QA/Meetings/20131007

Attendees

• adamw (108)
• Viking-Ice (44)
• handsome_pirate (33)
• roshi (24)
• tflink (20)
• kparal (9)
• adamw_webchat (8)
• jreznik (7)
• zodbot (5)
• dgilmore (4)
• pwhalen (2)
• sgallagh (2)
• mkrizek (2)

Agenda

• Previous meeting follow-up
• Fedora 20 Beta status
• Test Day status and help wanted
• Possible elongated schedule for Fedora 21
• Open floor

Previous meeting follow-up

• tflink or someone competent to talk to releng about where to file bugs in the cloud .ks and whether it needs to be packaged in the distro - dgilmore will be making sure cloud kickstarts get added to

   spin-kickstarts package

• adamw to talk to anaconda devs and get a build done for beta tc1 with live keyboard bug fixed - there was already a build ready that went into TC1. anaconda team was happy with that build for TC1
• spstarr to work on iSCSI install testing - spstarr not present at the meeting, did not look like testing had been done for TC1
• adamw to look into nightly fail - was already resolved soon after the meeting

Fedora 20 Beta status

• Quite a few required tests had not yet been run for TC1, we needed to get to those for TC1/TC2
• TC1 ARM images were DOA, dgilmore and pwhalen were dealing with it
• We decided it would be a good idea to build TC2 soon

Test Day status and help wanted

• 2013-10-08 would be virtualization test day, and Thursday 2013-10-10 would be GNOME test day
• adamw and handsome-pirate volunteered to help out with graphics test week
• Volunteers were still welcome to help run future Test Days

Possible elongated schedule for Fedora 21

• Per ticket #1178, FESCo wanted solid justifications for possibly providing extra development time for F21
• viking-ice suggested pushing to have anaconda development cycle adjusted to stabilize earlier in the cycle (would probably need a longer cycle to achieve)
• tflink wanted to work on test automation and blocker bug app
• handsome-pirate wanted to work on validation test result tracker app
• adamw suggested that all those with solid ideas draw up a detailed and realistic proposal for getting somewhere concrete within an elongated release cycle and add them to the FESCo ticket, perhaps after list discussion

Open floor

• viking-ice suggested a plan to assign 'QA team' members to specific packages or package groups for testing and triage tasks; further discussion needed

Action items

• adamw to co-ordinate with anaconda team to get a new anaconda build done and roll TC2
• tflink and handsome_pirate to write up detailed proposals for projects to be worked on in a possible longer f21 schedule and post / link to them in the trac ticket

