Tuesday, 2026-09-22

opendevreviewMerged openstack/election master: Exclude TC from PTL leaderless projects  https://review.opendev.org/c/openstack/election/+/100253107:29
opendevreviewMerged openstack/election master: [election-tool] Fix on tc/ptl announcement rendering  https://review.opendev.org/c/openstack/election/+/100606907:31
opendevreviewMerged openstack/election master: Update tc-election-summary.py: reflect 2027.1 results  https://review.opendev.org/c/openstack/election/+/100607307:31
opendevreviewElod Illes proposed openstack/openstack-manuals master: [www] Setup 2027.1 Indri and add project data to Hibiscus  https://review.opendev.org/c/openstack/openstack-manuals/+/100657708:51
croelandtzigo: so, do you think https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/CP7BTNWGYHSAKZ6HB4PBGPWPQU2LRKTZ/ should be discussed at PTG? I see nothing in https://etherpad.opendev.org/p/oct2026-ptg-os-tc yet, but I feel this issue (that already hit Debian) will hit everybody soon enough13:35
zigoYeah, we may talk about this indeed.13:42
zigoAlso about pkg_resources that's gone from setuptools >= 82, and that I had to patch everywhere !13:43
croelandtyeah that is the main issue with Paste*13:43
croelandtit seems "easy" to fork it but long term we might have to move awayfrom it13:43
croelandtzigo: do you think you will take part in the PTG and be there to discuss that?13:44
zigoI hope so. When's the PTG ?13:44
zigoOctobre, that I know.13:44
zigoHave you noticed that absolutely *ALL* XStatic packages are using pkg_resources ?13:45
croelandt12th to 16th13:45
croelandtI *think* the TC usually has sessions on Monday/Friday13:46
croelandtzigo: can you link to one of your patches for an Xstatic package?13:46
zigoSure.13:47
fungiit looks like xstatic itself relied on pkg_resources up until https://github.com/xstatic-py/xstatic/commit/391413a merged a year ago, but the latest release of xstatic (1.0.3) is 3 years old13:50
fungiso until there's a new xstatic release, we're stuck with requirements/constraints pointing at a version that relies on pkg_resources13:52
zigoLucky, the patches are almost all the time the same, and I was able to ask our local AI to write a script to automate fixing all of this. So it took me an afternoon for all of them.13:52
zigoFor XStatic, it's fairly easy to fix.13:52
zigoThough as fungi just wrote: it would need a release for all packages.13:53
fungisome of which we don't control13:53
zigozigo@k:zbuz>_ ~/salsa/openstack/xstatic $ ls | wc -l13:53
zigo4713:53
fungi(of note in this case, xstatic is not an openstack deliverable)13:53
zigoThat's not a small amount of packages we're talking about.13:53
zigoAlso to some worries: all of the Angular stuff must go, at some point.13:54
zigoIt has unadressed security issues.13:54
croelandtfungi: no commits in a year for xstatic13:55
croelandtI think for all these packages, we want to figure out whether they are maintained/active13:55
TheJuliazigo: ++13:55
croelandtand what we should do with them13:55
croelandtif Paste is no longer maintained and we want to move away, what should be the replacement? etc.13:55
croelandtsame for xstatic and angular13:56
fungimoving horizon off xstatic and angular is probably close to a from-scratch rewrite13:56
zigoWell, moving away from xstatic, I'm not sure. What will you replace it with? npm ? That's even worse !13:57
zigoI think it's ok to continue on XStatic, it has served us well so far.13:57
zigoJust Angular, they know about it, but it's a LOT of work.13:58
TheJuliaMaybe in a consistent way though?13:58
zigoThe only thing that could help would be throwing more hands at it.13:58
fungii have other concerns with xstatic, in that we've entrenched shipping vendored copies of javascript libraries we didn't create, in many cases old versions of them with known security vulnerabilities we don't/can't track13:59
TheJuliaVision perhaps may be first thing to reach before “throwing more hands at it”13:59
croelandtso maybe the web stuff should be a topic of its own13:59
croelandtPaste/PasteDeploy does affect ~20 projects and could be another topic14:00
fungithe idea was that downstream distributors would replace the vendored example versions with their own packaged securely-patched versions of those libs instead, but in practice that rarely happens14:00
