| *** ralonsoh_ is now known as ralonsoh | 08:41 | |
| *** dasm is now known as Guest1929 | 14:57 | |
| *** Guest1929 is now known as dasm | 15:02 | |
| knikolla | I won’t be able to make the meeting today and JayF agreed to chair it. Thank you JayF! | 16:31 |
|---|---|---|
| JayF | o/ | 16:31 |
| spotz[m] | o/ | 17:05 |
| spotz[m] | Today is video? | 17:05 |
| JayF | tc-members: Meeting in 55 minutes; agenda and zoom link at https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee | 17:06 |
| JayF | spotz[m]: yep | 17:06 |
| gmann | I think yes | 17:06 |
| spotz[m] | Thanks, Tuesday and Wednesday I'm in out of meetings and lose track. Thursday is starting to be as bad. Then we start over with the time change in a month or so:) | 17:11 |
| gmann | slaweq: seems like one extra file got added in this https://review.opendev.org/c/openstack/governance/+/892670 | 17:40 |
| slaweq | Ups. That's a mistake. I will remove it today. Thx | 17:42 |
| gmann | thanks | 17:42 |
| opendevreview | Slawek Kaplonski proposed openstack/governance master: 2022 User Survey TC Question Analysis https://review.opendev.org/c/openstack/governance/+/892670 | 17:57 |
| JayF | tc-members: 3 minutes, zoom link is in https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee | 17:58 |
| JayF | #startmeeting tc | 18:01 |
| opendevmeet | Meeting started Tue Sep 5 18:01:09 2023 UTC and is due to finish in 60 minutes. The chair is JayF. Information about MeetBot at http://wiki.debian.org/MeetBot. | 18:01 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 18:01 |
| opendevmeet | The meeting name has been set to 'tc' | 18:01 |
| JayF | Hi all, welcome to the weekly meeting of the OpenStack Technical Committee | 18:01 |
| JayF | A reminder that this meeting is held under the OpenInfra Code of Conduct available at https://openinfra.dev/legal/code-of-conduct | 18:01 |
| JayF | Today's meeting agenda can be found at https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee | 18:01 |
| JayF | #topic Roll Call | 18:01 |
| JayF | o/ | 18:01 |
| gmann | o/ | 18:01 |
| rosmaita | o/ | 18:01 |
| JayF | o/ | 18:02 |
| slaweq | o/ | 18:02 |
| spotz[m] | o/ | 18:03 |
| dansmith | o/ | 18:03 |
| JayF | #topic Gate Health Check | 18:03 |
| JayF | Dan reports better than average gate health. A few recurring failures but mostly steady from last week. Cinder is trying to fix an issue and will discuss more details at their meeting tomorrow. | 18:06 |
| fungi | not going to interrupt the call, but we had a zuul bug which came in with our automated upgrade on saturday which caused changes with zuul configuration for projects with deprecation warnings to not get tested. it was fixed yesterday a few hours after the problem was noticed | 18:08 |
| JayF | #note Devstack has removed support for Ubuntu Focal. Ubuntu Focal is not in the testing interface for this cycle. Projects using focal-based jobs need to move those jobs to newer, support devstack. | 18:09 |
| fungi | also yesterday opendev updated the default ansible version to 8 on the zuul executors for the openstack tenant. if any related problems are identified, please reach out | 18:09 |
| JayF | #note OpenDev zuul executors using Ansible 8 by default now. | 18:09 |
| dansmith | JayF: to be clear, not better than average.. better than recent, more like average :) | 18:09 |
| fungi | and zuul is switching from facebook-re2 to google-re2 so is deprecating lookarounds in regular expressions (and has added a negate option for regexes as an alternative). these warnings are non-blocking and informational for now | 18:10 |
| JayF | #topic OpenStack Elections | 18:10 |
| JayF | #link https://governance.openstack.org/election/ | 18:10 |
| fungi | traditionally, the i18n team's electorate consisted (almost entirely) of non-code contributors, for example | 18:14 |
| fungi | non-code contributor additions cutoff has always been when the governance repo gets tagged (so that the election officials have a reference point for what entries are valid) | 18:15 |
| JayF | yeah, that was last week :( | 18:15 |
| JayF | #note Call for extra-ACs was not done timely enough to prod PTLs to update their lists. We will call for this update to happen at the beginning of the cycle. | 18:16 |