IRC Log

adamw	#startmeeting Fedora QA meeting	15:00
zodbot	Meeting started Mon Oct 7 15:00:18 2013 UTC. The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot.	15:00
zodbot	Useful Commands: #action #agreed #halp #info #idea #link #topic.	15:00
adamw	#meetingname fedora-qa	15:00
zodbot	The meeting name has been set to 'fedora-qa'	15:00
adamw	#topic Roll call	15:00
* adamw wonders what's ahoyhoy in czech		15:00
adamw	ahojhoj, right?	15:00
* pwhalen is here		15:01
* kparal arrives		15:01
* mkrizek is here		15:01
mkrizek	adamw: right :)	15:01
adamw	ahojhoj it is, then	15:01
adamw	ahojhoj from praha	15:01
* jreznik says ahoj to everyone :)		15:01
jreznik	better to say ahoj, ahoj :)	15:02
adamw	but less burns-ian	15:02
* tflink is		15:03
tflink	here	15:03
jreznik	adamw: yep	15:03
adamw	tflink: aha, was wondering where you lot had got to	15:03
adamw	don't you just hate north americans who never seem to be around	15:03
adamw	roshi around?	15:04
tflink	interesting how that changes as you cross the pond	15:04
roshi	yes	15:04
* roshi here		15:04
adamw	tflink: i've got huge piles of stuff to do while i'm here, sadly	15:05
adamw	but today i've mostly just been wandering around prague being a tourist	15:05
adamw	i imagine viking will be showing up as we go along	15:05
adamw	#topic Previous meeting follow-up	15:05
adamw	"tflink or someone competent to talk to releng about where to file bugs in the cloud .ks and whether it needs to be packaged in the distro" - tflink, any news?	15:06
tflink	yes, cloud kickstarts need to be merged into spin-kickstarts	15:06
tflink	instead of forming a cloud-kickstarts package. IIRC, dgilmore was planning to do this	15:07
adamw	ok, is someone in charge of that?	15:07
adamw	welp, we need to make sure it happens by GA, i guess	15:07
adamw	#info "tflink or someone competent to talk to releng about where to file bugs in the cloud .ks and whether it needs to be packaged in the distro" - dgilmore will be making sure cloud kickstarts get added to spin-kickstarts package	15:07
tflink	so bugs like the i386 AMI one would be filed against spin-kickstarts	15:07
adamw	#info bugs in cloud kickstarts should be filed against spin-kickstarts	15:08
* Viking-Ice here for a brief moment		15:08
adamw	hi viking	15:08
adamw	if you're only here briefly, anything in particular you wanted to bring up?	15:08
Viking-Ice	well the schedule stuff but I'll probably miss that	15:09
adamw	i threw that on as i understand from tflink that fesco wanted people to pitch them specific proposals for work that could be done in a longer cycle	15:09
Viking-Ice	as well as I dont know how smart workflow it is to have Fedora bugs block RHEL bugs	15:09
adamw	what bug is that?	15:09
adamw	often, rhel qe clones bugs we find in fedora, for RHEL	15:10
adamw	when you clone, the original blocks the clone by default	15:10
adamw	and sometimes they don't remove that	15:10
adamw	one reason i hate using clones, but eh	15:10
* handsome_pirate staggers in		15:10
Viking-Ice	bug 947815 the blocker bug is open so it's fine ( for now )	15:10
Viking-Ice	but this could have been classic ACCESS DENIED	15:10
adamw	i usually figure the rhel folks know what they're doing there	15:10
adamw	i think it's more of a problem when rhel bugs block fedora bugs	15:11
adamw	if you come across any of those where the rhel bug is private, give one of the rh folks a yell and we'll try and get it opened up or remove the block or something	15:11
adamw	shouldn't ever be a case where fedora bugs depend on bugs that are rh-private for rhel reasons, i don't think (though sometimes fedora security bugs are embargoed, which is ok)	15:12
adamw	if the fedora bug *blocks* a rhel bug, that's not really our problem - that's up to rhel people to deal with, we can just deal with the fedora bug\	15:12
* handsome_pirate wonders why a RHEL bug would block a Fedora bug		15:12
adamw	handsome_pirate: it's usually just a mistake or a cloning artifact or something when it happens	15:13
adamw	so it can usually be fixed	15:13
adamw	Viking-Ice: did you have any particular ideas or concerns about the schedule stuff?	15:14
Viking-Ice	adamw, mostly that we need it to you know work on stuff	15:16
Viking-Ice	but fesco seemed to only give crap about what tflink and dgilmore where working on from qa/releng perspective	15:17
adamw	Viking-Ice: any particular 'stuff' you wanted to put on the list?	15:17
handsome_pirate	Viking-Ice: What do you mean?	15:17
Viking-Ice	adamw, qa community member work stuff like	15:17
handsome_pirate	Viking-Ice: What are you talking about, FESCo giving tflink crap? I wasn't there, so a summary would be cool	15:17
Viking-Ice	handsome_pirate, fesco was only giving crap ( only interested in ) the work tim and dennis were doing	15:18
jreznik	well, that was my concern - that we need something to back up delaying schedule, not to do it just for "we can" but to have a real stuff done	15:18
