| spotz[m] | Ok just catching up here, I reported the emails and system tasks issues yesterday morning. I had clicked the renew membership before the 30 days and had verified matching emails in LFX prior to yesterday. I know a few people who got the you need the LFX account email, there was no deadline for when it had to be done by though | 14:58 |
|---|---|---|
| fungi | the specific problem is that lfx kept crashing partway into loading the electorate, so each time wes tried it (unbeknownst to him) sent ballots to the first ~thousand voters in the rolls before crashing | 15:00 |
| spotz[m] | Ahh that makes sense | 15:00 |
| fungi | that's why some people had multiple ballots/vote buttons (once for each attempt) while others had none (they weren't in the first thousand) | 15:00 |
| fungi | for example we confirmed that bauzas should get a ballot but was not because he was closer to the end of the list, so once lf gets the bug fixed he should get one | 15:02 |
| bauzas | hah, glad to hear I was on the end of the list, when I was young, every morning when arriving on the class, I was afraid to be late in case the teacher was already asking who is there and was listing by the surname :) | 15:04 |
| bauzas | I'd have preferred to have a surname like "zauba" :) | 15:04 |
| fungi | yeah, i have no idea how these are sequenced | 15:06 |
| gouthamr | tc-members: a gentle reminder that our weekly IRC meeting will happen here in ~50 minutes. Please find the agenda here: https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee | 16:09 |
| clarkb | bauzas: the new ballot emails should be out nwo and they believe it was successful this time around if you want to check | 16:56 |
| noonedeadpunk | it looks fine this time | 16:58 |
| * gouthamr doesn't notice a difference :P | 16:59 | |
| noonedeadpunk | I didn't got 5 ballots :D | 16:59 |
| gouthamr | ah haha, i got one ballot and then much later a bunch of emails last time | 16:59 |
| noonedeadpunk | I got ballots to amount of emails in LF page | 17:00 |
| noonedeadpunk | each was votable | 17:00 |
| bauzas | clarkb: thanks, looking | 17:00 |
| gouthamr | ah i see | 17:01 |
| gouthamr | meeting time | 17:01 |
| gouthamr | #startmeeting tc | 17:01 |
| opendevmeet | Meeting started Tue Jan 13 17:01:11 2026 UTC and is due to finish in 60 minutes. The chair is gouthamr. Information about MeetBot at http://wiki.debian.org/MeetBot. | 17:01 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 17:01 |
| opendevmeet | The meeting name has been set to 'tc' | 17:01 |
| bauzas | clarkb: yay, got it \o/ | 17:01 |
| spotz[m] | I only got one ballot and only one task in the system | 17:01 |
| gouthamr | Happy New Year! Welcome 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 |
| gouthamr | Today's meeting agenda can be found at https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee | 17:01 |
| gouthamr | #topic Roll Call | 17:01 |
| noonedeadpunk | o/ | 17:01 |
| spotz[m] | o/ | 17:01 |
| gtema | o/ | 17:02 |
| gouthamr | courtesy-ping: frickler, cardoe, mnasiadka, bauzas | 17:03 |
| gouthamr | noted absence: t o n y b | 17:03 |
| cardoe | o/ | 17:03 |
| bauzas | o/ | 17:03 |
| bauzas | (still here :) ) | 17:03 |
| bauzas | happy new year to everyone | 17:03 |
| mnasiadka | o/ | 17:03 |
| spotz[m] | Yeah I was going to say bauzas and frickler have been here:) | 17:03 |
| bauzas | (well, got water meter leakage and power outage on the same first week of 2026 but I'm pretty sure the rest of the 2026 days will be then better :) ) | 17:04 |
| frickler | o/ | 17:04 |
| gouthamr | bauzas: wow, you're done with the worst | 17:04 |
| bauzas | (and gastro too) | 17:04 |
| gouthamr | alright, lets get started.. | 17:05 |
| gouthamr | Welcome back everyone, it's been a long time without meetings.. and understandably so, there were a bunch of holidays. I hope everyone had some downtime through it all | 17:05 |
| gouthamr | #topic Past AIs | 17:05 |
| gouthamr | we've got to rehash a bunch of context to get to these :) | 17:06 |
| gouthamr | we took one regarding updates to the technical vision document. i plan to get cracking on this soon, will add it to the tracker so we can monitor progress | 17:07 |
| gouthamr | i also reached out to jimmymcarthur and allison regarding to request a deadline extension for submitting user survey changes | 17:07 |
| gouthamr | i didn't hear back, i'll poke them again.. at the very least, some of the questions or answer choices were outdated for the TC and project teams.. it'd require some time from multiple folks/teams | 17:09 |
| gouthamr | we had an AI regarding AlmaLinux package support for OVS on non-v3 processors | 17:10 |
| gouthamr | wasn't sure if anyone was pursuing this through the holidays, or if it needs a newer push | 17:10 |
| spotz[m] | I'm pretty sure Jonathan was on PTO when this first came up | 17:11 |
| noonedeadpunk | we were trying to add it to CI in OSA at least | 17:12 |
| noonedeadpunk | re | 17:12 |
| mnasiadka | noonedeadpunk: are there any NFV SIG rebuilds done for Alma? | 17:12 |
| gouthamr | ack, i think mnasiadka had a workaround in the works: https://review.opendev.org/c/openstack/devstack/+/970281 | 17:13 |
| gouthamr | this is for devstack only, ofc.. just a stop gap | 17:13 |
| noonedeadpunk | mnasiadka: no, not for v2 | 17:13 |
| noonedeadpunk | for v3 default ones were working I think | 17:14 |
| mnasiadka | Yup, we’d need somebody to rebuild and maintain ovs/ovn to make it sustainable | 17:14 |
