄ 24 hour ⁄IP.

------------------------
6 - Reverse lookups

The web RDDS access offers advanced searchability capacities.
The following functions are available :

= Direct queries =

* Partial match query on domain name, administrative, technical, and billing contact name and address, registrant name and address, registrar name including all the sub-fields described in EPP (e.g., street, city, state or province, etc.).
* Exact match query on registrar id, name server name, and name server’s IP glue records
The result of direct queries is the object queried (contact, domain, ...)

= Reverse queries =

* Partial match query on domain name, administrative, technical, and billing contact name and address, registrant name and address, registrar name including all the sub-fields described in EPP (e.g., street, city, state or province, etc.).
* Exact match query on registrar id, name server name, and name server’s IP glue records including IPv6 queries.
The result of reverse queries is the list of objects of a given type linked with the result object (list of domains with a given contact result, or name server result,...)

This powerful tool is limited in access :
* Captcha system avoids scripting of the interface.
* Direct queries are open to every user but the number of result objects is limited to 1,000 answers for 1 query.
* Reverse queries can only be done by registrars on the extranet interface, and the number of result objects is limited to 10,000 answers for 1 query. The interface cannot be used more than 100 times a day.

------------------------
7 - Interconnectivity and synchronization with other systems

= SRS =

Data updated by the SRS are immediately visible in the RDDS with no further synchronisation needed. Rate limitation is applied both on SRS and RDDS service to avoid any load on the database. SRS and RDDS are partly in the same network zone, both RDDS servers and EPP SSL reverse proxies being in the public network zone (DMZ).

= Main database =

Hot standby of the production database is automatically taken into account by the RDDS Oracle Transparent Network Substrate configuration. Therefore if database are migrated in hot standby due to failure of part of the system, the RDDS service is automatically swapped to the new architecture.

= Rate limiting database =

No standby is implemented on the rate-limiting database. In case of failure, a standard global limitation is applied while, replacement of the database is operated.

= Monitoring =

Monitoring is operated through probes and agents scanning systems with a 5 minutes period. The monitoring system gets snmp data from all servers described in the RDDS architecture and also from dedicated Oracle monitoring agent for the database.
Hot standby is not implemented on monitoring agents.

------------------------
8 - Performance and scalability

The Registry’s RDDS offers high level production SLAs and derives from the branch of systems that have evolved over the last 12 years to successfully operate a set of french ccTLDs.

The Registry’s RDDS is used to publish .fr, .re, .yt, .pm, .tf, .wf TLDs information. It is used by more than 800 registrars in parallel managing more than 2 millions domain names and by a large user community.

AFNIC’s RDDS is designed to meet ICANN’s Service-level requirements as specified in Specification 10 (SLA Matrix) attached to the Registry Agreement.

As described in Question 31 (Technical Overview) in relation to each of the phases of the TLD’s operations, the following transaction loads are expected on the WHOIS servers : 6 requests⁄hour on average for both launch phase and on going operations.

AFNIC’s WHOIS systems can serve up to 10,000 requests⁄min on load balanced service to be compatible with the launch and growth scenario described in Question 31 (Technical Overview).

The targeted TLD objective being around 1,000 domain names with a provision for 6 requests⁄hour on average, this ensures that enough capacity is available to handle the launching period, as well as demand peaks and unexpected overhead.

Capacity planning indicators are set up to anticipate exceptional growth of the TLD.
Technologies used enables quick upgrade on all fields :
* Servers : virtual resizing to add CPUs or disk space if resource is available on the production ESX servers. If not, 2 spare additional ESX servers can be brought live if additional performance is needed.
* Servers (alternate) : additional servers can be added and taken into account immediately through dns round robin algorithm.
* Database : database capacity has been greatly oversized to avoid need of replacement of this physical powerful server. Precise capacity planning will ensure that sufficient delay will be available to acquire new server if needed. A threshold of 40% of CPU use or total storage capacity triggers alert for acquisition.

------------------------
9 - ICANN Bulk access compliance

The Registry Operator will provide both data escrow and ICANN bulk access in a same process.
Data escrow generates data on a daily basis. One file per week is kept for ICANN access to bulk data.

------------------------
10 - RFC compliance

The system has been launched compliant with RFCs. Mechanisms are in place to ensure that on going maintenance and new functional delivery stay compliant with RFCs.

= Delivery process =

The RDDS evolutions are developed on the development environment.
The development process implies strict coding rules and use of shared best practices. Pair programming is standard practice. Unit test are developed prior to function development to ensure resiliency of the produced code.

Delivery process take place in four steps :
* 1st step : RDDS validation and RFC compliance is checked through automated tools. A 100% compliance signal must be received to be able to proceed to second step.
* 2nd step : delivery to the pre-production environment. The development is delivered on the preproduction environment. This environment is available for internal testing team.
* 3rd step : delivery to the sandbox environment. This sandbox environment is opened for registrar where they have two accounts to validate their clients before production activation.
* 4th step : the new release is delivered in production.

= Format validation =

RDDS rfc compliance is reached through a specific RDDS checker which is use for non-regression test before each new release.

= Cross checking =

Whois cross checking partnership is established with .at Registry operator to validate in sandbox environment prior to delivery in production through mutual agreement.

= Whois Output =

* Output of a whois query is 100% similar to the whois query example available in RFC 3912.

------------------------
11 - Resources

------------------------
11.1 - Initial implementation

The initial implementation effort is estimated as follows :

Database Administrator 0.03 man.day
Network Administrator 0.03 man.day
System Administrator 0.03 man.day
Software Developer 0.10 man.day
Software Engineer 0.05 man.day

------------------------
11.2 - On-going maintenance

On-going maintenance on the RDDS module includes mainly integration of new policy rules, privacy law evolutions, evolution of contracts, infrastructure evolution, failover testing.

The on-going maintenance effort per year is estimated as follows, on a yearly basis :

Operations Specialist 0.15 man.day
System Administrator 0.05 man.day
Software Developer 0.05 man.day
Software Engineer 0.10 man.day

Table of Contents

1 - Global description
2 - Data associated with a domain name
2.1 - Technical data
2.2 - Administrative data
3 - Full domain name lifecycle overview
4 - Basic create⁄update⁄delete life cycle
4.1 - create
4.2 - update
4.2.1 - technical update
4.2.2 - administrative update
4.2.3 - context update
4.3 - delete⁄restore
5 - Transfer
6 - Renewal and auto-renewal
7 - Grace period and refund
8 - Resources allocated
8.1 - Initial implementation
8.2 - On-going maintenance

------------------------
1 - Global description

Domain names represents the core technical part of the Domain Name System. The lifecycle of a domain name can have both technical impacts, when it relates to technical data associated with the domain name, and administrative impact when related to the registrant of the domain name.

The following diagrams and descriptions will bring detailed answers to the question of the lifecycle of the domain name in regards to both these aspects

------------------------
2 - Data associated with a domain name

To clearly understand the lifecycle of the domain name, we must first give an exhaustive description of the data involved in the various operations to be made.