sgallagh	handsome_pirate: "Give a crap" vs. "giving crap". I love English.	15:18
sgallagh	Viking-Ice: And no, we just were expecting that those two would be representing their respective teams. It was not exclusive.	15:19
adamw	right, the idea is to nail down specific proposals, not just 'we got stuffs to do'	15:19
Viking-Ice	adamw, as well as we really need to push anaconda development cycle on a different timeline	15:19
Viking-Ice	so it's more or less ready when we branch	15:19
Viking-Ice	for the branched release	15:20
adamw	sounds like a slightly different topic, we'd have to discuss that with anaconda team i guess	15:20
Viking-Ice	yes but to be able to do so we need that delay	15:20
Viking-Ice	( to be able to sync that stuff properly )	15:20
handsome_pirate	Viking-Ice: Do you have a link to the logs?	15:20
* handsome_pirate wants to read up on this		15:20
Viking-Ice	it's there in the meetbot page	15:21
jreznik	or meeting minutes are sent to -devel	15:21
* handsome_pirate will look		15:21
adamw	Viking-Ice: i see, yeah, that links up	15:22
adamw	ok, i'll try and work that in when the topic comes up, and maybe you can write a post to test@ / anaconda-devel-list@	15:22
Viking-Ice	adamw, if we are going to be able to handle 3 ( or more ) products we need to put anaconda to be on different schedule so we can free up time to focus on those products	15:22
handsome_pirate	Oh, the whole Fedora as Three Products thing	15:23
Viking-Ice	in relation to that ( or more ) products	15:23
* adamw feels like he's been on the moon for a week		15:23
Viking-Ice	in anycase internal meeting so carry on I'll catch up when i get back	15:24
adamw	alrighty	15:25
adamw	getting back to follow-up...	15:25
* handsome_pirate notes that there still isn't a QA person on FESCo		15:25
* handsome_pirate will brb		15:25
adamw	#info "adamw to talk to anaconda devs and get a build done for beta tc1 with live keyboard bug fixed" - there was actually already a build ready at the time, and that went into TC1. i did check if they were happy with the build going into TC1, and they were.	15:25
adamw	handsome_pirate: i think no-one wants to stand. i know i don't. :P	15:26
adamw	" spstarr to work on iSCSI install testing" - anyone heard from him?	15:26
* adamw checks the matrix		15:26
tflink	he was talking about doing it, but I didn't hear anything more	15:28
adamw	there's no result for beta tc1	15:28
adamw	#info "spstarr to work on iSCSI install testing" - spstarr not around, looks like no result filed for tc1	15:29
adamw	#info "adamw to look into nightly fail" - was already resolved when I got to it, I did a pre-TC1 live image test which worked fine	15:29
adamw	#topic Fedora 20 Beta status	15:29
adamw	we're missing quite a lot of tc1 tests	15:30
adamw	would be nice if folks could get around to some of those (including me)	15:30
adamw	#info quite a few required tests have not been run for TC1, we need to get to those	15:30
adamw	looking at the blocker list, looks like we have some changes to anaconda to check out - could be about time for a TC2 with a new anaconda build, see if the DVD size is fixed	15:32
jreznik	DVD size should be fixed, I hope	15:33
jreznik	gimp-help langpacks were about 600 MB	15:34
adamw	#action adamw to co-ordinate with anaconda team to get a new anaconda build done and roll TC2	15:34
dgilmore	adamw: ill have to make some spins-kickstarts changes to work around a dracut bug on arm images	15:34
handsome_pirate	adamw: I've run for FESCo and didn't get it	15:35
handsome_pirate	adamw: I d/l the DVD iso and it was borked, need to redownload	15:36
handsome_pirate	adamw: Checksum didn't pan out	15:36
pwhalen	I'll be completing the rest of the ARM matrices today. The images themselves are doa due to the issue dgilmore mentions . I've filed a bug, added to blockers	15:36
adamw	sorry...looks like i lagged out there	15:41
adamw	#chair kparal tflink	15:41
zodbot	Current chairs: adamw kparal tflink	15:41
adamw	#info TC1 ARM images are DOA, dgilmore and pwhalen are dealing with it	15:42
adamw	dgilmore: are you ready to roll with the fix whenever we go for a TC2?	15:42
dgilmore	adamw: well i think the fix is reverting some things in dracut	15:43
dgilmore	i have a workaround	15:43
dgilmore	need to add dracut-generic-config to the arm images	15:43
adamw	ah, i see	15:44
adamw	alrighty, sounds like that's about it for Beta	15:46
adamw	moving on..	15:46
adamw	#topic Test Day status and help wanted	15:49
adamw	roshi: this was sort of for you, just to check in on where we stand with test days and see if you wanted to ask for people to help run some	15:49
roshi	for sure	15:49
roshi	tomorrow we have Virtualization test day - so if you can make it, that would be great	15:49
kparal	gnome test day on thursday should be run by lbrabec and garretraziel	15:50
