[dkelson]

Cherry-pick of #11215 for 3.18.

On 3.18.2, hammer users with organization/location defaults can't manage template inputs. hammer template-input create fails with a 422 ("This template is locked", "Input type is not included in the list") and hammer template-input list returns a 500. The organization_id/location_id that hammer adds get picked as the nested parent, so the template is looked up by the location's id.

This needs theforeman/foreman_remote_execution#1067 alongside it, or the foreign input sets API loses its parent template. 3.18 ships REX 16.5.x, which has no stable branch I could target. Could someone cut a 16.5.z with it, or tell me where to send it?

Tested by applying both changes to a 3.18.2 / Katello 4.20 install: template-input create and list work with organization/location ids, and so do the foreign input set, report template and audit lists.

[lzap]

Okay, first question: is this technically viable for 3.18? @evgeni @adamruzicka

Second question is for the backporting process: how to do this properly (I assume we would need 5.0 and 3.19 as well)? @ogajduse @Odilhao

[evgeni]

I don't think there are any more 3.18 releases planned where this could be shipped. (But that part is a question to Odilon and Ondrej)

Whether it's a good idea to backport this to 3.18, idk. "It's probably fine" is what my brain says, but… Don't trust it too much.

[ogajduse]

I don't think there are any more 3.18 releases planned where this could be shipped. (But that part is a question to Odilon and Ondrej)

Whether it's a good idea to backport this to 3.18, idk. "It's probably fine" is what my brain says, but… Don't trust it too much.

With the current workload, I do not expect any more releases within the 3.18 series.

[adamruzicka]

is this technically viable for 3.18?

Iirc this required changes in plugins as well. We'd have to pick them all and coordinate the releases so that all of them go out at the same time

[lzap]

Audit and reports is owned by @andreilakatos @kmalyjur @Lukshio @MariaAga @pondrejk @rmynar @pnovotny

[evgeni]

#11215 (comment) seems to indicate "no, don't pick"