15:01:45 #startmeeting Fedora QA meeting 15:01:45 Meeting started Mon Sep 4 15:01:45 2017 UTC. The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:01:45 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:01:45 The meeting name has been set to 'fedora_qa_meeting' 15:01:50 #meetingname fedora-qa 15:01:50 The meeting name has been set to 'fedora-qa' 15:01:53 #topic Roll Call 15:01:58 morning folks, who's around? 15:02:15 * tenk here 15:04:19 * kparal is here 15:04:36 ahoy, tenk and kparal 15:05:12 hello adamw, i am back to japan so can only stay for this one (already midnight here) 15:08:10 ah, indeed :/ 15:08:26 hey, isn't japan a work-all-night kinda place? take your laptop to the izakaya and stick with us :P 15:08:58 well, we're a little thin, but never mind, let's blow through things for the record 15:09:03 #topic Previous meeting follow-up 15:10:29 welp, these are all for sumantro, but i believe i know the status 15:10:47 #info "sumantro to check in with design team and make sure wallpapers will be available in time for Beta and Final" - sumantro says he did this, and they promised they'll be in shape this time 15:11:23 #info "sumantro to send out reminder about LxQT test day" - he did that, and the event has concluded 15:11:50 #info "sumantro to contact i18n team to make sure their regular test days are being planned, and atomic/cloud team" - he definitely checked in with g11n team and they're on the same page, not sure about atomic/cloud 15:12:25 any other stuff for follow-up? 15:12:40 2s 15:13:11 for the other testday do we have a status / something i can do? 15:13:24 i check the issue on pagure but nothing new 15:13:41 which one, sorry? 15:13:47 i'm mostly leaving test days to sumantro these days 15:14:47 all, ok will try to contact him again (probalby busy with folck and is nommination ;) 15:15:16 flock* 15:16:26 yeah, i think he may be travelling 15:16:46 i know he was blocking on the change maintainers for some test days 15:16:57 on the feature actually being ready for testing 15:17:12 ok thx 15:18:53 #topic Fedora 27 status 15:19:02 soooo, checking in on f27...my understanding of current status is: 15:19:26 #info Beta freeze is at 00:00 UTC tonight (00:00 on 2017-09-05) 15:19:31 openqa doesn't seem to reboot into the installed system, but rather boots from the dvd again 15:19:36 kparal: for uefi installs, yes 15:19:44 ok 15:19:50 kparal: it's https://bugzilla.redhat.com/show_bug.cgi?id=1487856 15:20:30 somebody broke stuff hard with shim 15:20:32 #info currently composes are succeeding, but with one major bug - https://bugzilla.redhat.com/show_bug.cgi?id=1487856 prevents all UEFI installs from booting properly - and several other blocker bugs (e.g. FreeIPA issues) 15:20:44 kparal: pjones renamed a bunch of packages and files to allow for 64-on-32 booting 15:21:07 then we, uhh, used a process of proactive discovery to figure out what other things needed changing in line with those changes 15:21:19 nice term :) 15:21:22 (by which i mean we kept firing composes and prodding things) 15:21:57 #info further testing beyond openQA may have been hampered by the frequent compose failures and broken composes till now 15:22:39 in fact uefi bit is broken only the dvd compose it seems 15:22:43 #info for now, the planned schedule still holds, which means we need to do a lot of testing to get up to the Beta criteria by Beta go/no-go on 2017-09-14 15:22:59 wow 15:23:00 I could install WS on uefi vm 15:23:14 dominicpg: *install* works 15:23:25 the boot manager entry that's written points to the wrong executable 15:23:40 what happens after that *may* depend a bit on the machine (it's to do with the UEFI fallback path) 15:23:58 in openqa, the installed system doesn't boot, it just boots back to the installer 15:24:09 in other circumstances it's possible the UEFI fallback path might kick in and work 15:25:23 yes, that's right. I noticed the loop ( install works but boot failed ) only when tried latest - 20170901n1 - spin . At the same, WS build installs and boots . 15:25:38 live, or netinst? 15:25:43 live 15:26:03 live worked well. so the problem seems to be only on server variant. 15:26:43 .hello alciregi 15:26:44 alciregi: alciregi 'Alessio Ciregia' 15:26:53 dominicpg: there's no reason at all that should be the case 15:27:00 dominicpg: the bootloader install code does not differ at all between the variants 15:27:15 so, not sure what's going on there, could be some difference in test process. in any case, we definitely know what the bug is... 15:28:18 yes, I agree and understand. anyway I will do test it again on my kvm vm 15:28:40 thanks 15:28:52 anyone got anything else on f27? 15:28:53 :) 15:30:01 I don't 15:30:03 rgr 15:30:38 btw what's the plan for system-config-date 15:30:51 dominicpg: there's going to be another anaconda build with the intended fix for the uefi bug soonish (I hope), so let us know how things go for you with the next compose after that... 15:30:58 I couldn't install s-c-kickstart due to an ugly dependency 15:31:08 ok 15:31:12 is there a bug filed? 