zigofungi: I very much agree. It's even worse for Angular: I used to rely on a patched version in Debian, ie: libjs-angularjs, though it has been removed from Debian because no upstream. So I had to re-embbed and use the unpatched version from XStatic, otherwise Horizon would have been kicked away from Debian.14:01
croelandtzigo: so maybe this should be part of the Horizon PTG topics14:01
croelandtand we'd keep cross-project issues for the TC PTG?14:02
zigoFor Angular and XStatic ? Maybe. But the pkg_resources is not only about XStatic and Horizon though.14:02
croelandtyeah14:02
zigoThere's countless of that shit everywhere.14:02
croelandtso pkg_resources is about 20 projects14:02
croelandtthat definitely should be TC14:02
TheJuliaTc likely still needs to engage to convey across the captured issues and challenges it sees14:02
croelandtbut "hey Angular is one of the 83238 dead JS frameworks" might be more of a Horizon-only thing14:03
TheJuliaOtherwise horizon is going to think “don’t rock the boat, and keep on keeping on”14:03
* croelandt is just trying who should talk about what when and where14:03
TheJuliayes and ?everywhere?14:03
zigoMany sphinx projects:14:04
zigo- sphinxcontrib.apidoc14:04
zigo- sphinxcontrib-programoutput14:04
zigo- sphinxcontrib-actdiag14:04
fungizigo: i started trying to add warnings to the readme files, like you see at https://pypi.org/project/XStatic-JQuery-Migrate/ but that effort stalled and needs to be revived14:04
zigo- sphinxcontrib-blockdiag/14:04
zigo- sphinxcontrib-nwdiag14:04
zigo- sphinxcontrib-seqdiag14:04
croelandtsee, you're probably one of the most informed people on this topic :)14:07
croelandtso you should definitely have a bit of time to explain the issue at PTG14:07
zigoI wrote referenced all the patches I could find on my laptop's filesystem ... :)14:09
TheJuliaJust putting this out there, it wouldn't be bad for the TC to discuss web-ey stuff early on and try to assert vision. The reason I note that is we whipped up an python/npm hybrid adminy-web ui for Ironic, which has a specific base use case, "operator on the data center floor" or "operator in the NOC" sort of use case. And to attest to how advanced some AI's now are, another contributor raised just the idea to their prompt's to 14:19
TheJuliabrainstorm and it argued with them and made them a semi-working proof of concept as well. It would get us dangerously close to a perception of vibe coding because agents are so good at it, but its a direct human application layer interface, its not stable long lived APIs sort of interactions, so the overall use case and even human interaction modeling likely shouldn't tried to be kept as software for humans != software.14:19
opendevreviewGrzegorz Grasza proposed openstack/security-doc master: Add OSSN-0111: Keystone MFA enforcement gaps and replay weaknesses  https://review.opendev.org/c/openstack/security-doc/+/100674214:37
JayFI mean, we haven't even reached consensus in Ironic that it even fits within our scope. It might be early for the TC to get involved. 16:06
clarkbthere is already a baseline of expectations in the supported platform doc too (though I have no idea if they remain relevant in the js/web space)16:07
gouthamrtc-members: a gentle reminder that this week's IRC meeting will be hosted here in ~45 minutes16:15
gouthamragenda is here: https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee#Agenda16:15
TheJuliaJayF: Oh, yes, there is still discussion to have there, but there is a wider discussion regarding webapps there which the tc ought to have. I'm saying they ought to be mindful of the latter, ironic is just an example of where we've got one wip posted and someone else has another wip they have created which likely follows the python bits + npm stuffs for the actual user interaction.16:33
TheJuliaTo that point, the npm stuff at least got initially charted by ?mkrotscek? ages and ages and ages ago.16:34
JayFyeah, I think there is potentially room for that, but bluntly I'm a little more careful of gathering scope given how many teams we have near minimum levels of maintenance16:35
TheJuliaThe other aspect, is the world is evolving, its okay that it is but to say the status-quo with angular/xstatic is... not really working for the end user side of it16:35
