- db-wg - mailman.ripe.net

Archiving the "Action List"
by Peter Hessler 04 Jun '25

04 Jun '25

As seen on the website, the Database Working Group has an Action List, instead of an Action Movie. https://www.ripe.net/community/wg/active-wg/db/action-list/ From a basic perusal of them, this seems to be a basic tracking system of TODOs and Proposals. Notice it says "last updated 7 May 2009". I propose we simply archive this list. If anything is still relevant, please bring it up as if it were new. Peter Hessler, On behalf of the DB WG Co-Chairs

4 3

status levels
by denis walker 03 Jun '25

03 Jun '25

Colleagues Maybe I asked too many questions in one email, so I'll take it bit by bit. Is it reasonable to limit the number of levels of SUB-ALLOCATED PA to 4? That is the most we have now in the RIPE Database. Or do you have another number in mind...maybe even 1? Should there be a limit on the number of levels in a LEGACY hierarchy? Currently it is infinite. cheers denids

5 17

NWI proposal: adjusting contact method requirements and adding new methods
by Leo Vegoda 03 Jun '25

03 Jun '25

Hi, A while back Cynthia, Dmitry and I looked at the possibility of adding new contact methods. Dmitry presented this at RIPE 88: https://ripe88.ripe.net/archives/video/1347/ I've tried to put this into a more complete NWI and offer it for discussion. The TLDR is: 1) Align the requirement levels for contact methods 2) Let people use new contact methods, like Signal We hope that by offering the methods people want to use, the contact data in the RIPE Database can be as useful as is possible. Kind regards, Leo ## Problem statement The requirement level for contact methods for person and role objects are inconsistent. This is both confusing and inconvenient. This is a proposal to add flexibility to the contact methods offered at the database level. It is not a proposal to change policy requirements for contact methods. This proposal is limited to admin-c and tech-c contact types. Our world offers more contact methods than were available when the currently supported methods were chosen for the RIPE Database. User acceptance of these methods differ by country and generation. Whether this proposal is accepted or not, it is good to review the rationale for these rules and have an updated rationale published. 1. Person and role objects have four contact methods. A postal address is mandatory for both. A phone is mandatory for person objects but not role objects. E-mail is mandatory for role but not person objects. 2. Postal mail is less useful than it was 20 years ago. Some countries are reducing the frequency of postal deliveries. And geographically distributed teams find postal mail less useful. 3. The telephone is becoming less useful, too. Call volumes are dropping and popular psychology publications explain why calls are less likely to be answered: - https://www.statista.com/statistics/551035/finland-call-volume-by-type-of-c… - https://www.bbc.com/news/articles/ckg8jllq283o - https://www.psychologytoday.com/intl/blog/un-numb/202405/why-gen-zers-keep-… ## Proposal We should provide contacts with more flexibility. We should also add new options that were not available when the RIPE Database's contact methods were last reviewed. The three elements below follow on from NWI-15, which is based on the RIPE Database Requirements Task Force recommendations for what should be published about resource holders. The registry needs to retain the full set of required data. The DBTF report did not mention contacts methods for the resources. 1. Align the contact method requirements between role and person objects used in admin-c and tech-c attributes. Make all contact methods optional for the public RIPE Database as long as one is chosen. There is no point publishing a contact method that will not be used. That would just waste the time of people trying to make contact. 2. Add support for new contact methods in the RIPE Database, including SIP (RFC 3969), Signal, Telegram, and WhatsApp. 3. Add support for RFC 8605 (ICANN Extensions for the Registration Data Access Protocol) and show the HTTPS intermediary and/or private URI where one is available. Making these changes will require updates to the RPSL and RDAP publication methods. The RIPE NCC should manage the specifics of the design and implementation. ## Recommendation for Impact Assessment The impact assessment provided by the RIPE NCC should note any contact method requirements specified in policy. It should also note where the policy does not specify a contact method requirement. The RIPE NCC might want to note the impact of enforcing policy requirements when more user choice is offered at the database layer.