------------------------
2.1 - Technical data

A domain name is a technical label used for Domain name resolution. To be effective, it is associated with nameservers -server hosting the configuration of the domain name -, IPv4 and IPv6 addresses - to identify on the network servers independently of the DNS, DNSsec signature information - delegation signer and cryptographic algorithm used-.
Less directly related to the technical basic configuration are :
* = clientHold = label : relates to the DNS or not DNS-publication status of the domain name.
* = auth_info = : a protection code linked with the domain and used by the owner to unlock some operations
* = client*Prohibited = : a list of status activated by the registrar to lock the domain name and prevent some operations
* = server*Prohibited = : a list of status activated by the registry service provider to lock the domain name and prevent some operations

------------------------
2.2 - Administrative data

A domain name has to be managed by his owner. Therefore it comes associated with a list of operational and administrative contacts that can be used to get in relation with the domain name owner or technical staff. The most important are administrative contact, technical contact, billing contact, and of course registrant contact. The last contact object is the registrar object which shows which registrar is in charge of domain name operations at the registry level.

Both these administrative and technical data are modified and used in the lifecycle and we will now describe this in detail.

------------------------
3 - Full domain name lifecycle overview

We have chosen to illustrate the registration lifecycle through a state diagram
This state diagram is joined as a separate file.

[see attached diagram Q27_3_global_lifecycle.pdf]
Diagram : Global Lifecycle
Description : Considering the wide range of the states and transition, the choice has been made to present a linear scenario going through all available operations. In this scenario, impact on registrar objects, registrant objects, domain objects, host objects are described at each step. Also statuses and forbidden operations at each step are indicated.
The following domain states have been introduced to describe the lifecycle major steps :
* registered : the domain name is registered, published in the Registration Data Directory Services (RDDS) but not in the DNS (clientHold label is set or there is no host information)
* active : the domain name is registered, published in the RDDS and in the DNS
redemption : the domain name is registered, published in the RDDS but not in the DNS. It will be - deleted if no action is taken by the registrar.
* locked : specific operations as transfer or delete have been forbidden by the registrar.
Impact on expiry dates has also been indicated though adequate formulas.

All aspects of the registration lifecycle are covered by standard Extensible Provisioning Protocol (EPP) RFCs and the EPP implementation is described in Question 25 (EPP).

------------------------
4 - Basic create⁄update⁄delete life cycle

The basic life cycle is described below without explanation of add grace period. The behavior of add grace period is described in chapter 7.

------------------------
4.1 - create

A domain name is created through a standard EPP domain:create command.
Administrative data linked with the creation are registrant contact, admin contact and technical contact, period before renewal.
Technical data linked with the creation are nameservers host objects, IP address for glue records, auth_info code.
The state of the domain name is REGISTERED if no host objects have been filled.
The state of the domain name is ACTIVE if host objects have been filled.
The state of the domain name can exceptionally be PENDING during the operation if a technical issue makes it asynchronous.
Otherwise this operation is real time and there is no delay elements to be considered.

Elements needed to create a domain are contacts (mandatory), host objects (optional) and auth_code (mandatory).
It can then be managed through domain:update commands.

------------------------
4.2 - update

domain:update commands enables a wide range of fields updates

------------------------
4.2.1 - technical update

Part of the fields of the update enables to update technical configuration. It enables nameserver, IP address, and dnssec options management. It is also used to remove a technical configuration..

The state of the domain name is REGISTERED if no host objects have been filled or have been removed.
The state of the domain name is ACTIVE if host objects have been filled.
The state of the domain name can exceptionally be PENDING during the operation if a technical issue makes it asynchronous.

------------------------
4.2.2 - administrative update

It is used to freely modify the various contacts linked with the domain name : administrative, technical, billing, and registrant contact.
The state of the domain name is not modified if only these fields are used.
The state of the domain name can exceptionally be PENDING during the operation if a technical issue makes it asynchronous.

------------------------
4.2.3 - context update

It is used by the client to modify status of the domain name and⁄or to modify the auth-info code linked with the domain name.
The status that can be changed are the following : clientHold, clientTransferProhibited, clientUpdateProhibited, clientDeleteProhibited, clientRenewProhibited.
The clientHold flag enables to remove the domain name from publication temporarily without deleting its technical configuration.
The other client*Prohibited statuses prevent the corresponding operation to be used.
The state of the domain name is REGISTERED if status is updated to clientHOLD.
The state of the domain name is LOCKED if status is updated to clientTransferProhibited.
The state of the domain name can exceptionally be PENDING during the operation if a technical issue makes it asynchronous.

------------------------
4.3 - delete⁄restore

Deletion can be used only by the registrar in charge of the domain name. It brings the domain name in Redemption grace period for a period of 30 days. It can be restored at any time during this period without any changes to the data. Deletion remove the domain name from the DNS publication service.
The state of the domain name is DELETED during redemption period.
The redemption period lasts 30 days. The domain is destroyed at the end of this period and a notification is sent.

------------------------
5 - Transfer

The transfer is described below without explanation of transfer grace period. The behavior of transfer grace period is described in chapter 7.

A transfer can be initiated only by an incoming registrar and using the auth_info that the owner has given him. This standard mechanism acts as a security and associates the triggering of transfer to the acceptance of the owner of the domain.
The transfer operation can be triggered only if the domain is not protected by a clientTransferProhibited lock.

[see attached diagram Q27_5_transfer_lifecycle.pdf]
Diagram : Transfer lifecycle
Description : Transfer operation includes various steps with impact on both outgoing and incoming registrars.

The outgoing registrar receive a transfer notification and can technically accept or reject the registrar change. Rejection can only be done in specific cases described in ICANN consensus policies.
If the outgoing registrar accepts the transfer, the operation is accepted immediately.
If the outgoing registrar does not validate the transfer, the operation is automatically accepted after 5 days.
If the outgoing registrar rejects the transfer, the operation is automatically cancelled and both registrars are notified of the rejection.
When the transfer succeeds, both registrars are notified through their EPP notification queue.

A reverse transfer can be asked by the losing registrar. The documents and cases where this cancellation of the transfer can be asked follow ICANN consensus policies on transfers. In case of disputes, the ICANN TDRP (Registrar Transfer Dispute Resolution Policy) is followed.

The state of the domain name is PENDING during the operation.

------------------------
6 - Renewal and auto-renewal

Domain:renew command is used by the registrar to increase the period of registration. If a domain name is registered for less than 10 years it can be renewed for a period up to 10 years at any time. The expiry date is updated.
The domain:renew command can be sent at any phase of the lifecycle (exception of add grace period is described in next chapter).

The registry lifecycle works with auto-renewal mechanisms. If a registrar do not renew or delete the name when it reaches the expiration date, a one year auto-renew period is added. As for other commands, a grace period is linked with this action (see chapter 7)

[see attached diagram Q27_6_grace_period_renew_autorenew_lifecycle.pdf]
Diagram : Grace Period renew⁄autorenew lifecycle
Description : This renew⁄autorenew lifecycle sum up impact of operations on domain name availability and statuses.

------------------------
7 - Grace period and refund

= Grace period =

The Grace Period mechanism refers to a specified period following an operation or change of status in which the operation may be reversed and a credit may be issued to the Registrar.

= Redemption Grace Period =

The Redemption Grace Period has been described in the delete⁄restore chapter.
During this period, domain name is still registered and can be reactivated through domain:restore command. No specific refund is linked with this period.

= Create - Add Grace Period (AGP) =

The implemented AGP is a five-day period following the domain:create command of a domain name.
The Registrar may delete the domain name at any time during this period and receive a full credit for the registration fee from the Operator. Once a domain name is deleted by the registry at this stage, it is immediately available for registration by any registrant through any Registrar.

