QA/Meetings/20100222

From FedoraProject

Jump to: navigation, search

Contents

Attendees

People present (lines said)

  • adamw (101)
  • jlaska (98)
  • maxamillion (32)
  • kparal (28)
  • Oxf13 (28)
  • wwoods (10)
  • zodbot (4)
  • OldFart (4)
  • skvidal (2)
  • jskladan (1)
  • nirik (1)

Regrets:

Agenda

Previous meeting follow-up

  1. adamw noted that the Privilege_escalation_policy is no longer a draft and has been reviewed by FESCO
  2. adamw thanked folks for joining last weeks Color Management test day (see recap).

Fedora 13 Alpha test status

Alpha Test Compose verification has completed. From the wiki pages, there were 26 bugs identified during testing.

Alpha RC#1 is available for testing. Instructions for downloading and providing feedback available at:

Current list of unresolved bugs blocking the Fedora_13_Alpha_Release_Criteria ...

  • RHBZ #566948 NEW - Input devices don't work
  • RHBZ #566979 NEW - Failed dependencies and login problem; basic video driver installation, F13 Alpha RC1.
  • RHBZ #566991 NEW - Incorrect repourl during install
  • RHBZ #566995 NEW - /etc/selinux/targeted/contexts/files/file_contexts: Multiple different specifications for /var/lib/cobbler/webui_sessions(/.*)? (system_u:object_r:httpd_cobbler_content_rw_t:s0 and system_u:object_r:public_content_rw_t:s0).
  • RHBZ #566460 ASSIGNED - kernel 2.6.33 strips coredump when using pipe in core_pattern
  • RHBZ #567319 ON_QA - GDM starts endlessly

The team discussed current Alpha test efforts, including the list of blocking bugs above. The severity of several bugs was discussed, and different workarounds were suggested for the SELinux issues.

Jkeating noted he was available to spin up any new composes on demand. New compose requests should go into the existing ticket (see ticket#3319).

The group concluded that the Alpha is at risk for the scheduled Go/No_go meeting on Wednesday.

Open discussion - <Your topic here>

development/13 bug procedures

Are there documentation updates needed for filing and closing development/13 bugs? Adamw noted this was on the agenda for tomorrows Bugzappers meeting.

Defining the QA team

A topic we may want to visit in the future ... define what it means to be on the QA team, and how to gain membership. With the current no frozen rawhide implementation, critical path package updates need someone from rel-eng and/or QA to validate the fixes prior to acceptance into the 'updates' repo. Maxamillion offered to send a first draft of what it means to be a QA team member.

Security Spin

Maxamillion looking for people to help draw up test cases for tools on the security spin. Created SecuritySpin:QA_Brainstorm wiki page to track ideas. To assist, please take a look at the email thread. Jlaska suggested pulling in the Package-x-generic-16.pngsectool test day hosts (see Test_Day:2009-09-01_Sectool).

Upcoming QA events

Action items

  1. maxamillion to draft up some proposed policy docs for qa FAS group membership
  2. jlaska / maxamillion to co-ordinate with sectool test day hosts tmraz, mbarabas and pvrabec who may be able to contribute

