16:01:38 #startmeeting F27-blocker-review 16:01:38 Meeting started Mon Sep 25 16:01:38 2017 UTC. The chair is pschindl_. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:01:38 Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:01:38 The meeting name has been set to 'f27-blocker-review' 16:01:46 #meetingname F27-blocker-review 16:01:46 The meeting name has been set to 'f27-blocker-review' 16:01:54 #topic Roll Call 16:02:00 .hello2 16:02:01 jkurik: jkurik 'Jan Kurik' 16:02:06 .hello2 16:02:07 frantisekz: frantisekz 'František Zatloukal' 16:02:31 Hi all 16:02:33 .hello2 16:02:34 sgallagh: sgallagh 'Stephen Gallagher' 16:02:50 * kparal is here 16:03:12 #chair jkurik kparal sgallagh frantisekz 16:03:12 Current chairs: frantisekz jkurik kparal pschindl_ sgallagh 16:03:50 #topic Introduction 16:03:57 Why are we here? 16:04:08 #info Our purpose in this meeting is to review proposed blocker and nice-to-have bugs and decide whether to accept them, and to monitor the progress of fixing existing accepted blocker and nice-to-have bugs. 16:04:10 #info We'll be following the process outlined at: 16:04:11 The big question indeed... 16:04:12 #link https://fedoraproject.org/wiki/QA:SOP_Blocker_Bug_Meeting 16:04:19 #info The bugs up for review today are available at: 16:04:21 #link http://qa.fedoraproject.org/blockerbugs/current 16:04:23 #info The criteria for release blocking bugs can be found at: 16:04:30 #link https://fedoraproject.org/wiki/Fedora_27_Alpha_Release_Criteria 16:04:32 #link https://fedoraproject.org/wiki/Fedora_27_Beta_Release_Criteria 16:04:34 #link https://fedoraproject.org/wiki/Fedora_27_Final_Release_Criteria 16:04:36 42 16:04:52 #info 3 Proposed Blockers 16:04:59 #info 8 Accepted Blockers 16:05:06 #info 3 Proposed Freeze Exceptions 16:05:21 So let's start with proposed blockers: 16:05:30 #topic (1495204) [Modular Server] TypeError: list indices must be integers or slices, not str 16:05:32 #link https://bugzilla.redhat.com/show_bug.cgi?id=1495204 16:05:34 #info Proposed Blocker, anaconda, NEW 16:06:01 so, this is server, so I'd keep on the proposed list, just skip it for now 16:06:24 ok, let's move then 16:06:28 we can deal with it on modular server beta meetings, which will surely come 16:06:34 #topic (1487956) Plasma severely broken if plasma-lookandfeel-fedora is not installed manually 16:06:37 #link https://bugzilla.redhat.com/show_bug.cgi?id=1487956 16:06:39 #info Proposed Blocker, plasma-workspace, MODIFIED 16:07:24 pschindl_: you said you have some info about how often this happens 16:07:34 when I tried it in a VM, it happened 16:07:44 * sumantrom[m] is out with flight delayed , Will look into logs later. 16:07:55 and having no option to reboot or power off correctly definitely violates the criteria 16:08:28 .hello2 16:08:29 dominicpg: dominicpg 'None' 16:08:31 I met it one hour ago. It is the case which is in the last comment (lbrabec added it) 16:08:49 I'll +1 it for the logout/shutdown problems. 16:08:58 We have a specific criteria for that, I think 16:09:04 We discussed it and lbrabec worked on it after I left. 16:09:14 pschindl_: the last comment only said he encountered it 16:09:21 and it can be fixed with an updated package 16:09:36 pschindl_: so how come openqa doesn't hit it? 16:09:52 Can we confirm that the fix is just a new Requires and has no other changes? 16:10:17 * kparal checking 16:10:23 https://openqa.fedoraproject.org/tests/146262#step/_graphical_wait_login/21 16:10:59 sgallagh: https://src.fedoraproject.org/rpms/plasma-workspace/c/35e512aa0e34835051e59d7bf46c2a1ab2839132?branch=master 16:11:02 I don't know. It just works there. And it is the same way I followed in my case. 16:11:52 pschindl_: well, it booted to the desktop, that's not a problem 16:11:58 the problem is shutting it down 16:12:04 which it seems openqa doesn't do 16:12:19 ok. I wasn't able to log in. That was my problem. 