| bauzas | noonedeadpunk: do we plan to have specific nodepools dedicated to Alma or would that be shared ? | 17:14 |
| mnasiadka | (For v2) | 17:14 |
| noonedeadpunk | and without v2 I'd agree it makes limited sense to add one more EL distro to images | 17:14 |
| noonedeadpunk | bauzas: so idea was that we can spawn alma across the board as there should not have been requirements | 17:15 |
| noonedeadpunk | *hardware requirements | 17:15 |
| bauzas | do we plan to supersede some of the jobs we have and run them on top of alma ? | 17:15 |
| noonedeadpunk | I kinda wonder if it makes sense to build ovs/ovn from source.... | 17:15 |
| bauzas | or would that be additional jobs ? | 17:15 |
| noonedeadpunk | in terms of execution time | 17:16 |
| mnasiadka | bauzas: the effort to add alma is purely to support x86-64-v2 which we have a lot in OpenDev resources - it can use the older HW generations of providers (I think like RAX non-flex). I think we didn’t have a decision if that should replace CentOS Stream jobs that are running on v3 - but first we would need that working in a sustainable model (currently devstack jobs compile ovs and ovn on runtime) | 17:16 |
| mnasiadka | Do we have any connection to the Alma folks if they are interested in helping? I’m happy to fire off an email - I just need addresses. | 17:17 |
| bauzas | mnasiadka: I understand and appreciate the need to move to alma for some of the old hardware we have, hence my question | 17:17 |
| noonedeadpunk | huh, kinda don't see much difference between rocky and alma even though ovn is built | 17:17 |
| bauzas | I just don't want to run more jobs on those legacy nodes if they're already at capacity | 17:17 |
| noonedeadpunk | https://zuul.opendev.org/t/openstack/runtime?job_name=devstack-platform-almalinux-purple-lion-ovn-source&project=openstack/devstack&branch=master&pipeline=check | 17:17 |
| noonedeadpunk | https://zuul.opendev.org/t/openstack/runtime?job_name=devstack-platform-almalinux-purple-lion-ovn-source&project=openstack/devstack&branch=master&pipeline=check | 17:17 |
| noonedeadpunk | https://zuul.opendev.org/t/openstack/runtime?job_name=devstack-platform-rocky-red-quartz&project=openstack/devstack&branch=master&pipeline=check | 17:17 |
| clarkb | right we already had centos 10 stream and rocky linux 10. But both require running on x86-64-v3 which our largest source of resources doesn't support. Alma has v2 packages for the base os so thei dea was we could leverage that to effectively double our quota for EL testing | 17:18 |
| clarkb | but no there isn't any special pool for anything we're just selecting locations based on the ability to boot on the hardware | 17:18 |
| fungi | what we didn't count on is that for some packages we need, they don't build their own and instead reuse the ones from centos | 17:19 |
| gouthamr | ack, ping jonathanspw for awareness that we need help, but if he isn't looking, perhaps spotz[m] can connect him here again? | 17:19 |
| bauzas | clarkb: thanks for the explanation | 17:19 |
| spotz[m] | I'll see him in 2 weeks if worst comes to worse | 17:19 |
| noonedeadpunk | I think our request/intention would be to have various SIGs to be existing for v2 | 17:20 |
| gouthamr | ack, we can check back in regarding this | 17:21 |
| gouthamr | we had a last AI regarding unmaintained branches.. i think specifically if we could sunset unmaintained/zed globally | 17:21 |
| * gouthamr tries to dig up activity regarding this | 17:22 | |
| gouthamr | we've acted on the branch cleanup until zed: | 17:23 |
| gouthamr | #link https://review.opendev.org/q/topic:%22unmaintained-eol%22 | 17:23 |
| fungi | zed was the bridge between our slurp "dress rehearsal" (yoga) and and our first real slurp release, since we don't do unmaintained branches for not-slurp, the eol of yoga implies there's probably limited reason to hold onto zed | 17:23 |
| gouthamr | ^ yes, poking elodilles to see if this is a good time/reason to clean that up as well | 17:24 |
| frickler | yoga isn't eoled yet | 17:24 |
| gouthamr | ah, here's a thread on the ML for reference: | 17:25 |
| gouthamr | #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/N24CYAJMLVJMAJMUE4RA5U7KOL7LABOW/#BN7QIMBV6VPXKZS3TXBZ5SCNYPLPZZSN ([stable][unmaintained] Move old Unmaintained branches to EOL) | 17:25 |
| gouthamr | we can follow up on this async too.. nothing concerning at the moment, it's still the same situation just slightly better with CI and resources being less and less burdened with this cleanup | 17:26 |
| fungi | i did get clarification from elodilles in irc that he meant through yoga | 17:27 |
| gouthamr | ah, no patch for it yet.. but it's coming? | 17:27 |
| fungi | rather than just through xena | 17:27 |
| fungi | yeah, he's going branch at a time | 17:27 |
| gouthamr | ++ | 17:27 |
| fungi | working his way forward | 17:28 |
| gouthamr | that's all the AIs we were tracking through the break... a reminder that next week, we'll use our APAC friendly meeting time.. since many of you are here, please do add this to your calendars :) | 17:29 |
| gouthamr | mnasiadka will be hosting it, and it's at 0800 UTC instead of 1700 UTC | 17:29 |
| bauzas | good reminder | 17:29 |
| gouthamr | #topic While we were away | 17:29 |
| gouthamr | being out for a few weeks and coming back to a pile of emails :) | 17:30 |