6 9

Proposal: Privacy in RIPE database objects
by Rodolfo García Peñas (kix) 27 May '25

27 May '25

Hi, I would like to make a proposal regarding the database so that privacy can be included in the objects. Among the main objectives of the database [1] are "To provide accurate registration information of these resources in order to meet a variety of operational requirements" and "Facilitating coordination between network operators (network problem resolution, outage notification etc.)", and therefore the database contains information to facilitate these actions. However, this information can be used to obtain information about individuals and entities in order to carry out unwanted actions (network recognition, phishing, etc.). To prevent network information and contact information from being obtained, objects with anonymized information are stored in the database, role objects are used instead of person objects, non-real information is used, or inetnum objects are not created. Some mailing threads related to these issues have been seen on the mailing list [2][3]. My proposal is that the database offers privacy regarding the information it contains, allowing real information to be stored while protecting LIRs and end users by providing filtered information. To this end, I propose using a system similar to "domain privacy" in DNS registries to hide desired information: - Use the option to hide sensitive information in mandatory fields. - Provide an option to maintain contact with the person responsible for the object. Example: person: Privacy protected address: Privacy protected address: Privacy protected e-mail: Privacy protected (contact link) phone: Privacy protected (contact link) remarks: For mail filtering or abuse problems contact with the abuse mailbox nic-hdl: ABCD1-RIPE mnt-by: Privacy protected (contact link) created: 2008-11-13T15:53:02Z last-modified: 2024-09-23T13:25:48Z source: RIPE One proposal is to use a Boolean model for each field (or for all fields in the object) to indicate whether the information should be "protected" or "public." Another example option would be to include levels such as "public" (publicly accessible), "LEA" (accessible by LEAs), "LIR" (accessible by other authenticated LIRs), and "private" (fully private). This information could be included by default at the user or LIR level using the LIR portal. For email or telephone communication, the platform could act as a mail intermediary or by providing anonymized email addresses, facilitating coordination between network operators/users. IMO, first, we must reach a consensus on whether this option is useful for database management, and then we can discuss implementation details. Regards, Rodolfo García (kix) PS. Thanks Jordi, Miguel for your help. [1] https://docs.db.ripe.net/What-is-the-RIPE-Database/Purpose-and-Content-of-t… [2] https://www.ripe.net/ripe/mail/archives/address-policy-wg/2023-September/01… [3] https://mailman.ripe.net/archives/list/db-wg@ripe.net/thread/MEOS4MD7ZYLVOA… The content of this email reflects personal opinions and does not represent the views of any organization.

4 6

Publishing End-Site Prefix Lengths
by Randy Bush 26 May '25

26 May '25

i received a suggestion to mention this on the ripe wg list. similar to geofeed, it allows one to document the prefix length one is using. Publishing End-Site Prefix Lengths https://datatracker.ietf.org/doc/draft-ietf-opsawg-prefix-lengths/ This document specifies how to augment the Routing Policy Specification Language (RPSL) inetnum: class to refer specifically to prefixlen comma-separated values (CSV) data files and describes an optional scheme that uses the Resource Public Key Infrastructure (RPKI) to authenticate the prefixlen data files. randy

2 2

RIPE89 Database Working Group Minutes
by Peter Hessler 16 May '25

16 May '25

Hello Database Working Group, The draft minutes from our session at RIPE89 has been published on the RIPE website, available at https://www.ripe.net/community/wg/active-wg/db/minutes/ripe-database-workin… We'd like to thank Alena from the RIPE NCC for preparing the draft. Please review the minutes, and send any necessary corrections to the working group mailing list. Peter Hessler, On behalf of the DB WG Co-Chairs

1 1

Regarding the issue of empty AS-SET in databases
by liu haoran 15 May '25

15 May '25

As the administrator of AKIX , we have observed a recurring issue where newly joined members frequently submit AS-SET objects without populating the mandatory members attribute in IRR databases. From both operational and RPSL (Routing Policy Specification Language) specification perspectives, an AS-SET containing no members represents an invalid configuration, as it fundamentally defeats its purpose of aggregating Autonomous Systems (AS) for routing policy management. We formally propose that IRR database maintainers implement mandatory validation during AS-SET creation to enforce: 1. Require at minimum one valid AS number in the members field 2. Reject AS-SET submissions containing empty/null members attributes 3. Provide clear error messaging specifying the validation requirements This technical enforcement would align with RFC 2725 (Routing Policy System Security) recommendations while significantly improving routing registry data quality and operational reliability for Internet Exchange ecosystems.