15:31:17 oh yeah, i see it 15:31:31 https://bugzilla.redhat.com/show_bug.cgi?id=1483667 15:31:34 yes ^ 15:32:31 what about bug 1487305: 15:32:39 ops, sorry 15:32:59 also, I think it's not good idea to have the first user in %wheel when created using initial setup. filed https://bugzilla.redhat.com/show_bug.cgi?id=1487855 15:33:46 dominicpg: I'll update both bugs 15:33:58 thx 15:34:58 guess I can close https://bugzilla.redhat.com/show_bug.cgi?id=1482881 against build 20170815n1 15:35:07 not witnessed such error anymore . 15:35:27 alciregi: hmm, that looks ba 15:35:28 bad* 15:35:31 thanks for filing it 15:36:08 alciregi: rasbpi 3 is a supported platform for f27, so that could be nominated as a beta blocker 15:36:48 folks on fedora-arm channel suggested me to file the bug 15:38:00 though I had installed f26 when it released, I didn't notice *cosmetic* bug - https://bugzilla.redhat.com/show_bug.cgi?id=1482438 , sigh . this is on track. thx 15:38:19 alciregi: i nominated it as a blocker, thanks again for the test 15:38:45 adamw: you are welcome 15:38:49 dominicpg: heh, that's a funny one :) 15:38:52 good catch 15:39:04 :D 15:39:37 alrighty then 15:41:18 #topic Kernel rebase testing plan 15:41:36 so these next two are ideas from sumantro that came up at flock, i was hoping he'd be here to talk about them, but i can run through them 15:41:43 so here's the skinny on this one, as I understand it 15:42:10 #info new upstream kernel releases are pushed into stable Fedora releases, at present. This can be a source of regressions 15:42:42 #info Plan is to run a Test Day event for each new kernel release to try and identify and fix problems before they get out into the stable releases 15:43:02 even minor releases? 15:43:33 #info sumantro will be co-ordinating this with the kernel team, the Test Day will involve running the kernel test suite (see https://fedoraproject.org/wiki/KernelTestingInitiative ) and generally checking for issues 15:43:42 kparal: no, i mean the 4.x releases, not 4.x.y 15:43:48 so 4.13, 4.14, 4.15 etc. 15:44:16 i believe this means we get an event around each 3 months, so watch out for those! 15:44:31 sounds reasonable, if we can get enough attendance and sumantro is willing to do test days for these 15:44:49 however, a similar approach can be also done with bodhi - just disable autokarma and let it sit there for two weeks or more 15:45:04 +1 15:45:13 the issue with bodhi is, we don't know for sure what people are testing 15:45:24 and sumantro thinks having an event will get more folks involved 15:45:31 oh, and the other thing is, the timing here will be different 15:45:48 it'll happen long *before* the kernel is actually proposed as an update, back in the time where we can actually get things fixed upstream 15:46:26 ok, let's try it 15:46:53 yup 15:47:26 #topic Activity meeting plan 15:48:07 #info sumantro says that many of the newer folks he's mentoring see this meeting as more of an administrative/discussion meeting, and don't really want to jump in with practical questions etc. 15:49:02 #info so he proposes having a regular meeting (I think the idea was monthly) focused on practical testing activity, selecting specific things that need testing at the time and focusing on those 15:49:32 with the benefit that people can work together in real time, an event should drive participation, and we can take advantage of it to drive testing to needed areas (e.g. focusing on some neglected validation tests) 15:50:00 not the principle of testing day? 15:50:56 well it'd be similar to a test day, i guess, but it'd be recurring on some kind of dependable schedule, and for the stuff we need to do on an ongoing basis (validation testing, update testing) - i think 15:51:10 but it's a good question, would be good to have sumantro's answer 15:52:46 any other comments / questions on that one? 15:53:00 again, sounds like a good idea 15:53:12 let's see how many people joins 15:53:37 indeed, no specific action required here, just wanted to give a heads-up on the ideas - just watch out for the new events :) 15:53:41 #topic Open floor 15:53:50 so, one note, i'll be trying to put the 'no more alphas' changes into production this week 15:54:05 things have been a little chaotic in f27 lately, but i'm hoping i'll be able to get back to that 15:55:28 any other notes / thoughts? 15:55:39 i am not sure but i don t have seen heroes of fedora for Fedora 26 Final, i miss it or the place of publication change? 15:55:53 it hasn't been done yet 15:55:59 i think sumantro volunteered at flock 15:56:03 so hopefully we'll get one up soon 15:56:14 #info Heroes of Fedora 26 is not done yet, sumantro should be working on it soon 15:56:30 generating it is a bit more work than it should be, the tools are kinda janky 15:57:08 it's not a blocker for F27 15:57:10 XD 15:57:27 ;-) 15:59:44 hehe 15:59:57 fedora 27 release is delayed due to insufficient appreciation awarded for f26 15:59:58 :P 16:00:50 alrighty, sounds like we're done here 16:00:53 thanks for coming, folks 16:01:02 we're going to attempt to run a blocker review meeting in #fedora-blocker-review, if we can get enough folks 16:01:05 so join us there! 16:01:07 #endmeeting