| gouthamr | #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/J54CRB2OYIWPHSD55FFZGFYZLJWFE6BF/ | 17:30 |
| gouthamr | (H' release name is Hibiscus) | 17:30 |
| spotz[m] | We need flowers now | 17:30 |
| gouthamr | ^ following up on that, allison shared some foundation marketing themes: | 17:31 |
| gouthamr | #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/CLCQWVDFM3NIHGHOADVCZVJNXEBVGXE3/ (Marketing/Messaging Themes for Hibiscus) | 17:31 |
| gouthamr | ^ its an opportunity for project teams and the TC to share relevant material with the foundation | 17:31 |
| gouthamr | i'll reach out to her and seek details on the priorities either on the ML or during a future TC meeting as she suggested | 17:32 |
| gouthamr | we have some more detail regarding the Digital Sovereignty topic here: | 17:33 |
| gouthamr | #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/6MJQ35FSDLRTWYY5SSFU3V5NRMPCRAFW/ (Supporting Digital sovereignty) | 17:33 |
| spotz[m] | If we're doing anything new and exciting for security that's always important with all the digital soveignty stuff going on these days | 17:33 |
| spotz[m] | see:) | 17:33 |
| gouthamr | diablo_rojo's running a meeting soon: January 20th at 7:00 UTC | 17:33 |
| gouthamr | please do respond to the email thread, i suspect an etherpad will be shared at some point too | 17:34 |
| gouthamr | finally, there's been some engagement from folks already on this.. bitergia dashboards are going away tomorrow (January 14, 2026): | 17:35 |
| gouthamr | #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/L63X443VSEOSTOQNG57GEDYSF3BX744L/ (Bitergia dashboards are going away, welcome LFX soon-ish) | 17:35 |
| gouthamr | a nice list of changes has already been requested on the thread | 17:36 |
| bauzas | I was just becoming used to Bitergia dashboards, lol :) | 17:37 |
| spotz[m] | Yeah | 17:38 |
| sean-k-mooney | im not too sure how vocal to be on that thread as i dont use the metrics dashboard often | 17:38 |
| sean-k-mooney | although the new one dont really cover the usecase i orgianlly had but its not a huge deal if it doesnt | 17:38 |
| gouthamr | teams found review metrics useful to quantify the involvement of people and curate the core reviewers list | 17:39 |
| sean-k-mooney | yep | 17:39 |
| sean-k-mooney | that was the primary thing i used to use it for, that and just overall partispation | 17:40 |
| sean-k-mooney | if i wanted to know things like how many bugs were fixed or how many feature were deliver we coudl get that in several other ways | 17:40 |
| sean-k-mooney | btu review partisitpaion not so much | 17:41 |
| gouthamr | another thing of late, among the action items we took for bridging-the-gap, the review timelines - mean time for first review, time to close etc. | 17:41 |
| spotz[m] | Well other people besides us use the metrics, so being able to do participation and how folks are participating is important at least for my uses | 17:41 |
| sean-k-mooney | one thing that is perhaps misleading is the "contibutions out side of work hours" view | 17:42 |
| sean-k-mooney | mainly because how does it knwo what is outside of work hours for a given contibutor | 17:42 |
| spotz[m] | Ha I was going to ask if that required you to put in your work hours or timezone, but like for me I work all over the place time wise during the day | 17:43 |
| fungi | we suspect it's based on the timezone declared in your git commits | 17:44 |
| sean-k-mooney | i mean to do it propelry it woudl but i dotn have that set anywhere and if you assume based on contry that would be quite incorrect | 17:44 |
| fungi | in which case it thinks i live in "utc" | 17:44 |
| clarkb | and then assuming work hours are 8am-5pm or whatever. We opened an issue indicating this is probably not accurate enough to rely on | 17:44 |
| sean-k-mooney | oh well that also goign to be wrong | 17:44 |
| fungi | but the other thing about that which i find glaringly terrible is that they want to imply that if you have a lot of people contributing outside business hours then that's a risk to the project, and inherently unhealthy for some reason | 17:45 |
| sean-k-mooney | again not a big deal but some of the metics seem less suited for globally mainteind software | 17:45 |
| sean-k-mooney | one of the pros of not haveing the review data perhaps is that it will dicusage companies for settign reveiw targets as bonus critia | 17:47 |
| bauzas | well, in an opensource world, workdays aren't a good metric to track | 17:47 |
| sean-k-mooney | anyway this is proably a littel off topic for the meeting | 17:48 |
| bauzas | I know that most of us are company-backed for working on openstack, but there is a world for on-off contributors working on their free time | 17:48 |
| gouthamr | yes got a couple of other items to cover today :) | 17:48 |
| gouthamr | #topic Rehashing SIGs | 17:48 |
| gouthamr | think this is a good one to t o n y b when he reads scrollback | 17:48 |
| mnasiadka | bauzas: and even I am also company-backed - we’re a global community and a lot of things I do are outside of business hours (because my business hours are also a bit spread here and there, not a contiguous block) | 17:49 |
| gouthamr | we cleaned up a bunch of metadata regarding SIGs, but, not the outdated info about SIGs themselves.. we had a question on the ML pertaining to a project maintained by the Ansible SIG: | 17:49 |
| gouthamr | #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/7QPVDSMZZXSFPS22T4QWRLILUZBJGINH/ (Ansible Collections question) | 17:49 |
