16:00:42 <adamw> #startmeeting Fedora QA meeting
16:00:42 <zodbot> Meeting started Mon Feb 27 16:00:42 2017 UTC.  The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:42 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
16:00:42 <zodbot> The meeting name has been set to 'fedora_qa_meeting'
16:00:45 <adamw> #meetingname fedora-qa
16:00:46 <zodbot> The meeting name has been set to 'fedora-qa'
16:00:49 <adamw> #topic Roll call
16:00:54 <adamw> morning folks, who's around for the QA meeting?
16:00:58 * pschindl_ is here
16:01:08 * prakash is here
16:01:09 * sumantrom is here
16:01:16 * garretraziel is here
16:01:34 * Southern_Gentlem here
16:01:51 * coremodule is here.
16:01:51 * roshi is here
16:02:29 * nirik is not here.
16:03:10 * siddharthvipul is here
16:04:04 <adamw> morning everyone
16:04:16 <siddharthvipul> good morning :)
16:04:22 <siddharthvipul> night here though
16:04:29 <adamw> PST is the one true timezone!
16:04:34 <sumantrom> morning adamw :D
16:04:35 <prakash> good morning :) :P
16:04:41 <adamw> see, sumantro knows ;)
16:04:54 <siddharthvipul> :p
16:05:09 <adamw> alrighty, let's get going then
16:05:14 <adamw> #topic Previous meeting follow-up
16:05:21 <adamw> what didn't adamw do last week
16:05:26 * handsome_pirate shows up for the first time in forever
16:05:45 <adamw> morning pirate
16:06:18 <adamw> #info "adamw to follow up with releng on compose failure" - i forget precisely which compose failure that was, there were a few, but anyhow, we worked through several and composes are now happening again
16:06:30 <adamw> #info "adamw to look into Kerberos credential cache Change and see if it's significant" - whoops, haven't done that yet
16:06:35 <adamw> #action adamw to look into Kerberos credential cache Change and see if it's significant
16:06:47 <adamw> next up, we have: "roshi and cmurf to work with Atomic WG to try and get the PRD and wiki space updated, so we can then think about how to derive requirements and test processes from it"
16:06:58 <adamw> roshi: how's that going?
16:07:21 * tflink shows up late
16:07:27 * adamw brb
16:07:40 <roshi> it's up
16:07:44 <roshi> comments welcome :)
16:07:47 * roshi finds the link
16:08:32 <roshi> https://fedoraproject.org/w/index.php?title=Atomic/PRD
16:09:30 <roshi> cloud/atomic meetings are on wednesday, so any questions I can bring up there
16:11:04 <adamw> awesome, thanks
16:11:36 <adamw> #info Atomic PRD (draft?) is up at https://fedoraproject.org/wiki/Atomic/PRD , please review and send along any comments you have
16:11:38 <roshi> or we can influx the meeting with QA folks and we can all ask there :P
16:11:39 <adamw> grr
16:11:40 <adamw> #undo
16:11:40 <zodbot> Removing item from minutes: INFO by adamw at 16:11:36 : Atomic PRD (draft?) is up at https://fedoraproject.org/wiki/Atomic/PRD , please review and send along any comments you have
16:11:57 <adamw> #info "roshi and cmurf to work with Atomic WG to try and get the PRD and wiki space updated, so we can then think about how to derive requirements and test processes from it" - Atomic PRD (draft?) is up at https://fedoraproject.org/wiki/Atomic/PRD , please review and send along any comments you have
16:12:08 <adamw> alright, the other action item was also Atomic PRD-related so we can probably skip that
16:12:33 <adamw> #topic Fedora 26 status
16:12:58 <adamw> so f26 has been kind of on a rollercoaster lately
16:13:26 <adamw> lots of compose problems and a python bug killing the installer. but the last couple of days seem to have had full-ish composes that pass most tests
16:14:56 <roshi> sweet
16:15:07 <adamw> #info we had a couple of weeks of missing or broken composes, but we seem to have better results in the last few days
16:15:09 * kparal is late
16:15:13 <adamw> morning kparal
16:15:27 <adamw> #info Fedora 26 Branch point is tomorrow (2017-02-28)
16:15:54 <sumantrom> adamw , this is an awesome news :)
16:15:56 <adamw> for those who don't know, that means tomorrow is the day F26 diverges from Rawhide and we'll start getting daily F26 builds; Rawhide at that point is rolling towards F27.