| spotz[m] | I know I asked for the repos to be expanded to include goverance, maybe we could have it expanded to include translations? | 18:16 |
| fungi | spotz[m]: there isn't one, that's the problem | 18:17 |
| JayF | #action gmann to ensure that a call for extra-ACs is included in future election/governance operational documentation to ensure it's difficult to miss in the future. | 18:18 |
| fungi | i18n contributors can be extracted from the translation platform, but requires a translation manager who has the privileged access to export that activity data | 18:19 |
| fungi | it is and has been a manual process which the i18n ptl typically performed | 18:20 |
| frickler | re "Devstack has removed support for Ubuntu Focal", nova moved first with bumping libvirt reqs beyond what is in focal, blocking devstack and others's gate. I've already complained to them about bad timing doing this the last minute before library freeze | 18:20 |
| JayF | #action JayF to contact i18n SIG about extra-ACs | 18:20 |
| JayF | #topic Open Discussion and Reviews | 18:21 |
| JayF | #link https://openinfra.dev/ptg/ | 18:21 |
| JayF | #note Please register as an attendee for the PTG | 18:21 |
| JayF | #link https://review.opendev.org/q/projects:openstack/governance+is:open | 18:21 |
| JayF | User Survey analysis and improvements are patches in governance good to review. | 18:22 |
| frickler | why is registering considered so important? they didn't even do double-opt-in last time I checked | 18:22 |
| JayF | I am unsure; the foundation organizes PTGs and I'm passing along requirements laid out by them -- including preregistration. | 18:23 |
| frickler | but your wording makes it sound like the TC endorses those reqs | 18:23 |
| spotz[m] | Frickler do you want to join the call? https://us06web.zoom.us/j/87108541765?pwd=emlXVXg4QUxrUTlLNDZ2TTllWUM3Zz09 | 18:24 |
| JayF | frickler: we are in zoom meeting this week, I am proxying your comments but discussion in both places at the same time is not trivial | 18:24 |
| frickler | nope. certainly not on zoom, but also not otherwise | 18:24 |
| frickler | feel free to defer my questions until the meeting is done if you prefer that | 18:25 |
| JayF | frickler: re: PTG, I encourage people to do the things that are requested by the folks doing logistics because I assume they wouldn't put an unneeded requirement in there. The details of how a PTG is organized is not something I have time to understand completely. | 18:26 |
| fungi | frickler: for the libvirt bump, should the tc have been reminding teams sooner that focal is not in the pti for 2023.2 and teams can/should drop their focal jobs? | 18:27 |
| frickler | fungi: well focal not being in the PTI doesn't mean that projects are not allowed to support it | 18:28 |
| fungi | i guess i'm unclear on what the responsibility is to teams for supporting things that aren't covered in the pti but other teams want to test anyway | 18:28 |
| frickler | I think a general reminder would be in order to do bumps affecting a lot of projects early in the cycle | 18:28 |
| clarkb | I argued in an email to the openstack-discuss list recently that not only should they be done in the cylce but they should be done knowing that things will break and then work through it from there | 18:29 |
| clarkb | rip bandaids off and then move forward | 18:29 |
| frickler | being more strict against projects with broken CI and/or a backlog of release relevant patches might also be a good idea, yes, but somewhat orthogonal to the other thing | 18:30 |
| JayF | #endmeeting | 18:38 |
| opendevmeet | Meeting ended Tue Sep 5 18:38:44 2023 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 18:38 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/tc/2023/tc.2023-09-05-18.01.html | 18:38 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/tc/2023/tc.2023-09-05-18.01.txt | 18:38 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/tc/2023/tc.2023-09-05-18.01.log.html | 18:38 |
| JayF | gmann: who to email to get the video I rec'd in the cloud? | 18:38 |
| gmann | JayF: allison Price aprice can help on this ^^ | 18:39 |
