Skip to main content

Last Call Review of draft-ietf-regext-epp-delete-bcp-08
review-ietf-regext-epp-delete-bcp-08-dnsdir-lc-weber-2024-11-28-00

Request Review of draft-ietf-regext-epp-delete-bcp
Requested revision No specific revision (document currently at 09)
Type Last Call Review
Team DNS Directorate (dnsdir)
Deadline 2024-11-28
Requested 2024-11-14
Authors Scott Hollenbeck , William Carroll , Gautam Akiwate
I-D last updated 2024-11-28
Completed reviews Dnsdir Last Call review of -08 by Ralf Weber (diff)
Secdir Last Call review of -08 by Carl Wallace (diff)
Artart Last Call review of -08 by Jiankang Yao (diff)
Genart Last Call review of -08 by Linda Dunbar (diff)
Dnsdir Telechat review of -09 by Ralf Weber
Assignment Reviewer Ralf Weber
State Completed
Request Last Call review on draft-ietf-regext-epp-delete-bcp by DNS Directorate Assigned
Posted at https://mailarchive.ietf.org/arch/msg/dnsdir/wXnfOcfStUAT8Smqzftja8V4GFQ
Reviewed revision 08 (document currently at 09)
Result Almost ready
Completed 2024-11-28
review-ietf-regext-epp-delete-bcp-08-dnsdir-lc-weber-2024-11-28-00
Moin!

I am an assigned DNS Directorate reviewer for draft-ietf-regext-epp-delete-bcp.

For more information about the DNS Directorate, please see
https://wiki.ietf.org/en/group/dnsdir

Overall the draft does a great job explaining the problem space of host object
deletion in EPP and uses a great structure for that.

From a DNS perspective however I think there are some things that could be
improved. First I think we need a more precise definition of what renaming
means for the DNS referral answer. If e.g the name and not the IP that has
different impact on DNS resolution than if a there is a name that is not
resolvable or a renaming and giving an IP that actually responds. This should
IMHO be added to the cases in section 5. I am not even sure on what the results
are for all the cases, but would offer to help with text/examples on that once
that is clear for all the cases.

Second on section 5.1.3.4 the text on if I should run an resolver or an
authoritative answer for the sacrifical name server host and what the answer
actually is. Now there might have been discussions about that that I am not
aware of and there may be a reason for the text, but if so it should be noted.
My rather simple view on this if I had to run such a server I would answer
REFUSED to all the queries I got.

Not related to this draft, but I am happy that the so far proposed DELEG
solutions will solve the problem as there no longer is a need for a host
object, but as we are not there yet this document is badly needed, so please
keep up the good work and if there is anything I can help please reach out.

So long
-Ralf