16:15:56 <roshi> party time!
16:16:14 <adamw> it's also a point on the Change schedule and means we can start setting up F26 test  days
16:16:32 <sumantrom> adamw , yes ! :D
16:16:34 <coremodule> We can do that.
16:16:50 <sumantrom> coremodule yes ! :)
16:17:00 <prakash> great :)
16:17:17 <siddharthvipul> great..test days! My time has come
16:17:25 <adamw> haha :)
16:17:29 <adamw> good to see everyone's excited
16:17:44 <handsome_pirate> so, that means Wednesday I'll be switching to F26 :)
16:17:49 * handsome_pirate woots
16:17:51 <adamw> we still have several blockers outstanding, so i'll start sending out blocker status mails soon
16:18:27 <adamw> alpha freeze is 2017-03-07, so we still have a bit of time to go.
16:18:49 <adamw> anyone got anything to bring up about f26?
16:20:28 <adamw> apparently not!
16:20:54 * roshi has nothing
16:21:12 <adamw> #topic Fedora 27 Changes: No More Alphas
16:21:44 <adamw> so, there's a Change proposal for F27 that would have a pretty big impact on us: https://fedoraproject.org/wiki/Changes/NoMoreAlpha
16:21:47 <adamw> have folks read it?
16:21:59 * handsome_pirate has
16:22:20 <kparal> let's also cancel Finals, and we'll stay on Betas eternally
16:22:26 * kparal read it
16:22:46 <adamw> kparal: i like it!
16:22:46 * roshi did as well
16:22:50 <roshi> as well as the discussion
16:22:51 <garretraziel> I've also read it
16:23:07 <handsome_pirate> Yeah, I'm not quite sure how blockers will work, though
16:23:08 <adamw> cool. so, what do you think? good idea? bad idea? needs more wokr?
16:23:25 <adamw> any particular problems anyone can see around the idea?
16:23:43 <roshi> I like the idea in general, but I think we're going to have to be prepared, at least for F27 to fudge the rules abit as we get things tuned
16:24:05 <roshi> as we iron out edge cases we'll inevitably find
16:24:05 <garretraziel> I was surprise by sheer amount of negativity in that thread
16:24:12 <roshi> as was I
16:24:35 <adamw> handsome_pirate: the basic idea is there would never be any 'blockers' because nothing that made things fall below alpha quality would ever be allowed to land in 'the real distro': it's not true CI, but package builds that broke stuff would never get tagged for the true Rawhide repo. though of course there's a lot of detail in the implementation
16:25:00 <handsome_pirate> So, what do we do with all the Alpha criteria?
16:25:08 <handsome_pirate> also, brb
16:25:19 <roshi> implement them into tests for gating
16:25:39 <Southern_Gentlem> basicly rawhide stays at alpha quaility  +!
16:25:43 <roshi> with some manual spot checking of rawhide, I'd imagine
16:25:45 <adamw> that's the *idea*, yeah
16:25:52 <kparal> 👏
16:25:57 <Southern_Gentlem> +1
16:26:00 <adamw> we'd have to check how much of the alpha criteria are currently covered with automated tests
16:26:07 <sumantrom> +1
16:26:21 * kparal started using gnome-characters
16:27:34 * Southern_Gentlem isnt happy that f26 has already pushed a week
16:27:41 <adamw> Southern_Gentlem: ? eh?