= Auto-renew Grace Period =

The auto-renew add grace period is implemented. If during this 45 days period the domain is deleted by the incoming registrar, the ʹdomain:renewʹ command is refunded.

= Renew Grace Period =

The renew grace period is implemented. If during the 5 days period following explicit renew bye the registrar, the domain name is deleted, the renew is then refunded.

= Transfer Grace Period =

The transfer grace period is implemented. If during the 5 days period following a transfer the domain is deleted, the transfer is then refunded.

= AGP Limits Policy =

If too many deletions take place during the AGP from a given registrars, a financial penalty is applied.
The Add Grace Period Limits Policy allows a registrarʹs account to be debited each month for all AGP deletions that exceed the greater of either:
* 50 domain names, or
* 10% of net new adds for the previous month

------------------------
8 - Resources allocated

Four categories of profiles are needed to run the Registry’s Technical Operations : Registry Operations Specialists (I), Registry Systems Administrators (II), Registry Software Developer (III) and Registry Expert Engineers (IV). These categories, skillset and global availability of resources have been detailed in Question 31 (Technical Overview of Proposed Registry). Specific workload for this question is detailed below.

------------------------
8.1 - Initial implementation

The set up of a precise lifecycle implies actions by developers, assisted by a senior staff member expert in internet technologies and RFCs to apply proper configuration for the given TLD. Compliance is strictly tested.

The initial implementation effort is estimated as follows :

Software Developer 0.20 man.day
Software Engineer 0.20 man.day

------------------------
8.2 - On-going maintenance

On-going maintenance on the lifecycle includes mainly integration of new policy rules.
The on-going maintenance effort per year is estimated as follows, on a yearly basis :

Software Developer 0.20 man.day
Software Engineer 0.20 man.day

Table of contents:

1. Application overview
2. Single point of contact
3. Anti-abuse policy
3.1. Definitions
3.2. Policies and procedures
3.3. Policies for handling complaints regarding abuse
3.4. Orphan glue records
3.5. Controls for ensuring the proper access to domain functions
4. Whois Abuse Prevention Policies
4.1. Measures for promoting Whois accuracy
4.2. Protection against unfair use of Whois service
4.2.1 Protection against Data Mining
4.2.2 Prevention of unauthorized data modification
5. Resources for policy implementation

1. Application overview

The registration policy for the .frogans gTLD registry conforms to its following characteristics:

- The OP3FT (the Applicant) will be the sole and unique holder of domain names in the .frogans gTLD registry;

- The OP3FT will not commercialize any domain names in the .frogans gTLD registry to third parties;

- The maximum number of domain names in the .frogans gTLD registry will be 1,000.

The OP3FT will use domain names in the .frogans gTLD registry for addressing the computers that are dedicated to the functioning of the Frogans layer on the Internet. Typically, the domain names in the .frogans gTLD registry will be composed of informative terms such as “service1.frogans” for example.

To the extent that the OP3FT is both the Registry Operator of the .frogans gTLD and the unique registrant of domain names in the .frogans gTLD registry, there will be neither abusive nor malicious domain name registrations in the .frogans gTLD registry, nor will there be orphaned registrations or erroneous or abusive data in the Whois.

The OP3FTʹs objective in answering question 28 is to provide a thorough explanation of its policies and procedures to minimize abusive registrations and other activities that have a negative impact on Internet users.

2. Single point of contact

For that which concerns an implementation plan to establish and publish on its Web site a single abuse point of contact responsible for addressing matters requiring expedited attention and providing a timely response to abuse complaints concerning all names registered in the TLD through all registrars of record, including those involving a reseller, the OP3FT will publish on its Web site (www.op3ft.org) the contact information of the person in charge of receiving all questions and requests concerning domain name registrations in the .frogans gTLD registry.

This single point of contact will comprise a Web interface offering the possibility for Internet users to report any abuse (phishing, spamming, trademark abuse etc.) concerning a name registered in the gTLD .frogans registry. This contact Web interface will be available on the OP3FT’s Web site.

This single point of contact will enable a quicker and better management of complaints. Complaints will be addressed by filling out a form through online services on the OP3FTʹs Web site.

A dedicated team will be in charge of handling these complaints in a due time. All requests should be acknowledged and processed within 24 hours. According to the nature of the reported abuse (phishing, spamming, trademark abuse, etc), an appropriate action will be taken by the OP3FT.

Internet users will be given access to all necessary information regarding remedies to abusive registrations on the OP3FTʹs single point of contact Web page. This Web page will also contain links to all relevant organizations addressing these issues.

3. Anti-abuse policy

By implementing the anti-abuse policy described below, the OP3FT will also contribute to and protect the integrity, security and stability of the DNS.

3.1. Definitions

The OP3FT will use the following definitions of what constitutes abuse in the .frogans gTLD registry.

From ICANN (online presentation of Registration Abuse Policy, available at http:⁄⁄www.icann.org⁄en⁄resources⁄policy⁄background⁄rap):

- registration abuse: “In general, the term covers a broad variety of illegal or illegitimate behaviors considered contrary to the intent and design of normal domain registration processes. Registration abuse often involves malicious actors trying to register in ways that avoid lawful authorities or conceal a registrantʹs identity. Registration abuse can also enable other kinds of abuses, such as phishing and spam.”

From the Public Interest Registry (Domain Name Anti-Abuse Policy, available at http:⁄⁄www.pir.org⁄why⁄anti_abuse_policy):

- Spam: The use of electronic messaging systems to send unsolicited bulk messages. The term applies to e-mail spam and similar abuses such as instant messaging spam, mobile messaging spam, and the spamming of Web sites and Internet forums. An example, for purposes of illustration, would be the use of email in denial-of-service attacks.

- Phishing: The use of counterfeit Web pages that are designed to trick recipients into divulging sensitive data such as usernames, passwords, or financial data.

- Pharming: The redirecting of unknowing users to fraudulent sites or services, typically through DNS hijacking or poisoning.

- Willful distribution of malware: The dissemination of software designed to infiltrate or damage a computer system without the ownerʹs informed consent. Examples include, without limitation, computer viruses, worms, keyloggers, and trojan horses.

- Fast flux hosting: Use of fast-flux techniques to disguise the location of Web sites or other Internet services, or to avoid detection and mitigation efforts, or to host illegal activities. Fast-flux techniques use DNS to frequently change the location on the Internet to which the domain name of an Internet host or name server resolves. Fast flux hosting may be used only with prior permission of PIR.

- Botnet command and control: Services run on a domain name that are used to control a collection of compromised computers or ʺzombies,ʺ or to direct denial-of-service attacks (DDoS attacks).

- Distribution of child pornography.

- Illegal Access to Other Computers or Networks: Illegally accessing computers, accounts, or networks belonging to another party, or attempting to penetrate security measures of another individualʹs system (often known as ʺhackingʺ). Also, any activity that might be used as a precursor to an attempted system penetration (e.g., port scan, stealth scan, or other information gathering activity).

- Unlawful content or any content that contravene public order according to French law and in particular to Law on the Freedom of the Press of 29 July 1881 (crimes against humanity apology⁄promotion or contestation, incitement to discrimination, hatred or violence).

