Friday, 2020-01-17

*** mattw4 has quit IRC00:25
openstackgerritmelanie witt proposed openstack/grenade master: DNM: testing grenade zuulv3 in nova  https://review.opendev.org/70300500:28
*** openstackgerrit has quit IRC00:57
gmannefried: no, those are failing on scenario tests with no valid host. I did not check root cause but not related to network things.02:12
*** rfolco has joined #openstack-qa02:57
*** rfolco has quit IRC03:02
*** rh-jelabarre has quit IRC03:22
*** openstackgerrit has joined #openstack-qa03:55
openstackgerritGhanshyam Mann proposed openstack/tempest master: Install Tempest & plugins always on py3  https://review.opendev.org/70301103:55
*** vishakha has joined #openstack-qa04:25
*** raukadah is now known as chandankumar04:46
*** udesale has joined #openstack-qa04:49
*** dustinc is now known as dustinc|PTO04:52
*** rpittau|afk has quit IRC05:34
*** evrardjp has quit IRC05:34
*** evrardjp has joined #openstack-qa05:34
*** rpittau|afk has joined #openstack-qa05:34
*** surpatil has joined #openstack-qa05:48
*** soniya29 has joined #openstack-qa05:49
*** udesale has quit IRC05:56
*** udesale has joined #openstack-qa05:56
*** soniya29 has quit IRC06:00
*** gcheresh has joined #openstack-qa06:03
*** gcheresh has quit IRC06:11
*** gcheresh has joined #openstack-qa06:18
*** SurajPatil has joined #openstack-qa06:36
*** SurajPatil has quit IRC06:37
*** surpatil has quit IRC06:40
*** soniya29 has joined #openstack-qa06:42
*** surpatil has joined #openstack-qa06:49
openstackgerritSoniya Murlidhar Vyas proposed openstack/tempest master: Added fix to 'test_delete_saving_image' testcase  https://review.opendev.org/70280806:56
*** gcheresh has quit IRC07:08
*** lee1 has joined #openstack-qa07:13
*** ramishra has quit IRC07:18
*** trident has quit IRC07:18
*** lifeless has quit IRC07:19
*** mtreinish has quit IRC07:19
*** tbarron has quit IRC07:19
*** ianw has quit IRC07:19
*** mjturek has quit IRC07:19
*** lyarwood has quit IRC07:19
*** zigo has quit IRC07:19
*** rnoriega_ has quit IRC07:19
*** brtknr has quit IRC07:19
*** openstackstatus has quit IRC07:20
*** paras333 has joined #openstack-qa07:27
*** slaweq has joined #openstack-qa07:35
openstackgerritParas Babbar proposed openstack/tempest master: Fix test_stamp_pattern to remove dependency of user supplied volume name  https://review.opendev.org/69859807:38
openstackgerritSoniya Murlidhar Vyas proposed openstack/tempest master: Added fix to 'test_delete_saving_image' testcase  https://review.opendev.org/70280807:55
*** ramishra has joined #openstack-qa08:05
*** trident has joined #openstack-qa08:05
*** lifeless has joined #openstack-qa08:05
*** mtreinish has joined #openstack-qa08:05
*** tbarron has joined #openstack-qa08:05
*** ianw has joined #openstack-qa08:05
*** mjturek has joined #openstack-qa08:05
*** zigo has joined #openstack-qa08:05
*** rnoriega_ has joined #openstack-qa08:05
*** brtknr has joined #openstack-qa08:05
openstackgerritParas Babbar proposed openstack/tempest master: Fix test_stamp_pattern to remove dependency of user supplied volume name  https://review.opendev.org/69859808:12
*** paras333 has quit IRC08:13
*** tesseract has joined #openstack-qa08:16
*** rpittau|afk is now known as rpittau08:28
*** tosky has joined #openstack-qa08:29
*** iurygregory has joined #openstack-qa08:32
*** dmellado has quit IRC08:34
*** dmellado has joined #openstack-qa08:35
*** lee1 is now known as lyarwood08:36
*** gfidente has joined #openstack-qa08:38
*** jpena|off is now known as jpena08:48
*** ralonsoh has joined #openstack-qa08:50
*** iurygregory_ has joined #openstack-qa09:01
*** iurygregory has quit IRC09:03
openstackgerritLuigi Toscano proposed openstack/devstack stable/stein: Fix MySQL log collection  https://review.opendev.org/70303909:03