IRC transcript

jlaska #startmeeting Fedora QA Meeting 16:00
zodbot Meeting started Mon Feb 22 16:00:18 2010 UTC. The chair is jlaska. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00
zodbot Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:00
jlaska adamw: you betcha 16:00
jlaska #meetingname fedora-qa 16:00
zodbot The meeting name has been set to 'fedora-qa' 16:00
jlaska #topic Gathering in the lobby 16:00
* Oxf13 16:00
* jskladan is here 16:00
jlaska Oxf13: jskladan: welcome 16:00
* kparal joins 16:00
adamw ello 16:01
* jlaska tips hat to adamw & kparal 16:01
* jlaska needs a new formal greeting method 16:01
adamw wave of the cane? 16:01
adamw the natural, jauntier counterpart to the hat tip 16:01
jlaska adamw: ooh, good choice 16:01
* wwoods appears 16:02
jlaska I believe we also get 100% of maxamillion today 16:02
jlaska wwoods: heyo! 16:02
jlaska I have a hard stop after 30 minutes today, so if we run longer I could use someone to continue chairing 16:02
jlaska I kept the agenda light today ... but suggestions are always encouraged 16:03
jlaska #topic Previous meeting follow-up 16:03
jlaska This should be easy ... 16:03
jlaska we had nothing listed from last weeks meeting 16:03
jlaska anything not listed that people would like to update the team on? 16:03
adamw i'm sure i did lots of important things last week 16:04
adamw so, yeah, I'm awesome 16:04
jlaska adamw: I think we have a t-shirt for that 16:04
adamw oh, we had the test day. it went well. 16:04
jlaska #info color management test day recap - http://lists.fedoraproject.org/pipermail/test/2010-February/088574.html 16:05
jlaska adamw: you had some updates on the privilege escalation front last week as well? 16:05
* jlaska doesn't see
Warning (medium size).png
This page is a draft only
It is still under construction and content may change. Do not rely on the information on this page.
anymore
16:06
adamw yeah, it's now officially a policy. yay. 16:06
adamw so we can get started on writing up some privesc tests if we want. 16:06
adamw any volunteers? :D 16:06
jlaska adamw: let's add that to the future TODO list 16:07
adamw ok 16:07
jlaska #info privilege escalation is approved as an official policy - https://fedoraproject.org/wiki/Privilege_escalation_policy 16:07
jlaska #idea develop test cases to validate the privilege escalation policy 16:08
jlaska alrighty, let's dive into Alpha 16:08
jlaska #topic F-13-Alpha test status 16:08
jlaska So you probably saw Hurry's mail that there is an Alpha release candidate available for test 16:08
jlaska #info F-13-Alpha-RC1 available for test - http://lists.fedoraproject.org/pipermail/test-announce/2010-February/000023.html 16:09
Oxf13 it seems kinda brown paper baggy 16:09
jlaska Oxf13: it sure does :( 16:09
wwoods oh no, reall? 16:09
jlaska so ... I just wante dto spend a few moments making sure we all know what's up ... and what the next steps are 16:10
* adamw doesn't really work over the weekend, so can you provide an executive summary? 16:10
* adamw claiming not executive rank but executive attention span 16:10
jlaska so, we got some weekend testing on the RC, which resulted in new additions to the blocker list 16:10
jlaska #info F13Alpha blocker bugs - https://bugzilla.redhat.com/showdependencytree.cgi?id=538273&hide_resolved=1 16:11
jlaska adamw: best spot to visit is the test result wiki pages ... 16:11
adamw gadzooks 16:11
jlaska https://fedoraproject.org/wiki/Test_Results:Fedora_13_Alpha_RC1_Install#Test_Matrix 16:11
jlaska has the highlights 16:11
adamw looks quite...failish. 16:11
Oxf13 I'd say we've got our work cut out 16:12
jlaska yeah, I'm not at all clear why the change in fortune from TC2 16:12
Oxf13 vastly newer anaconda no? 16:12
jlaska Oxf13: well, afaict it's not all anaconda 16:12
Oxf13 sure 16:13
adamw jlaska: the input device breakage is the x server change from hal to udev for input device configuration 16:13
adamw i knew about that but I didn't know anaconda was actually using the hal system for some kind of configuration 16:13
jlaska that's https://bugzilla.redhat.com/show_bug.cgi?id=566948 16:13
Oxf13 even the stuff that is "anaconda" might not all be anaconda 16:13