TheJuliayeah, the reality is to address some things, we're going to have to become more comfortable with evolving forward because we can't really "just do nothing" or "try to keep the tech debt because that is easier", because in the end, its a huge cost16:36
fungibut also, openstack was built by the people who wanted to use it, unless that continues i don't think it can be sustainable16:37
funginaive recommendaitons to "evolve" aren't actionable, sadly16:38
TheJuliaBuilt by people who wanted to use it, but now we're in a diffrent age of usage where users are no longer creators/maintainers16:39
fungievolution is an uncontrolled reaction to changing conditions, so by definition we can't choose to evolve16:39
TheJuliaWell, I'm broadly trying to champion evolution becuase I've had people double down on just piling techdebt as the needful thing to do when in reality, its really not from a big picture standpoint.16:40
fungiwe will either evolve or perish, but evolution is messy and not designed16:40
TheJuliafair16:40
TheJuliaand very true16:40
JayFI kinda disagree with this framing?16:40
JayFWe're literally in the technical committee chat room -- we can't pretend like the community can't have directionality put on it16:41
JayFif I were to extend the metaphor I might say the TC are intelligent designers, but I really would prefer not16:41
fungiwhat we design isn't evolution though, they're different things. we will evolve, hopefully in ways compatible with what we design16:42
gouthamrthe TC can set an aspiration and endorse community initiatives... skyline is/was one such thing - didn't pan out because the biggest backer dropped out and its now truly someone's hobby work16:43
JayFgouthamr: that's my big concern, and in context of the stuff TheJulia has proposed in ironic, it's extra scary because that's being proposed in-tree16:44
gouthamr#TIL about ironic's web ui... is this something in the works?16:44
JayFgouthamr: ironic-ui is a horizon plugin that's existed a long time16:44
TheJuliaMaybe more to the point, there are efforts/thoughts/ideas on the periphery based upon use case challenges where existing tools/patterns may not match. That itself is evolution happening, the key is build consensus16:44
JayFgouthamr: I'd call it barely maintained16:44
TheJuliagouthamr: And there was also ironic-webui ages ago as a pure js api client assembly16:44
JayFThe openapi push, if we can get it across a finish line, will help with these kinda problems too16:45
fungidid/does bifrost also have a webui?16:45
TheJuliagouthamr: but use case wise, one thing doesn't always solve all use cases or patterns so we've got a few people going "hey, maybe we do something with a targeted use case"16:45
JayFfungi: bifrost is ansible in a dress, more or less16:45
fungii hope it's a sequined dress at least16:45
JayFfungi: so not really, but you COULD install horizon+ironic-ui and point it at ironic, assuming you deployed with keystone16:45
TheJuliaA frosty bridge of ansible16:45
JayFfungi: they're a bunch of sparkly red sequins that if you put under a microscope are fedora shaped (/s)16:45
fungiouch16:45
gouthamrack, which is the state with all of horizon unfortunately, because we've been limping along with old dependencies and are limping along while upstream packages keep going out of support. we could probably ask what the team's plans are - but likely they're to KTLO rather than an ambitious rewrite 16:46
TheJuliaBut to the point, digging through horizon is not the needful for DC ops who are doing inventory or breakfix activities either. We could, but do they really need to tie it in, and even then we would want to do it in a modern form so we can engage with the camera and do things like barcode/information capture16:46
TheJuliaso it boils down to different groups of users, and thats not a defect, that is an artifact of success16:47
TheJuliaJayF: hey now, bifrost got started at HPE ;)16:48
JayFTheJulia: to be clear: my questions are shaped "does this belong in tree in ironic?" (with my answer being "oh I hope not"), and "does this belong in openstack?" (with my primary concern being bus factor for contributors)16:48
JayFTheJulia: that was more a nod to the control over ansible ;)16:48
TheJuliaSo if anything, it whatever the hplinux logo was16:48
JayFif bifrost is ironic in a dress, the dress is ansible, and almost certainly you'd have to call that an RH design ;) 16:48