16:12:42 that would be a different problem I think 16:12:54 this is just about splash screen, alt+f2, and reboot dialog 16:13:05 and unlocking 16:13:14 anyway, +1 blocker 16:13:41 "Shutting down, logging out and rebooting must work using standard console commands and the mechanisms offered (if any) by all release-blocking desktops. " 16:13:43 https://fedoraproject.org/wiki/Fedora_27_Beta_Release_Criteria#Shutdown.2C_reboot.2C_logout 16:13:57 ack 16:14:02 +1 to block 16:14:03 ack 16:14:06 yeah, blocker, +1 16:14:19 ah, ok. Then I'm talking about something else. I just know that when I left lbrabec was trying to install plasma-lookandfeel-fedora to fix it so I thought that it's the same problem, sry for confusion. 16:14:28 Then this isn't covered in openqa. 16:14:35 yeah, I see 16:14:59 pschindl_: well, then, we should report the login screen problem as well :) 16:15:25 Only if you can reproduce it consistently, please :) 16:15:27 proposed #agreed - 1487956 - AcceptedBlocker (beta) - This bug violates the criterion: "Shutting down, logging out and rebooting must work using standard console commands and the mechanisms offered (if any) by all release-blocking desktops." 16:15:38 ack 16:15:48 sgallagh: I hope not, it happened to me just once :) 16:15:52 ack 16:15:54 ack 16:16:07 ack 16:16:28 #agreed - 1487956 - AcceptedBlocker (beta) - This bug violates the criterion: "Shutting down, logging out and rebooting must work using standard console commands and the mechanisms offered (if any) by all release-blocking desktops." 16:16:36 #topic (1492858) The whole log is full of messages "tracker-store[5035]: Could not insert FTS text: constraint failed" and tracker-store process continue consume 100% CPU 16:16:39 #link https://bugzilla.redhat.com/show_bug.cgi?id=1492858 16:16:41 #info Proposed Blocker, tracker, NEW 16:16:59 The bug is reported against rawhide (not F27). Do we know this is an issue on F27 as well ? 16:17:45 that's probably a mistake, I assume 16:18:05 ok 16:18:06 but there are no details, no package versions, no debugging, so it's hard to say anything about this 16:18:14 from my POV, it's happening to a single user 16:18:26 Insufficient information, move on. 16:18:27 so -1 for the moment, until we see more reports about this 16:18:42 -1 16:18:42 I would be -1 to block and ask for more details 16:19:21 -1 16:19:41 I have seen tons of "Could not create FTS insert statement: no such tokenizer: TrackerTokenizer" in logs 16:20:09 but no functional difficulties . 16:20:30 just mentioned, if someone had experienced similar :) 16:20:32 proposed #agreed - 1492858 - RejectedBlocker (beta) - This bug doesn't seem to violate any criterion and it wasn't reproduced with F27. 16:20:51 patch 16:20:52 anyway -1 atm 16:20:58 I think it was with F27 16:21:02 dominicpg: Have you seen it on F27? 16:21:19 yes in f27 16:21:32 ok 16:21:57 will try to reproduce and file a bz 16:22:05 proposed #agreed - 1492858 - RejectedBlocker (beta) - This lacks any proper details and seems to be related to a single setup with particular files. Please re-propose if it affects multiple people and has serious consequences. 16:22:25 ack 16:22:26 dominicpg: just comment in the existing one 16:22:27 ack 16:22:34 ok kparal 16:22:41 ack 16:23:15 and probably change the version to F27 16:23:52 #agreed - 1492858 - RejectedBlocker (beta) - This lacks any proper details and seems to be related to a single setup with particular files. Please re-propose if it affects multiple people and has serious consequences. 