| gouthamr | think gtema and noonedeadpunk had some concerns about this recently, and got stuff resolved | 17:51 |
| gouthamr | would you be willing to help? | 17:52 |
| gtema | I hope it to die sometime. But ok, will try to find some time for review | 17:52 |
| gouthamr | perhaps we can take the opportunity to ask if there's interest to re-form this SIG and better maintain the ACO repo | 17:52 |
| gouthamr | haha, yeah, i think we need more maintainers gtema :( there seem to be users, maybe they can be convinced to help here | 17:53 |
| gtema | you should have people interested in reviewing. Just reforming the SIG into the NEW SIG will bring nothing | 17:53 |
| gouthamr | for one, the SIG is the maintainer, unless we move the project to a project team | 17:54 |
| gtema | there is no project team and the delivery is not the official one (license) | 17:54 |
| gtema | and the SIG is "empty" | 17:54 |
| gouthamr | your concern that SIGs are a bit less responsive is valid, and that's one thing we're working on, sloly | 17:54 |
| gouthamr | yes | 17:55 |
| clarkb | what is the impact to ansible users if this goes away? | 17:55 |
| gtema | I myself do not depend on ansible (don't believe it is the right tool to manage openstack resources) and thus not active on it | 17:55 |
| mnasiadka | I think me and noonedeadpunk are very interested for this deliverable to live | 17:56 |
| sean-k-mooney | i know there is some usage of it in some of the vmware migrtion tooling | 17:56 |
| sean-k-mooney | so os-migrate uses ansibel collection to create teh openstack resouces | 17:56 |
| mnasiadka | (Or rather Kolla and OSA are very interested) | 17:56 |
| gouthamr | interesting | 17:56 |
| gtema | interesting indeed - official os projects depend on inofficial os project | 17:56 |
| sean-k-mooney | https://github.com/os-migrate/os-migrate/blob/main/requirements.yml#L8 | 17:57 |
| sean-k-mooney | am no | 17:57 |
| clarkb | every openstack project depends on non openstack projects (the linux kernel, kvm, qemu, mysql/mariadb, python, the list is long) | 17:57 |
| sean-k-mooney | so os-migrate is not offially part of openstack is it? | 17:57 |
| gouthamr | sean-k-mooney: he meant OSA and Kolla | 17:57 |
| sean-k-mooney | kolla and osa i dont thnk have a depency today but i coudl be wrong | 17:57 |
| spotz[m] | I thought it was more for reducing duplication of efforts | 17:58 |
| sean-k-mooney | ack its been a few years since i contirbtued to kolla-ansible and it didnt use the ansible collection at the tiem btu that could have changed | 17:58 |
| clarkb | opendev uses it minimally to bootstrap new cloud regions when we get access to them. But I think we would/could write a python script that uses the sdk to do the same thing without too much trouble. But there may be much larger impacts to others and I think that is worth considering particularly if groups like kolla and osa are willing to get more involved | 17:58 |
| mnasiadka | Kolla uses openstack.cloud in basically all roles that deploy a service (for keystone interaction and so on) | 17:58 |
| sean-k-mooney | its not listed as a depency https://github.com/openstack/kolla-ansible/blob/master/requirements.yml | 17:59 |
| noonedeadpunk | osa also uses it | 17:59 |
| gtema | sorry, need to run. I am open to have a dedicated discussion on that (yet again), but nothing will change unless someone start contributing (also reviews) | 17:59 |
| sean-k-mooney | but ok it used to use osc | 17:59 |
| mnasiadka | It’s in kolla-toolbox - https://github.com/openstack/kolla/blob/master/docker/kolla-toolbox/requirements.yml | 17:59 |
| clarkb | sean-k-mooney: I don't think ansible collections are managed like python deps | 17:59 |
| noonedeadpunk | and I think the main problem making it project deliverable is the license | 17:59 |
| sean-k-mooney | i think in anycase if it is usign ti then that a factor | 17:59 |
| noonedeadpunk | as it's not Apache 2.0 | 17:59 |
| noonedeadpunk | I try to review from time to time, but really not with good periodicity | 18:00 |
| sean-k-mooney | clarkb: it depend on if you are declaring them in a glaxay role or not there are a few ways to specify them | 18:00 |
| gouthamr | we're a minute past | 18:01 |
| gouthamr | lets wrap up this meeting here | 18:01 |
| gouthamr | does anyone want to note anything else for the minutes today? | 18:01 |
| fungi | related to a foundation board discussion late last year, i'll be posting to openstack-discuss later today with a draft goal for the foundation staff to "take openstack to the next level in 2026" which we're working to finalize in order to present to the board of directors at an upcoming meeting. i'll tag the tc (and others) in subject for visibility | 18:01 |
| fungi | #link https://lists.openinfra.org/archives/list/foundation-board@lists.openinfra.org/thread/YRBIUTUBDBSHFUNVRX76PGPU6TKDZROU/ Take Things to the Next Level (reference for prior context) | 18:01 |
| gouthamr | ack ty fungi | 18:03 |
| gouthamr | alright, thank you all for participating | 18:03 |
| gouthamr | i'll close with the reminder that next week's meeting will be at 0800 UTC and hosted by mnasiadka here.. | 18:04 |
| gouthamr | #endmeeting | 18:04 |
| opendevmeet | Meeting ended Tue Jan 13 18:04:09 2026 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 18:04 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/tc/2026/tc.2026-01-13-17.01.html | 18:04 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/tc/2026/tc.2026-01-13-17.01.txt | 18:04 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/tc/2026/tc.2026-01-13-17.01.log.html | 18:04 |