jlaska right 16:13
jlaska howabout desktop validation, anyone have input on that front? 16:14
jlaska kparal, you tried a live image earlier today? 16:14
kparal jlaska: I did try the today's image 16:14
kparal jlaska: there's some problem with selinux, just reported: https://bugzilla.redhat.com/show_bug.cgi?id=567319 16:15
OldFart I tried a desktop and after clicking the install to hard drive nothing happened, then tried in a terminal and nothing happened and then it was very late so went to bed. 16:15
kparal jlaska: I will do the desktop validation with selinux disabled, that should work 16:15
jlaska kparal: that sounds F13Alpha worthy 16:15
OldFart I had selinux disabled. 16:15
jlaska OldFart: Clyde, welcome! 16:16
adamw well 16:16
adamw first i'd want to know if the nightlies are f13 or f14 16:16
adamw since they diverge now 16:16
kparal adamw: nirik said it's f13 16:16
adamw ah k 16:16
nirik yes, f13 16:17
jlaska okay, so we have a lot of FAIL so far with RC1. And it seems clear that we'll need an RC2 if we want to meet the alpha release criteria 16:17
jlaska sound accurate? 16:17
kparal yes 16:17
Oxf13 absolutely 16:18
adamw yeah, it seems pretty no-brainery that we'll need to delay the alpha 16:18
kparal but we don't have much time to go/nogo meeting 16:18
jlaska is there any additional testing we can perform in the meantime? 16:18
OldFart Later today I will try the desktop and see if I can capture to data on whats happening. 16:19
wwoods I don't think we need a meeting if it doesn't meet basic acceptance criteria 16:19
jlaska wwoods: it'll certainly be a quick meeting :D 16:19
wwoods (does it?) 16:19
adamw i think we hold the meeting anyway and just say 'well, duh, no.' :) 16:19
adamw anyway, seems like everyone agrees 16:20
adamw oh, and i'd agree kparal's bug feels like a blocker 16:20
jlaska is there anything we can do in the meantime, or are we blocked awaiting new test images? 16:20
kparal adamw: it blocks this: https://fedoraproject.org/wiki/QA:TestCases/Install_Source_Live_Image 16:20
kparal which is an Alpha criterion 16:20
adamw kparal: well, more properly it blocks the criterion 'The installer must boot (if appropriate) and run on all primary architectures from default live image...' but yeah 16:21
kparal jlaska: the desktop validation testing should be possible, as I mentioned, just boot with selinux=0 or enforcing=0 16:21
adamw kparal: well, that could possibly invalidate the results 16:21
adamw if selinux caused problems with any of the tests, you wouldn't be able to tell that way 16:21
kparal hmm 16:21
adamw so i'm not sure it's good testing to do 16:21
kparal that sounds valid 16:22
kparal the testing should be done with selinux enabled, right 16:22
jlaska unless otherwise specified, yes 16:22
kparal even desktop testing 16:22
Oxf13 it should, unless selinux is so horked that you get all failures 16:23
Oxf13 failing every test because of selinux doesn't help as much 16:23
jlaska imo ... it's okay to workaround it temporarily to see what else lurks 16:23
jlaska but not something to rely on for the release 16:23
adamw Oxf13: as i said, in that case, i'd rather wait till selinux is fixed then test 16:23
adamw Oxf13: than test without it 16:23
adamw Oxf13: because that's the state we would release in 16:23
Oxf13 adamw: I agree but disagree 16:24
kparal  :) 16:24
adamw well i'm worried this would happen 16:24
jlaska adamw: it should be tested that way prior to release 16:24
adamw jlaska: yeah 16:24
Oxf13 if we waited to test, only to find other things that were obviously broken, we've just lost that much time in getting them fixed 16:24
kparal it's true we can find more errors not related to selinux, if we test now even without selinux 16:24
adamw Oxf13: fair point 16:24
adamw so, how about this 16:24
adamw if you do the desktop validation with selinux disabled you can log *failures* but not successes 16:24
OldFart I would like to get by the x server issue by using the desktop to test other anaconda features. 16:25
Oxf13 while positive results could not be taken as authoritative, negative results have value. 16:25
jlaska adamw: oh definitely 16:25
Oxf13 adamw: jynx (: 16:25
