| mnaser | gmann, JayF: i was the second ever chair after dhellman and when gmann came in, i figured since i left dhellman in there, kept a little emiratus in there | 00:40 |
|---|---|---|
| mnaser | but thats just the history lesson behind it :> | 00:40 |
| gmann | yeah | 01:24 |
| JayF | I just am pointing my mind at things like, e.g. the pypi maintainer cleanup -- and I lean in the direction of less people with access is better. | 01:49 |
| *** elodilles_pto is now known as elodilles | 06:35 | |
| knikolla | tc-members: reminder, weekly meeting in ~1 hour. This is the one on Zoom. | 16:59 |
| diablo_rojo | Figures when I would be able to participate in a meeting on IRC it would be the zoom week :/ | 17:51 |
| diablo_rojo | knikolla: items for the newsletter was the more pressing item I wanted to bring up - if you can do a callout, I would appreciate it! https://etherpad.opendev.org/p/newsletter-openstack-news | 17:53 |
| knikolla | #startmeeting tc | 18:00 |
| opendevmeet | Meeting started Tue Aug 1 18:00:54 2023 UTC and is due to finish in 60 minutes. The chair is knikolla. Information about MeetBot at http://wiki.debian.org/MeetBot. | 18:00 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 18:00 |
| opendevmeet | The meeting name has been set to 'tc' | 18:00 |
| knikolla | #link https://us06web.zoom.us/j/87108541765?pwd=emlXVXg4QUxrUTlLNDZ2TTllWUM3Zz09#success | 18:01 |
| rosmaita | ty | 18:01 |
| knikolla | #topic Roll call | 18:01 |
| knikolla | o/ | 18:01 |
| slaweq | o/ | 18:02 |
| gmann | o/ | 18:02 |
| rosmaita | o/ | 18:02 |
| dansmith | o/ | 18:02 |
| JayF | o/ | 18:03 |
| rosmaita | #link https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee#Next_Meeting | 18:03 |
| knikolla | #topic follow up on past action items | 18:05 |
| knikolla | #link https://review.opendev.org/c/openstack/project-team-guide/+/843457 | 18:05 |
| knikolla | #topic Release notes guidelines for SLURP/NON-SLURP cadence | 18:07 |
| knikolla | #action rosmaita to review guidelines patch and poke at automating it | 18:08 |
| knikolla | #topic Unmaintained status replaces Extended Maintenance | 18:09 |
| dansmith | knikolla: can we move gate health back up to earlier in the meeting like it used to be, or make sure we don't get into these 40 minute discussions in "past action items"? | 18:40 |
| dansmith | I have something to discuss re: gate health today and I'm wondering if we're going to have time | 18:41 |
| knikolla | makes sense, will do it starting from the next meeting. thanks for suggesting. | 18:42 |
| dansmith | thanks | 18:43 |
| knikolla | #topic gate health check | 18:45 |
| fungi | not to interrupt the discussion, just a heads up that zuul has recently (as of our rolling upgrade last saturday) grown a throughput improvement insofar as that it can detect and act on job failures earlier than before (at the time of the first task failure rather than waiting until the job completes), so that may help speed up deep gate queues and gate reset propagation down the line. | 18:49 |
| fungi | there's some pending ui support to expose that rarly-failure state a little better, but at the moment it just looks like zuul has a premonition that the job will fail (because it already knows it's in the process of failing) | 18:49 |
| fungi | s/rarly/early/ | 18:50 |
| fungi | the keystone activity during devstack setup seems likely related to lack of token caching and insistence in devstack making lots of independent calls to openstackclient | 18:52 |
| knikolla | #endmeeting | 19:00 |
| opendevmeet | Meeting ended Tue Aug 1 19:00:07 2023 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 19:00 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/tc/2023/tc.2023-08-01-18.00.html | 19:00 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/tc/2023/tc.2023-08-01-18.00.txt | 19:00 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/tc/2023/tc.2023-08-01-18.00.log.html | 19:00 |
| knikolla | thanks all! | 19:00 |
| dansmith | fungi: I think I saw that in action today, where a build was marked as failed with no failed jobs, and then later a failed job appeared | 19:01 |