tflink	what about NA timezones?	15:50
adamw	#info tomorrow (2013-10-08) is virtualization test day, and thursday (2013-10-10) is GNOME test day; please come if you can	15:50
adamw	isn't garretraziel NA?	15:51
tflink	brno	15:51
kparal	adamw: you'll meet him soon enough :)	15:51
roshi	I'll be here for NA timezones on thursday	15:51
kparal	roshi: thanks	15:51
kparal	also the gnome developers promised their presence	15:51
adamw	kparal: sigh, me getting confused again	15:52
adamw	roshi: any more coming up later that need help?	15:53
roshi	for the graphics test week, we need volunteers to run the scheduled days (requested here: https://fedorahosted.org/fedora-qa/ticket/406)	15:53
roshi	we have test days throughout the coming weeks	15:53
adamw	i can probably take gfx test week	15:54
adamw	should be just after i get back to canada	15:54
kparal	how about wayland? that one will need a ton of preparation and I haven't seen it in the request for help	15:54
roshi	that's part of the gfx test week	15:54
roshi	but yes, we need someone to run that - or we need to split it off and reschedule it	15:54
roshi	until someone takes it	15:55
handsome_pirate	roshi: I'll get it	15:55
roshi	ok, great	15:55
kparal	handsome_pirate: there are no previous test cases I think, so they need to be written from scratch	15:56
kparal	best to consult this with gnome devs	15:56
handsome_pirate	kparal: Roger	15:56
handsome_pirate	kparal: I'll start on that this afternoon	15:56
handsome_pirate	Actually	15:56
kparal	I think that wayland test day can't be done without a heavy participation of gnome devs	15:57
handsome_pirate	kparal: I've started poking at Gnome devs	15:57
adamw	most likely, and x devs.	15:57
adamw	ok, moving on, since time is tight...	15:57
adamw	#topic Possible elongated schedule for Fedora 21	15:57
tflink	#link https://fedorahosted.org/fesco/ticket/1178	15:58
adamw	so as mentioned earlier, fesco is/was wanting specific proposals of work that could be done in a longer release schedule	15:58
adamw	#info fesco is wanting specific proposals of work that could be done in a longer release schedule: see https://fedorahosted.org/fesco/ticket/1178	15:59
handsome_pirate	roshi: I want to make sure of the date for Wayland: 2013-10-25	15:59
adamw	we can certainly add ideas/proposals to that ticket; is there anything we want to kick around?	15:59
adamw	we have viking's proposal from earlier about using a longer release to try and shift anaconda development to stabilize earlier, but that'd really need anaconda dev buy-in obviously	15:59
handsome_pirate	adamw: I'm +1 for that one	16:00
adamw	are there any other ideas? work on any of the tooling we have underway at the moment, blocker bug app etc?	16:00
roshi	handsome_pirate: yep - it's already on the wiki and fedocal	16:00
adamw	tcms stuff?	16:00
tflink	adamw: that had been in the back of my mind	16:00
handsome_pirate	adamw: Yeah, time to work on tools would be really amazing	16:00
tflink	i got reamed out a little at the last fesco meeting for not having a specific plan, so we'd need more specifics	16:01
adamw	well, sounds like we need to come up with fairly realistic and specifically timed-out proposals	16:01
adamw	rght	16:01
tflink	I didn't realize they were expecting more than proposals for automation or I'd have mentioned it earlier	16:01
adamw	and it'll help our credibility if we can realistically commit and stick to anything we propose	16:01
handsome_pirate	The question becomes, what do we want to focus on?	16:02
handsome_pirate	Also, assignments	16:02
adamw	shall i give you two action items?	16:02
adamw	yeah\	16:02
handsome_pirate	adamw: Might as well	16:02
handsome_pirate	tflink: I'll ping you this afternoon about it	16:03
adamw	is there one specific thing you'd like to work on above others, and can commit realistically to getting something concrete done on in, say, 3 months?	16:03
handsome_pirate	Well, it is afternoon now	16:03
adamw	if so, write up a good solid proposal for it, post it to the list and link from the ticket, i'd say	16:03
handsome_pirate	adamw: Testtrack	16:03
handsome_pirate	The test results tracking app	16:04
* Viking-Ice jumps back in		16:07
adamw_webchat	sorry folks, i'm lagging super-bad	16:09
adamw_webchat	someone else want to take over chair to close out the meeting?	16:09
tflink	sure	16:09
adamw	#action tflink and handsome_pirate to write up detailed proposals for projects to be worked on in a possible longer f21 schedule and post / link to them in the trac ticket	16:10
adamw_webchat	there, got that through	16:10
adamw_webchat	if anyone else has concrete and achievable ideas for work to be done in a longer schedule, please add them too, of course	16:10
tflink	maybe we should collect them outside the trac ticket first	16:11
Viking-Ice	I'll try to throw on paper the anaconda stuff and get that discussion going with the anaconda developers	16:11