* adamw call of nature, brb 16:25
kparal adamw: we can provide PASS results, but we just have to know not to rely it until re-tested again with selinux on 16:25
jlaska kparal: let's use the WARN or FAIL state for that now 16:26
jlaska that'll let us know it was tested, and it doesn't meet the stated objectives in the test 16:26
kparal alright 16:26
jlaska I think I described WARN for that scenario 16:26
jlaska either way ... I think we all agreed we can't rely on testing with selinux disabled 16:26
jlaska alright ... so next steps on the Alpha time ... 16:27
jlaska #topic F-13-Alpha - next steps 16:27
jlaska we have a batch of F13Alpha blocker bugs, we have 2 days until the go/no_go meeting 16:27
jlaska A reminder of the goal ... the F13Alpha list should contain no unMODIFIED bugs 16:28
jlaska is that accurate? 16:28
Oxf13 I'm on hand to make trees like a madman as we get fixes 16:28
Oxf13 we're going to want to keep a good mapping of potential fixes to bodhi tickets 16:29
skvidal Oxf13: trees are good - I like oak trees personally 16:29
jlaska Oxf13: how would you like that type of feedback? email/irc/tickets? 16:29
* skvidal stops being silly 16:29
Oxf13 jlaska: any of hte above works. 16:29
jlaska Oxf13: okay 16:29
Oxf13 jlaska: doing it in the ticket is probably a good canonical location for tree compose requests 16:29
jlaska #info requests for new trees should take place in the open RC ticket (https://fedorahosted.org/rel-eng/ticket/3319) 16:30
adamw back 16:30
kparal so, F13Alpha blocker list may contain modified bugs (not fixed) to let F13Alpha GO? 16:30
Oxf13 kparal: yes, if they are modified and there is a build in koji for them 16:31
Oxf13 we can compose the tree before the build makes it all the way through bodhi and shows up in a public repo 16:31
kparal Oxf13: but the F13Alpha will contain all the fixes, right? 16:31
kparal ok 16:31
adamw jlaska: i thought we were trying to avoid it this cycle? 16:31
adamw the criteria says there shouldn't be any open bugs. really that'd just involve closing them if a build known to fix the bug was in place for the compose, though. 16:31
Oxf13 adamw: I thought we were trying to avoid bugs this cycle too (: 16:31
adamw i mean, avoid having open ones when releasing :) 16:32
jlaska adamw: that's correct 16:32
Oxf13 by the time we release, bodhi could have automunged the bugs 16:32
jlaska we can do builds and composes ... but at release ... these puppies need to be CLOSED 16:32
Oxf13 since we're using bodhi this time around we can let it decide when to close the bug(s) 16:32
adamw okay anyway, cross that bridge when we get to it 16:32
jlaska so ... let's talk about that bridge 16:33
jlaska we don't have any blocker review meetings between now and go/no_go 16:33
jlaska is there anything QA can do to help the blocker bugs along? 16:33
adamw jlaska: well, a ton of them are 'needsretesting' anaconda aren't they? 16:33
jlaska yup, I think we can knock some of those out with the current set (using VNC or text-mode installs) 16:34
jlaska #help Bugs listed as MODIFIED can be verified using RC1 - https://bugzilla.redhat.com/showdependencytree.cgi?id=538273&hide_resolved=1 16:34
jlaska anything else to consider on these bugs? 16:35
jlaska otherwise, I'll need to turn the #chair over to someone else for the remainder of the agenda 16:36
jlaska any takers on #chair? 16:36
adamw i'm not so sure about https://bugzilla.redhat.com/show_bug.cgi?id=566995 16:36
adamw if it's not actually causing any problems in itself 16:36
jlaska I wasn't either ... until it spits out to the console during install a *LOT* 16:37
jlaska so I raised it to get some feedback on what's impacted by this ... if it's just noise or something other 16:37
adamw ah k 16:37
adamw well, dwalsh is cc'ed so not much we can do there 16:38
adamw so, I dunno, what's the plan from here? 16:39
adamw when do we envisage doing an rc2? when the big bugs from the blocker list are fixed? 16:39
jlaska that's my impression 16:40
adamw are we working on the assumption we'll be delaying the alpha now? 16:40
jlaska the phrase I used is ... the alpha is 'at risk' 16:40
jlaska  :) 16:40
jlaska #chair adamw 16:40