*** rcernin_ has joined #openstack-qa09:04
*** rcernin has quit IRC09:04
openstackgerritAlfredo Moralejo proposed openstack/devstack master: [dnm] testing centos-8 bringup  https://review.opendev.org/68861409:09
*** lucasagomes has joined #openstack-qa09:16
fricklerianw: I'd like to see your opinion on these two when you have time https://review.opendev.org/702960 https://review.opendev.org/70257009:29
*** dtantsur|afk is now known as dtantsur10:06
*** ccamposr__ has quit IRC10:18
*** rcernin_ has quit IRC10:34
*** ccamposr has joined #openstack-qa10:41
*** bhagyashris has joined #openstack-qa11:01
*** soniya29 has quit IRC11:01
*** dviroel has joined #openstack-qa11:10
*** bhagyashris has quit IRC11:12
*** ccamel has joined #openstack-qa11:16
*** camelCaser has quit IRC11:16
openstackgerritStephen Finucane proposed openstack/devstack master: inc/python: Remove ability to mark packages as non-Python3  https://review.opendev.org/68758411:19
openstackgerritStephen Finucane proposed openstack/devstack master: stackrc: Change default Python 3 version  https://review.opendev.org/68758811:19
*** bhagyashris has joined #openstack-qa11:25
*** soniya29 has joined #openstack-qa11:27
*** surpatil has quit IRC11:33
*** pcaruana has joined #openstack-qa11:59
*** paras333 has joined #openstack-qa12:00
*** paras333 has quit IRC12:01
*** paras333 has joined #openstack-qa12:02
openstackgerritVictoria Martinez de la Cruz proposed openstack/devstack-plugin-ceph master: Updates script to use latest version of Ceph  https://review.opendev.org/67672212:17
openstackgerritParas Babbar proposed openstack/tempest master: Fix test_stamp_pattern to remove dependency of user supplied volume name  https://review.opendev.org/69859812:18
*** rfolco has joined #openstack-qa12:27
*** iurygregory_ is now known as iurygregory12:40
*** dmellado has quit IRC12:41
*** dmellado has joined #openstack-qa12:44
*** udesale_ has joined #openstack-qa12:45
*** jpena is now known as jpena|lunch12:46
*** udesale has quit IRC12:48
*** rh-jelabarre has joined #openstack-qa12:51
openstackgerritAlfredo Moralejo proposed openstack/devstack master: [dnm] testing centos-8 bringup  https://review.opendev.org/68861412:53
openstackgerritSlawek Kaplonski proposed openstack/tempest master: Fix test_subnet_details scenario test case  https://review.opendev.org/70307213:13
*** gfidente has quit IRC13:27
*** ricolin has joined #openstack-qa13:30
*** dave-mccowan has joined #openstack-qa13:38
*** jpena|lunch is now known as jpena13:39
*** gfidente has joined #openstack-qa13:42
*** pcaruana has quit IRC13:45
*** soniya29 has quit IRC13:47
*** yamamoto has joined #openstack-qa13:59
*** nweinber__ has joined #openstack-qa14:11
*** pcaruana has joined #openstack-qa14:22
*** paras333 has quit IRC14:30
*** dave-mccowan has quit IRC14:32
*** dave-mccowan has joined #openstack-qa14:34
*** paras333 has joined #openstack-qa14:37
paras333afazekas:  gmann : Hey Guys, whenever you get a chance, Can you take a look on this patch https://review.opendev.org/698598 I have updated according to the review provided:)14:59
efriedmasayukig: https://review.opendev.org/#/c/702553/ is passing, would you please merge it?15:04
efriedbetween that and https://review.opendev.org/#/c/702960/ nova's gate is ill :(15:05
openstackgerritLee Yarwood proposed openstack/devstack-gate master: WIP Remove g-api from subnodes  https://review.opendev.org/70309915:09
gmannefried: he might be sleeping. afazekas for you - https://review.opendev.org/#/c/702553/15:09
efriedthanks gmann15:09
