The ROID is generated by the Registry for each type of object supported (e.g., 
Domain Name in RFC 5731, Host in RFC 5732, Contact in RFC 5733).  A thin 
registry will generate the ROIDs for the Domain Name and Host objects.  A thick 
registry will add support for the Contact objects and the Contact ROID.  See 
section 2.8 of RFC 5730 
(https://datatracker.ietf.org/doc/html/rfc5730#section-2.8) and the EPP 
Repository Identifiers registered in 
https://www.iana.org/assignments/epp-repository-ids/epp-repository-ids.xhtml 
used in the ROID suffix.  The ROID needs to be a globally unique value for each 
type of object leveraging the EPP Repository Identifier IANA Registry, 
generated by the server, and returned to the client in the object info response.

--

JG

[cid87442*image001.png@01D960C5.C631DA40]

James Gould
Fellow Engineer
jgo...@verisign.com<applewebdata://13890C55-AAE8-4BF3-A6CE-B4BA42740803/jgo...@verisign.com>

703-948-3271
12061 Bluemont Way
Reston, VA 20190

Verisign.com<http://verisigninc.com/>

From: Marco Schrieck <marco.schri...@internetx.com>
Date: Friday, January 31, 2025 at 12:09 AM
To: "regext@ietf.org" <regext@ietf.org>
Subject: [EXTERNAL] [regext] Re: [Ext] Clarification on ROID Usage for 
Registrars in Thin Registry RDAP Implementations


Caution: This email originated from outside the organization. Do not click 
links or open attachments unless you recognize the sender and know the content 
is safe.

Hi Rubens,

No its the other way. More and more gTLDs will get thin.

Ciao
Marco

On 31 January 2025 00:31:45 CET, Rubens Kuhl <rubensk=40nic...@dmarc.ietf.org> 
wrote:

Don’t the remaining thin registries allow for thin to thick migration on a 
granular basis ?


Rubens



Em 30 de jan. de 2025, à(s) 18:24, Marco Schrieck 
<marco.schri...@internetx.com> escreveu:

Hi

Yes the registry should generate them, but what with thin Registries.

Only we as registrar have the contact data. So we don't have a registry 
generated roid.

I can generate on but the rdap profile checker require me to register the 
suffix at iana.

Eg
1234567-IX

If i don't register it we got an error the we use uncorrect value.
"Globally unique identifier not registered in EPPROID"

At the moment we let it empty and got a warning instead of error.

Marco

On 30 January 2025 22:00:50 CET, "Andrew Newton (andy)" <a...@hxr.us> wrote:

Marco,

Assuming you are talking about the 2019 gTLD profile, the registry
should be generating the ROIDs because it is also required of them.

-andy

On Thu, Jan 30, 2025 at 1:52 PM Marco Schrieck
<marco.schri...@internetx.com> wrote:

 As I understand it must be registered there. But as i can oversee it, there 
are only Registries.

 Marco


 On 30 January 2025 19:42:38 CET, James Mitchell <james.mitch...@iana.org> 
wrote:

 My understanding is the EPP ROID is an identifier that is composed of a suffix 
that identifies the registry/repository to which an object belongs. That 
repository identifier should be registered in the IANA registry of EPP 
Repository Identifiers at 
https://www.iana.org/assignments/epp-repository-ids<https://secure-web.cisco.com/1jnLscMClteISaGbgSKBEm0AfyV-xz7CyAKrhih8RYKbk2tFw7c38E_SncT6PLtv8E1xK38elzTP2Cpd1bplnxEzxBUiRyWJp-N3aYh-qOsC-YN462UBPdk5h-nzmq8eQGyFKFb4QLL-AgUigPiznLbiUs6wwQ1F1RdQ2dKP7RqHLl8mZ8eggfcVpzLESXIa20dwKueeaEfcwAPoEtgyXQMyRxM6Rsn-ZYEBMcpMdZR-63oCrRJ26Dz2VlRdrkwxthv92QzFFkgfNW0ZbDdLzEm0FEvROBl1AWIoI1FVWJk4bYZQAr9VRGRy-BxLyxtTu/https%3A%2F%2Fwww.iana.org%2Fassignments%2Fepp-repository-ids>.
 I can’t speak to whether a registration is necessary for your use case – that 
would appear to be up to those writing the (test) requirements.



 Thanks,

 James



 From: InterNetX - Marco Schrieck <marco.schri...@internetx.com>
 Organization: InterNetX GmbH
 Date: Thursday, January 30, 2025 at 7:14 AM
 To: "regext@ietf.org" <regext@ietf.org>
 Subject: [Ext] [regext] Clarification on ROID Usage for Registrars in Thin 
Registry RDAP Implementations



 Hi All,



 I am writing to seek guidance on the handling of ROID (Repository Object 
IDentifier) in RDAP implementations for registrars together with thin registry 
models (where registrars hold domain/contact/host data). Our organization acts 
as a registrar and is working to comply with the RDAP profile outlined in RFC 
7483 and related 2019 updates.

 Context

 ·         We operate in a thin registry environment where the registry 
delegates RDAP queries to registrars.

 ·         Our implementation uses registrar-generated identifiers (not ROIDs), 
as the registry does not assign or store ROIDs.

 ·         During RDAP testing, we encountered errors such as “globally unique 
identifier not registered in EPPROID”, suggesting a mismatch between our 
identifiers and ROID expectations.

 Questions



 1.    RFC 7483 §10.2.4 mentions roid as optional. For registrars in thin 
models:

 2.

 3.    Is it acceptable to use registrar-generated handles (e.g., UUIDs) 
instead of ROIDs in RDAP responses?

 4.

 o    Are there best practices for mapping internal registrar IDs to RDAP 
handle or roid fields?

 o

 o    Did the 2019 discussions formalize any extensions (e.g., custom JSON 
fields) for registrars to bypass ROID requirements?

 o

 o    How do we resolve errors like “identifier not registered in EPPROID” if 
ROIDs are registry-managed but unavailable to registrars?

 §





 Thank you for your insights.


 Ciao
 Marco





 --

 InterNetX GmbH

 Johanna-Dachs-Str. 55 • 93055 Regensburg • Germany

 Tel. +49 941 59559-0



 internetx.com • internetx.com/linkedin • internetx.com/twitter



 Geschäftsführer:

 Elias Rendón Benger (CEO), Lars Krämer

 Amtsgericht Regensburg, HRB 7142

________________________________

 regext mailing list -- regext@ietf.org
 To unsubscribe send an email to regext-le...@ietf.org

________________________________

regext mailing list -- regext@ietf.org
To unsubscribe send an email to regext-le...@ietf.org
_______________________________________________
regext mailing list -- regext@ietf.org
To unsubscribe send an email to regext-le...@ietf.org

_______________________________________________
regext mailing list -- regext@ietf.org
To unsubscribe send an email to regext-le...@ietf.org

Reply via email to