16:28:02 <Southern_Gentlem> look at the schedule
16:28:13 <Southern_Gentlem> not releasing till 6/13
16:28:35 <adamw> so i can see several possible impacts of this on us. first, we're likely gonna be involved in monitoring this setup. as it's not true CI, there's going to need to be human participation in figuring out exactly why things are getting rejected, making sure people know what they need to be fixing
16:28:56 * handsome_pirate returns
16:29:10 <Southern_Gentlem> and of course i have an event the weekend before
16:29:26 <adamw> oh, that came from https://pagure.io/fesco/issue/1669 .
16:29:50 <adamw> second, of course, it means there won't be any more Alpha validation cycle, so no Alpha RC, go/no-go meeting, etc. which saves us some work
16:30:04 <adamw> but we'll still want to be doing the non-automated validation tests on nightly composes during that time
16:30:48 <adamw> third, it might impact the way the overall release schedule is drawn up - branch point, for instance - so we'll have to follow that and adjust our own schedule as necessary
16:31:44 <sumantrom> adamw then from f27 from when will the bodhi activation point begin?
16:32:15 <adamw> sumantrom: that's the kinda thing that would change, i'm not sure if it's decided exactly when it would be. i think dennis was thinking we could move branching and bodhi activation a bit later
16:32:37 <adamw> (fwiw, dennis is the main guy behind this change, i put my name on it for a bit of moral support but it's mostly his work)
16:32:38 * handsome_pirate isn't so sure moving branch later would be a good idea
16:34:14 <sumantrom> okay adamw , but if the bodhi activation is deffered , how are we planning to go ahead with test days ?
16:34:19 <adamw> handsome_pirate: might be good input to send to dennis
16:34:32 <roshi> if rawhide is alpha quality, I'm not sure it would make a difference
16:34:33 <sumantrom> will there be enough time to test *all* the major changes?
16:34:43 <roshi> we'd just do test days on rawhide
16:35:07 <adamw> sumantrom: exactly, that's the kind of thing we'd have to think about re-arranging: basically, figure out whether the QA event is truly related to the other event it's attached to, or if we should split them up. the key thing for Test Days is really the Change process, so we'll have to see if the Change process dates move with this
16:35:20 <handsome_pirate> roshi:  but, are we going to be putting this much effort into ensure rawhide is alpha quality all the time?
16:35:39 <roshi> that's what the gating is for - it's supposed to be automated
16:35:59 <sumantrom> roshi then we have to also reaggrange the code completion time period for each of the changes
16:36:40 <adamw> right, that'll be the interesting question - do the Change process dates change
16:36:50 <adamw> i don't think that's been figured out yet
16:37:05 <roshi> but I don't htink it'll affect test days
16:37:24 <roshi> since ideally we work with the people pushing the change, and the test day is to help them with their changes
16:37:50 <adamw> what we're basically doing at present is starting test days after the 'code complete' point
16:38:08 <adamw> on the basis that we want to be testing changes that are fundamentally 'done'
16:38:17 <adamw> we can re-examine that assumption, or just follow wherever that point moves on the cycle, i guess
16:40:04 <adamw> #info qa consequence 1: involvement in overseeing the process that decides whether to push things to Rawhide in this system
16:40:27 <adamw> #info qa consequence 2: Alpha RC testing, go/no-go etc. drop out of the cycle
16:41:06 <adamw> #info qa consequence 3: will need to re-examine the qa schedule and see if some events that are directly or indirectly related to the Alpha schedule need to change
16:41:14 <adamw> that sound about right? any other notes on this change?
16:41:19 <roshi> lgtm
16:41:39 <tflink> that sounds about right to me
16:42:59 <adamw> alrighty
16:43:01 <adamw> #topic Fedora 27 Changes: Bodhi Non-RPM Artifacts
16:43:09 <adamw> this one's maybe less significant, but i thought worth bringing up anyway
16:43:10 <sumantrom> sounds allright!