The OP3FT is committed to creating and implementing policies and procedures that prevent abusive registrations and other activities that have a negative impact on Internet users.

Conforming to the industry best practices presented in the Registration Abuse Policies Issues Report (ICANN 2008), the OP3FT will offer a wide range of effective safeguards to prevent abusive uses of registry and domain names, keeping in mind that the anti-abuse policy will be an internal procedure that the OP3FT pledges to follow as being the sole and unique registrant of all domain names in the .frogans gTLD registry.

3.2. Policies and procedures

This section provides a description of policies and procedures that define malicious or abusive behavior, capture metrics, and establish Service Level Requirements for resolution, including service levels for responding to law enforcement requests.

Although, there will be no possibility for the abusive or malicious registration of domain names in the .frogans gTLD registry, the OP3FT will take internal measures to ensure that registrations in the .frogans gTLD registry will be carried out by authorized and qualified personnel who will respect a procedure for preventing any abusive registration, as defined in section 3.1 above.

The OP3FT will develop strict internal rules for allowing only the registration of domain names in the .frogans gTLD registry that respect third party rights and the criteria set by ICANN, notably those in Module 3 of the Application Guidebook (AGB) and in the Registration Abuse Policies Issues Report (available at http:⁄⁄www.icann.org⁄en⁄resources⁄policy⁄background⁄rap) and according to the Domain Name Anti-Abuse Policy of the Public Interest Registry (PIR).

The OP3FT will carefully implement internal rules preventing various misconducts and abusive registration behaviors such as but not limited to using a name that:

- Appears on ICANNʹs list of reserved names;

- Corresponds to a protected geographic name;

- Corresponds to a trademark or other intellectual property for which the OP3FT has neither rights nor legitimate interests;

- Corresponds to an existing TLD;

- Corresponds to a Non-Governmental Organization (OGN);

- Is contrary to general principles of international law for morality and public order as defined in article 3.5.3 of the AGB;

- Is contrary to the criteria developed in the Uniform Domain-Name Dispute-Resolution Policy (UDRP) (each registration will be notably subject to a study for conformity regarding the criteria detailed in paragraphs 4(a), 4(b) and 4(c) of the UDRP so as to determine if there will be a risk of abuse);

- Is contrary to the criteria developed in the Registration Abuse Policies Issues Report; each registration will be subject to a study for conformity with this report and to the findings of ICANN working groups (GNSO) on this subject; each registration will be subject to the RAP (Registration Abuse Policy) upon that policyʹs finalization.

With respect to the last rule in the above list, the OP3FT will not:

- Use the .frogans gTLD, its registry or any related tool to violate laws or regulation, register deceptive and⁄or offensive domain names,

- Use the .frogans gTLD, its registry or any related tool to trick people into believing they are connected to a legitimate Web site (i.e. phishing),

- Edit, host or harbor pure gripe sites,

- Use the .frogans gTLD, its registry or any related tool to register domain names for which the OP3FT has neither rights nor legitimate interests, reproducing trademarks and brands or any other illicit behavior in regard of third parties rights, including intellectual property rights,

- Use computers or any other automated means to generate continuous or repeated requests to register domain names or use a Whois service,

- Use computers or any other automated means to slow down or alter online services, including DNS lookups.

Being both the registry operator and the sole and unique registrant of domain names in the .frogans gTLD registry, the OP3FT will have full hand on registrar activities and will formally prevent bad practices, such as name spinning, domain tasting, front running and other scams.

As such, OP3FT is dedicated to enforcing the respect of third parties rights, network and Internet tool integrity and is committed to make use of the .frogans gTLD in compliance with ICANN policies.

Concerning the particular abuses below:

- Cache poisoning:

One of the main authentication issue encountered on the DNS is the cache poisoning issue. This directly affect data flow at the DNS service level without having to corrupt or modify data in the registry database.

The answer to this issue is the implementation and deployment of DNSSEC. The registry back-end service provider of the OP3FT, AFNIC (Association Française pour le Nommage Internet en Coopération), already manages DNSSEC-enabled zones successfully: on September 29, 2010, AFNIC, finished adding its 6 ccTLDs key materials (DS records) into the IANA root zone, ending with .FR after extensive tests with its other TLDs. Since then, related DNSSEC operations and monitoring are spread inside the organization, alongside all other standard day to day operations, so that DNSSEC is a core service enabled by default.

- Domain name Sniping (grabbing):

Domain name sniping refers to the practice of trying to re-register potentially interesting domain names immediately after they are deleted.

The OP3FT being the sole and unique registrant of domain names in the .frogans gTLD registry, this type of abuse will not occur.

- Domain name tasting:

Domain name tasting is a practice used by registrants to check if a domain name is of interest or not, exploiting the 5-days Add Grace Period (AGP) during which a newly created domain name may be deleted with a refund of the domain fee. Domain name tasting is dealt with through the implementation of AGP.

However, as the OP3FT is the sole and unique registrant of domain names in the .frogans gTLD registry, this type of abuse will not occur.

3.3. Policies for handling complaints regarding abuse

The OP3FT will process any complaint with the greatest care and according to the ICANNʹs criteria, notably to delete any domain name registration which has been shown to represent an abusive registration, for example, through the reception of a detailed notice showing evidence of an infringement of rights through the registration of a domain name in the .frogans gTLD registry.

The OP3FT will implement this policy despite its conviction that such abuses will not occur in the .frogans gTLD registry given that:

- The OP3FT will be the sole and unique holder of domain names in the .frogans gTLD registry;

- The OP3FT will not commercialize any domain names in the .frogans gTLD registry to third parties.

In the execution of this policy, the OP3FT will process each complaint within 24 hours and will take all the necessary steps to offer a satisfactory response to the complainants.

Should immediate action be taken by competent authorities, the OP3FT is committed to comply with their demands without delay. The OP3FT will work closely with these authorities. This may concern the following cases (but not limited to):

- Court orders;

- Inquiries from law enforcement bodies (e.g, OCLCTIC - The Office central de lutte contre la criminalité liée aux technologies de lʹinformation et de la communication, which is the French Police unit specialized in cybercrime);

- Anti-phishing groups (e.g, CERTs).

3.4. Orphan glue records

According to the definition found in the ʺSSAC Comment on the Orphan Glue Records in the Draft Applicant Guidebook”, a glue record becomes an ʺorphanʺ when the delegation point NS record (the ʺparent NS recordʺ) that it references is removed while retaining the glue record itself in the zone. Consequently, the glue record becomes ʺorphanedʺ since it no longer has a parent NS record. In such a situation, registrars and registrants usually lose administrative control over the record, and the recordʹs attribution to a certain registrar may become unclear, which makes it a potential vector for abuse.

Concerning the question of proposed measures for removal of orphan glue records for names removed from the zone, when provided with evidence in written form that the glue is present in connection with malicious conduct, such a situation shall not occur for domain names in the .frogans gTLD registry given that:

- The OP3FT is both the registry operator and the sole and unique registrant of domain names in the .frogans gTLD registry; and

- The OP3FT will have full hand on registrar activities.

In any event, the glue record policy in effect for the .frogans TLD record avoids this situation entirely by disallowing orphan glue records altogether. This corresponds to policy #3 mentioned in section 4.3 (page 6) of the SSAC document mentioned above. The technical implementation within the Registry and its associated zone generation process ensures this by the following measures:

- Any host object which is a glue record can be created only if the domain name exist and is sponsored by the registrar creating the host.