| JayF | re: "outside business hours" -- I wonder if there's an even nastier bug around that: if it's somehow measuring contribution as happening at merge/approval time, instead of at gerrit-change-request submission time. That would certainly cause a headache. | 18:04 |
| sean-k-mooney | i guess there are 2 datasouces there | 18:05 |
| sean-k-mooney | one is the commit metadta abotu autor/commit time | 18:05 |
| sean-k-mooney | but then serpeatly there is data form a merge commit if there is oen or gerrit review | 18:05 |
| sean-k-mooney | i.e. when i +2" vs when when the patch was written vs when it merged | 18:06 |
| sean-k-mooney | i was confused by the gerrit merged metrics too | 18:07 |
| JayF | yeah, I think the data is just not great | 18:07 |
| sean-k-mooney | "merged a changeset" | 18:07 |
| JayF | I'm really disappointed, TBH, that this was announced with such small notice and it's not really good at all for our use cases. | 18:07 |
| sean-k-mooney | is that +w? | 18:08 |
| sean-k-mooney | because if a human is ever submiting a change and they are not part of core infra working around a zuul bug | 18:08 |
| sean-k-mooney | that a problem | 18:08 |
| sean-k-mooney | if tis +w then its fine but in our worklow all submitions should be done by zuul | 18:09 |
| fungi | it's because lf has a "project health" dashboard and is not keen on paying a third party to host another one, and the cut-off is when our bitergia contract that's not going to get renewed expires | 18:09 |
| clarkb | https://github.com/linuxfoundation/insights/issues/688 is the issue fwiw | 18:09 |
| fungi | think of it as more growing pains from the lf merger | 18:10 |
| JayF | fungi: yeah, while I understand that, I also know that this is costing me some time to re-orient my processes, and to direct my team to update their info so we're properly affiliated. That's not cheap/free either. | 18:10 |
| sean-k-mooney | so is there dashboard intened to be an opesource verion of githubinsigtes i.e. https://github.com/linuxfoundation/insights/pulse | 18:11 |
| sean-k-mooney | it feels like that is more what they are tryign to replciate | 18:12 |
| JayF | the github insights are ... exceedingly not useful ... I hope not :( | 18:12 |
| JayF | the difference between a high score leaderboard and a data warehouse | 18:12 |
| clarkb | I can't speak for them but that hasn't been the impression I've gotten. they've listened to the feedback we've pvodied particularly with differences for gerrit | 18:12 |
| clarkb | *we've provided | 18:13 |
| JayF | I suspect some of these complaints have as much to do with how good stackalytics (when it was maintained) and biterg are | 18:13 |
| sean-k-mooney | ack | 18:13 |
| JayF | they are trying to curate an experience when we've had basically raw data querying for 15 years | 18:13 |
| sean-k-mooney | im wondering if it was more insight for marketign/adoption or insight for the oss proejct ot under stand there contributor dynmaics | 18:14 |
| sean-k-mooney | i guess its a bit of both | 18:14 |
| fungi | yeah, the growing pains are mostly that this dashboard was not developed originally to be useful for community project maintainers, it was meant for lf management primarily to gauge project health. but they're keen to extend it to support the openstack community's use case | 18:14 |
| JayF | I wish the extend happened before we lost the alternative. | 18:14 |
| sean-k-mooney | fungi: ya that the impression i got when i looked at it and what i ment by marketing above | 18:14 |
| clarkb | re raw data querying I think a lot of that can be done via gerrit now too | 18:14 |
| JayF | Taking away a hammer and replacing it with a paper towel roll is no fun | 18:15 |
| clarkb | (for the gerrit specific stuff anyway) | 18:15 |
| sean-k-mooney | clarkb: i was not really sure how much of this historcal data was aviable in its api but i guess the dashboard ahs to get the data form somewhere | 18:15 |
| JayF | I'm more than a little miffed that the affiliation data appears to have been lost again too, but I only discovered that this morning | 18:16 |
| sean-k-mooney | i think stackalitics used to ingest the eventstream (the same one zuul consumse) | 18:16 |
| JayF | literally this email hit the list the *day after* I wrote all the internal documentation for setting up affiliation :( | 18:16 |
| sean-k-mooney | JayF: thats a cahgne in where aht is pulled form | 18:16 |
| sean-k-mooney | JayF: for stackalitics there was a file in repo that had it but that raised some data proection concerns | 18:17 |
| sean-k-mooney | i belive LFX is pulling it form a linux foundation accoutn if you have one | 18:17 |
| clarkb | sean-k-mooney: I think all of the data should be available in gerrit. (We know of three changes that don't reindex which we believe is due to sql data inconsistencies that didn't play nice with the notedb transition so there are three exceptions) | 18:17 |
| fungi | lf uses some sort of third party that guesses contributor affiliation mappings, i think, not clear on the details but i believe it can be adjusted in the lf profile. the other problem though is lf considers contributors' affiliations to be sensitive personal data they can't publicly disclose | 18:17 |