mnasiadkaI’d say one thing is what horizon team is doing, second thing is that mostly the $project teams have no clue what to do with their -ui deliverable and they just merge translation patches and try to make sure the ui plugin doesn’t get awfully broken cycle to cycle16:49
JayFmnasiadka: that's what I'm counting for "barely maintained" for ironic-ui16:49
ChristianBerendt[m]Hi, everyone. Just a quick test to see if the Matrix bridge is working properly on my site.16:51
clarkbI'm old enough to remember when you'd be shown the door immediately for taking a camera into the datacenter. Doesn't matter if that camera was on company issued hardware either16:51
TheJuliaI think part of the issue which gets dug to the surface is horizon sort of takes a different model on rbac and interactions, and we basically ignore that and I'm not sure how we *really* reconcile that back to reality as well or even give the dependency issues... where does one even really get started anyway. 16:51
fungiChristianBerendt[m]: it's working great!16:51
clarkbChristianBerendt[m]: yes your test message came through16:51
ChristianBerendt[m]Thanks16:51
gouthamrw00t, welcome, ChristianBerendt[m] 16:51
TheJuliaclarkb: blame vendors encoding passwords in 3d barcodes16:52
gouthamro.O16:52
JayFclarkb: I'm pretty sure $downstream here operates similarly :)16:52
fungiclarkb: i remember having to hunt high and low to find a cell phone that didn't include a camera so i wouldn't violate our data center policies16:52
JayFclarkb: the idea of using a mobile app to manage DC issues is mildly hilarious given our security model :P 16:52
JayFfungi: hate to tell you but I brought camera phones into HS more than a dozen times, but maybe they were nicer to customers ;) 16:53
JayF(you did work there, right? I'm not misremembering)16:53
clarkbto be clear I'm sure it is useful to some orgs. I'm just surprised given my background. We had hdd shredders accessible in the DC too because those were not allowed to exit the doors either16:53
fungiJayF: yes, though i think the camera policy there got relaxed in later years16:53
clarkbone door was a turnstile and the bay doors required multiple people to authorize opening (or you'd set off the fire alarms)16:54
TheJuliafungi: yeah, In my last DC job, that challenge is why we tossed the requirement. Cabinets kept locked/restricted access controls + a trusted dc ops team were good enough mitigating controls for our auditors. And... then when we needed photos for cooling warranty/repairs... it just made life easier.16:54
fungiit was also my job to maintain the iso/iec mapping for our facility security policy documents, so i probably followed them more closely than other staff members16:56
fungiat least if i didn't know about people violating policies, i didn't have to lie to the auditors16:57
TheJuliaheh, Similar role of meeting with all the auditors. I remember an auditor asking me about details of why and I think he gave up on the question realizing how stupid it was when I was at the 45 minutes of a process )16:58
clarkboutside of the DC was an entirely different story. We would occasionally dumpster dive the hardware dev locations and you'd often get your hands on a brand new unreleased CPU you could stick in a workstation that lived at your desk16:58
gouthamr(in 2 minutes, you folks will be interrupted)16:58
gouthamr:P sorry17:00
gouthamr#startmeeting tc17:00
opendevmeetMeeting started Tue Sep 22 17:00:46 2026 UTC and is due to finish in 60 minutes.  The chair is gouthamr. Information about MeetBot at http://wiki.debian.org/MeetBot.17:00
opendevmeetUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.17:00
opendevmeetThe meeting name has been set to 'tc'17:00
gouthamrWelcome to the weekly meeting of the OpenStack Technical Committee. A reminder that this meeting is held under the OpenInfra Code of Conduct available at https://openinfra.dev/legal/code-of-conduct.17:01
gouthamrToday's meeting agenda can be found at https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee#Agenda17:01
gouthamr#topic Roll Call17:01
mnasiadkao/17:01
ChristianBerendt[m]o/17:01
noonedeadpunko/17:01
gouthamrcourtesy-ping: spotz[m] frickler cardoe mnasiadka gtema 17:03
gouthamrgrr, sorry mnasiadka 17:03
cardoeo/17:03
mnasiadkagouthamr: no worries, I’m always late :)17:03
cardoeI'm always on time and then ADHD kicks in and I switch to another tab17:03
gtemaSorry, I am on PTO till end of the month17:04