adamw_webchat	right, if it needs discussion and refinement, we can work on that on the list	16:11
tflink	so we present a more coherant proposal to fesco	16:11
adamw_webchat	but we only have two days till fesco meeting	16:11
tflink	yep	16:11
* handsome_pirate will get on it after lunch		16:11
adamw_webchat	vik: right, that one's going to need buy-in from them of course, it's a no-go if they don't want to do it	16:11
Viking-Ice	I'm pretty sure we can floor any kind of product plans if they will not buy into it	16:12
adamw_webchat	heh, true	16:12
tflink	#chair adamw_webchat	16:12
zodbot	Current chairs: adamw adamw_webchat kparal tflink	16:12
tflink	is there anything else we'd like to cover? we're already over time	16:12
Viking-Ice	we need to somehow find time for that and our largest time consumer is anaconda	16:12
adamw	#topic open floor	16:13
adamw	and as tflink said, make it snappy :)	16:13
Viking-Ice	so the general Idea I have surrounding qa community member is to merge the role of tester and triager	16:13
Viking-Ice	since the bottom line is that both skill sets are more or less required ( for proper reporting and triaging )	16:14
adamw	the thing with triage is that it seems like no-one actually wants to do any...	16:14
Viking-Ice	as things stand now yes	16:14
Viking-Ice	there seems to be little interest in it	16:15
Viking-Ice	but we should be able to change that	16:15
Viking-Ice	by merged these roles ( hopefully )	16:15
adamw	but sure, so far as the whole group membership thing goes, that's a sensible approach	16:16
adamw	so what exactly is the idea in concrete terms? people who read through Join and post up self-introductions can get triagers group membership, or what?	16:16
Viking-Ice	well we need to rotated qa community members between components	16:16
Viking-Ice	so you know people wont just stick themselves to the most popular component	16:16
adamw	it's a bit hard to tell people what to do	16:17
adamw	but sure, i like the idea of trying to get qa folks to take on some triage work	16:17
Viking-Ice	well people sign up to a community member role and he gets assigned a week task ( or something ) testing and triaging x component	16:18
Viking-Ice	then next week he rotates to the next one	16:18
roshi	who would manage all that?	16:18
roshi	setting up roles, notifying people	16:18
roshi	etc	16:18
roshi	?	16:18
roshi	or are you thinking some form of automated tool?	16:19
Viking-Ice	you mean creating the schedule	16:19
roshi	yeah	16:19
Viking-Ice	yeah we need to automate the schedule creation	16:19
adamw	well, and then making sure any of it actually happens...	16:19
roshi	getting new things added to the schedule and whatnot	16:19
tflink	week-long rotations seem like they would be a bit on the fast side - might it be better to leave people in a place for longer to gain more experience?	16:19
Viking-Ice	tflink, not sure that will work	16:19
roshi	I would think 2 week min if that route was taken	16:19
Viking-Ice	we need to cover components on full products	16:20
roshi	can you define "components" and "full products?"	16:20
Viking-Ice	group of components make up x product	16:20
roshi	(just so I know exactly what you're meaning - me being a new guy)	16:20
Viking-Ice	so for example I first task would be to cover the components that make up the core/baseOS then extend to the next level on top of that	16:21
Viking-Ice	etc	16:21
roshi	ok	16:21
Viking-Ice	I mean our	16:22
Viking-Ice	there is no I in team	16:22
roshi	do we have a list of all that, or is that something that should be created?	16:22
adamw	we have the critpath list	16:22
Viking-Ice	yup we can start with that	16:22
* handsome_pirate wonders if anyone has considered how much extra work this is going to be		16:22
adamw	we used to try and cover that with triage under bz	16:22
adamw	well, it's not like we'd expect 100% coverage right away or anything, i do like the idea of trying to at least get some kind of systematic coverage of different components	16:23
adamw	but we're a long way over time at this point	16:23
adamw	Viking-Ice: it'd be nice if you'd give us a day's advance notice of big ideas so we can work them into the meeting schedules :P	16:23
Viking-Ice	adamw, well perhaps we should just schedule a brain farting session on etherpad and take something to build upon from that	16:24
adamw	blue sky thinking!	16:25
handsome_pirate	Ponies!	16:25
handsome_pirate	Unicorns!	16:25
Viking-Ice	so let's close this meeting since the general idea is out there and try to hold a spesific qa member meeting to start working out the larger vision	16:25
handsome_pirate	Rainbows!	16:26
adamw	sure, we could do that sometime	16:26
adamw	Viking-Ice: sounds like a plan	16:26
* adamw sets fuse		16:26
adamw	thanks for comin, folks	16:27
adamw	#endmeeting	16:27

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