| JayF | That is *the* problem. | 18:17 |
| JayF | Especially when I work for a place where part of the deal is ensuring GR gets credit for the people they employ to work on OSS. | 18:18 |
| JayF | It's actively harmful to my team. | 18:18 |
| clarkb | LFX still attributes the activity to the company if the affiliation is set up aiui | 18:18 |
| fungi | yeah, it just doesn't give individual affiliations | 18:18 |
| JayF | my teams' are not being reported right, as I emailed the mailing list, I'm seeing both "GR-OSS" and "G-Research" affiliations in the list | 18:19 |
| JayF | and the lack of detailed data makes it impossible for me to troubleshoot directly | 18:19 |
| sean-k-mooney | if you use your company email for the git then it woudl be there also | 18:19 |
| sean-k-mooney | in the git history and the git view of the contibutions | 18:19 |
| JayF | using company email for git is an antipattern and hostile to the developer IMHO :) | 18:19 |
| sean-k-mooney | oh i know i stopped doing that when i left intel beucase it was too much of a pain to have to change again if i ever leave redhat | 18:19 |
| JayF | I think of it as a personal identity thing, I'm not "$company Software Developer" when I contribute to OpenStack, I'm me. And that reputation will last long after that company is the one paying me to do work. | 18:20 |
| sean-k-mooney | i mean yes there is the protfilo part of it too. as an assied we dont actully have an affilation in gerrit today | 18:21 |
| sean-k-mooney | you can put it in the about me filed | 18:21 |
| sean-k-mooney | but there is no affilaiation field | 18:21 |
| clarkb | correct affiliation would be one of the things yo ucannot directly query gerrit for | 18:21 |
| sean-k-mooney | for bitergia i think it pulled it form the opendev/openinfra profiel | 18:21 |
| sean-k-mooney | https://opendev.org/x/stackalytics/src/branch/master/etc/default_data.json is what stackaltics used | 18:23 |
| sean-k-mooney | the pros and cons fo that was it actully supprted start and end dates but also your eamil was public in yet another place | 18:24 |
| clarkb | re stackalytics using the event stream it actually led to them hitting connection limits due to flaky connectivity and bad ssh implementations. Instead of fixing it they just started creating a new account every time they hit the limit. I'm not sure if htey ever fixed the ssh implementation | 18:24 |
| sean-k-mooney | really | 18:24 |
| clarkb | the gerrit api lets you query for changes since $time which is probably better for systems like this. Just query once an hour for things since an hour ago | 18:25 |
| sean-k-mooney | im not sure why they need more hten one connection but ok | 18:25 |
| fungi | the stackalytics solution had two problems around affiliation: 1. you needed a stackalytics core reviewer to approve additions/changes, and 2. the service only picked that up if you restarted it and waited for it to reindex all changes and rebuild the mappings in memory | 18:25 |
| sean-k-mooney | that and the instace we used were never offical either right | 18:25 |
| sean-k-mooney | as in they were never part fo open infra they were hosted by mirantis and other that found it useful | 18:26 |
| clarkb | sean-k-mooney: the connection would crash then they would start a new one without ever stopping the old one properly. If that happened frequently enough in a short period you'd hit the 96 connection limit and instead of fixing it in code they just made a new account and reconnected to start over | 18:26 |
| sean-k-mooney | clarkb: so i shoulnt host my onw copy :) | 18:26 |
| sean-k-mooney | clarkb: ack i never really looked at the code much | 18:27 |
| clarkb | it is possible it eventually got fixed too I'm not sure | 18:27 |
| sean-k-mooney | i hesistate to suggest this but its kind of a shame that zuul cant provide this info :) | 18:27 |
| JayF | sean-k-mooney: vibe code openalytics time, except with a better gerrit data fetch strategy /s | 18:27 |
| sean-k-mooney | i mean if i ever realy needed thisfor soem reason im sure claude-code coudl spit out a cli that would dump the data i wanted | 18:28 |
| sean-k-mooney | but part of havign the hosted solcution was they cashed the data so that it dint put load on teh upstream gerrit server | 18:29 |
| sean-k-mooney | JayF: on the topic of documetinting affilcation was that internal or did you update teh project teams guide/upstream docs | 18:29 |
| JayF | sean-k-mooney: internal, but I am happy to share the marekdown with you | 18:30 |
| JayF | actually I think it's public | 18:30 |
| sean-k-mooney | im not sure how many people even know that the affilation can be configured. | 18:30 |
| fungi | though they only cached it to memory, changes to affiliations and repository lists required day-long outages to restart and rebuild the world | 18:30 |
| JayF | https://github.com/G-Research/gr-oss/blob/master/docs/attribution.md#openstack-configuration I think this is public | 18:30 |
| sean-k-mooney | nope | 18:31 |
| JayF | dang | 18:31 |
| sean-k-mooney | its fine | 18:31 |
| sean-k-mooney | i just dont know how many contibuts actully knwo about this and how to set it up | 18:31 |
| JayF | here's the snippet https://www.irccloud.com/pastebin/UHJiABNo/openstack-affiliation-configuration.md | 18:31 |
| sean-k-mooney | it probably shoudl be part of https://docs.openstack.org/contributors/ somewere too if tis not already coverd | 18:34 |
