17:01:04 #startmeeting F21 Beta Go/No-Go meeting 17:01:04 Meeting started Fri Oct 24 17:01:04 2014 UTC. The chair is jreznik. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:01:04 Useful Commands: #action #agreed #halp #info #idea #link #topic. 17:01:05 #meetingname F21 Beta Go/No-Go meeting 17:01:05 The meeting name has been set to 'f21_beta_go/no-go_meeting' 17:01:19 hi again! 17:01:22 #topic Roll Call 17:01:27 .hello roshi 17:01:28 roshi: roshi 'Mike Ruckman' 17:01:47 * kparal lurks 17:02:02 .hello sgallagh 17:02:04 sgallagh: sgallagh 'Stephen Gallagher' 17:02:25 * satellit listening 17:03:06 waiting for a moment... 17:03:09 * pwhalen lurks 17:03:10 ahoy 17:03:15 #chair roshi kparal sgallagh adamw 17:03:15 Current chairs: adamw jreznik kparal roshi sgallagh 17:04:18 #topic Purpose of this meeting 17:04:20 #info Purpose of this meeting is to see whether or not F21 Beta is ready for shipment, according to the release criteria. 17:04:21 #info This is determined in a few ways: 17:04:23 #info No remaining blocker bugs 17:04:24 #info Release candidate compose is available 17:04:26 #info Test matrices for Beta are fully completed 17:04:27 #link http://qa.fedoraproject.org/blockerbugs/milestone/21/beta/buglist 17:04:29 #link https://fedoraproject.org/wiki/Test_Results:Fedora_21_Beta_RC1_Install 17:04:31 #link https://fedoraproject.org/wiki/Test_Results:Fedora_21_Beta_RC1_Base 17:04:32 #link https://fedoraproject.org/wiki/Test_Results:Fedora_21_Beta_RC1_Desktop 17:04:33 #link https://fedoraproject.org/wiki/Test_Results:Fedora_21_Beta_RC1_Server 17:04:56 #topic Current status 17:05:29 so we can use some of QA's Highly Professional Tools to check status (har) 17:05:38 today, we have RC1 but there's a list of proposed blocker bugs we have to go through 17:05:53 adamw: any hint how to use that hi-tech tool? 17:06:02 jreznik: well, first ze blockers 17:06:02 https://qa.fedoraproject.org/blockerbugs/milestone/21/beta/buglist 17:06:19 it's already above :) 17:06:36 * kalev cringes at the list of blockers. 17:07:40 * Corey84 does too 17:08:02 6 bugs in the queue of proposed blocker bugs 17:08:10 willing to help with the multipath as i use lvm 17:08:11 #info 6 bugs in the queue of proposed blocker bugs 17:08:46 #info RC1 is available for release validation 17:08:52 1154743 that is 17:09:16 let's move on to the mini blocker review before we can say go/no-go 17:09:36 roshi: may I ask you to lead it? to fulfil all process needs :) 17:09:36 ayup 17:09:39 For certain definitions of "mini" 17:09:44 sure thing 17:10:05 thanks! 17:10:07 first off 17:10:13 #info http://fedoraproject.org/wiki/Fedora_21_Beta_Release_Criteria 17:10:27 #topic (1155026) Missing addon results in a traceback when %addon header options are specified 17:10:30 #link https://bugzilla.redhat.com/show_bug.cgi?id=1155026 17:10:32 #info Proposed Blocker, anaconda, MODIFIED 17:11:47 so this one we determined should be fixable with updates 17:12:09 I'll note that this is an accepted freeze exception and marked as fixed in the upcoming anaconda build, whenever that comes 17:12:10 but you used *that* word 17:12:22 the important thing isn't the version of anaconda in the *installation environment* but the version installed to the system, because that's what initial-setup uses when it runs. so long as the user gets a fixed anaconda from updates while installing it'll be fine 17:12:38 -1 17:12:44 -1 17:12:52 -1 17:12:59 as the bug doesn't affect live images and we don't ship a DVD with KDE on it any more, I couldn't think of a scenario where it matters what version of the anaconda package is on the media 17:13:21 * danofsatx has arrived 17:13:36 proposed #agreed - 1155026 - RejectedBlocker - As this bug can be fixed with updates there's no reason to block for it. 17:13:36 and Workstation uses gnome-initial-setup which is unaffected 17:14:13 right 17:14:13 ack 17:14:17 ack 17:14:31 #agreed - 1155026 - RejectedBlocker - As this bug can be fixed with updates there's no reason to block for it. 17:14:42 * jreznik would really prefer force it in anaconda and no initial setups involved at all, we are not oem :) 17:14:47 #topic (1155576) TypeError: must be string, not None 17:14:47 #link https://bugzilla.redhat.com/show_bug.cgi?id=1155576 17:14:47 #info Proposed Blocker, anaconda, POST 17:14:52 jreznik: i tried to sell that once, it didn't fly. there was some kind of good reason... 