noonedeadpunk:D same for me17:04
gouthamrah, ack both of you.. i thought summer's wrapped up here in the Northern Hemisphere :) 17:05
gouthamrbut that's some well deserved break from the release no doubt17:05
frickler\o17:05
gouthamrty for joining, let's get started.. 17:05
gouthamr(notes spotz[m] had mentioned being on a plane elsewhere)17:06
gouthamr#topic Last Week's Action Items17:06
gouthamrfirst off, i'd like to welcome ChristianBerendt[m] to the TC! thanks for signing up to serve, congratulations on your election17:06
ChristianBerendt[m]ty happy to be here17:07
gouthamri'd also like to welcome back frickler cardoe and gtema.. gtema returns to the TC after a year away.. 17:07
gouthamrthank you all for nominating yourself, and congratulations on your re-election17:08
gouthamr(missed dansmith in the courtesy ping)17:08
cardoeSpeaking of which... Great work this past cycle at chair gouthamr17:08
gouthamrty cardoe, also a good segue into that topic17:09
gouthamrwe're on the third business day since we merged the election results per my math17:09
* cardoe put the ball on the t-ball stand for gouthamr 17:09
gouthamr:P17:09
gouthamrif anyone's interested to nominate themselves, today's the final day.. 17:10
gouthamrget your nomination into this folder:17:10
gouthamr#link https://opendev.org/openstack/governance/src/branch/master/reference/tc-chair-candidates/2027.1 17:10
gouthamr#link https://governance.openstack.org/tc/reference/tc-chair-elections.html 17:11
cardoewhere's your name in there? :D17:11
mnasiadkaI think cardoe wanted to ask - gouthamr are you willing to continue?17:12
gouthamri was hoping someone would volunteer :)17:13
* noonedeadpunk does not see much hands17:13
gouthamrhah :) if you're still thinking, take your time.. 17:14
gouthamrsince we'll be decision-pending, we'll hash out the rest of the tc-chair election process outside this meeting... we can discuss on this channel and gerrit. i'm hoping we'll get folks running/winning/nominated asap17:14
gouthamrwe've got some work to do when we have a chair, but do think through a liaison role or two if you'd like17:15
gouthamrwe need a VMT liaison17:15
gouthamrall the existing DPL "tc-liaison"s are on the TC17:16
gouthamrso there's no need to replace them, unless there is interest to do so17:16
gouthamr#link https://opendev.org/openstack/governance/src/branch/master/reference/projects.yaml (search "tc-liaison")17:17
gouthamranother liaison role is here:17:18
gouthamr#link https://opendev.org/openstack/governance/src/branch/master/reference/popup-teams.rst (search "TC Liaison") 17:18
gouthamrfor this though, the liaison doens't have to be a TC member17:18
gouthamri think with that, we're done talking about elections17:19
gouthamrand related action items17:19
gouthamroh, i forgot, we also need an election team liaison, and if we can help, it should be someone other than spotz[m] 17:20
gouthamrwe can nominate all of these next week17:21
gouthamras we close this out, let me thank ianychoi, slaweq, spotz[m] and fungi for all the great job on the 2027.1 elections!17:21
fungi<- supporting cast17:22
gouthamrwe had an action item regarding recapping security work that happened during this release17:23
gouthamrJayF took an action item yesterday to share a blurb with aprice[m] on behalf of the VMT, and I am working on a list of contributors (bug reporters, fixers, reviewers) that contributed. fungi is working on the contributor list for this release, and i was suggesting incorporating this data into that17:24
fungiyeah, i spoke with aprice about it today too17:25
gouthamr++17:25
fungiwe're still working out how to incorporate it, but should be able to include that in some form17:25
gouthamrty fungi 17:25
fungi(the security contributors thanks list i mean)17:25
gouthamri don't know if we saw any other project team specific highlights in the meantime after our call last week.. but, the message we'll be sharing across these should capture the mood 17:25
gouthamrfungi was sharing some of the release contribution stats here earlier, and you can poke him if you have questions17:26
fungiyeah, i was asked by the staff to pinpoint what parts of openstack the recent increases in contributor and merged change counts came from, so working on that now17:27
gouthamrgreat17:28