openstackgerritAlfredo Moralejo proposed openstack/devstack master: [dnm] testing centos-8 bringup  https://review.opendev.org/68861415:24
*** yamamoto has quit IRC15:28
gmannfrickler: i agree with efried on 702960/.15:29
gmannthere is patch to remove g-api form subnode but i do not know why it was there at first place so cannot tell any side effect of that but doing protection on devstack side is not so bad15:30
gmann"touch DB stuff if DB service on node"15:30
*** yamamoto has joined #openstack-qa15:44
*** yamamoto has quit IRC15:45
*** yamamoto has joined #openstack-qa15:45
*** udesale_ has quit IRC16:01
*** udesale has joined #openstack-qa16:03
*** udesale has quit IRC16:10
openstackgerritLee Yarwood proposed openstack/tempest master: Add boot from encrypted snapshot test  https://review.opendev.org/69554916:13
*** slaweq has quit IRC16:16
*** mattw4 has joined #openstack-qa16:20
openstackgerritAlfredo Moralejo proposed openstack/devstack master: [dnm] testing centos-8 bringup  https://review.opendev.org/68861416:29
*** udesale has joined #openstack-qa16:37
*** yamamoto has quit IRC16:40
*** paras333 has quit IRC16:41
*** rpittau is now known as rpittau|afk16:53
openstackgerritRajat Dhasmana proposed openstack/tempest master: Refactor methods to be used by other tests  https://review.opendev.org/70249316:54
*** dtantsur is now known as dtantsur|afk16:58
*** lucasagomes has quit IRC16:59
*** udesale has quit IRC17:00
*** stephenfin has joined #openstack-qa17:15
openstackgerritStephen Finucane proposed openstack/devstack-gate master: Stop installing g-api on subnodes  https://review.opendev.org/70312917:15
stephenfinfrickler, gmann, (artom, efried): ^17:15
stephenfinAs an alternative to https://review.opendev.org/#/c/702960/17:16
stephenfinI'm still onboard with https://review.opendev.org/#/c/702960/ too though, fwiw, given the follow-up comments17:16
stephenfinand https://review.opendev.org/703129 probably needs some input from the glance folks17:16
artomstephenfin, I dunno - talk with clarkb, he was saying g-api on the subnode makes sense for grenade upgrade testing17:17
efriedstephenfin: note https://review.opendev.org/#/c/703099/17:17
stephenfingdi17:17
stephenfinThanks, efried17:17
efrieddifferent approach though. lyarwood care to comment?17:17
*** yamamoto has joined #openstack-qa17:17
stephenfinwe need to combine the two17:18
stephenfinlyarwood fixed the tests, I fixed the code17:18
efriedbut lyarwood abandoned because apparently some part of having g-api on subnode could be legit.17:18
efriedthough I don't claim to understand the nuances.17:19
stephenfinhow can we prove that, I wonder?17:19
openstackgerritStephen Finucane proposed openstack/devstack-gate master: Stop installing g-api on subnodes  https://review.opendev.org/70312917:19
efriedMy priority is merging https://review.opendev.org/#/c/702960/ asap to get the gate moving. Then we can debate the other thing at leisure.17:19
efriedand by "we" I mean "people who know things"17:20
artomI'm with efried - the devstack change isn't insane AFAICT, and people know know more seem to agree17:20
clarkbartom: well it was one reason it may make sense at least17:20
clarkbartom: I think we should ask the glance team if that is somethng they were trying to test with grenade17:20
clarkbefried: ^17:20
*** chandankumar is now known as raukadah17:20
stephenfincan we just revert the original patch and debate all this next week, as efried suggested in the thread?17:21
clarkbas a general rule that is why you might find "odd" things installed on the subnode. But we should still confirm it was itnentional17:21
stephenfinon openstack-discuss, maybe. There's clearly some nuance to be worked out17:21