17:16:57 that one seems like user error to me 17:17:10 this one could lead to some confused users 17:17:17 * jreznik is not sure he understands this bug 17:17:28 it's not a user *error* 17:17:28 indeed but doesnt it still force at least one 17:17:36 jreznik: aiui, go into the root pw spoke then leave without entering a password 17:17:43 entering the root password spoke and not setting a password makes a crash 17:17:44 that's not really an error, you maybe just changed your mind 17:17:54 yeah 17:18:11 it's kinda borderline for me, but i think if this was the last blocker we'd probably fudge it, so to be consistent i'll say -1 blocker, +1 FE, +1 Final blocker 17:18:13 and a reasonable expectation is to be able to poke at things without setting anything 17:18:48 it's not nice behaviour but I don't see that need for Beta 17:18:53 do we have a volunteer to secretarialize? 17:18:54 it's a conditional violation so we have to make a subjective call on how 'significant' it is, i can probably be OK with this kind of bug being acceptable in beta but definitely should be fixed for final 17:18:57 roshi: i'm doing it already 17:19:03 sweet, thanks 17:19:06 but at least user should be forced even without root id think I'm with adamw 17:19:09 -1 blocker/+1 FE 17:19:19 -1 blocker / +1 FE 17:19:23 but definitely +1 final 17:19:59 yes, installer tracebacks leave a really bad impression, so I wouldn't mind if we mark it as a final blocker now 17:20:35 Corey84: you're required to create either a root pw or admin user, this is simply a crash, not a logic error 17:20:52 dc'd there (formerly Corey84) 17:21:24 proposed #agreed - AcceptedFreezeException AcceptedBlocker(Final) - While this bug is annoying, it's not quite enough to block for Beta. However, it is accepted as a blocker for Final. If a fix lands before Beat release, please pull it in. 17:21:55 ack 17:22:06 ack 17:22:08 ack 17:22:14 #agreed - AcceptedFreezeException AcceptedBlocker(Final) - While this bug is annoying, it's not quite enough to block for Beta. However, it is accepted as a blocker for Final. If a fix lands before Beat release, please pull it in. 17:22:24 #topic (1156354) ValueError: cannot modify protected device 17:22:24 #link https://bugzilla.redhat.com/show_bug.cgi?id=1156354 17:22:24 #info Proposed Blocker, anaconda, NEW 17:23:37 adamw and I already voted in the bug 17:23:47 seems like a pretty clear blocker to me too, sadly 17:23:57 indeed 17:24:08 +1 17:24:12 this is pretty +1, yeah. seems to be a complete showstopper for at least intel fwraid 17:24:39 +1 17:24:40 +1 17:25:04 proposed #agreed - 1156354 - AcceptedBlocker - This is a clear violation of the Beta firmware raid criteria: "The installer must be able to detect and install to hardware or firmware RAID storage devices." 17:25:33 ack 17:25:38 ack 17:25:57 ack 17:25:58 #agreed - 1156354 - AcceptedBlocker - This is a clear violation of the Beta firmware raid criteria: "The installer must be able to detect and install to hardware or firmware RAID storage devices." 