zodbot Current chairs: adamw jlaska 16:40
jlaska I don't have anything else on this topic ... just wanted to make sure all the issues were raised and we had an idea how to proceed 16:41
adamw okay 16:41
kparal on Live image the Rawhide repository is enabled, instead of Fedora repo. is that a bug? 16:41
jlaska so to summarize ... 16:41
adamw anyone else have anything to add on the alpha topic? 16:41
adamw kparal: seems like one to me, yeah - could be similar to https://bugzilla.redhat.com/show_bug.cgi?id=566991 ? 16:41
adamw but definitely file it and escalate as a blocker 16:42
kparal ok 16:42
jlaska kparal: good catch 16:42
jlaska the only other topics I had for today were the open-discussion items 16:43
adamw yeah 16:43
maxamillion w00t! I'm late but just in time for open-discussion 16:43
maxamillion  :P 16:43
adamw hey maxa 16:43
jlaska adamw: can you walk meetbot through those items? 16:44
maxamillion I just wanted to point out ---> http://fedoraproject.org/wiki/SecuritySpin:QA_Brainstorm 16:44
adamw okay okay everyone calm down! 16:44
adamw deep breaths! 16:44
adamw #topic open floor 16:44
wwoods harrumph! etc! 16:44
maxamillion adamw: hi hi :) 16:44
adamw now, children 16:44
adamw form an orderly queue 16:44
adamw jlaska suggests: 16:44
adamw #topic open floor: fedora 13 bug filing procedures 16:44
adamw but notes 'possible BugZapper topic' 16:44
adamw which, for me, it is - I have something about this on the BZ agenda for tomorrow 16:45
jlaska adamw: you raised this last week, just wanted to make sure it was captured somewhere 16:45
jlaska great! 16:45
adamw i'm just concerned about what we do with bugs filed on rawhide between a release and a branch point, since it's now not clear whether we should treat those as rawhide or f13 bugs, but we'll discuss that in bz tomorrow 16:45
maxamillion I do like me some procedures 16:45
adamw anyone have any comments on that / related concerns? 16:45
adamw okey dokey 16:46
adamw #topic open floor: defining FAS qa membership 16:46
maxamillion oooo, good one 16:46
adamw not sure where this one comes from; every time it's come up in the past we've noted that we don't actually use the FAS system for anything much so it's a no-op 16:46
adamw i guess someone has new concerns / proposals? 16:47
jlaska well, we have a real use case now where FAS qa membership provides a benefit 16:47
maxamillion adamw: its being used for packages in critical path for updates to F13 16:47
jlaska ^^^ yup 16:47
maxamillion adamw: those packages have to get karma from members of the FAS group in order to make it through bodhi 16:47
jlaska so, I just wanted to start this thought process ... and add to the collective team TODO list 16:47
adamw ahhh 16:48
adamw so, the question is, how do we define who gets on the list 16:48
maxamillion bingo 16:48
maxamillion also, what requirements need to be met for newcomers who want to join the FAS group 16:48
maxamillion  ?* 16:48
adamw we could take a cue from the bugzappers system for new members 16:48
adamw and grandparent in existing ones, i guess 16:49
maxamillion is anyone even a member of that group? 16:49
adamw i think like three people or something :)( 16:49
Oxf13 well, I'd say flush the group membership except for those who are regularly showing up at this meeting 16:49
wwoods actually the QA group was flushed a while ago 16:50
wwoods wasn't it? 16:50
Oxf13 from that point, those in the group can serve as a proxy for those who want to be in the group 16:50
adamw not everyone who contributes usefully on the list shows up to the meetings 16:50
* wwoods tried to flush the QA group about a year ago 16:50
jlaska it might need a re-flush 16:50
wwoods who's been adding people to it? 16:50
Oxf13 those who want can do the testing and provide the karma, a QA member can "sign-off" on that testing 16:50
Oxf13 eventually we'll grant full membership to those who prove that they are doing the right thing 16:50
maxamillion +1 16:50
kparal we should also create a short wiki document saying how and when to do ACK and when not to do it 16:51
maxamillion now the question is ... how to put that into verbage for the wiki? 16:51
wwoods you can't view *approved* group members? feh 16:51