| fungi | dansmith: that's precisely it. in a week or two it will look like https://matrix-client.matrix.org/_matrix/media/r0/download/acmegating.com/ljRghcEytLBWKYFWrkFcQWIu | 19:02 |
| dansmith | ah cool | 19:02 |
| fungi | with descriptive tooltips | 19:03 |
| fungi | so the progress bar for the failing job will go red while it continues to finish up | 19:03 |
| dansmith | yeah, nice | 19:03 |
| fungi | but more importantly, zuul knows at that point the change as a whole is failing and can proceed with whatever downline retesting that may imply | 19:04 |
| fungi | so hopefully at least a minor speedup in change processing | 19:05 |
| dansmith | at the moment all I can hope for is a minor speedup in failing the next job which will obviously fail because everything is failing :) | 19:07 |
| fungi | bingo! it's like you've been here a while | 19:08 |
| gmann | +1 | 19:08 |
| dansmith | heh | 19:08 |
| gmann | and can we add ability to stop the run by author if any quick job failing realize them no need to run it further as I need to debug current failure more | 19:09 |
| gmann | it can be done currently by pushing the change but if anyone need time to debug and do change/test locally then they will be able to stop the further run and give that node to other change | 19:10 |
| gmann | at least that will save a lot of nodes for the changes waiting in the queue | 19:14 |
| JayF | I'll note, re: DB performance, Ironic did some benchmarking which led us down some paths to significantly improve our API performance (and by extension; backend DB performance) https://github.com/openstack/ironic/tree/master/tools/benchmark | 19:15 |
| fungi | gmann: briefly abandoning and restoring the change can also do it | 19:15 |
| JayF | I'm not sure if there's anything useful or unique there which could be applied to Neutron or Keystone; but it's there | 19:15 |
| JayF | having the benchmark was super nice because we were able to clearly demonstrate when we improved vs not improved a thing | 19:16 |
| fungi | zuul should act on the abandon event and dequeue the change immediately | 19:16 |
| gmann | fungi: yeah but that does not seems right way, abandon is more of denying the changes to go furher. idk | 19:16 |
| JayF | https://ironicbaremetal.org/blog/the-search-for-performance/ is a blog we published about the process | 19:16 |
| fungi | zuul doesn't know anything about change owners or gerrit's acls, so you need some sort of action which will trigger an event zuul can act on | 19:17 |
| fungi | setting the change to wip state may now dequeue as well, worth checking | 19:18 |
| gmann | owner of that can be same as abandon/restore which is core group or author | 19:18 |
| gmann | ohk, i did not know about wip. let me try that | 19:19 |
| fungi | as in gerrit's wip toggle, though i suppose workflow -1 might do it too | 19:19 |
| gmann | trying both | 19:20 |
| fungi | i know those will keep zuul from trying to merge the change, but may not eject them from the gate, not sure | 19:20 |
| gmann | seems like not, I tried in this which still running https://review.opendev.org/c/openstack/tempest/+/890242 | 19:22 |
| knikolla | diablo_rojo: I still see the items we added last time here https://etherpad.opendev.org/p/newsletter-openstack-news, did those get sent out and therefore I can clean them up and add new ones? | 19:22 |
| dansmith | W-1 to kick it out would be awesome | 19:22 |
| gmann | yeah | 19:22 |
| fungi | gmann: oh! you were testing on a change that was in the check pipeline, not the gate | 19:55 |
| fungi | we tested on https://review.opendev.org/890065 just now and workflow -1 definitely kicked the change out of the gate | 19:55 |
| fungi | apparently zuul's continuous evaluation of the votes and other submit requirements on a change is fairly new (in the past year), so maybe just nobody's noticed it worked | 19:59 |
| diablo_rojo_phone | knikolla: they did not make it last time but I will make sure it's all added for this month! Feel free to add to the list through Thursday this week. | 20:13 |
| gmann | fungi: ohk, at least that will be helpful. any chance we can do same for check pipeline too ? | 20:45 |