- Any deletion of a domain name which have subordinate hosts can be done only when these hosts are deleted. If these hosts are used in delegations for other domain names in the .frogans gTLD registry, these delegations have to be removed to delete the host objects and then the domain name.

- If the sponsored registrar of the domain name cannot remove these delegations (e.g. explicit refusal or inactivity from subordinate hosts registrars), the sponsored registrar may use a specific procedure provided by the Registry Operator (OP3FT) in which the Registry Operator (OP3FT) will contact the registrar(s) used in delegating the host object(s) and will ask them to remove the delegation(s). The registrar(s) will then have 10 days to remove the delegation(s), after which time, if there is no removal of the delegation(s), the Registry Operator (OP3FT) will deactivate directly the DNS configuration of the domain name(s) concerned. At the end of the procedure, the Registry Operator (OP3FT) will contact the sponsored registrar to inform him that he can delete the host object(s) and the domain name.

3.5. Controls for ensuring the proper access to domain functions

The OP3FT will use the .frogans gTLD with the aim of ensuring the security, stability and reliability of the Frogans layer for the benefit of all Internet users.

Even though the OP3FT will be both the registry operator and the sole and unique registrant, it will implement very strict controls to ensure proper access to domain functions for the .frogans gTLD given that, according to ICANN requirements, any ICANN-accredited registrar may gain access to these domain functions. Controlling this access thus represents a critical issue for achieving the goal of the .frogans gTLD.

These controls will include multi-factor authentication at the registrar and registry operator levels (such as strong passwords, tokens or one-time passwords) from the registrant (the OP3FT) to process registration, update, transfer, and deletion requests.

4. Whois Abuse Prevention Policies

4.1. Measures for promoting Whois accuracy

RFC3912 specifies the Whois protocol and explains it as follows: ʺWhois is a TCP-based transaction-oriented query⁄response protocol that is widely used to provide information services to Internet users. While originally used to provide ʹwhite pagesʹ services and information about registered domain names, current deployments cover a much broader range of information services. The protocol delivers its content in a human-readable format.ʺ

The OP3FT understands that the availability of accurate information about registered domain names is a sensitive subject. This information shall contain registrant contact information, including administrative, technical and billing contact details.

In case of malicious or abusive activity, the Whois contact is usually the first and most important source of information. Whois accuracy is therefore a major step to counter malicious conduct. This information may be required by law-enforcement authorities to identify individuals and organizations responsible for domain names.

As being both the registry operator and the sole and unique registrant of domain names in the .frogans gTLD registry, the OP3FT pledges to keep the Whois for the .frogans gTLD registry correct, precise and complete. Moreover, a verification and updating process will be undertaken every year for ensuring the pertinence of the information posted in the Whois.

4.2. Protection against unfair use of Whois service

Whois Service gives access to sensitive data, including contact details of registrants. The OP3FT, as the .frogans gTLD registry operator, is committed to insure the protection of these data against abusive behaviors. Firstly, the OP3FT will implement technical measures to prevent data mining on the Whois, such as automated collection of registrants’ email addresses, that may result in spamming. Secondly, the OP3FT and its registry back-end service provider, AFNIC, will deploy all necessary means to secure access to its database, specifically by implementing procedures in order to prevent Unauthorized Data Modifications. These procedures will reinforce the security of both EPP and Web-based access to Whois data.

4.2.1 Protection against Data Mining

Users of the .frogans gTLD Whois database will commit to using the published data according to the laws and regulations in effect. Moreover, users shall respect the provisions of the French Data Protection Act. Violation of this act carries criminal penalties.

As users will access personal data, they must refrain from any collection, misuse or any act that could lead to the invasion of privacy or damaging the reputation of individuals.

The OP3FT can at any time filter the access to its Whois database services in case of suspicion of malevolent use.

Additionally, the OP3FT will implement the following methods:

- Captcha: users shall pass a Captcha before access is granted to the Web-based Registration Data Directory Services (RDDS).

- Rate-limiting: The registry has chosen limitation measures for the number of requests in order to prevent abuse in the use of personal data and to guarantee the quality of the service. By a transparent parameter adjustment policy, the registry guarantees quality of service to the punctual users and professionals. The rates and thresholds of this system are described in the registry use case of question 26.

- White list: The white list mechanism offers specific access for registrars to the port 43 whois considering that the incoming traffic must come from two pre-defined IP address. This white list access offers higher thresholds of rate limiting for the users.

4.2.2 Prevention of unauthorized data modification

Data modification is managed through strict authentication and access policies.

- SSL⁄TLS protocol is used on all interfaces with clients (both EPP and Web-based SRS).

- a password policy is applied both on the password itself (minimum length, mandatory digits and non-alphanumerical characters), and on the live length of the password;

- use of an SSL client certificate pre-installed by the registry for EPP access;

- IP authentication limited to two addresses.

The OP3FTʹs registry back-end service provider, AFNIC, will share its experience in the .fr ccTLD with the aim of ensuring effective, timely and sufficient Domain Data Access Control.

Further details on this subject can be found in the response to question 26.

5. Resources for policy implementation

For that which concerns the resources committed to the deployment and application of these policies, the OP3FT commits to assign, on a part-time basis, a specifically trained engineer who will be responsible for the registration of domain names in the .frogans gTLD registry. This engineer will be assisted by the OP3FTʹs legal team for ensuring that domain name registrations will be carried out in compliance with the applicable OP3FT policies, which themselves will remain compliant with ICANNʹs gTLD policies, and in order to process any complaints.

As to the technical processing system and policy implementation, these are provided by AFNIC in its capacity as technical registry back-end service provider for the .frogans gTLD registry. Their costs are included in the fixed fee agreed upon in the technical registry back-end service provider agreement between the OP3FT and AFNIC, signed on April 3, 2012, further detailed in the answer to question 47(a).

These technical functions for the protection of third-party rights and for the regulatory conformity of the .frogans gTLD are thus ensured without interruption and without discrimination.

Table of contents:

1. Application overview
2. Rights Protection Mechanism Description
2.1. Uniform Domain Name Dispute Resolution Policy and Uniform Rapid Suspension System
2.2. Trademark Claims service
2.3. Sunrise service
3. Resources

1. Application overview

The registration policy for the .frogans gTLD registry conforms to its following characteristics:

- The OP3FT (the Applicant) will be the sole and unique holder of domain names in the .frogans gTLD registry;

- The OP3FT will not commercialize any domain names in the .frogans gTLD registry to third parties;

- The maximum number of domain names in the .frogans gTLD registry will be 1,000.

To the extent that the OP3FT is both the Registry Operator of the .frogans gTLD and the unique registrant of domain names in the .frogans gTLD registry, there will be neither abusive nor malicious domain name registrations in the .frogans gTLD registry, nor will there be infringements upon the intellectual property rights of third parties.

The OP3FTʹs objective in answering question 29 is to provide a thorough explanation of its policies and procedures to minimize abusive registrations and other activities that have a negative impact on Internet users.