17:26:02 #topic (1154347) Anaconda fails to recognize local standard SATA disks after secure-erase 17:26:05 #link https://bugzilla.redhat.com/show_bug.cgi?id=1154347 17:26:08 #info Proposed Blocker, device-mapper-multipath, NEW 17:26:40 probably can't pitch this as just a "Really Secure Erase - Makes even your drive disappear!" can we? 17:27:17 it's all in the marketing, baby 17:27:37 I think randomuser could work some magic with it 17:27:46 lol 17:27:51 pass it off to the magazine, put some FPL sauce on it and we'd be good to go 17:27:58 secure erase == nwipe yes? 17:28:37 puh, this is probably related to his hardware 17:28:45 ^ +1 17:28:58 sounds like a dying h/w case to me 17:29:13 I'd say -1 blocker due to the fact that we haven't heard of this issue from other people, likely specific to his hardware 17:29:26 i dunno how many people would do a secure erase like this 17:29:30 it would be nice for someone else to test, though 17:29:49 yeah, some testing would be good I think 17:30:06 I will attempt in vm tonight /tommorrow 17:30:25 i use nwipe and lvm / luksLVM regularly 17:30:26 well, I guess the F20 anaconda could see those drives... 17:31:20 "The disks are no longer secure-erased. A complete linux system is installed on the drives, and even though the system has a complete/functional linux system installed, the Anaconda installer on the F21A DVD *STILL* will not recognize the drives." 17:31:42 hum, hadn't seen that 17:31:48 so it sounds like some other hardware specific issue, not related to secure erase at all 17:31:54 ^ 17:32:03 or anaconda even 17:32:03 still, maybe we should punt this for some input from anaconda folks 17:32:22 +1 punt 17:32:25 +1 final -1 Blocker 17:32:29 +1 punt 17:32:36 +1 punt 17:32:41 though I feel like if anaconda was missing a bunch of drives all the time we'd have heard about it... 17:32:52 (that's given that we're now slipping, with the previous bug - we have time to punt :>) 17:33:08 or maybe +1 FE so that if the anaconda folks can pull this fix in for the next build if they want, but punt blocker status? 17:33:29 ^ im cool with that 17:33:30 i can be +1 FE, sure, anaconda crashers are always bad 17:33:30 proposed #agreed - 1154347 - Punt - We're going to wait for some feedback from the anaconda team before determining blocker status. 17:33:39 * roshi amends 17:33:41 +1 FE, +1 punt 17:33:50 ^ 17:34:01 * jreznik is not very happy to grant +1 FE now like a blank FE without knowing what's the fix and how intrusive it will be 17:34:13 punt, -1 FE 17:34:56 the bug's clearly that they're being misidentified as multipath for some reason 17:35:14 * kalev goes to get some more blueberry soup. 17:35:16 proposed #agreed - 1154347 - AcceptedFreezeException Punt - We're going to wait for some feedback from the anaconda team before determining blocker status. However, if a fix can be found beforehand pull it in to the next compose. 17:35:32 wait, blueberry *soup*? 17:35:37 o.O 17:35:43 * roshi has never heard of such a thing 17:36:08 nom nom :) 17:36:25 ack 17:36:33 ack 17:36:48 kalev: Is that like blueberry pie without the crust? Is that a thing? That should totally be a thing. 17:36:56 http://www.myrecipes.com/recipe/chilled-blueberry-soup 17:36:57 sgallagh: almost, yes! 17:37:05 ack 17:37:15 #agreed - 1154347 - AcceptedFreezeException Punt - We're going to wait for some feedback from the anaconda team before determining blocker status. However, if a fix can be found beforehand pull it in to the next compose. 17:37:26 #topic (1156380) 21 Beta RC1 live images cannot be written with 'livecd-iso-to-disk --format --reset-mbr --efi' 17:37:29 #link https://bugzilla.redhat.com/show_bug.cgi?id=1156380 17:37:31 #info Proposed Blocker, livecd-tools, NEW 17:37:42 Obvious +1 blocker 17:37:46 yeah 17:37:57 I think we all ran into this at the same time this morning 17:38:10 +1 17:38:26 +1 17:38:37 +1 17:38:48 just to be clear, this isn't a bug in livecd-iso-to-disk, it doesn't look like 17:38:55 does it work with dd ? 17:39:03 it seems to be a bug in image generation - the image generation bits are also in livecd-tools package 17:39:12 yeah, i believe so 17:39:28 if it was the only thing left, I'd be inclined to fudge this by saying to use dd instead 17:39:38 but meh, we have the raid thing as well :( 17:40:04 proposed #agreed - 1156380 - AcceptedBlocker - This bug is a clear violation of the Beta boot criteria: "All release-blocking images must boot in their supported configurations." 