QA/Meetings/20131209

Attendees

• adamw (63)
• cmurf (54)
• kparal (17)
• nirik (7)
• roshi (7)
• greenlion (6)
• tflink (6)
• handsome_pirate (5)
• zodbot (4)
• jreznik (4)
• pwhalen (1)
• masta (1)
• mkrizek (1)
• pschindl (1)

Agenda

• Previous meeting follow-up
• Fedora 20 Final status
• Open floor

Previous meeting follow-up

• adamw to draft a new test case and matrix row for validating cloud image checksums - adamw still failed to get around to this, will try once more for next week

Fedora 20 Final status

• Test coverage was looking decent, just a couple of tests remain that hadn't been run since Beta
• Two accepted blockers were outstanding, and several proposed
• Next compose date depends to some extent on anaconda work
• Go/No-Go again on Thursday

Open floor

• tflink and roshi both looking for feedback on ideas: tflink's proposal for a taskotron task description format and roshi's proposal for validation test case 'roadmaps' for running multiple tests together

Action items

• adamw to draft a new test case and matrix row for validating cloud image checksums

IRC Log

adamw	#startmeeting Fedora QA meeting	16:00
zodbot	Meeting started Mon Dec 9 16:00:51 2013 UTC. The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot.	16:00
zodbot	Useful Commands: #action #agreed #halp #info #idea #link #topic.	16:00
adamw	#meetingname fedora-qa	16:00
zodbot	The meeting name has been set to 'fedora-qa'	16:00
adamw	#topic Roll call	16:01
* roshi is here		16:01
adamw	ahoyhoy, folks, who's on board the ship o' release validation fun?	16:01
* nirik is lurking around		16:01
* jreznik is here		16:01
* mkrizek is here		16:01
* cmurf is wearing is life vest, just in case.		16:01
cmurf	^h	16:01
* pwhalen is here		16:01
* greenlion looking in		16:02
* kparal late		16:02
cmurf	have we lost the relay?	16:04
adamw	hmm?	16:04
kparal	has he #chaired someone else?	16:04
adamw	yeesh, it's only been four minutes and you're already plotting the mutiny?	16:04
adamw	#plank kparal cmurf	16:05
adamw	what, meetbot, there is no "makes X walk the plank" macro? this is an UNACCEPTABLE OVERSIGHT	16:05
adamw	fine, then	16:05
adamw	#chair kparal cmurf	16:05
zodbot	Current chairs: adamw cmurf kparal	16:05
adamw	#topic Previous meeting follow-up	16:05
adamw	time to rename this section 'adamw sucks corner', folks	16:05
cmurf	i'm referring to the ping timouts, connection resets, host changes - people left and there was a pause	16:06
adamw	#info "adamw to draft a new test case and matrix row for validating cloud image checksums" - still didn't get around to this one, though somehow I found time to write a completely different test case. sorry. one more try	16:06
adamw	#action adamw to draft a new test case and matrix row for validating cloud image checksums	16:06
adamw	anything for follow-up that i missed?	16:06
cmurf	*shrug* one was obviously much more interesting than the other :-)	16:06
* pschindl is here		16:07
adamw	guess not	16:08
adamw	wb, tflink	16:08
tflink	not sure what happened there,	16:09
tflink	fun times :)	16:09
adamw	#topic Fedora 20 Final status	16:09
adamw	tflink: well, we used the opportunity to re-assign all release blockers to you.	16:09
tflink	adamw: I thought that the idea was _not_ to slip more :-P	16:09
adamw	=)	16:09
cmurf	right	16:10
adamw	so, we still have a couple of tests that haven't run since Beta: http://testdays.qa.fedoraproject.org/testcase_stats/Category_Installation_validation_testing.html	16:10
jreznik	that's not nice from you adamw!	16:10
adamw	hehe	16:10
cmurf	no more release blockers short of injuries that cause scaring	16:10
adamw	i really need to do the hw raid one	16:10
adamw	if people can help get https://fedoraproject.org/wiki/QA:Testcase_Kickstart_Hd_Device_Path_Ks_Cfg and https://fedoraproject.org/wiki/QA:Testcase_Kickstart_File_Path_Ks_Cfg done that'd be great, yes I know they're a pita	16:10
cmurf	short of a firmware bug it really oughta just work like installing to a single drive	16:10
adamw	cmurf: yeah, we all know the theory ;)	16:11
cmurf	right so not a big deal waiting until the last second to test it	16:11
adamw	then we have a few from tc2, tc3 and tc4 to work through, but we'll probably have RC1 by then	16:11
adamw	or at least tc6...	16:12
* roshi crosses his fingers for RC		16:13