2. Rights Protection Mechanism Description

2.1. Uniform Domain Name Dispute Resolution Policy and Uniform Rapid Suspension System

The OP3FT undertakes to fully implement the provisions of the Uniform Domain Name Dispute Resolution Policy (UDRP - available at www.icann.org⁄en⁄help⁄dndr⁄udrp) as well as the Uniform Rapid Suspension System (URS - available in the Application Guidebook dated January 11, 2012, pages 310 to 320 at http:⁄⁄newgtlds.icann.org⁄en⁄applicants⁄agb) for ensuring that the .frogans gTLD complies to the same rules as those in effect for other gTLDs. This full implementation ensures the protection of rights holders.

As the .frogans gTLD registry operator, the OP3FT will use an internal policy for the registration of domain names in the .frogans gTLD registry that includes the provisions of the UDRP and the URS: each registration will be notably subject to a study for conformity regarding the criteria detailed in paragraphs 4(a), 4(b) and 4(c) of the UDRP and 1.2.6 of the URS.

As the registrant of domain names in the .frogans gTLD registry, the OP3FT will also sign the applicable agreements (e.g. the UDRP) with the registrar(s) of the .frogans gTLD.

AFNIC (Association Française pour le Nommage Internet en Coopération), which provides the registry back-end services to the OP3FT, will apply the UDRP and provide assistance, if necessary, for carrying out decisions rendered by ICANN-approved dispute resolution service providers (UDRP and URS).

2.2. Trademark Claims service

Concerning the service for processing complaints related through rights protection mechanisms (RPMs), the OP3FT will process any complaint with the greatest care and according to the ICANNʹs criteria, notably to cancel any domain name registration which has been shown to represent an abusive registration, for example, through the reception of a detailed notice showing evidence of an infringement of rights through the registration of a domain name in the .frogans gTLD registry.

The OP3FT commits to assign, on a part-time basis, a specifically trained engineer who will be responsible for the registration of domain names in the .frogans gTLD registry. This engineer will be assisted by the OP3FTʹs legal team for ensuring that domain name registrations will be carried out in compliance with the applicable OP3FT policies and in order to process any complaints.

However, such complaints will not occur for domain names registered in the .frogans gTLD registry, given that:

- The OP3FT will be the sole and unique holder of domain names in the .frogans gTLD registry;

- The OP3FT will not commercialize any domain names in the .frogans gTLD registry to third parties.

For further information, please refer to the answer to question 28, section 3.3.

2.3. Sunrise service

ICANN requires that the registry operator offer Sunrise registration services for a minimum of thirty (30) days during the pre-launch phase and notice must be provided to all trademark holders in the Trademark Clearinghouse (TMCH) if someone is seeking a sunrise registration.

Given the following characteristics of the .frogans gTLD, being that

- the OP3FT will be the sole and unique holder of domain names in the .frogans gTLD registry,

- the OP3FT will not commercialize any domain names in the .frogans gTLD registry to third parties and

- the maximum number of domain names in the .frogans gTLD registry will be 1,000,

the organization of a Sunrise registration service does not appear to the OP3FT to provide the Internet community with tangible value. However, the OP3FT will organize the appropriate Sunrise registration services if ICANN so requires in the final Registry Agreement.

3. Resources

The OP3FT will assign, on a part-time basis, a specifically trained engineer who will be responsible for the registration of domain names in the .frogans gTLD registry. This engineer will be assisted by the OP3FTʹs legal team for ensuring that domain name registrations will be carried out in compliance with the applicable OP3FT policies and in order to process any complaints.

Additionally, a technical system for processing the domain name registration policy in the .frogans gTLD registry will be implemented by AFNIC as the provider of registry back-end services to the OP3FT.

All resources necessary for the development, deployment and application of an internal domain name registration policy for respecting third-party rights, conforming to ICANN policies (the UDRP and the URS, in particular) have been integrated into the OP3FT budget.

Further details concerning the costs of these resources are provided in the answer to question 47(a).

These technical functions for the protection of third-party rights and for the regulatory conformity of the .frogans gTLD are thus ensured without interruption and without discrimination.

Table of Contents

1 - Background
2 - Organization of security
2.1 - The place of Security in AFNIC’s processes:
2.2 - Security Coordination
2.3 - Assignment of responsibilities
2.3.1 - Organizational chain of responsibility
2.3.2 - Relations with the authorities and groups of specialists
2.4 - Independent security review
2.5 - Relations with third parties
2.5.1 - Risk Management
2.5.2 - Security of sensitive areas
2.5.3 - Sensitive external sites
2.5.4 - Security assurances for domain name registrants
3 - Registry Asset Management
3.1 - Responsibilities for Registry assets
3.1.1 - Inventory of assets
3.1.2 - Qualification of support assets
3.1.3 - Ownership of assets
3.1.4 - Good and fair use of assets
3.2 - Guidelines for the classification of information
4 - Security related to human resources
4.1 - Roles and Responsibilities
4.2 - Background checks conducted on security personnel
5 - Physical and environmental security
5.1 - Secure areas
5.2 - Hardware security
6 - Operations Management and Telecommunications
6.1 - Procedures and responsibilities related to operations
6.2 - Scheduling and acceptance testing of the system
6.3 - Protection against malicious and mobile code
6.4 - Back-up
6.5 - Security management of network services
6.6 - Monitoring operation of the System
7 - Access Control
7.1 - Business requirements for access control
7.2 - Control of network access
7.3 - Control of access to operating systems
8 - Acquisition, development and maintenance of information systems
8.1 - Cryptographic measures
8.2 - Management of technical vulnerabilities
9 - Managing incidents related to information security
9.1 - Managing improvement and incidents related to information security
10 - IT Disaster Recovery Plan
11 - Integrating audits of the information system

------------------------
1 - Background

The security policy is designed to ensure proper management of the risks that may significantly impact the services provided, the contexts in which they are implemented, and the key personnel involved in operating the Registry. It also defines security level for the scalability ⁄ responsiveness to security incidents, the Registry Data integrity and the confidentiality of personal data of domain name owners.

The Information Security Policy is reviewed at least once a year.

------------------------
2 - Organization of security

------------------------
2.1 - The place of Security in AFNIC’s processes:

AFNIC has set up a Quality Management System (QMS) following the European Framework for QUality Management (EFQM) excellence model. It describes AFNIC’s activities as a series of business processes. Security Process called “ENSURE SECURITY AND BUSINESS CONTINUITY” is one of the cross-business-processes supporting process. It is designed to be compliant with the ISO 27001 norm.
Ensuring security and business continuity mainly consists in defining and controlling how to :
* Supervise the governance of security,
* Apply security measures into the concerned operational fields,
* Manage the risks that could negatively impact the Registries operations.

The implementation of the AFNICʹs ISMS (Information Security Management System) is performed in the framework of the Security process with a view to obtaining ISO 27001 certification by 2014.

------------------------
2.2 - Security Coordination

The overall responsibility for security rests with the CEO. He is assisted in this role by the AFNIC Security Manager (ASM).

Strategic supervision is ensured in a concerted manner by the AFNIC Security Council (ASC) chaired by the AFNIC CEO. The purpose of the ASC is to assist and ensure that the conditions are conducive to attaining the security objectives that fall within the scope of the current strategy.

The ASC further supports the development of security practices at AFNIC through the supporting of operation business functions in implementing security policies, business continuity plans, and staff awareness activities. In carrying out its assignment, the ASC may refer at any time to the Executive Management for advice or a decision on the security of AFNIC and TLD.

