Wednesday, 2026-02-25

d34dh0r53#startmeeting keystone15:02
opendevmeetMeeting started Wed Feb 25 15:02:02 2026 UTC and is due to finish in 60 minutes.  The chair is d34dh0r53. Information about MeetBot at http://wiki.debian.org/MeetBot.15:02
opendevmeetUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.15:02
opendevmeetThe meeting name has been set to 'keystone'15:02
d34dh0r53Reminder: This meeting takes place under the OpenInfra Foundation Code of Conduct15:03
d34dh0r53#link https://openinfra.dev/legal/code-of-conduct15:03
d34dh0r53#topic roll call15:04
d34dh0r53admiyo, bbobrov, crisloma, d34dh0r53, dpar, dstanek, hrybacki, lbragstad, lwanderley, kmalloc, rodrigods, samueldmq, ruan_he, wxy, sonuk, vishakha, Ajay, rafaelwe, xek, gmann, zaitcev, reqa, dmendiza[m], dmendiza, mharley, jph, gtema, cardoe, deydra15:04
gtemao/15:04
d34dh0r53hi gtema , good pto?15:04
gtemastill there ;-)15:05
gtemausing time to work on keystone-rs15:05
d34dh0r53nice15:05
d34dh0r53#topic review past meeting work items15:06
d34dh0r53#link https://meetings.opendev.org/meetings/keystone/2026/keystone.2026-02-11-15.05.html15:06
d34dh0r53nothing to review from the last meeting15:06
d34dh0r53#topic liaison updates15:06
d34dh0r53no updates from me15:06
gtemanothing from me15:07
d34dh0r53#topic specification Secure RBAC (dmendiza)15:08
d34dh0r53#link https://governance.openstack.org/tc/goals/selected/consistent-and-secure-rbac.html#z-release-timeline_15:08
d34dh0r532026.1 Release Timeline15:08
d34dh0r53Update oslo.policy in keystone to enforce_new_defaults=True15:08
d34dh0r53Update oslo.policy in keystone to enforce_scope=True15:08
d34dh0r53I just saw dmendiza a minute ago15:09
d34dh0r53guess he's not around15:10
d34dh0r53#topic specification Secuirty Compliance Testing (dmendiza)15:10
d34dh0r53#link https://review.opendev.org/c/openstack/devstack/+/95796915:10
dmendiza[m]🙋‍♂️15:10
d34dh0r53👋15:10
dmendiza[m]Sorry, was afk15:11
dmendiza[m]Right, so RE: SRBAC, looks like the functional tests are non-voting15:11
dmendiza[m]which sucks, because I broke some stuff  not too long ago and didn't notice because non-voting == non-caring15:11
gtemaevery day something is broken, who cares ;-)15:12
dmendiza[m]We also have some duplication in SRBAC flags for testing15:12
dmendiza[m]I submitted a few patches but things are failing all over the place15:13
dmendiza[m]#link https://review.opendev.org/c/openstack/keystone/+/97670915:13
dmendiza[m]#link https://review.opendev.org/c/openstack/keystone-tempest-plugin/+/97671015:13
dmendiza[m]For that second patch, note that SRBAC test failures are currently at 36 failures15:14
dmendiza[m]Anyway, I need to find some time to fix this stuff15:14
dmendiza[m]There was also a patch from gmaan for adding SRBAC unit tests as well?15:14
dmendiza[m]let me see if I can find it15:15
dmendiza[m]#link https://review.opendev.org/c/openstack/keystone/+/88643415:15
dmendiza[m]also failing15:15
d34dh0r53I just workflow'd the keystone-tempest-plugin patch15:15
dmendiza[m]Dave Wilde (d34dh0r53): I appreciate it, but it won't go anywhere because the Depends-On patch is failing.15:16
d34dh0r53yep, I know, but at least it's done for when the other one passes15:16
gtemaand why is this protection job now failing?15:18
gtema"AttributeError: module 'keystone.common.policies.base' has no attribute 'RULE_SYSTEM_READER_OR_OWNER'" - was there something renamed?15:18
dmendiza[m]Possibly?  Like I mentioned earlier, the protection job is non-voting so who knows how long it has been failing --- changes like that would not have been caught15:19
gtemaack15:20
d34dh0r53#link https://review.opendev.org/c/openstack/keystone/+/76047815:21
d34dh0r535 years :o15:21
dmendiza[m]heh15:22
d34dh0r53I always though that protection-functional was a superset, but I guess that's not the case15:22
bbobrovi think that's about when they stopped working on openstack15:23
gtemaI am so terribly frustrated over all the different jobs and those different assumptions that for me personally it does not worth of any use - full rewrite is necessary. Everything is leaking heavily15:23
d34dh0r53yeah, there's 10+ years of cruft15:25
dmendiza[m]Hoping to get some upstream time on Friday, so I'll keep digging at this15:26
dmendiza[m]For the Security Compliance testing -- no progress since last week.15:27
d34dh0r53thanks dmendiza 15:27
d34dh0r53#topic keystone-rs15:27
d34dh0r53#link https://github.com/openstack-experimental/keystone15:28
gtemacardoe should be happy, I am finishing implementation for the k8s token review auth15:28
gtemaI have also implemented the raft based distributed storage that would help making things much more smooth15:29
gtemaalso now ensuring local development with the k8s is usable with skaffold - let the tests run in the real k815:30
gtemathey are also executed in the CI same way15:30
gtemamy zero-trust concept passed the internal review in the company with people saying - go for it15:31
gtemaand the k8 auth makes pretty much sense in combination with grpc and mtls communication between services15:32
gtemathat's it so far15:32