adamw	for RC, ignoring for now the proposed blocker list, we have https://bugzilla.redhat.com/show_bug.cgi?id=1027947 and https://bugzilla.redhat.com/show_bug.cgi?id=983110	16:13
jreznik	The kde one looks like we have a hit	16:15
adamw	well, not entirely	16:15
adamw	just arbitrarily turning off glib2 mainloop integration for a chunk of kde startup is a pretty big change to throw in at this point	16:15
adamw	and right now i think they're discovering some consequences of it in #fedora-kde...	16:16
adamw	but yeah, it mostly looks like we've got a few tests to clean up while we wait on devs for an rc1	16:16
cmurf	isn't it kinda late for such big changes?	16:16
adamw	and folks, if you have blocker bugs in you, *please* get them out today if at all possible =) i know this is the time when we get some spare minutes to do freeform testing and spot things we missed before, but try and get it done asap	16:17
kparal	can't create new bugs in bugzilla	16:17
cmurf	i went through that phase for beta, i'm ready for this goose to lay its egg man	16:18
adamw	cmurf: yes, but we gotta fix it _somehow_. the build was mostly meant as a debugging test rather than as the fix, but they floated the possibility of pushing it if it fixes the bug and they can't come up with a more specific fix	16:18
kparal	I wanted to report some, but ...	16:18
* handsome_pirate stumbles in late		16:18
nirik	kparal: I understand they are going to push a hotfix for that. not sure when tho	16:18
cmurf	re: new BZ bugs, is this isolated or can no one create bugs?	16:19
nirik	its the top level report a bug thing timing out...	16:20
kparal	someone at #fedora-admin claimed it works for him	16:21
kparal	doesn't for me	16:21
adamw	oh good, looks like dlehman is on PTO today. just what we need for blocker fixing.	16:21
nirik	kparal: it loads here after 1-2minutes.	16:21
adamw	kparal: i thought i left specific instructions for you to be bound and gagged outside the reach of a keyboard	16:21
* adamw knows a workaround but ain't telling kparal		16:22
cmurf	no kidding	16:22
* cmurf snickers		16:22
adamw	if you're called kparal, look away now	16:22
* kparal gets proxy error		16:22
adamw	for everyone else, you can go to http://bugz.fedoraproject.org/(packagename) then use the 'report a new bug against this package' thing	16:23
adamw	(or you could just, you know, memorize the parameters to enter_bug.cgi)	16:23
handsome_pirate	Yay	16:23
handsome_pirate	New workflow	16:23
kparal	https://bugzilla.redhat.com/enter_bug.cgi?product=Fedora&component=anaconda doesn't work for me either	16:23
roshi	kparal, I think adamw just turned your access off :p	16:24
cmurf	doesn't work = doesn't load?	16:24
kparal	probably	16:24
kparal	cmurf: proxy error after several minutes	16:24
cmurf	it's hanging for me atm	16:24
cmurf	what are "DE tools"?	16:25
cmurf	context is "Open created LUKS device using DE tools"	16:25
greenlion	cmurf, dolphin in my case	16:26
cmurf	is that included in live desktop?	16:26
greenlion	probably works with nautilus too	16:26
greenlion	yes, KDE Live	16:26
adamw	okay, so is there anything else to discuss about f20 status?	16:27
cmurf	kparal reproduced 1008732 with live desktop though, not live kde	16:27
cmurf	adamw nothing other than the obvious	16:27
adamw	oh, yes	16:27
adamw	#agreed everybody panic	16:27
cmurf	right	16:27
adamw	=)	16:28
tflink	adamw: EINSUFFICIENTCOFFEE	16:28
handsome_pirate	adamw: What if I'm already panicking?	16:28
greenlion	why panic? just delay release to january, or february... :)	16:28
jreznik	set panic mode	16:28
adamw	handsome_pirate: panic harder	16:28
cmurf	i have 1tbsp of soymilk left so yes coffee panic mode is about to happen, and it's like -10C outside	16:28
handsome_pirate	adamw: I'm trying!	16:28
cmurf	-16C, yikes	16:29
roshi	cmurf, you don't drink your coffee black?	16:29
adamw	obviously he hasn't been working hard enough	16:29
cmurf	nope, comes out of the press pot like syrup it needs to be cut	16:29
roshi	lol	16:29
cmurf	besides i drink it because i actually like coffee not because i need it	16:30
cmurf	anyway!	16:30
* masta looks in		16:31
cmurf	blockers?	16:31
adamw	#topic open floor	16:31
adamw	not that it hasn't been open floor for ten minutes...	16:31
adamw	but do we have anything else?	16:31
