Friday, 2022-10-07

*** rcastillo|ruck is now known as rcastillo03:47
ianwabout to restart gerrit with sha256:1da94630c56884365842c12ad47c336f5f5d3cdee979fb7249da531e63f8fd26" which is what 859885 pushed03:56
ianw#status log restarted gerrit  @ 3.5.3 (
opendevstatusianw: finished logging03:59
*** soniya29|ruck is now known as soniya2905:29
*** ysandeep|out is now known as ysandeep05:43
fricklerchecking upgrading devstack to use jammy I noticed that it currently simply inherits the nodeset from the base job. which brings up the question when we would consider switching the base job. would the beginning of this cycle be a good time or shall we wait another six months like we did with bionic->focal?05:51
*** anbanerj is now known as frenzyfriday06:38
*** ysandeep is now known as ysandeep|brb06:52
*** ysandeep|brb is now known as ysandeep07:06
*** arxcruz is now known as arxcruz|rover07:06
*** jpena|off is now known as jpena07:17
*** ysandeep is now known as ysandeep|lunch09:29
*** dviroel|afk is now known as dviroel10:33
*** dviroel is now known as dviroel|biab10:34
*** rlandy|out is now known as rlandy10:40
*** ysandeep|lunch is now known as ysandeep11:01
fungifrickler: i agree, switching to jammy real soon would be preferable11:28
fungimaybe worth putting on the meeting agenda to work out a timetable for announcements and such11:30
fricklerfungi: o.k., I'll start with a patch for base-test, then11:30
fungisounds good--thanks!11:30
*** dviroel|biab is now known as dviroel12:26
fricklercomparing base to base-test, these ara-related vars are missing from the latter, is there a reason for that or should I sync?
fricklerhas been this way for 3y , not sure if it would matter for the jammy test. guess I'll sync in order to be on the safe side12:35
opendevreviewDr. Jens Harbott proposed opendev/base-jobs master: Switch the base-test job to use ubuntu-jammy
fricklerthis leaves base-test using the test-prepare-workspace-git role, hope that's ok12:44
fungii think we no longer need any ara-related vars in our base job, that presumably dates from the days before zuul-web's console view12:56
fricklerfungi: so just drop them from base&-minimal instead?13:12
fungii would. since they're not included in base-test we've effectively already tested that they're not needed. probably worth doing as a separate cleanup change though and not incorporating into the jammy testing patch13:13
fungifrickler: do we need to rebase 848463 after 858376 or squash them?13:18
fungii think zuul is going to ignore 848463 because the queue fix isn't present there yet13:19
fungiso it won't be able to merge before 858376 which fixes the zuul config13:19
funginevermind, looks like it merged in july and i had stale review info in my gertty because i stopped subscribing to devstack-gate after we declared it unsupported13:20
opendevreviewDr. Jens Harbott proposed opendev/base-jobs master: Switch the base-test job to use ubuntu-jammy
opendevreviewDr. Jens Harbott proposed opendev/base-jobs master: Drop ara related vars from the base jobs
fricklerfungi: ds-gate should be fine now except that kopecmartin|sick wanted to find a job that uses it and still works, so that we can restore some functional testing. but maybe retirement will be faster13:23
*** dasm|off is now known as dasm13:24
fungiespecially since there's been a spate of broken em branches in openstack which has raised the possibility of at least dropping all their integration tests (many of which are probably using dg)13:27
*** rlandy is now known as rlandy|biab14:18
*** ysandeep is now known as ysandeep|afk14:20
*** ysandeep|afk is now known as ysandeep|out14:20
*** rlandy|biab is now known as rlandy14:48
fungiinfra-root: config-core: artifact signing certificate rotation patch would be good to get merged today (i've already given the #openstack-release channel):
clarkbfungi: you can probably go ahead and approve that one? its something we do regularly and is expected15:08
clarkbI haven't seen any complanits about the new default version of ansible. Has anyone else noticed problems?15:13
clarkbfrickler: ++ to moving forward wtih bumping our default test node to jammy15:13
fungiokay, single-core approving it in that case15:16
fungitaking a lunch outing, but will be back in an hour or so15:17
*** marios is now known as marios|out15:23
opendevreviewMerged openstack/project-config master: Replace old Zed cycle key with 2023.1/Antelope
*** dviroel is now known as dviroel|lunch15:29
clarkbI'm upgrading to linux 6 right now. Wish me luck. I think this has the amd speedups too, though I bet they are all so small that I won't notice15:44
clarkbwoo package install of some firmware fails15:52
clarkbafter manually moving some files around so that cpio could successfully write new content where old content existed I have rebooted and it seems functional16:02
clarkbthere is a new jitsi meet release. I think tonight's daily infra-prod-meetpad run will pick it up. More a heads up than anything else. Might be worth retesting on monday16:09
*** jpena is now known as jpena|off16:11
fricklerinteresting zuul message that I haven't seen before (next to last one), maybe related to changed gerrit behavior?
clarkbfrickler: that seems likely. I believe that message is generated by gerrit for the user clicking the submit button16:28
clarkbfrickler: looking at it more closely I think it is explaining why it carried the +2 code review over making the chagne eligible for merging.16:48
fricklerclarkb: yes, that's what I understood, too. I mostly stumbled about it because the formatting in the mail notification is kind of weird
clarkbfungi: I've managed to give some proper thought and left comments. I didnt' come up with better label names but did have suggestions for the behavior of the label itself16:55
*** dviroel|lunch is now known as dviroel16:55
fungigood enough16:56
clarkbfungi: did you see the recent mm3 performance problems thread on the mm users mailing list?17:03
clarkbI'm not terribly worried but it is probably somethign we'll need to monitor after moving the initial domains17:03
fungii didn't, since i've only been occasionally looking at their archives, not subscribed17:04
clarkbah, tldr is someone with ~350 lists and some lists with 1500 users ends up with mailman taking up to an hour to flush all messages out to the local mail server17:05
fungiwe definitely have lists with that many subscribers17:09
clarkbya unfortunately difficult to test without flooding people.17:11
clarkbI guess we could make a test list with 1500 different email addresses for ourselves17:11
clarkband then just file away/delete the 1500 emails we get17:11
fungior we just deal with it. i'm honestly not that concerned17:12
fungiopenstack-discuss is up to 1771 subscribers (openstack-dev was much higher before we retired it though)17:13
fungimy records say it had 5396 when i shut it down17:14
fungiopenstack-announce has 337917:14
fungithat's probably the highest subscriber count of any of the mailing lists we host at this point17:15
clarkbit might also be in their case that you need to have hundreds of lists all active at one. When we probably have 2-3 17:15
fungialso likely depends on system resources17:16
fungiour current mm2 server is pretty laggy because it's got 4gb ram and is also using around 1gb of swap17:17
fungiwhen we make a new server we should probably give it 8gb ram just because of all the parallel mta processes17:18
fungiduring heavy periods it seems to dip heavily into swap on the existing server17:19
clarkbfungi: should I go ahead and update the mm3 change?18:27
fungiclarkb: for the existing comments? i'm still working on it, about caught up now18:40
opendevreviewJeremy Stanley proposed opendev/system-config master: Add a mailman3 list server
opendevreviewJeremy Stanley proposed opendev/system-config master: WIP fork the maxking/docker-mailman images
*** dasm is now known as dasm|off19:32
*** dviroel is now known as dviroel|out21:24
opendevreviewJeremy Stanley proposed opendev/system-config master: Add a mailman3 list server
opendevreviewJeremy Stanley proposed opendev/system-config master: WIP fork the maxking/docker-mailman images

Generated by 2.17.3 by Marius Gedminas - find it at!