Skip to main content

Early Review of draft-ietf-ivy-network-inventory-software-04
review-ietf-ivy-network-inventory-software-04-yangdoctors-early-rahman-2026-09-01-00

Request Review of draft-ietf-ivy-network-inventory-software
Requested revision No specific revision (document currently at 04)
Type Early Review
Team YANG Doctors (yangdoctors)
Deadline 2026-08-28
Requested 2026-07-21
Requested by Daniele Ceccarelli
Authors Bo Wu , Cheng Zhou , Qin Wu , Mohamed Boucadair
I-D last updated 2026-07-06 (Latest revision 2026-07-06)
Completed reviews Yangdoctors Early review of -04 by Reshad Rahman
Opsdir Early review of -04 by Chongfeng Xie
Assignment Reviewer Reshad Rahman
State Completed
Request Early review on draft-ietf-ivy-network-inventory-software by YANG Doctors Assigned
Posted at https://mailarchive.ietf.org/arch/msg/yang-doctors/lnmDT2OGIXcjxM3VkA7CnDQxFwU
Reviewed revision 04
Result Almost ready
Completed 2026-09-01
review-ietf-ivy-network-inventory-software-04-yangdoctors-early-rahman-2026-09-01-00
This is review of -04 by Reshad Rahman.

In general the document and YANG both look good but I believe there are a few
issues which need to be discussed/addressed.

Issues/comments/questions
=========================

General

- I’m not a big fan of the term “non-physical”. One suggestion is to change
“ne-non-physical”  to “ne-virtualized” and keep “ne-software”, “ne-virtual”
(rename to “ne-vm”) and “ne-container” based off ne-virtualized (another
suggestion is “ne-softwatre-based” from Claude). If the WG feels that
“non-physical” is more appropriate, I’ll defer to the WG consensus but there
doesn’t seem to have been any discussions (on the mailing list) on this topic.

Section 6
- The name of the module is ietf-network-inventory-sw-ext, having -ext suffix
is unusual. Rename to something along the lines of
ietf-network-inventory-software.

- The description of ne-container says “Runs under Docker/K8s”. I believe
Docker can be an example but there could in theory be other container types.
And container doesn’t mean K8s is used either, I’d remove any mention of K8s.

- We only have “software-installed” and “software-activated” as
software-status. I can think of “software-install-failed”,
“software-activation-failed”, “software-activation-in-progress” etc.

- What should activation-time be if the software is not activated. Add some
text to the description. And/or add when-stmt to activation-time so that it’s
in the schema only if status is software-activated.

- Is activation-time “mandatory” if status == software-activated? If yes, add a
must statement to status and/or update the description of activation-time.

- activation-time mentions “was rebooted into”, so does it change on every
reboot? Please clarify that in the description.

- For containers, my understanding is that the revision is mutable. So 7.9.2 is
just a tag and tags can be reassigned. You probably need an image digest for
containers. And since we’re here, an image registry too to indicate where the
image came from. So for containers I’d have name, tag (not revision), digest
and registry for the image. I’d like to hear from people with expertise in this
area.

Section 8
- XPath “/nwi:network-elements/network-element/component/component/“ is
incorrect.

Appendix A
- Does that example validate with yanglint? Note that leaf activation-time has
no prefix but status and installation-time do.

Minor comments/questions + nits
===============================

Section 1.2
Please add the terms defined in draft-ietf-ivy-network-inventory-yang being
reused here

Section 2
Typo Platformos -> platforms

Section 6
- Did you consider adding a container in software-info-grouping to wrap the 3
leaf nodes?

- Did you consider adding data nodes to explicitly indicate the currently
installed/activated software. This can be “extracted” from the current data but
having it explicit would be more human-friendly.