17:40:40 kalev: the criteria explicitly say that's not acceptable, we've been over it before and specifically decided we want all supported usb writing methods to work at beta 17:40:49 ack 17:40:50 fair enough 17:40:51 ack 17:41:49 #agreed - 1156380 - AcceptedBlocker - This bug is a clear violation of the Beta boot criteria: "All release-blocking images must boot in their supported configurations." 17:41:51 #topic (1156378) SELinux denies package install for rolekit 17:41:54 #link https://bugzilla.redhat.com/show_bug.cgi?id=1156378 17:41:57 #info Proposed Blocker, selinux-policy, MODIFIED 17:42:47 +1 blocker 17:43:06 nice that there's already a fix too 17:43:11 * roshi likes those blockers 17:43:12 +1 17:43:15 +1 blocker 17:44:10 proposed #agreed - 1156378 - AcceptedBlocker - This bug is a clear violation of the server roles beta criteria: "Release-blocking roles and the supported role configuration interfaces must meet the core functional Role Definition Requirements to the extent that supported roles can be successfully started, stopped, brought to a working configuration, and queried." 17:44:26 ack 17:44:41 ack 17:44:45 ack 17:44:47 #agreed - 1156378 - AcceptedBlocker - This bug is a clear violation of the server roles beta criteria: "Release-blocking roles and the supported role configuration interfaces must meet the core functional Role Definition Requirements to the extent that supported roles can be successfully started, stopped, brought to a working configuration, and queried." 17:44:54 that's it for the proposed blockers 17:45:13 there's 2 proposed FEs if we want to look at those as well 17:45:28 ack 17:45:32 we are on it, why not 17:45:35 not necessary for go/no-go purpsoes 17:45:41 but if there's only 2 17:45:45 it's not neccesary but 2 17:45:46 that's what I was thinking... 17:46:09 sure, let's -- it's always nice to get some extra polish by pulling in FE fixes 17:46:17 #topic (1155026) Missing addon results in a traceback when %addon header options are specified 17:46:20 #link https://bugzilla.redhat.com/show_bug.cgi?id=1155026 17:46:22 #info Proposed Freeze Exceptions, anaconda, MODIFIED 17:46:30 we already looked at this one 17:46:37 yep 17:47:15 no need for FE since updates fix it 17:48:29 well, we could wave the FE wand at it to cover them fixing it along with the blocker bugs, so the fix is in updates 17:48:33 unless I misunderstand something, I guess 17:48:34 well, we _could_ +1 FE it, just to make sure the update is available 17:48:37 right 17:48:56 just say +1, i'll write something plausible. :P 17:49:02 +1 FE 17:49:05 works for me 17:49:06 +1 FE 17:49:39 agree there 17:49:58 proposed #agreed - 1155026 - AcceptedFreezeException - It would be good to get the fixed package into the repo for release. 17:50:25 ack 17:50:46 ack 17:51:54 #agreed - 1155026 - AcceptedFreezeException - It would be good to get the fixed package into the repo for release. 17:51:57 #topic (1155756) NoSuchGroup: None 17:51:59 #link https://bugzilla.redhat.com/show_bug.cgi?id=1155756 17:52:02 #info Proposed Freeze Exceptions, anaconda, POST 17:52:50 how common is not putting the packages you want in your kickstart? 