| dansmith | not being in the PTI means devstack and nova don't have to support it, which means anyone that depends on nova and/or devstack that want to test focal have to make their own arrangements, IMHO | 18:39 |
| JayF | ack, I'll drop her a line, thanks | 18:39 |
| gmann | allison@openinfra.dev | 18:39 |
| gmann | cool | 18:39 |
| dansmith | since you're *allowed* to run jobs other than what is specified in the PTI, I don't think we can force break people at the beginning of the cycle or anything either so I'm not really sure how much better we can be | 18:40 |
| aprice[m] | JayF i can email it out once it has processed | 18:40 |
| JayF | aprice[m]: ack; I don't know what to do with it after that point either | 18:40 |
| dansmith | other than lots of communication, lots of pushing fixes for people, an extra cycle of best-effort support, etc (all of which was done) | 18:40 |
| gmann | yeah, centos was good example (at 1-2 yeasr back) where its testing was heavily dependendent on TripleO or other tool and not on devstack so devstack does not support any distro does not means it is undtested | 18:40 |
| gmann | *untestable | 18:41 |
| fungi | changing libvirt minimums just before the requirement freeze was pretty much the latest it could happen in the cycle, so maybe the point is to argue for an earlier requirements freeze? | 18:41 |
| dansmith | fungi: if it applies to all requirements then fine (but I don't want to do that) | 18:42 |
| fungi | (since nova/libvirt isn't the only place this can crop up, it just happens to be the most recent example) | 18:42 |
| dansmith | can't be special for nova/libvirt, IMHO | 18:42 |
| JayF | Some amount of this is a communications issue; we should make it clear in the PTI that distributions not listed there will not be expected to have support in devstack by the end of the cycle. | 18:42 |
| gmann | yeah it is for all other req. OVS i remember in past for neutron | 18:42 |
| fungi | dansmith: right, i literally mean moving the requirements freeze time in the release schedule | 18:42 |
| JayF | It's ambiguous today in the sense that it's left to the project; but if we made a stronger statement there maybe folks would have a different expectation? | 18:42 |
| gmann | fungi: dansmith but that shorten the time of development and new features. I prefer project be more active on these kind of unguaranteed support from tooling and be prepared themselves in advance when PTI is updated | 18:43 |
| fungi | if the current requirements freeze is coming so late in the cycle that projects changing their requirements just before that creates a serious problem for other projects, then adjusting the schedule to have a longer requirements freeze in the future seems like the path forward | 18:44 |
| gmann | PTI updates are good notification of what they can get for sure and what they need to arrange by their own if that does not work | 18:44 |
| fungi | right, my point is that if projects want requirements to stop changing so late in the cycle, then they need to balance that pain against the pain of having a longer freeze | 18:45 |
| gmann | yeah | 18:45 |
| JayF | fungi: I'd argue that the pain of the earlier freeze is felt by a distinct group of people than the folks it makes things easier for | 18:46 |
| JayF | fungi: like, people pained by freezing requirements for longer are a completely separate group than the people getting benefit from freezing it for longer | 18:46 |
| JayF | so the incentives around that are a little rough | 18:46 |
| fungi | yes, that's why it's the tc's role to balance the different expectations of those separate constituents | 18:47 |
| fungi | (and why the contributors to all those projects are the ones who elect the tc members) | 18:47 |
| JayF | fungi: When I say that; I mean more that I (we?) should try to find an angle where we can get the behavior we want with incentives aligned for everyone. | 18:50 |
| fungi | sure, or at least try to find a defensible middle ground between everyone's conflicting expectations | 18:51 |
| JayF | Conflicting expectations is a great way to put it. | 18:51 |