------------------------
2.3 - Assignment of responsibilities

------------------------
2.3.1 - Organizational chain of responsibility

The application of security measures to the SRS, DNS, Whois, and other Information Systems is the responsibility of the CTO (head of the Information Systems Division).
The implementation of security measures for staff and premises is the responsibility of the CFO.
The implementation of security measures with respect to legal obligations and registry policies is the responsibility of the Registryʹs Legal Affairs and Policies Director.
The application of security measures relating to damage to the Registryʹs image is the responsibility of the Marketing and Innovation Director.
All the collaborators must be aware of their responsibility concerning the security of resources and information they are accessing, manipulating, publishing.

------------------------
2.3.2 - Relations with the authorities and groups of specialists

AFNIC has an agreement with the French National Agency for the Security of Information Systems (ANSSI). Against this background, the two structures cooperate on security issues that may affect AFNIC services related to its Internet business and risk management in this area.
They cooperate within the framework of two programs on the resilience of the Internet in France :
* Cooperation between the operators of vitals infrastructures in order to improve their capacity to respond to major crises affecting several operators at the same time: the Internet critical services (IP Routing and DNS) are now included in the nomenclature;
* Cooperation to assess the resilience of the French .fr TLD and more generally all the TLDs operated by AFNIC for use by the public.

------------------------
2.4 - Independent security review

Security audits must be conducted by independent organizations twice a year on global and ⁄ or specific issues related to security. No independant assessment report is yet available (see attached document Q30a.pdf)

------------------------
2.5 - Relations with third parties

------------------------
2.5.1 - Risk Management

Risk studies are conducted using the EBIOS methodology (Expression of Business needs and Identification of Security Objectives, in French). This method was designed in 1995 by the French National Agency for Information Security. It is currently used to identify the worst-case scenarios that could affect registry activity. That leads Afnic to design and apply mitigation measures to enhance the protection against these worst-case scenarios.

The control of the effectiveness and efficiency of mitigation measures is performed by the AFNIC’s Security Council all along the year.

------------------------
2.5.2 - Security of sensitive areas

All sensitive areas are under control. That means that access must be controlled and could be restricted to authorized personnel only.
Co-contractors may be requested to sign a confidentiality agreement if required by the sensitivity of information and data they need to know and⁄or use. They only have access to critical technical facilities if accompanied, and never work on production systems.

------------------------
2.5.3 - Sensitive external sites

All security must be applied to protect AFNIC’s resources on external sites. That can be made by private zones and access control to them managed by AFNIC itself.

------------------------
2.5.4 - Security assurances for domain name registrants

The Registry guarantees the following for registrants :
* The continuous availability of operations on its portfolio of domain names, in accordance with the SLA on the SRS
* The continuous availability of information related to the domain, on condition that the registrant uses the services provided to carry out the operations in question,
* The confidentiality of the registrantsʹ personal data (except where other special conditions apply related to the policy of the registry)
* The confidentiality of non-public data relating to the domain and ⁄ or its portfolio of domain names,
* The confidentiality of the transactions with the Registryʹs system,
* The integrity of the information related to its domain name,and published in the WHOIS and the DNS.

------------------------
3 - Registry Asset Management

The security of the registryʹs assets is ensured by the staff assigned to the registryʹs production operations and management activities.
Considering the network connectivity provided by third party, AFNIC’s property begins at the service delivery point.

------------------------
3.1 - Responsibilities for Registry assets

------------------------
3.1.1 - Inventory of assets

Assets used in the run of critical services are identified, qualified, and managed under the guidance of the present policy. Assets considered are staff, infrastructure, software, connectivity, data and providers.

------------------------
3.1.2 - Qualification of support assets

The assets contributing to the Services are classified in 3 main categories :
* Computer Systems and Telecommunications : Hardware and Software; Communications Channels; Outsourced Services;
* Organizations : Staff; Corporate departments;
* Physical locations for business : Offices; Hosting Datacenters;

------------------------
3.1.3 - Ownership of assets

Registry data belong to the Registry owner. They are subject to the rules of the contract with ICANN, plus the applicable legal and ⁄ or legislative rules depending on the context in which the registry is implemented

------------------------
3.1.4 - Good and fair use of assets

All the registry operations and services must be used by third party in accordance with the contractual rules defined by the owner and the operator of the TLD.

------------------------
3.2 - Guidelines for the classification of information

The data used or produced in the context of the Registry are classified in the 3 following categories :

= Critical information = : it can⁄must be accessed⁄showed only by accredited persons. Disclosure or alteration may result in significant damage but repairable.

= Reserved information = : Information is limited to persons, entities or authorized partners. Disclosure or alteration may result in significant harm.

= Internal Information = : Information is available to staff of AFNIC and authorized partners. Disclosure or alteration may perturb the normal functioning of the company, without lasting consequence.

------------------------
4 - Security related to human resources

------------------------
4.1 - Roles and Responsibilities

There are 2 categories of staff :

* Technical staff : These personnel have access to resources according to defined rights.
* Administrators in charge of administering production resources. They can access all the production resources and data.
* Technicians in charge of the operation, maintenance and monitoring of the production system. They have limited rights of access to production resources. They can access certain resources on request and when accompanied by an administrator.
* Experts in charge of the design and development of production resources. They only have access to the production resources on request and when accompanied by a technician and ⁄ or an administrator.
* Non-technical staff :
* Administrative staff and managers (excluding production).

------------------------
4.2 - Background checks conducted on security personnel

French law applies to all staff. The contract they sign with their employer contains sufficient provisions in terms of professionalism and ethics for the activity involving the TLD. Same rules are applicable a Data Center level.

------------------------
5 - Physical and environmental security

------------------------
5.1 - Secure areas

AFNIC production sites are secured at the means of access to them. The DATA CENTER sites must meet the standards of industrial and environmental security compatible with the constraints implied by their activity. The layout of the premises must be such that access is restricted only to authorized personnel at entry points selected and controlled by AFNIC.

------------------------
5.2 - Hardware security

The Data centers that host AFNIC services ensure at least Tier 3 levels of resilience.

------------------------
6 - Operations Management and Telecommunications

AFNIC controls the operation of all the resources used to deliver essential services with the exception, of course, of outsourced services such as certain DNS servers.
AFNIC operates dark fiber connections between its sites. The terminals are owned by AFNIC. They are operated by AFNIC personnel.

------------------------
6.1 - Procedures and responsibilities related to operations

Operating procedures are documented and kept up to date on the intranet of the IT team.
Access to the applications, servers and databases must be defined and kept up to date for each staff member.
Access privileges are defined in order to respect the security rules associated with the classification of information.
Operations related to DNSSEC are subject to even more stringent security regulations and require respecting the DPS procedure.

------------------------
6.2 - Scheduling and acceptance testing of the system

The test, pre-production and production phases must be clearly specified. Any production launch must be announced to the registrars at least 2 month before it applies.

------------------------
6.3 - Protection against malicious and mobile code

All the entry points to the production servers are filtered by the firewall, which applies the filtering policy common to all the procedures, whether they involve a human operator or an automated process.