d34dh0r53wow, that sounds awesome, maybe I'll try to compile it and play around with it on Friday15:33
d34dh0r53thanks gtema 15:33
d34dh0r53#topic open discussion15:33
bbobrovI would like to get your attention on https://bugs.launchpad.net/keystone/+bug/2140492 and the 2 patches related to it15:34
bbobrovRight now a user can create an app cred with secret "1" or "securesecret" right from the docs. Compliance people are not happy.15:34
bbobrovCompliance people are also unhappy that i cannot force everyone to rotate the secrets.15:34
bbobrovThe patches add 2 opt-in settings to fix this. Feedback welcome.15:35
gtemaack15:35
cardoegtema: awesome news15:35
cardoeI haven't had a chance to run skyline against keystone with some more debugging (my backlog is so long right now... :/)15:36
cardoeI just want to get to a place with bbobrov and gtema where we can merge the fix to that keystone function so that its impossible for it to return None when the code base requires it not to return None.15:36
gtemaas pretty much for everyone ;-) That's why I took PTO to work without stress on keystone-rs15:37
cardoegtema: as you know that code wouldn't compile if it was written in Rust since its possible to return a None when its defined as never returning None.15:37
gtemasure - all for rust - XD15:37
bbobrovcardoe: for your issue the most important thing is to have it easily reproduced in keystone only, without skyline15:38
gtemaand with python we only see breakages now DAILY15:38
cardoeYeah I gotta figure out what the payload that skyline is sending to keystone.15:38
cardoeCause it'll happen with a curl call once I figure out that payload.15:39
bbobrovif you manage to get the actual token, you could cheat and decode it with the fernet keys from keystone to look at the payload15:39
gtemaand it is not going to be very easy, since it passes the mod_auth_oidc and mapping engine first15:39
bbobrovor there is no fernet token yet?15:39
gtemait's not that skyline itself sends some trash15:39
gtemanope, from what I understand this happens in the authentication/re-authentication15:39
gtemawhen user switches the project scope or something similar - don't know the detail for sure, but the failure is exactly in the authentication code15:40
gtemaat least it looks to me like that15:40
cardoeyeah so basically if I authenticate with skyline and click the drop down to pick a project... keystone throws a 50015:42
cardoeskyline has a python backend API service that's doing some extra queries that the user might not have access to.15:43
gtemahmm, could it be the /v3/auth/projects call then?15:43
cardoeYeah15:43
cardoeUnfortunately I need to flog the real people that are seeing this and having it happen engage rather than proxying through me.15:44
gtemawell, we should get some debug log including what the initial http call exactly was15:44
cardoeI'll get you guys the debug logs.15:44
gtemathanks15:45
d34dh0r53👍️15:45
d34dh0r53let's move on to bug triage15:45
d34dh0r53#topic bug review15:45
d34dh0r53#link https://bugs.launchpad.net/keystone/?orderby=-id&start=015:45
d34dh0r53first one is sunbeam15:46
d34dh0r53#link https://bugs.launchpad.net/keystone/+bug/214057915:46
gtemaehm, why do we have to deal with it?15:47
d34dh0r53I was just thinking that this looks like an ubuntu bug15:47
d34dh0r53marked it as invalid for keystone15:48
d34dh0r53next up15:48
d34dh0r53#link https://bugs.launchpad.net/keystone/+bug/214171315:48
d34dh0r53#undo15:48
opendevmeetRemoving item from minutes: #link https://bugs.launchpad.net/keystone/+bug/214171315:48
d34dh0r53nevermind, that's flagged security, I'll review15:49
d34dh0r53that's it for keystone15:50
d34dh0r53#link https://bugs.launchpad.net/python-keystoneclient/?orderby=-id&start=015:50
d34dh0r53nothing new in python-keystoneclient15:50
d34dh0r53#link https://bugs.launchpad.net/keystoneauth/+bugs?orderby=-id&start=015:50
d34dh0r53keystoneauth is good15:50
d34dh0r53#link https://bugs.launchpad.net/keystonemiddleware/+bugs?orderby=-id&start=015:50
d34dh0r53no new bugs in keystonemiddleware15:51
d34dh0r53#link https://bugs.launchpad.net/pycadf/+bugs?orderby=-id&start=015:51
d34dh0r53pycadf is good15:51
d34dh0r53#link https://bugs.launchpad.net/ldappool/+bugs?orderby=-id&start=015:51
d34dh0r53nothing new in ldappool15:51
d34dh0r53#topic conclusion15:51
d34dh0r53thanks everyone, that's all I have15:51
gtemaare you going to join Friday?15:52
d34dh0r53yes15:53
gtemagood15:53
d34dh0r53see you then15:53
gtemayupp, see you. Thanks15:54
d34dh0r53#endmeeting15:54
opendevmeetMeeting ended Wed Feb 25 15:54:23 2026 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)15:54
opendevmeetMinutes:        https://meetings.opendev.org/meetings/keystone/2026/keystone.2026-02-25-15.02.html15:54
opendevmeetMinutes (text): https://meetings.opendev.org/meetings/keystone/2026/keystone.2026-02-25-15.02.txt15:54
opendevmeetLog:            https://meetings.opendev.org/meetings/keystone/2026/keystone.2026-02-25-15.02.log.html15:54
*** croeland1 is now known as croelandt22:39

Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!