| fungi | for the present case, it seems like changing a requirement has unexpectedly impacted work for some projects. the release schedule includes a requirements freeze precisely to mitigate this concern, and the change happened before that freeze, so either the freeze is too short, or the expectation those projects have is unreasonable, or not all requirements are created equal and there's a need | 18:54 |
| fungi | for more than one kind of requirements freeze... | 18:54 |
| frickler | maybe the underlying issue is a different one. if projects affected by the current situation would have amended their testing according to the PTI, there would be no issue | 18:58 |
| dansmith | for sure. | 18:58 |
| frickler | so having the TC or whoever is delegated that task would check earlier in the cycle the status of all testing, that could also be a solution | 18:59 |
| dansmith | meaning have the TC police everyone's jobs? | 19:00 |
| dansmith | I mean, you're allowed to have a focal job if you want it right now... | 19:00 |
| frickler | the question is whether that job is a significant part of a project's CI | 19:01 |
| frickler | the TC already defines requirements by writing the PTI, but it mostly doesn't check whether that is followed, except maybe for unit tests via centrally managed templates | 19:01 |
| dansmith | right, but someone could have a focal-based job that doesn't depend on nova (or devstack for that matter) and be fine.. I'm just saying, determining "significant part of a project's CI" is non-trivial | 19:05 |
| frickler | well, if that's the only functional test, then it becomes pretty obvious, otherwise I agree. see this may need some discussion with the project team to decide rather than just be solved by a quick look at .zuul.yaml | 19:06 |
| frickler | but also I think well over 50% of all projects will have a hard time doing any sensible functional tests without keystone,glance,nova and neutron | 19:08 |
| fungi | JayF pointed out on the call that ironic does standalone functional testing (in addition to testing with devstack), in its mission to support independent deployment use cases | 19:19 |
| fungi | that's one path for projects, though it does almost certainly involve development work to put together the testing | 19:20 |
| fungi | swift's functional tests are similarly independent of devstack, if memory serves | 19:20 |
| JayF | well, bifrost does. I'm not sure any of those jobs vote on Ironic :) | 19:20 |
| JayF | gmann: knikolla: I have access to the recording. Unsure what the next steps are to get this uploaded. | 19:28 |
| gmann | JayF: we need to upload it to this channel but not sure if you have upload permission there, may be knikolla can do it tomorrow or today if he is available https://www.youtube.com/channel/UCBuGwBXOmWHydSE09RM84wQ | 19:31 |
| JayF | I almost certainly do not | 19:31 |
| knikolla | JayF: Send me your youtube account’s email address and I can add you to the channel | 19:33 |
| JayF | knikolla: you have an email with both the video info and my personal google acct email | 19:34 |
| JayF | knikolla: that being said; in 8-9 days we know if I'm still on the TC, so if you can do the thing, maybe wait to grant access? | 19:34 |
| opendevreview | Ian Y. Choi proposed openstack/election master: Add template for CIVS opt-in https://review.opendev.org/c/openstack/election/+/893591 | 19:36 |
| knikolla | Fair point. I can upload it. Thanks for sending me the recording. | 19:36 |
| knikolla | And thanks for covering for me :) | 19:36 |
| JayF | it happens. Now I know how to run a zoom meeting. And I 100% didn't do the intro once before, and once after starting recording | 19:36 |
| JayF | <.< >.> | 19:36 |
| opendevreview | Ghanshyam proposed openstack/election master: Add template for adding Extra AC https://review.opendev.org/c/openstack/election/+/893810 | 22:45 |
| opendevreview | Ghanshyam proposed openstack/election master: Add template for adding Extra AC https://review.opendev.org/c/openstack/election/+/893810 | 22:48 |
| gmann | tc-members ^^ adding extra AC process and email template in election | 22:49 |
| gmann | spotz[m]: rosmaita: I just remember that we have Extra AC deadline in release schedule since starting itself https://releases.openstack.org/caracal/schedule.html#c-extra-atcs | 22:50 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!