Each development must apply security rules and recommendations on the development of application.
The Web access must be protected against the most common (Script kiddies, SQL injection …)

------------------------
6.4 - Back-up

Registry data are stored and secured using the real-time replication mechanisms of the production Database Management System (production DBMS).
In addition, a physical backup of the entire database must be performed at the same time as the back-up of the other components of the SRS.
To be compliant with the ICANN requirements, a data escrow deposit must be performed every day between 0:00 am end 12:00pm

------------------------
6.5 - Security management of network services

A strict partitioning into zones must be implemented in order to avoid interconnections between the external production, administration and backup networks.

Any internal and external attempts to access production servers must pass through a Firewall. They are recorded in a log file for later analysis. The detection of malicious code based on a regularly updated list must be performed at this level.

An intrusion detections system must be installed and running between firewall and production servers.

------------------------
6.6 - Monitoring operation of the System

Automated monitoring must be implemented. It must cover the hardware, software systems and production applications.
Any failure must be subject to a specific alert sent to the staff:
* on duty during office hours;
* on standby outside office hours;

------------------------
7 - Access Control

------------------------
7.1 - Business requirements for access control

Access to the information system requires prior identification and authentication. The use of shared or anonymous accounts must be avoided. Mechanisms to limit the services, data, and privileges to which the users have access based on their role at AFNIC and the Registry must be implemented wherever possible.

------------------------
7.2 - Control of network access

The internal network must be partitioned to isolate the different services and applications and limit the impact of incidents. In particular it is highly desirable to isolate services visible from the outside in a semi-open zone (DMZ). Similarly, access to the wireless network must be controlled and the network must be subject to appropriate encryption.

------------------------
7.3 - Control of access to operating systems

The production servers must be confined in secure facilities. Access must be restricted to authorized personnel only. The personnel in question are the members of the operating teams and their managers, IT personnel and those of the Security Manager.

------------------------
8 - Acquisition, development and maintenance of information systems

------------------------
8.1 - Cryptographic measures

Cryptographic measures must be implemented to secure the exchanges :
* between the workstations of technical staff and the access proxies to production servers;
* between the Registrars and the EPP server;
* between the DNS master servers and the resolution servers;
* to upload the records of the Escrow Agent.

------------------------
8.2 - Management of technical vulnerabilities

The technical configuration of hardware and software used must be subject to up to date documentation.
The changes in technical configurations must be constantly monitored and documented.
Security alerts involving updates and ⁄ or patches to production systems must be constantly monitored.
Application procedures must be documented and updated based on the recommendations of the designers of a component.

------------------------
9 - Managing incidents related to information security

------------------------
9.1 - Managing improvement and incidents related to information security

The crisis management procedure serves to mobilize at a sufficiently high echelon, all the appropriate levels of responsibility for taking decisions on the actions required to resolve the crisis and return to normal.
Each security incident must be analyzed under the cover of the Security Council and the recommendations, if any, are applied, checked and evaluated as required by the QMS.

------------------------
10 - IT Disaster Recovery Plan

The risk analysis must produce some inputs for the elaboration of a disaster recovery plan. That plan has to be established and regularly tested in order to maintain or recover Registry activity and make critical services available at the required SLA after an interruption or a crash of critical services of the Registry.

------------------------
11 - Integrating audits of the information system

Security audits are performed annually. They are launched on the initiative of the CTO or upon request from the ASC. They are carried out by independent bodies and relate to one or more of the essentials services of the Registry.

The ASC and the ASM control the implementation and the efficiency of these measures in the framework of S3 process (see section 2.1).

Alain MARTEL	Member of the Board of Directors
Alexis TAMAS	Member of the Board of Directors
Amaury GRIMBERT	President of the Board of Directors

New gTLD Application Submitted to ICANN by: OP3FT

String: frogans

Originally Posted: 13 June 2012

Application ID: 1-1270-9301

Applicant Information

1. Full legal name

2. Address of the principal place of business

3. Phone number

4. Fax number

5. If applicable, website or URL

Primary Contact

6(a). Name

6(b). Title

6(c). Address

6(d). Phone Number

6(e). Fax Number

6(f). Email Address

Secondary Contact

7(a). Name

7(b). Title

7(c). Address

7(d). Phone Number

7(e). Fax Number

7(f). Email Address

Proof of Legal Establishment

8(a). Legal form of the Applicant

8(b). State the specific national or other jursidiction that defines the type of entity identified in 8(a).

8(c). Attach evidence of the applicant's establishment.

9(a). If applying company is publicly traded, provide the exchange and symbol.

9(b). If the applying entity is a subsidiary, provide the parent company.

9(c). If the applying entity is a joint venture, list all joint venture partners.

Applicant Background

11(a). Name(s) and position(s) of all directors

11(b). Name(s) and position(s) of all officers and partners

11(c). Name(s) and position(s) of all shareholders holding at least 15% of shares

11(d). For an applying entity that does not have directors, officers, partners, or shareholders: Name(s) and position(s) of all individuals having legal or executive responsibility

Applied-for gTLD string

13. Provide the applied-for gTLD string. If an IDN, provide the U-label.

14(a). If an IDN, provide the A-label (beginning with "xn--").

14(b). If an IDN, provide the meaning or restatement of the string in English, that is, a description of the literal meaning of the string in the opinion of the applicant.

14(c). If an IDN, provide the language of the label (in English).

14(c). If an IDN, provide the language of the label (as referenced by ISO-639-1).

14(d). If an IDN, provide the script of the label (in English).

14(d). If an IDN, provide the script of the label (as referenced by ISO 15924).

14(e). If an IDN, list all code points contained in the U-label according to Unicode form.

15(a). If an IDN, Attach IDN Tables for the proposed registry.

15(b). Describe the process used for development of the IDN tables submitted, including consultations and sources used.

15(c). List any variant strings to the applied-for gTLD string according to the relevant IDN tables.

16. Describe the applicant's efforts to ensure that there are no known operational or rendering problems concerning the applied-for gTLD string. If such issues are known, describe steps that will be taken to mitigate these issues in software and other applications.

17. (OPTIONAL) Provide a representation of the label according to the International Phonetic Alphabet (http://www.langsci.ucl.ac.uk/ipa/).

Mission/Purpose

18(a). Describe the mission/purpose of your proposed gTLD.

18(b). How do you expect that your proposed gTLD will benefit registrants, Internet users, and others?

18(c). What operating rules will you adopt to eliminate or minimize social costs?

Community-based Designation

19. Is the application for a community-based TLD?

20(a). Provide the name and full description of the community that the applicant is committing to serve.

20(b). Explain the applicant's relationship to the community identified in 20(a).

20(c). Provide a description of the community-based purpose of the applied-for gTLD.

20(d). Explain the relationship between the applied-for gTLD string and the community identified in 20(a).

20(e). Provide a description of the applicant's intended registration policies in support of the community-based purpose of the applied-for gTLD.

20(f). Attach any written endorsements from institutions/groups representative of the community identified in 20(a).

Geographic Names

21(a). Is the application for a geographic name?

Protection of Geographic Names

22. Describe proposed measures for protection of geographic names at the second and other levels in the applied-for gTLD.

Registry Services

23. Provide name and full description of all the Registry Services to be provided.

Demonstration of Technical & Operational Capability

24. Shared Registration System (SRS) Performance

25. Extensible Provisioning Protocol (EPP)

26. Whois

27. Registration Life Cycle

28. Abuse Prevention and Mitigation

29. Rights Protection Mechanisms

30(a). Security Policy: Summary of the security policy for the proposed registry

© Internet Corporation For Assigned Names and Numbers.