| JayF | This is a summary of already-existing docs | 18:34 |
| JayF | as usual in openstack, our obsession with getting it in the right repo means you need a decoder ring to find how to set it up | 18:34 |
| sean-k-mooney | well its not somuch an obsession of gettign it in the right rep more so i know we had an internal doc at intel for seting it up too becuase we could never find the upstream one | 18:36 |
| fungi | sean-k-mooney: as for how many contributors knew they could set affiliations in their openinfra profiles, in recent cycles it was about 50% of individuals, though those individuals were owners of around 80% of the merged changes so it was the more-involved contributors | 18:36 |
| sean-k-mooney | i woudl expect it to be somewhre in https://docs.openstack.org/contributors/common/accounts.html proably | 18:36 |
| sean-k-mooney | fungi: ya i was refelcting on the fact i know i did it many years ago | 18:37 |
| sean-k-mooney | btu when we had a new hier join i dont knwo fi this was something we even mentioned | 18:37 |
| fungi | right, and also we have no idea if the affiliations they set are being kept up to date, of course | 18:37 |
| sean-k-mooney | and if i was new to workign on opentsack i dont knwo if this is something that i woudl discover by my self | 18:37 |
| JayF | sean-k-mooney: lets be real: how accessible is openstack already to unassisted new devs? | 18:38 |
| JayF | this is the least of that problem | 18:38 |
| sean-k-mooney | ya i dont know i havent really contibuted to enough other comunites to say | 18:39 |
| fungi | you could ask that of open source as a whole, for that matter | 18:39 |
| sean-k-mooney | right for example i dont knwo how i woudl even start contibutiing to python itself or k8s for example | 18:40 |
| sean-k-mooney | unless its a standalon github repo its often not trivial ot work that out | 18:41 |
| fungi | and there are many axes of "accessible" as well (accessible for people with specific needs? welcoming and not filled with hostile/toxic leadership? et cetera) | 18:42 |
| sean-k-mooney | on a releated note. i got the email for the elections after the restart and i still dont see a voting option | 18:43 |
| sean-k-mooney | im not sure if that because my account is assofted with a diffent email then the voting invite althouh i have added my work email to it as well | 18:43 |
| fungi | sean-k-mooney: you also needed to reestablish your individual membership under the new foundation no later than six months prior to the election | 18:46 |
| sean-k-mooney | by the way i see "voting required" brifly while the page is loadign but when it compeltes it not there | 18:46 |
| sean-k-mooney | fungi: i did that last summer | 18:46 |
| fungi | oh, you received a ballot e-mail? | 18:46 |
| sean-k-mooney | yes | 18:46 |
| sean-k-mooney | this i belive is the 7th | 18:46 |
| fungi | Subject: Invitation to Participate in: OpenInfra Foundation 2026 Individual Director Election | 18:47 |
| clarkb | someone else had a similar problem and the issue was they were logged into lfx using an account that wasn't affiliated with the email address the ballot was sent to | 18:47 |
| sean-k-mooney | it was sent to my work email yep "Invitation to Participate in: OpenInfra Foundation 2026 Individual Director Election" | 18:47 |
| clarkb | if you have multiple lfx accounts this may be the problem | 18:47 |
| sean-k-mooney | i did not have any when i got the first email | 18:47 |
| sean-k-mooney | i then created one not assocated to my work email and added my work email to it | 18:47 |
| sean-k-mooney | but it still does not show up | 18:48 |
| clarkb | sean-k-mooney: for me it shows up in my activities list on the main dashboard | 18:49 |
| sean-k-mooney | i dont see an activites list. what is the main dashbaord https://openprofile.dev/? | 18:50 |
| clarkb | sean-k-mooney: ya that page. At the top it has my name and some of my affiliation info then below that is a horizontal bar space with a list of activities in it and one is the election | 18:51 |
| sean-k-mooney | ya not so much for me https://imgur.com/a/nbx6lu5 that middle row is where i appres while the page is loadignbut nothigh after that | 18:53 |
| clarkb | sean-k-mooney: ok ya it is missing for you. The what brings you here today is below my activities list bar | 18:54 |
| clarkb | can you double check the email address that recieved the ballot shows up in https://openprofile.dev/edit/email-management ? | 18:54 |
| sean-k-mooney | sure i can check. it show us in https://openprofile.dev/edit/connected-accounts | 18:56 |
| sean-k-mooney | no its not in email managment | 18:56 |
| sean-k-mooney | i added my work google acount as a login method so i was expect it to use that let me see if addign it in email-managemnt works | 18:57 |
| clarkb | sean-k-mooney: it is possible that you may need them to add the address. I ended up needing that done for my account for some reason | 18:58 |
| sean-k-mooney | i updated https://imgur.com/a/nbx6lu5 | 18:58 |
| sean-k-mooney | so when i go to add it apprently there is another lfx account created that i can merge it with | 18:59 |
| sean-k-mooney | btu i did have one before so i wonder if they auto created them? | 18:59 |
| clarkb | sean-k-mooney: yes I think one of the things they did in this process was stage accounts for those who woudl receive ballots that didn't already have one | 18:59 |