17:53:07 doesn't that kinda violate what a kickstart is for? 17:53:25 no, it's just using a partial kickstart 17:53:28 it's anaconda folks proposing it as FE, they know what they are doing -- I'd be +1 FE 17:53:55 say you want to specify some other attribute of install like the passwords and keyboard layout and storage config, but leave package selection up to the person running the install 17:54:20 +1 FE, not a blocker but would be good to fix 17:54:22 ah, that makes sense then 17:54:28 Still kind of an edge case... 17:54:41 +1 FE 17:54:49 I'm -1 FE. It's not a common enough behavior to break freeze on, IMGO 17:54:51 *IMHO 17:54:55 +1 FE 17:55:12 sorry sgallagh the +1s have it 17:55:21 -1 FE + 1 final rather 17:55:22 * roshi is +/- 0 on it 17:55:33 You will rue the day! or something 17:55:39 haha 17:55:50 hm, i could be argued into it by sgallagh 17:55:58 * adamw wonders what the fix looks like 17:55:59 well, we're 3+/2- 17:56:11 not exactly consensus 17:56:29 but still majority 17:56:40 or, behind door three: just leave this alone til Wednesday and get back to the Go/No-go meeting :) 17:56:54 so do we want to wait for fix? 17:57:00 The point of Freeze Exception as I understand it is to fix issues that are either serious but not blockers or else things that cannot easily be resolved by an update. While this is *kind of* the latter, we still have Final to fix it. 17:57:01 roshi: yep 17:57:10 it's adding a single if: line 17:57:23 sgallagh: well, it's considered in the context of the milestone in question 17:57:36 sgallagh: if we considered Final an 'update' to Beta we'd never accept anything as a freeze exception for Beta, would we? 17:57:51 I didn't phrase that well 17:58:08 lol 17:58:22 https://lists.fedorahosted.org/pipermail/anaconda-patches/2014-October/014297.html is the fix, I believe. 17:58:31 I basically mean that to me, this is a case of "Stick it in (un)Common Bugs" and don't break freeze. 17:58:44 But as I've been outvoted already... *shrug* 17:58:50 i don't see how anythting could Possibly Go Wrong 17:58:55 well, having the fix in for Beta means that the fix gets better testing 17:59:00 sgallagh: we usually aim for broad consensus here, not just a straight majority vote 17:59:08 ok 17:59:08 and even if there's a fallout, we'll be in a better position for Final 17:59:19 and yeah, kalev's point is why i kinda want this one, because this is the kind of use case we want to catch further bugs in 17:59:25 kalev: Unless the fallout causes Beta to slip further... 17:59:38 well, in that case it's trivial to revert -- just a single if() 17:59:44 yeah, sgallagh usually our votes aren't split :) if they are we go for more discussion/bribes/andstuff 18:00:04 Bribes, did you say? ;-) 18:00:14 * jreznik accepts any bribe to vote with sgallagh 18:00:17 this also enables further testing on the custom path that adamw described earlier: setting keyboard layout in the installer, but leaving package selection empty 18:00:19 * roshi said nothing :) 18:00:27 err, setting keyboard layout in the ks 18:00:28 jreznik: Have a single M&M 18:00:40 I can lean +1 for this for the more testing 18:00:55 though it feels kinda edge-casey to me in and of itself 18:01:21 anyway, I don't feel particularly strongly either way 18:01:24 i'm swayed by the fix looking *really* safe 18:01:26 just playing the devil's advocate here :) 18:01:32 it literally just checks if something exists before doing stuff to it 18:01:38 yeah 18:01:48 True enough 18:01:56 more a edge case indeed but big enough to cause ripples im sure 18:02:12 you coming to the light side over here sgallagh ? 18:02:14 Though I don't really have a sense of what the side-effect is of not having groups selected there. 18:02:54 well, logically speaking, the change can only result in the groups being unselected *less* often. regardless of all other considerations. 18:03:02 for clarity, votes again? then we can be done with this one 18:03:05 so if you're worreid about that, the patch is making things better. ;) 18:03:17 +1 18:03:30 +1 18:03:40 Ok, I'm convinced that it's sufficiently unlikely to cause harm. +1 18:03:41 +1 18:03:45 +1 18:03:45 +1 18:03:51 wohoo, consensus! 18:04:13 And the steak dinner I was just promised in a PM had *nothing to do with it* 18:04:15 <.< 18:04:17 >.> 18:04:31 proposed #agreed - 1155756 - AcceptedFreezeException - As the fix is really simple and this helps facilitate more testing, please pull this in the next compose. 