16:24:08 ack 16:24:10 actually it's also proposed as a FE 16:24:21 So let's move to proposed FEs 16:24:33 we can start with this one :) 16:24:47 -1 FE 16:24:59 #topic (1492858) The whole log is full of messages "tracker-store[5035]: Could not insert FTS text: constraint failed" and tracker-store process continue consume 100% CPU 16:25:01 #link https://bugzilla.redhat.com/show_bug.cgi?id=1492858 16:25:03 #info Proposed Freeze Exceptions, tracker, NEW 16:25:03 Any change here is likely to be part of a stack that risks destabilization 16:25:05 :) 16:25:13 I would be -1FE until we have more details and we know the impact 16:25:21 I'd like to propose we simply unpropose all proposed FEs which have no signs of a fix being worked on or ready. because otherwise it's a big waste of time 16:25:34 we can approve a FE any time, we can always gather 3 people on irc 16:25:50 so it seems pointless to discuss something that has very little chance of having a fix in time anyway 16:26:11 kparal: I like the idea 16:26:21 kparal++ 16:26:21 sgallagh: Karma for kparal changed to 4 (for the f26 release cycle): https://badges.fedoraproject.org/tags/cookie/any 16:26:33 yes, make sense 16:26:35 ok, so I'll just unpropose this one and we can move on 16:26:45 pschindl_: I'll add #info 16:26:46 New policy for FEs in general? They only can be proposed if a fix is already available? 16:27:06 or a dev is working on it. or we know a dev wants to work on it 16:27:15 or there's at least some communication between us and the dev 16:27:18 OK 16:27:26 something more than vacuum :) 16:27:39 kparal: Now you're just trying to change ALL the rules ;-) 16:28:20 adamw might dislike this, but I think it's a better use of time. we can review the approach once he's back 16:28:23 for the moment, I'd just use it 16:28:46 #info unproposing as FE because there's no sign of a fix coming yet 16:28:57 #topic (1495156) weather is still loading in notification area 16:28:59 #link https://bugzilla.redhat.com/show_bug.cgi?id=1495156 16:29:01 #info Proposed Freeze Exceptions, gnome-weather, NEW 16:29:51 the same thing here, unpropose until a fix is ready 16:29:53 So unpropose? :) 16:29:55 pschindl_: Thank you for not proposing that as a blocker :) 16:30:19 we can still do that! probably final, though :) 16:30:27 critical information missing - should I take my umbrella today? 16:31:00 Yeah, I could get behind a final blocker here 16:31:11 Do we want to just re-propose for that? 16:31:32 ah, not really. if this was the last blocker, would you slip for it? 16:31:47 At Final? Maybe 16:31:49 I would, I would :) 16:31:51 Definitely not at Beta 16:32:04 Let's just unpropose it for now. 16:32:12 ok 16:32:12 sgallagh, you surprise me sometimes :) 16:32:19 once we have Beta Freeze lift, we do not need the FE, so it might be fixed between Beta and Final even there is no FE 16:32:20 It's quite visible bug, so we can propose it again later. 16:32:20 pschindl_: info a move on 16:32:29 kparal: I will attempt to improve so that I surprise you most of the time 16:32:37 good 16:32:50 #info there is no response from dev yet, so unproposing for now. 16:33:03 #topic (1494598) Deprecation of liblvm API and python bindings 16:33:05 #link https://bugzilla.redhat.com/show_bug.cgi?id=1494598 16:33:07 #info Proposed Freeze Exceptions, lvm2, POST 16:33:46 honestly, I don't like this much 16:33:52 it's lvm that's changing 16:33:59 just in order to add a warning 16:34:04 that doesn't need to be on the compose medium 16:34:10 it can be as a regular update 16:34:15 Yes 16:34:18 -1 FE 16:34:18 should not this wait for the Beta Freeze Lift ? 16:34:31 -1 FE 16:35:10 -1 FE 16:35:11 -1 FE 16:35:29 -1 FE 16:36:25 proposed #agreed - 1494598 - RejectedFreezeException - This bug would mean a change in critical component. We think that update after release should be enough. 