*** tosky has quit IRC17:21
efriedwhere's frickler geographically? If past EOB, I sure don't want to wait another day to get things moving.17:22
stephenfinrevert seems like the approach so, assuming others use a similar strategy to nova17:22
stephenfin(if it breaks stuff, pull it out pronto and rework at our leisure)17:22
efriedI'd be happy to propose a revert, adding to the spread of patches addressing the issue. yoctozepto, would you be behind that?17:24
efriedI mean, it's not like we're at feature freeze or anything, but that doesn't mean it's okay to block nova's gate indefinitely while we come up with a perfect fix.17:26
openstackgerritStephen Finucane proposed openstack/devstack master: Revert "Run Glance initialization when Glance is enabled, not just registry"  https://review.opendev.org/70313117:26
*** yamamoto has quit IRC17:26
yoctozeptowell, folks, I fixed the issue, not happy with doing 2 reverts17:26
efriedI don't disagree. The actual-revert would only be for the sake of being able to defer addressing frickler's objection to your other proposed fix.17:27
stephenfinyeah, but it does buy us time to debate that properly17:27
yoctozeptothere is nothing to debate from my pov17:27
yoctozeptonova, cinder and neutron can do it and glance not17:28
yoctozeptosounds ridiculous17:28
gmannefried: he (frickler ) is berlin based if i am not wrong so not sure he will be online.17:28
stephenfinbut there is from others, so that debate has to be had17:28
stephenfinwe can't cram something in without giving them a chance to come around17:28
stephenfinso let's try talk frickler around next weekend, like decent people17:29
gmanndisabling g-api in greande subnode need glance team input which is not so active i will say so that will take time.17:29
stephenfin*week17:29
stephenfinagreed17:29
yoctozeptowell, you can always revert the fix if it's not the right one17:29
efriedRight, so with my PTL hat on, if frickler isn't around to debate and/or remove his -1, we need to merge *something* to get nova's gate moving. Whatever that something is, we'll have to agree on with those present. Perhaps that means overruling frickler's -1 and merging that patch anyway. Or perhaps it means a revert of the original.17:29
gmannhe will agree i think, it is just matter of when he come back online17:29
stephenfinefried: it's always a revert in nova land17:30
efriedI don't have a technical hat in this ring, so I really don't care which patch we merge.17:30
efriednor do I especially care to participate in the longer discussion, happy to leave that for people who can spell grenade etc.17:31
gmannyoctozepto: revert is safest and fastest option here to unblock nova.17:31
efried^ and unrevert later once we have agreement17:32
stephenfinagreed17:32
gmannyeah.17:32
stephenfinI don't disagree with yoctozepto's follow-up patch. I just don't want to overrule frickler when they're not here to defend themselves17:32
gmann+117:32
gmannwe have to unblock gate as priority than what is best possible way to solve the occurred issue.17:33
stephenfinyoctozepto: Are you okay with that. Once everyone is back on Monday we can get the confusion sorted out, unrevert the patch and merge your fix17:33
stephenfinnova's immediate issues are fixed and no one feels they were pushed aside17:33
gmannstable queens and stable rocky is also blocked for neutron and that is another complex situation on how to fix it as best, so let go one by one.17:34
*** evrardjp has quit IRC17:34
*** evrardjp has joined #openstack-qa17:34
yoctozeptoyeah, I'm not going to be super stubborn on this (but I can!), I just feel like the situation went ridiculous due to hasty review from frickler, and I don't mean it in a bad way - I did bad -1s myself and blocked legit work17:35