gouthamrfor a release recap, tune into an episode of OpenInfra Live's new podcast that look at this through the lens of the project maintainers and the community leads on Oct 1st: 17:28
gouthamr#link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/22VSNDEOVFJVALUOJTDUJ4Q63K3FLSVU/17:28
gouthamrother than that:17:30
gouthamrI did notice some action on a past AI:17:30
gouthamr#link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/TZ3B7CXWJHE7T5UVIS6RUC4TL6QQ2QLP/ ([tc][barbican][horizon] Reviving barbican-ui: proposing a dedicated core reviewer group)17:30
gouthamrmharley[m]: ty for starting that thread; noonedeadpunk was any TC action required on this?17:30
gouthamrguess not, your questions were relevant, ty for posing them.. and you'll probably brainstorm on the biweekly meeting of the barbican team next week17:33
gouthamr#link https://etherpad.opendev.org/p/barbican-weekly-meeting (barbican meeting on Sep 28) 17:33
gouthamralright, that's a wrap on action items... anything else to mention here?17:33
noonedeadpunkI don't think it was needed at the moment17:34
noonedeadpunk(TC attention)17:34
gouthamr++17:34
gouthamr#topic TC PTG17:35
gouthamrit's time to plan our agenda for the Indri/2027.1 PTG (Oct 12-16 2026)17:36
gouthamrthe schedule website is live as of today:17:36
gouthamr#link https://ptg.opendev.org/ptg.html (PTG Schedule)17:36
gouthamr#link https://etherpad.opendev.org/p/oct2026-ptg-os-tc (TC PTG Etherpad)17:37
gouthamr^ please add your topics there between now and early Oct 17:37
gouthamrwe'll try and schedule our sessions as we've done in the past - beginning and end of the PTG period so we can seed discussions and reflect on what project teams, SIGs, pop up teams, working groups have done through the week17:39
gouthamrif there are any other ideas, let me know. we do have a good amount of time to inform and draw audiences that we'd like17:39
gouthamrfor example, operators. In the last ptg, we tried a session with the Operators' Radio Hour folks... i was hoping that project teams could find themes there to drive into their own PTG sessions.. think we should encourage that again - it'll need some organizing from PTLs/liaisons17:41
gouthamri'll suggest this on the ML17:42
cardoeyeah good idea17:42
gouthamrfeel free to use the etherpad as a placeholder while you gather thoughts too ^17:43
ildikovgouthamr: I'm about to send an email to openstack-discuss to collect questions and discussions topics for Ops discussions at the PTG17:44
gouthamras usual, we'll remind project organizers to be mindful and schedule cross project sessions at the beginning of the week17:44
gouthamrildikov: nice! would you have a common ops hour like last time? lmk if you see any cross-cutting themes to push into the TC's time17:45
gouthamranything else about the PTG for this week?17:46
ildikovgouthamr: I've been chatting with Chris Morgan about this and also brought up the question on the OpenStack Ops Matrix channel. The current thinking is to decide based on what topics people want to discuss. As if there are more questions specific to particular project teams, then it might be easier to schedule time with them during their sessions.17:46
gouthamr++17:46
ildikovHaving an Ops Radio Hour session/common Ops session is also an option17:47
ildikovor folks might end up with the combination of the above17:47
mnasiadkaKolla is only planning to schedule an operator hour this PTG - but happy to schedule something together with Ops community instead/in addition to17:48
ildikovIn either case, input from the developer community on what topics/questions they're interested in would be very helpful and highly appreciated!17:48
ildikovmnasiadka: noted, thank you for pointing that out!17:48
gouthamryeah, it might need some prep to bring something to a project team directly - a bug, an ML discussion or a well articulated topic in their etherpad is welcome..17:49
gouthamri picked up some feedback from nova maintainers last cycle - they were hoping for concrete takeaways.. i think its someone's burden to connect people when teams expect folks to just come to them with a flushed out idea 17:49
gouthamrs/idea/problem17:50
ildikovyeah, we'll be doing our best to circulate the questions and topics in advance to the PTG17:51
gouthamr++17:51
ildikovI think some of the disconnect is the scope, operators see OpenStack clouds as a whole, and not every challenge can be mapped to a single OpenStack project/service17:51
