15:00:53 #startmeeting keystone 15:00:53 Meeting started Wed Jan 17 15:00:53 2024 UTC and is due to finish in 60 minutes. The chair is d34dh0r53. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:53 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:53 The meeting name has been set to 'keystone' 15:01:05 #topic roll call 15:01:19 admiyo, bbobrov, crisloma, d34dh0r53, dpar, dstanek, hrybacki, knikolla[m], lbragstad, lwanderley, kmalloc, rodrigods, samueldmq, ruan_he, wxy, sonuk, vishakha, Ajay, rafaelwe, xek, gmann, zaitcev, reqa, dmendiza[m], mharley,jph 15:01:26 o/ 15:01:54 🙋 15:03:18 #topic liaison updates 15:03:28 nothing from VMT nor Release Management 15:04:42 #topic review past meeting work items 15:04:54 #link https://meetings.opendev.org/meetings/keystone/2024/keystone.2024-01-10-15.01.html 15:05:16 no updates, still a WIP to add these sections to the docs 15:05:25 #action d34dh0r53 Look into adding/restoring a known issues section to our documentation 15:05:34 #action d34dh0r53 add https://bugs.launchpad.net/keystone/+bug/1305950 to the known issues section of our documentation 15:05:53 #topic specification OAuth 2.0 (hiromu) 15:06:03 #link https://review.opendev.org/q/topic:bp%252Foauth2-client-credentials-ext 15:06:05 #link https://review.opendev.org/q/topic:bp%252Fenhance-oauth2-interoperability 15:06:07 External OAuth 2.0 Specification 15:06:09 #link https://review.opendev.org/c/openstack/keystone-specs/+/861554 15:06:11 OAuth 2.0 Implementation 15:06:13 #link https://review.opendev.org/q/topic:bp%252Fsupport-oauth2-mtls 15:06:15 OAuth 2.0 Documentation 15:06:17 #link https://review.opendev.org/c/openstack/keystone/+/838108 15:06:19 #link https://review.opendev.org/c/openstack/keystoneauth/+/838104 15:08:26 looks like hiromu isn't around today 15:08:28 moving on 15:08:46 #topic specification Secure RBAC (dmendiza[m]) 15:08:56 #link https://governance.openstack.org/tc/goals/selected/consistent-and-secure-rbac.html#z-release-timeline_ 15:08:58 2024.1 Release Timeline 15:09:00 Update oslo.policy in keystone to enforce_new_defaults=True 15:09:02 Update oslo.policy in keystone to enforce_scope=True 15:09:04 #link https://review.opendev.org/c/openstack/keystone/+/902730 15:09:06 #link https://review.opendev.org/c/openstack/keystone-tempest-plugin/+/903713 15:09:09 Still loking for reviews on the Phase 1 patch: ^^^^ 15:09:40 Working on the tempest test changes, but it's a bit messy since I have to reorganize some large classes. 😒 15:10:12 I'll review the phase 1 patch today, on my list of ToDo's 15:10:23 thanks d34dh0r53 15:12:27 np 15:12:32 next up 15:12:44 #topic specification Add schema version and support to "domain" attribute in mapping rules (gtema) 15:12:52 #link https://review.opendev.org/c/openstack/keystone-specs/+/748042 (merged) 15:12:54 #link https://review.opendev.org/c/openstack/keystone/+/739966 15:13:31 pls reviews 15:13:42 ack, I'll add that to my list 15:13:44 we are ready for next steps and need that change to be processed 15:13:50 thanks gtema 15:15:07 #topic open discussion 15:15:19 I don't have anything 15:16:56 i have 15:17:08 i would like to talk about project tags and access to them by various actors. Today project tags are there for project admins. They can change them, via a separate API or via a project update API. 15:17:41 We would like to have project tags, that would be writeable only by "system admins". It will allow our system admins to enable certain features for projects based on existence (or nonexistence) of the tags. 15:18:32 the tags are also company-specific and probably should not be done by adding new resource options 15:19:06 would adding tag prefixes or tag namespaces be an option? 15:19:26 would this change be even in scope of keystone, or should it rather be implemented elsewhere? 15:21:49 tags are not only writable by admins - there are clouds where domain admins can manage project tags 15:22:30 yes, that is true. But i would like to designate certain tags to be writeable only by system admins. 15:22:50 ahh, I see, a permissions flag for certain tags 15:25:16 dmendiza[m]: would this be possible with a custom policy? 15:27:58 Hmm... it would be tough to do, I think ... the policy engine works on an endpoint (URL) and can possibly use metadata about the object to make a decision. 15:28:28 Not sure what could be done in a tag to differentiate it from other tags 15:28:34 they only way to do it via policies is to delegate the decision to a remote server (via http: rule) 15:29:59 hmm, ok 15:30:50 another way to implement this would be via a new endpoint, something like /v3/projects/project_id/system-tags/, but it is a stretch, because it asks for a /domain-tags/ after that :) 15:31:10 right 15:31:49 I think this needs a spec as it's an API change as proposed here 15:32:24 honestly, i am not even sure that it should be in scope of keystone; maybe i should just go with yet another microservice 15:32:54 which is also not great, because i have to protect it with all the oslo stuff 15:34:49 could i ask if it would be possible to merge the fix here https://bugs.launchpad.net/keystone/+bug/2030061? 15:38:17 jrosser: I'll take a closer look at the fix for that later today, it does need tests though 15:38:59 ok so on that, andrew works on my team and we put a bunch of time to debug and make the bug / example fix 15:39:13 however we certainly do not have the skills to make a test for it 15:40:03 bbobrov: yeah, it's a grey area 15:42:11 jrosser: ack 15:42:49 moving on for the sake of time 15:43:17 bbobrov: it's up to you, we're certainly willing to look at a spec for adding that ability 15:43:25 #topic bug review 15:43:46 #link https://bugs.launchpad.net/keystone/?orderby=-id&start=0 15:44:02 looks like a new bug for keystone 15:44:05 #link https://bugs.launchpad.net/keystone/+bug/2049559 15:45:02 any volunteers to look at that one? 15:45:36 i think it was a deliberate decision to hide the error message from the user 15:46:05 I think so too, but I'd have to check 15:46:37 and that the reason is transmitted only via notifications 15:47:36 yeah, this is by design, just found the release note stating that 15:47:41 I'll update the bug 15:49:19 next up 15:49:30 #link https://bugs.launchpad.net/python-keystoneclient/?orderby=-id&start=0 15:49:40 no new bugs for python-keystoneclient 15:49:54 #link https://bugs.launchpad.net/keystoneauth/+bugs?orderby=-id&start=0 15:50:10 keystoneauth has no new issues 15:50:21 #link https://bugs.launchpad.net/keystonemiddleware/+bugs?orderby=-id&start=0 15:50:37 keystonemiddleware is also goo 15:50:39 d 15:50:50 #link https://bugs.launchpad.net/pycadf/+bugs?orderby=-id&start=0 15:51:02 nothing new in pycadf 15:51:13 #link https://bugs.launchpad.net/ldappool/+bugs?orderby=-id&start=0 15:51:24 ldappool is good to go 15:51:28 #topic conclusion 15:51:41 Thanks folks! 15:51:48 I don't have anything 15:52:14 #endmeeting