18:04:36 sgallagh: don't talk about dinner, /me is really getting hungry 18:04:38 Ack 18:04:41 ack 18:04:46 ack 18:04:48 ack 18:04:57 #agreed - 1155756 - AcceptedFreezeException - As the fix is really simple and this helps facilitate more testing, please pull this in the next compose. 18:05:11 * roshi hands the meeting back to jreznik 18:05:23 * jreznik thanks roshi for great job 18:05:31 #topic Go/No-Go decision 18:05:46 Obviously, blockers remain. No-Go. 18:05:50 as we have accepted unresolved blockers, let's go straight to go/no-go decision 18:06:03 and empty cloud images to boot :) 18:06:32 lol 18:06:33 ng 18:07:17 roshi: how do we track this empty cloud images? 18:07:23 not much on the cloud images personally but still an issue indeed 18:07:41 proposal #agreed Fedora 21 Beta RC1 is no-go due to unresolved accepted blocker bugs 18:07:43 did anyone hear from dgilmore yet? 18:07:44 ack 18:07:46 ack 18:07:57 qa votes no-go due to outstanding blockers, as per our policy on votes in this meeting 18:08:00 ack 18:08:19 #info qa votes no-go due to outstanding blockers, as per policy on votes in this meeting 18:08:24 jreznik: dgilmore was going to look into it - there's talk going on about it in cloud now 18:08:35 #agreed Fedora 21 Beta RC1 is no-go due to unresolved accepted blocker bugs 18:08:37 something with oz got borked or something 18:08:40 maybe we should have a blocker bug to track the cloud issues 18:08:48 I'll make one 18:08:58 and just put it there as a blocker 18:09:03 adamw: yeah, I don't like this virtual blocker nobody really can see 18:09:12 roshi: thanks 18:09:26 I won't let anyone forget the cloudy bits 18:09:36 -1 virtual blocker lets get tracker indeeed 18:09:50 what if you get eaten by a raptor driving a bus in the CLOUD? 18:10:31 if I get eaten by a cloud raptor, I'll make sure to give him the worst indigestion he's ever felt 18:11:18 roshi: you can also juice your blocker stats this way! file the bug and mark all the cloud tests as failing 18:11:24 instant multiplier 18:11:27 :D 18:11:35 ok, thank you guys for coming today 18:11:46 so, we reset with a full week slip, right? 18:11:47 I like being multiplier-ed 18:11:56 sounds about right 18:11:59 and thanks for all the effort you put into RC1 testing 18:12:02 adamw: Right, go/no-go for next Thursday 18:12:03 I'd say so 18:12:30 Yes, thanks very much to the QA folks who worked entirely too late last night. 18:12:31 #action jreznik to announce slip 18:12:50 +1 slip 18:13:18 CoreIT84: That's not really a vote. It's general policy if we vote no-go 18:13:25 #topic Open floor 18:13:39 anything else? otherwise I'll close this meeting 18:14:04 please folks follow up on blockers, verify fixes, file karma etc 18:14:10 as now I'd eat raptor, any raptor I could catch from my couch 18:14:14 it'd be good if we can get as many fixes as possible in today 18:14:27 jreznik: you don't have Raptor Hut in czech? 18:14:37 your raptor's delivered hot and fresh in 30 minutes or your money back 18:14:51 ah, one thing I wanted to mention 18:15:07 30 minuts, great delivery time! 18:15:14 phoronix already posted a new that we had another slip 18:15:20 who's the spy 18:15:26 damnit larabel, i know you're in here 18:15:31 might have been better to not send out an email yesterday about slipping 18:15:45 kalev: my fault, I wrote it in comments 18:15:47 because we weren't really slipping yet at that point; only postponing Go/No-Go decision 18:16:02 i mean, i really wanted to get on record how *AWESOME WE THINK PHORONIX IS* 18:16:10 right folks? we all love phoronix! phoronix is great. 18:16:35 the best! 18:16:57 heh, yesterday I read some PR about clouds - comparison of vps and they said "we used world-respected Phoronix Test Suite" 18:17:44 3... 18:18:19 2... 18:19:28 cloud tracking bug: 1156603 18:19:29 1... 18:19:35 thanks roshi 18:19:46 and thanks everyone again! 18:19:50 #endmeeting