cmurf	i do not	16:32
tflink	If anyone's interested in automation - comments and discussion appreciated: https://lists.fedoraproject.org/pipermail/qa-devel/2013-December/000556.html	16:32
roshi	I'm looking for feedback on an idea if people haven't already seen it: http://roshi.fedorapeople.org/testing-efficiently.html	16:33
roshi	I'll be sending an email out to test@ once I get it a bit more fleshed out	16:33
adamw	#info tflink and roshi both looking for feedback on ideas: https://lists.fedoraproject.org/pipermail/qa-devel/2013-December/000556.html and http://roshi.fedorapeople.org/testing-efficiently.html	16:33
adamw	thanks guys!	16:33
cmurf	speaking of the test matrix	16:34
cmurf	it seems that we get a TC and then a day later we get another anaconda	16:34
cmurf	and it's like… ok we have this TCx' (that's a prime) derivative outside the matrix	16:35
cmurf	and the potential for regressions this time of the process increases I think	16:35
adamw	cmurf: i didn't do another build right away because it looked like we'd be able to do an rc quicker	16:35
cmurf	yeah and i'm glad for that actually	16:36
cmurf	what i'm wondering is if there's a way to inhibit builds of anaconda until more stuff is fixed	16:36
cmurf	because with every new build it really means, esp with an RC, we need to retest the whole matrix	16:37
cmurf	in addition to any affected blockers	16:37
cmurf	so fix more bugs before rolling another anaconda?	16:37
cmurf	i'm not even doing a 10th of the work some folks are and it's a bit much to see only 1/2 the matrix get filled up before another anaconda appears and i'm like …. crap	16:38
adamw	well, it's kinda up to anaconda team's discretion when to build a new one	16:38
kparal	that's the basic dilemma. I believe we should use tools like http://testdays.qa.fedoraproject.org/testcase_stats/ more and not just blindly do all the test case again and again and again	16:38
cmurf	adamw: understood, but i think they might be under the imrpession they're doing us a favor to get bug fixes tested sooner	16:39
adamw	still, i don't quite see the practical difference between building anaconda 20.1, 20.2, 20.3, 20.4, 20.5, 20.6, and 20.7 with one fix each and building 20.1, then building 20.7 with all the fixes	16:39
cmurf	kparal good point	16:39
kparal	anaconda builds should not be delayed. we can delay TCs if we want	16:39
adamw	kparal: during TC phase that's already what we should be doing	16:40
kparal	frequent anaconda builds are actually helpful for us	16:40
adamw	kparal: it's only RCs where in theory we ought to run every test	16:40
cmurf	ahh ok	16:40
kparal	adamw: well, this approach is not really described anywhere, but it's good that you agree	16:40
adamw	kparal: yeah, that's my perspective too, the important question is how often we build TCs, not how often anaconda pushes builds out	16:40
cmurf	ok great so 2nd question is if there is a way now or could be in the future to support an overlay for the DVD ISO	16:41
cmurf	and netinst	16:41
cmurf	so that anaconda can be updated and remain resident	16:41
cmurf	like the live	16:41
kparal	cmurf: hardly. there's not even rpm	16:41
adamw	um. i have no idea? be one to ask bcl, i guess.	16:41
* nirik isn't sure what that gets you... not having to download a new one?		16:42
kparal	nirik: not waiting for a next compose	16:42
cmurf	not having to wait for a TC build with that version of anaconda in it	16:42
greenlion	updates.img?	16:42
nirik	ah. I guess.	16:42
kparal	updates.img is hard to create by users	16:42
nirik	too many things like that tho and you end up not testing the thing we are shipping tho.	16:43
cmurf	exactly	16:43
cmurf	and life install is gnome/kde only which takes a ton longer than a minimal install	16:43
cmurf	live, rather	16:43
cmurf	that's all i've got, meanwhile i'm testing some of these newly proposed blockers	16:43
cmurf	what's next? blockers?	16:44
* cmurf hears Jeopardy muzak in the background		16:45
adamw	blocker meeting next	16:46
adamw	top of the hour, in #fedora-blocker-review	16:47
adamw	everyone go get coffee :)	16:47
cmurf	OIC	16:47
* adamw sets quantum fuse, thanks for coming folks!		16:47
tflink	that's one long quantum fuse	16:54
adamw	#endmeeting	16:55

Generated by irclog2html.py 2.12.1 by Marius Gedminas - find it at mg.pov.lt!