ildikovwhile teams would like to understand how to make their projects better specifically17:52
gouthamrack, that's indeed where the TC can help17:52
ildikovso I hope a bit more time spent on prepping will help with bridging that17:52
ildikovgouthamr: that would be great!17:52
gouthamrack, and ty for that coordination.. its so much hard work and feels like herding cats, but its very valuable eventually and i'm very glad to see your energy for it :) 17:53
gouthamras the ideas fill up, please use this channel to brainstorm - we'll help with organizing the topics17:54
ildikovOf course! I'm happy to help and please always feel encouraged to give me feedback and ideas on what we could/should try  next to improve things, including communication.17:54
gouthamrty, anything else for $topic?17:55
gouthamri'll skip the "Continuing the discussion on Keystone situation" topic - it's being bounced across meetings, and is probably good to keep simmering on this so we can bring it to the PTG (i've added the topic there)17:56
gouthamr#topic A check on gate health17:56
mnasiadkaI don’t know if the earlier horizon plugins discussion should land in the TC slots or Horizon slots, but with my Magnum PTL hat I’d assume that it’s important to think about managing that properly :)17:56
gouthamrack, please do add it to the etherpad mnasiadka, tmazur will probably help refine thoughts for it17:57
clarkbover the weekend one of the zuul executors filled its root disk which caused it to fail jobs before they even started in a tight loop. YOu may haev noticed this as jobs failing with RETRY_LIMIT statuses. I pulled the executor out of the cluster and it has since been cleaned up and readded. Anyway if you see RETRY_LIMITS from late firday/early saturday pacific time then you can17:57
clarkbrecheck17:57
gouthamr++ ty cardoe 17:57
gouthamrgrr17:58
gouthamrtab complete, ty clarkb 17:58
cardoeI'll take free credit. :D17:58
gouthamrmaybe it was an executor on Rackspace :) 17:58
gouthamr#link https://review.opendev.org/q/topic:%222027-1-testing-runtime%22 (2027.1 runtime update) 17:59
clarkbit was but rackspace wasn't at fault. It appears that podman can leak layers in its storage system17:59
clarkband for whatever reason this particular executor leaked enough layers for it to become a problem17:59
clarkbwe're going to mitigate by resetting all of the podman system state during the next weekend restarts17:59
gouthamrouch, if there's a podman bug you'd raise, let me know.. 18:00
gouthamr^ gmaan hope this runtime update is going well - this would mean that we start seeing py3.14 jobs be voting on master 18:01
gouthamrwe're a bit over, any other notes for the minutes today?18:01
gmaanyeah, I removed the WIP from this and it is good to go https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/100170418:02
gouthamr++18:02
gouthamrty for working on that18:02
mnasiadkaSo off it goes, I’ll ignore the super-nit ;-)18:02
gouthamrhaha, it's in bold, and hard to ignore18:02
gouthamrlet's wrap it up here.. ty all for attending!18:03
gouthamr#endmeeting18:03
opendevmeetMeeting ended Tue Sep 22 18:03:10 2026 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)18:03
opendevmeetMinutes:        https://meetings.opendev.org/meetings/tc/2026/tc.2026-09-22-17.00.html18:03
opendevmeetMinutes (text): https://meetings.opendev.org/meetings/tc/2026/tc.2026-09-22-17.00.txt18:03
opendevmeetLog:            https://meetings.opendev.org/meetings/tc/2026/tc.2026-09-22-17.00.log.html18:03
mnasiadkaTell that to my Asperger ;-)18:03
dansmithgouthamr: sorry forgot to tell you I was out, my bad18:49
gouthamrdansmith: ack, np!18:51
ildikovgouthamr: mnasiadka: I just sent out my email to openstack-discuss on the Ops sessions. Sharing the etherpad to collect sessions here too for additional visibility: https://etherpad.opendev.org/p/ops-devs-qanda-202620:01
ildikovIt has instructions on how to "fill it out", but please let me know if any of it reads unclear20:01
gouthamrty ildikov, format seems fine.. i'll keep checking the pad22:06
-opendevstatus- NOTICE: The Gerrit service on review.opendev.org will be offline momentarily while we upgrade to a new patch release, but should return within a few minutes23:02
ildikovgouthamr: awesome, thank you!23:36

Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!