yoctozeptoso go whatever works and is accepted in the community :-)17:35
yoctozepto(and I get a free commit by reverting a revert next week)17:36
artomyoctozepto, it's not a personal thing against the quality of your fix. I totally get that you made a patch, saw it had unintended consequences, and quickly came up with a fix, only to have it summarily shut down. It sucks, I get it :)17:36
gmannreal issue is lack of jobs coverage on devstack and tempest gate where one fix trigger other issue.17:36
artomWe just want to get Nova merging code again17:36
yoctozeptogmann: true17:36
efriedI'm assuming frickler didn't realize this was a gate blocker before dropping a -1 and leaving for the dap :P17:36
efriedday17:36
yoctozeptoartom: who wouldn't? :D17:36
gmannefried: exactly.17:37
efriedgmann: we could make the n-l-m job voting in devstack17:37
gmannit is just matter of he is not around.17:37
yoctozeptoefried: true17:37
efriedit's there and running, just not voting.17:37
gmannefried: i am ok with that, we have same in tempest gate. i will +2 once you push that17:37
stephenfinI guess the real solution is to finish migrating to zuul v3 and kill d-g17:37
stephenfinWIP though17:37
gmannefried: wait, you said it is running already ?17:38
efriedyup, we definitely know we want to do that, but it turns out to be hard.17:38
artomstephenfin, the real solution is to put everything and containers and rnu k8s, amirite?17:38
gmanni think its not on devtack gate17:38
efriedum, I thought it was gmann17:38
efriedI thought I saw it (red) on the original patch...17:38
gmannok, adding n-v also work17:38
stephenfink8s will save humanity. Praise k8s.17:38
efriednope, I'm wrong17:38
efriedadding n-v doesn't tend to work, red nonvoting jobs get ignored.17:39
gmannyoctozepto: let's go for the revert and we will re-revert your patch for sure17:39
yoctozeptonah, it's not running there17:39
yoctozeptogmann: yeah, agreed17:39
yoctozeptoalready +117:39
stephenfingmann: Waiting on your approval https://review.opendev.org/#/c/703131/17:39
stephenfinIf there are no other devstack cores online, it might need to be a single +2 approval too /o\17:40
yoctozeptofolks, wait, you did not revert the 2nd change17:40
yoctozeptothis will fail17:40
stephenfinwdym?17:40
stephenfinah d7dfcdb4674daae8a294848b1de6fa87c5d7d4eb17:40
gmann+A, i thought it is not up yet17:40
gmannstephenfin: i fast approve.17:41
yoctozeptohttps://review.opendev.org/70270917:41
yoctozeptomhm17:41
yoctozeptofirst revert this one17:41
yoctozeptoor do a double revert17:41
efriedI'm not sure I understand why that's necessary, but if you say so.17:42
yoctozeptowell, if you leave it in place, then all jobs will fail due to glance never being initialized17:42
openstackgerritStephen Finucane proposed openstack/devstack master: Revert "Run Glance initialization when Glance is enabled, not just registry"  https://review.opendev.org/70313117:42
openstackgerritStephen Finucane proposed openstack/devstack master: Revert "Stop enabling g-reg by default"  https://review.opendev.org/70313717:42
yoctozeptothat was the point of it - to stop having g-reg around17:42
efriedoh, *other* jobs17:42
stephenfindone17:42
efriedyeah, I guess we should care about that :P17:42
stephenfinyoctozepto: good catch, thanks17:42
yoctozeptoefried: well, *yours* too :P17:43
yoctozeptostephenfin: yw17:43
efriedokay.17:43
yoctozeptogmann: now need two approvals17:44
gmannyoctozepto: efried stephenfin sure, waiting for the gate results now to confirm the other jobs results.17:45
yoctozeptosure17:45