kparal for QA members 16:51
jlaska kparal: +++1 16:51
kparal another policy task :) 16:52
Oxf13 we're going to need to do something very similar for the releng group 16:53
adamw so what's the next step on this? who wants to draft up some wiki pages? 16:53
maxamillion I can, but I imagine there will be some heavy editing needed 16:54
maxamillion but I wouldn't mind putting together the first draft 16:54
kparal great 16:54
adamw excellent 16:54
adamw #action maxamillion to draft up some proposed policy docs for qa FAS group membership 16:54
adamw that's everything on the list... 16:54
adamw so, maxamillion, your time to shine :P 16:54
adamw #topic open floor: security spin 16:54
maxamillion heh :D 16:55
maxamillion http://fedoraproject.org/wiki/SecuritySpin:QA_Brainstorm 16:55
maxamillion I'm trying to work out some bits here since almost all the tools on the spin are command line, I want to get some people to help write some test cases to make sure those bits aren't broken 16:55
maxamillion then once that piece is done and we can successfully do some manual testing I was wanting to move to a scripted test and *hopefully* toss it in the AutoQA ring 16:56
maxamillion but that's the large scale hope .... right now its just going to be an information gathering piece 16:56
adamw yep, like you wrote to the list 16:57
adamw so, anything in particular to discuss or is this just a call for help? 16:57
maxamillion more or less just a call for help if anyone has a few spare cycles and knows anything about any of the tools listed 16:57
maxamillion I don't know about all of them which is really my main motivation behind extending out to the group for help 16:58
adamw yeah, so get off your lazy butts people :) 16:58
maxamillion lol 16:58
jlaska maxamillion: there was a sectool test day a few releases back ... lemme remember who hosted that event so maybe they have ideas 16:58
jlaska https://fedoraproject.org/wiki/Test_Day:2009-09-01_Sectool 16:58
maxamillion jlaska: oh, that'd be awesome! 16:58
jlaska looks like ... tmraz, mbarabas and pvrabec 16:59
maxamillion oh goodness ... we don't appear to have sectool on the security spin :X 17:00
adamw #info maxamillion looking for people to help draw up test cases for tools on the security spin 17:01
maxamillion anyhoo, as far as QA efforts go there really wasn't a whole lot to add to what I posted on the list, I just thought I'd bring it up in case some of the people in the meeting hadn't gotten a chance to read through their email (I know I get backed up quite a bit at times) 17:01
adamw #action jlaska / maxamillion to co-ordinate with sectool test day hosts tmraz, mbarabas and pvrabec who may be able to contribute 17:01
adamw okay, that's that one 17:02
adamw anyone else have an open floor topic? 17:02
jlaska I already chewed up my allowance :) 17:04
maxamillion oh, I had a random side question that *might* have been covered and I'm just a tardy turd ... but what's the current libvirt/KVM QA plans, etc.? and where can I find docs on how to help with that? (I have hardware that does KVM now and I'm excited about it) 17:04
adamw i think our test plans are basically 'we do most of our testing in virtual machines so if it's broken we probably know about it' :P 17:04
adamw but I may be missing something 17:04
kparal and we have it as F13Beta criteria 17:05
maxamillion adamw: fair enough ... that's pretty much what my work flow is at this point as well 17:05
maxamillion ok, just thought I'd ask :) 17:05
Oxf13 I'm stepping out, I have a local engagement 17:06
adamw proper formal testing would be useful i guess 17:06
adamw i don't think we have any planned atm 17:06
adamw #topic open floor 17:06
adamw anything else? 17:07
maxamillion not I 17:08
jlaska nothing here 17:08
* adamw is busy watching youtube videos about insane japanese robot suits 17:08
maxamillion adamw: LOL 17:09
adamw alrighty then - thanks for coming everyone 17:09
adamw http://www.youtube.com/watch?v=G4evlxq34og , btw 17:09
jlaska adamw: thanks for the #chair assist 17:09
adamw aka HOLY SHIT IT'S THE FUTURE 17:10
adamw #endmeeting 17:10

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!