drop support for EPHEMERAL user type in mapping

[Steve Martinelli]

With the shadow users implementation, federated users are no longer emphemeral. Support for specifying this option in a mapping should be removed. The option should be ignored and should result in a log that indicates a timeframe for removal (2 cycles)

See this blueprint for details: https://blueprints.launchpad.net/keystone/+spec/shadow-users-newton

[Steve Martinelli (stevemar)]

patch: https://review.openstack.org/#/c/296639/

[Steve Martinelli (stevemar)]

Bumping this to Ocata, it's low priority for newton and can be fixed when the federation mapping engine is fixed up next release.

[OpenStack Infra (hudson-openstack)]

Change abandoned by Ron De Rose (<email address hidden>) on branch: master
Review: https://review.openstack.org/296639
Reason: Will create a new patch for this.

[Lance Bragstad (lbragstad)]

Automatically unassigning due to inactivity.

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/570706

[Vishakha Agarwal (vishakha.agarwal)]

Hi,

Is anyone working over it?? AS I am willing to work over it.

Thanks.

[sonu (sonu-bhumca11)]

Hi Vishakha,

Yeah I am working on it. I pushed a patch before.
ref : https://review.openstack.org/#/c/570706/

I will update it soon,

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/633553

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/633757

[Colleen Murphy (krinkle)]

I don't think we can reasonably remove support for this without breaking the V3 API.

[Lance Bragstad (lbragstad)]

I had a discussion with @dstanek a couple years ago when we were enhancing the mapping engine to handle the auto-provisioning case. At that time, he had ideas about implementing a versioning mechanism specifically for mappings.

I don't really remember the details and I don't think they were written down, but that might be something to consider if we don't want to wait for v4.

[Colleen Murphy (krinkle)]

If we added versioning for the mapping API we would still need to support the old style mapping and so we still couldn't drop cruft from the backend.

In my opinion, continuing to support both explicitly local users as well as non-local "ephemeral"-but-not-really-anymore users has benefits, regardless of versioning problems.

[Lance Bragstad (lbragstad)]

I don't have a strong preference to remove ephemeral types. I think the reasoning in comment #12 is sufficient to close this out as Invalid or Won't Fix.

I'll let others weigh-in on the discussion if they feel differently before closing this out.

[Lance Bragstad (lbragstad)]

We discussed this at the team meeting today [0].

For now, I think it's safe to mark this as incomplete until we have a deep dive on the shadow user work and it intersects with this work.

[0] http://eavesdrop.openstack.org/meetings/keystone/2019/keystone.2019-02-05-16.01.log.html#l-77

[OpenStack Infra (hudson-openstack)]

Change abandoned by Vishakha Agarwal (<email address hidden>) on branch: master
Review: https://review.openstack.org/633757

[OpenStack Infra (hudson-openstack)]

Change abandoned by Vishakha Agarwal (<email address hidden>) on branch: master
Review: https://review.openstack.org/633553

[Launchpad Janitor (janitor)]

[Expired for OpenStack Identity (keystone) because there has been no activity for 60 days.]

[OpenStack Infra (hudson-openstack)]

Change abandoned by "Gage Hugo <email address hidden>" on branch: master
Review: https://review.opendev.org/c/openstack/keystone/+/570706
Reason: Abandoning since there hasn't been any recent activity, if anyone wants to continue this work, please feel free to restore this or create a new change.