*** jpena is now known as jpena|off17:45
stephenfinand on that good note, it's 17:45 on Friday evening and I am outta here 🍻17:46
stephenfinhave a good weekend, all o/17:46
gmannstephenfin: you too.17:46
gmannyoctozepto: thanks17:46
efriedthanks for the help stephenfin17:48
gmannyoctozepto: now on other bug on EOLing py2 drama. I am going to try py3.6 install on xenial first. if that does not work, i will pin tempest and plugins with py2-supported version with corresponding u-c.17:48
yoctozeptogmann: makes sense17:48
yoctozeptothis really is drama17:49
gmannqueens is already EM so pin is all as per policy and rocky is supposed to be EM on feb 24th so it will not be worst to pin by 1 month early.17:49
yoctozeptoso maybe just pin it and off it goes?17:50
gmannstein onwards works well as use bionic there17:50
gmannwe need to trigger some logic to pin plugins, for tempest we have a global var17:51
*** gfidente has quit IRC17:51
yoctozeptowell, if we want to keep testing stables with master then the most future-proof would be separate node17:52
yoctozeptoor containerization17:52
yoctozeptothen tempest is completely outside the effects of distro it tests17:52
yoctozepto(except for kernel if container which is irrelevant for tempest)17:52
gmannyeah that will be more work. we need to test only non-EM branch with master and all EM should be with corresponding tag if master does not work.17:54
yoctozeptoyeah, but, as shown, those non-em stables could still be a pita ;-)17:55
gmannso here py2 drop is valid reason to not to use master. testing stable branch py2 with py2-supported tempest.17:55
gmannit is req today but any py2 non-compatible code makes master tempest non testable for stable branch (EM or non-EM).17:56
gmannI was planing to switch that by Feb mid but it trigger little early due to this bug17:56
yoctozeptolater we might be experiencing py3.x and (fast forward) py4 issues17:58
yoctozeptoso for this policy it really mandates more separation of tempest17:58
yoctozeptowe actually have a plan to run tempest in kolla-ansible17:59
yoctozeptobut always something more urgent :-)17:59
gmannok, Tempest and plugins kolla image is there right17:59
gmanni think it was running on some job, chandan did some time back but not sure18:00
gmannsome job in kolla18:00
yoctozeptonot sure either, not tested nowadays18:00
yoctozeptoit builds but no run in CI18:00
gmannohk18:02
yoctozeptogmann: so - are you going with the pin for rocky?18:09
gmannyoctozepto: yes, rocky and queens. but after py3.6 try18:09
yoctozeptoahh, ok18:10
yoctozeptoif it's easy, then why not18:10
*** ralonsoh has quit IRC18:11
*** dave-mccowan has quit IRC18:23
*** dtroyer has joined #openstack-qa18:29
*** pcaruana has quit IRC18:51
*** gcheresh has joined #openstack-qa19:22
*** tesseract has quit IRC19:32
*** gcheresh has quit IRC19:34
*** tosky has joined #openstack-qa19:41
*** openstackstatus has joined #openstack-qa20:01
*** ChanServ sets mode: +v openstackstatus20:01
*** dtroyer has quit IRC20:21
*** dtroyer has joined #openstack-qa20:28
*** rh-jelabarre has quit IRC20:36
*** nweinber__ has quit IRC20:36
*** yamamoto has joined #openstack-qa21:24
*** yamamoto has quit IRC21:29
*** dave-mccowan has joined #openstack-qa21:30
*** gcheresh has joined #openstack-qa21:44
*** brinzhang has quit IRC21:50
*** gcheresh has quit IRC22:06
*** rh-jelabarre has joined #openstack-qa22:33
*** slaweq has joined #openstack-qa22:50
*** slaweq has quit IRC23:14
*** dave-mccowan has quit IRC23:15
*** rh-jelabarre has quit IRC23:27
*** mattw4 has quit IRC23:57
*** dviroel has quit IRC23:57
*** rcernin_ has joined #openstack-qa23:59

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!