16:43:36 <adamw> we might have to reconsider elements of how fedora-easy-karma work and stuff, maybe update the karma guidelines
16:43:59 <adamw> and it will have an impact on the new openqa update testing stuff, not sure if it affects taskotron
16:44:05 <adamw> sorry:
16:44:36 <adamw> #info https://fedoraproject.org/wiki/Changes/BodhiNonRPMArtifacts
16:45:07 <tflink> it'll probably have a few consequences for taskotron but I think that we're ready for most of them
16:45:41 <adamw> cool
16:45:51 <adamw> #info this will have some impact on taskotron, but tflink thinks we're prepared
16:46:19 <adamw> #info it will also affect the openqa update testing scheduler, i am trying to get info on how fedmsgs for non-rpm updates will look
16:46:40 <adamw> #info we will likely have to adjust fedora-easy-karma and maybe update some wiki pages too
16:46:54 <adamw> any other thoughts / notes on this one?
16:47:52 <roshi> none here
16:48:19 <adamw> #topic Open floor
16:48:22 <adamw> alrighty then
16:48:31 <adamw> we got a bit longer for open floor, in case there's anything i missed
16:48:40 <kparal> I noticed this one: https://fedoraproject.org/wiki/Changes/ArmSupportInFmw
16:48:50 <kparal> we might want to organize a test day for it, perhaps
16:49:23 <kparal> might be a good opportunity for interns I think
16:50:24 <garretraziel> is it completely new thing or is it just integrated fedora-arm-installer?
16:50:42 <kparal> FMW is the tool we use to create bootable usb sticks
16:51:03 <kparal> it simply got extended to be able to write arm disk images
16:51:11 <kparal> show them in listing, etc, I assume
16:52:05 <garretraziel> yeah, but is code for ARM written from scratch or does it use existing arm installer?
16:52:47 <garretraziel> question is whether it writes installer image (like it does for x86_64) or whether it writes arm disk images
16:53:11 <adamw> "This means Fedora Media Writer will have to list the ARM images, offer their download and be able to unpack them reliably (as they're shipped LZMA-compressed - .xz)"
16:53:14 <adamw> so, sounds like the disk images.
16:53:23 <kparal> it's just dd, I believe, nothing else
16:53:55 <adamw> isn't there sometimes some extra bootloader setup involved?
16:54:09 <garretraziel> I do think so, yes
16:55:50 * roshi goes to make some coffee and a sandwhich before blocker review
16:56:28 <garretraziel> I'm asking because AFAIK, arm disk images are blocking, but arm installer images aren't
16:56:31 * adamw drafts a list reply
16:56:37 <adamw> garretraziel: yes, this is correct
16:57:30 <adamw> so clearly this 'self-contained' change has consequences
16:57:50 <adamw> #info https://fedoraproject.org/wiki/Changes/ArmSupportInFmw looks to have consequences for release blocking images, we should follow up and ensure it is properly tested
16:58:02 <adamw> does anyone want an action item for that?
16:58:14 <sumantrom> adamw I have replied to FransiscoD's email on classroom , can I volunteer for it ? if anyone wants they can join!
16:58:47 <adamw> sumantrom: i saw your reply, yep, please do go ahead and volunteer :)
16:59:16 <sumantrom> thanks , sure!  :)
16:59:46 <adamw> #info sumantrom points out that there is a call out for Fedora Classroom instructor volunteers, please do volunteer if you have time and expertise! https://lists.fedoraproject.org/archives/list/test@lists.fedoraproject.org/message/COTDSCJ6NOT6UZQ5SXN3Y7B2J7MYT7SD/
16:59:52 <adamw> alrighty, time for blocker review to start
16:59:57 <adamw> so see you over there shortly
17:00:00 <adamw> thanks for coming everyone!
17:00:02 * adamw sets fuse
17:00:28 * sumantrom thanks adamw for hosting the meeting ! :)
17:02:16 <pwhalen> for arm images the appropriate uboot is written to an offset (which varies depending on the board). the images are setup for the raspberry pi 2/3, so for those boards its just a simple dd
17:03:47 <pwhalen> adamw, happy to take any action item for testing FMW with arm images
17:05:39 <adamw> #endmeeting