| gmann | few of them in newsletter etherpad are not valid anymore, we should not add them in this month | 20:46 |
| fungi | what's the reason for dequeuing a change in check? those are independent so they don't wait on each other to complete anyway (or they require each other, so aren't going to complete any sooner if another one is dequeued) | 20:46 |
| gmann | fungi: same reason, if I see anyone integrated job failing early give me signal that this change need more debug/time/updates then I can at least stop other jobs still running and taking tiem | 20:48 |
| gmann | tc-members: FYI, board meeting in an ~11 min from now in case anyone want to join https://board.openinfra.dev/meetings/2023-08-01 | 20:49 |
| dansmith | this is just a regular one, not the tc-board sync right? | 20:49 |
| knikolla | ++ thanks for the ping gmann | 20:49 |
| knikolla | @dansmith yep | 20:49 |
| gmann | dansmith: yes, regular one | 20:50 |
| dansmith | ack | 20:50 |
| opendevreview | Brian Rosmaita proposed openstack/project-team-guide master: Add SLURP release notes strategy https://review.opendev.org/c/openstack/project-team-guide/+/843457 | 21:38 |
| TheJulia | knikolla, gmann, fungi: I just added the foundation-board address to the invite I have. If folks can confirm once it makes it through moderation if that works on their calendars, that might just be the easiest path forward. If not, I'll go add all add all of the directors one by one. | 22:12 |
| gmann | cool, let me check | 22:13 |
| fungi | i think jbryce moderates the board ml, if your post is stuck in moderation | 22:13 |
| TheJulia | gmann: I'm assuming, since it gets tagged as an isc file for an invite, that it will likely hit moderation, but Jonathan is typically pretty good at clicking that button quickly | 22:14 |
| fungi | looks like it went through | 22:14 |
| TheJulia | \o/ | 22:14 |
| TheJulia | cool | 22:14 |
| TheJulia | Well, in that case I'm going to go find some lunch | 22:14 |
| fungi | killer plan | 22:14 |
| TheJulia | Thanks everyone! | 22:14 |
| fungi | thank you too! | 22:14 |
| gmann | TheJulia: perfect, it worked, I got the invite with ical | 22:14 |
| fungi | in the future i can send copies of those to the foundation-board ml in addition to openstack-discuss, if it helps | 22:15 |
| gmann | ++ | 22:15 |
| TheJulia | fungi: ++, that would be awesome, since it is easy for me to miss stuff or lag by a period because I'm pulled in numerous directions | 22:16 |
| fungi | i'm all too familiar with that feeling | 22:17 |
| fungi | just didn't want to step on any toes | 22:17 |
| TheJulia | :) It is a team effort! | 22:18 |
| fungi | more like group therapy ;) | 22:18 |
| TheJulia | heh | 22:18 |
| TheJulia | Anyway! lunch! | 22:19 |
| * TheJulia re-engages cloaking device labeled "vacation" | 22:19 | |
| fungi | you're as terrible at vacation as the rest of us | 22:19 |
| JayF | I've been fussing at TheJulia for planning on taking this meeting from Maui for weeks, she can attest | 22:21 |
| JayF | the sand on those beautiful beaches is jealous of the attention zoom just got LOL | 22:21 |
| TheJulia | eh, more like 1.5 weeks since I booked my hotel/flights not even 2 weeks ago | 22:22 |
| JayF | 1.5 is still weeks :D | 22:22 |
| fungi | back to warm sands and tasty drinks! | 22:22 |
| JayF | mmm delicious sand | 22:23 |
| * TheJulia shimmers away as only a trekie can | 22:23 | |
| spotz[m] | I just logged in no one was still on | 22:26 |
| fungi | spotz[m]: meeting started ~1.5 hours ago | 22:27 |
| fungi | just wrapped up | 22:27 |
| spotz[m] | I was at a dinner in Cork, we just ended | 22:27 |
| fungi | almost didn't make quorum, just got attendance over the minimum threshold about an hour into the call | 22:28 |
| spotz[m] | Yeah I got on when I could. It was 90 minutes until we got dinner. First meeting I’ve missed:( | 22:30 |
| fungi | well, my point is, it wasn't just you who couldn't make it. you're in good company | 22:31 |
| fungi | also in cork. i wouldn't have bothered dialling in at all if that were me! ;) | 22:31 |
| spotz[m] | Hehe PTO is next Week:) | 22:33 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!