| clarkb | sean-k-mooney: there should've been an email in december prompting yout to set that account up. Alternative I think you can merge teh account which is what I did though the automated merge failed and i had to get them to do it on the backend for some reason | 18:59 |
| sean-k-mooney | right and they presumable used the primay email form thje opendev id for that | 19:00 |
| clarkb | if the automated merge fails for you let fungi or I know and we can proxy a request for merging for you | 19:00 |
| sean-k-mooney | it errors with " This email is associated with a different account. You may contact us to merge the accounts." | 19:01 |
| sean-k-mooney | but tis ment to send an email ill try it again and see | 19:01 |
| sean-k-mooney | i can see if i can loginto that account with google but if they prefreated the account for alot of peple that would be a lot of manual merging for them | 19:02 |
| clarkb | sean-k-mooney: I think that is exactly what I hit. So I think your two options are sending a requset to merge teh account which maybe you can do via that system somewhere or use the staged account as a second account | 19:02 |
| clarkb | I'm running the opendev team meeting now so distracted, but happy to help later | 19:02 |
| fungi | well, the two aren't mutually exclusive | 19:02 |
| clarkb | right you could use the account as a second account for now then merge them later | 19:03 |
| fungi | exactly | 19:03 |
| fungi | that's what i'd do, just in case it took lf folks longer than the voting window to merge them | 19:03 |
| sean-k-mooney | to do that i proably need to remove the ablity to login with google usign that ohter account | 19:03 |
| sean-k-mooney | currently i have both ya it logs me into the same accont if is log in with private browsign mode | 19:06 |
| sean-k-mooney | my account does not matter that much i just wonder if others are affected the same way | 19:06 |
| fungi | i keep two completely separate lf accounts on purpose, fwiw | 19:07 |
| sean-k-mooney | well i was planing to ony have a personal one becuase for things like this if i vote im obviousl doign that in a persona capasity | 19:08 |
| fungi | yeah, i have a personal one and a work one basically | 19:08 |
| sean-k-mooney | but i can see usign either a personal or work account for things like confernces | 19:08 |
| sean-k-mooney | ok that works | 19:11 |
| sean-k-mooney | i guess i can have both account for now | 19:11 |
| sean-k-mooney | it was in a semi setup state | 19:11 |
| sean-k-mooney | as in i did not actully have a user name assigned to the account for example | 19:11 |
| clarkb | sean-k-mooney: you were able to log into the other accoutn then? Do you see the election link in activities there? | 19:12 |
| sean-k-mooney | when i removed smooney@redhat.com as an SSO option in https://openprofile.dev/edit/connected-accounts form my sean@seanmoney.info acconnt yes | 19:12 |
| sean-k-mooney | priot to that the work google accoutn signed into my personal one | 19:13 |
| clarkb | got it. | 19:13 |
| sean-k-mooney | you have to be able to login to add an account under https://openprofile.dev/edit/connected-accounts but its kind of sus that they woudl alow one acocunt to alias a sperate accoutn like that | 19:14 |
| sean-k-mooney | for better or wors i started using firefox's contaienrs a few months back | 19:14 |
| clarkb | as long as the enrollment for the new login ensures you control that login it is probably safe from a security perspective but does lead to weird user experience when it shadows things | 19:14 |
| clarkb | we can probably find a way to provide feedback on that being weird | 19:15 |
| sean-k-mooney | so i can actully have diffent tabs signed into the same sight with differnt cookies | 19:15 |
| sean-k-mooney | so the weirdness stems form the fact that however htey created the acount on the backend it never sent any email or similar to notify me so the acocunt partly existign but wihout any password or username | 19:16 |
| sean-k-mooney | if my work account wasnt gamil based i proably woudl nto have been abot to login althogh if i knwo the acocutn existied i might have been able to reset the password | 19:17 |
| fungi | i'd chalk that up to using an identity provider integrated into my e-mail provider, sort of fails at separation of concerns | 19:18 |
| sean-k-mooney | clarkb: as an asside soem of the urls (all of them?) are borken in the vote | 19:18 |
| sean-k-mooney | i.e. https://openinfra.org/election/2026-individual-director-election has https://openinfra.org/election/2026-individual-director-election%3C/a%3E%3C/p%3E as the url which has the closing ancor and paragrah tags at the end | 19:19 |
| sean-k-mooney | https://imgur.com/a/RmQckP8 | 19:20 |
| sean-k-mooney | there is some markdown rednerign issue not that i cant work around it but just a bug in how the election was configured or rendered by the tool | 19:21 |
| clarkb | I'll pass that along | 19:22 |
| sean-k-mooney | its looks like its not correctly doing html escaping in whatever templating engine is being used | 19:23 |
| -opendevstatus- NOTICE: An update to one of our base jobs roles broke another base job role. This update has been reverted and jobs should be working again. | 20:59 | |
| fungi | https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/message/MPLKSGH2IKULPOA3XIYROV3WKB2DF43X/ is the post i promised earlier today during the meeting | 22:41 |
| gouthamr | ++ | 22:41 |
Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!