4 5

Hierarchical AS-SET names
by James Bensley 12 May '25

12 May '25

Hi all, I raised a request with ARIN to enforce the use of hierarchical AS-SET names moving forward, as is the case in RIPE, APNIC, and LACNIC. ARIN have opened a consultation process, which closes on the 23rd of May. If you would like to see ARIN also adopt the requirement for AS-SETs to have hierarchical names, please visit the following page and contact ARIN to record your support for this request: https://www.arin.net/participate/community/acsp/consultations/2025/2025-1/ Kind regards, James Bensley (he/him) Inter.link GmbH Boxhagener Str. 80, 10245 Berlin, Germany Email: hello(a)inter.link, Phone (general): (+49) 030577123821 Phone (mobile): (+49) 015792522412 Registry: Local court Charlottenburg, HRB 138876 Managing directors: Marc Korthaus, Theo Voss

1 0

RIPE Database Acceptable Use Policy Updated
by Edward Shryane 30 Apr '25

30 Apr '25

Dear colleagues, At the 183rd RIPE NCC Executive Board meeting on 21st March 2025, the following changes to the RIPE Database Acceptable Use Policy were approved: * Clarified that the RIPE NCC accounts by prefix, the size of which is determined by the RIPE NCC. This better reflects the reality of how we implemented the feature. * Clarified that the RIPE NCC may lift a permanent block of a user from querying the RIPE database on request, at our sole discretion and on a case-by-case basis. * Use the term "RIPE NCC Access" rather than "SSO". The minutes from the 183rd Executive Board meeting are available: https://www.ripe.net/about-us/executive-board/minutes/2025/183rd-executive-… The updated Acceptable Use Policy is available on the RIPE NCC website : https://www.ripe.net/manage-ips-and-asns/db/support/documentation/ripe-data… A diff of the changes is visible in the DB documentation repository : https://github.com/RIPE-NCC/ripe-database-documentation/blame/main/docs/25.… Previous discussion: https://mailman.ripe.net/archives/list/db-wg@ripe.net/thread/VNZW422LST24X6… Regards Ed Shryane RIPE NCC

1 0

Questions on "status:"
by denis walker 29 Apr '25

29 Apr '25

Colleagues I am working with the RIPE NCC to document the current rules for the "status:": attribute values. This is not an attempt to simplify or substantially change the values, although some aspects do need clarifying. There is no documentation anywhere on what the current rules are. This information only exists completely in the RIPE Database software and in the test suite for the database software. The goal is to describe all the current status rules and relationships in the RIPE Database documentation. Some of these rules have never been discussed by the community to the point of a consensus. Some are interpretations made by the database developers over the years (myself included) based on comments in various other discussions. I have some questions that need clarification from the community. I have listed them below. Some might seem obvious, but within the software some things are allowed and some things are not disallowed just because no one has ever specified what it should (not) be. If you have a strong opinion on any of these questions, please share it. Any questions that no one comments on, then as we have done in the past we will document 'an interpretation'. If the community accepts the resulting documentation, the software will be adjusted accordingly. cheers denis 1/ Is there a use case for AGGREGATED-BY-LIR more specific to AGGREGATED-BY-LIR? If so, to how many levels? 2/ Is there a use case for LIR-PARTITIONED PA more specific to ALLOCATED UNSPECIFIED? 3/ Is there a use case for LIR-PARTITIONED PA more specific to SUB-ALLOCATED PA? As the name suggests, it was introduced to partition an allocation. There are currently about 438 hierarchies with LIR-PARTITIONED PA more specific to SUB-ALLOCATED PA in the database. 4/ Is there a use case for LIR-PARTITIONED PA more specific to LIR-PARTITIONED PA? If so, to how many levels? There are currently about 83 hierarchies with LIR-PARTITIONED PA more specific to LIR-PARTITIONED PA in the database. They only exist to two levels. 5/ Should we say that SUB-ALLOCATED PA 'must' be defined in the RIPE Database? This is not clear in the address policy. As we move to a world where not everything must be defined in the database, we probably should be clear about what must be defined there. (Which may require an update to the address policy.) 6/ There are currently multiple levels of SUB-ALLOCATED PA in the database. The max that exists at the moment is 4 levels of SUB-ALLOCATED PA. Is that a reasonable level to set as a limit? Or should we not allow it at all? 7/ Should there be a limit on the number of levels in a LEGACY hierarchy?

1 0