16:36:45 patch 16:38:11 proposed #agreed - 149598 - RejectedFreezeException - This bug does not address any issue that cannot be resolved with an update after installation. 16:38:32 ack 16:38:35 ack 16:38:42 ack 16:38:58 ack 16:39:04 Sorry that took so long. Doorbell rang as soon as I typed "patch"... 16:39:21 #agreed - 149598 - RejectedFreezeException - This bug does not address any issue that cannot be resolved with an update after installation. 16:39:43 That's all from proposed blockers and FEs 16:39:53 Let's move to accepted blockers 16:40:04 let's go through those which are not verified and not modular server related 16:40:18 Looks like they are all marked ON_QA 16:40:29 Should we just ask for a compose and call it a day? 16:40:38 #topic (1494138) crashes in wayland/Xorg on vc4 with 4.13 with the Raspberry Pi 16:40:40 #link https://bugzilla.redhat.com/show_bug.cgi?id=1494138 16:40:42 #info Accepted Blocker, kernel, ON_QA 16:40:43 (A compose including the patch for KDE) 16:40:58 this one needs testing from somebody with rpi3 16:41:02 using RC2 16:41:24 sgallagh: this will be quick 16:41:28 ok 16:41:46 I will try this. 16:41:59 #action pschindl to test this bug 16:42:12 #topic (1170803) calls e2fsck on all ext volumes, provides no status indicator, and hangs indefinitely if e2fsck doesn't exit 16:42:14 #link https://bugzilla.redhat.com/show_bug.cgi?id=1170803 16:42:16 #info Accepted Blocker, python-blivet, ON_QA 16:42:29 I propose to just close this bug because there seems to be no one willing to test the fix 16:42:32 I feel like we've seen this one more than once 16:42:43 we can alwways reopen this if it occurs again 16:42:50 +1 16:43:45 +1 16:43:48 +1 16:43:58 #info there are no more reports, we'll just close this one, can be reopened if it will happen again. 16:44:10 And that's it :) 16:44:25 #topic Open floor 16:45:01 OK, so new compose request for the KDE thing and transfer most results from RC2? 16:45:07 (Since the only change will be a new dep) 16:45:11 as sgallagh already mentioned I would like to as QA to request RC asap, as we already have patches for all the blockers 16:45:32 +1 16:45:41 yes, I'll file a request right away 16:45:42 Oh, are there others not yet in a compose? 16:45:45 OK 16:46:03 kparal: Which others are going in? 16:46:31 some FEs 16:46:35 mate-desktop-1.19.0-4.fc27 for #1493739 16:46:58 Can we re-review those FEs? 16:46:59 systemd-234-7.fc27 for #1475570 16:47:04 we can 16:47:08 it's probably a good idea 16:47:13 I'd like to be comfortable that we aren't doing anything like UPDATING SYSTEMD 16:47:20 haha 16:47:30 :D 16:47:39 pschindl_: work, work 16:48:01 Or is that one a blocker fix? 16:48:07 actually, I was wrong, systemd update was in RC2 16:48:10 is that https://bugzilla.redhat.com/show_bug.cgi?id=1475570 ? 16:48:14 #topic (1491045) AArch64 install fails with 'The package 'grub2' is required for this installation.' 16:48:16 #link https://bugzilla.redhat.com/show_bug.cgi?id=1491045 16:48:18 #info Accepted Freeze Exceptions, anaconda, ON_QA 16:48:55 pschindl_: let's do it differently, I'll tell which just which ones to discuss 16:49:04 instead of going through all 16:49:15 kparal: I'll leave it on you :) 16:49:34 pschindl_: so, 1493739 16:49:34 Is this one already in a compose, then? 16:49:38 yes 16:49:42 ok 16:50:07 #topic (1493739) Mate-Compiz spin needs f27 default background for beta release 16:50:09 #link https://bugzilla.redhat.com/show_bug.cgi?id=1493739 16:50:11 #info Accepted Freeze Exceptions, mate-desktop, ON_QA 16:50:35 this one should be safe to include in RC3 16:50:44 Yeah, very limited in potential impact. 16:50:49 +1 16:51:00 +1 16:51:21 +1 16:51:21 +1 16:51:24 * nirik arrives late, votes +1 also 16:51:29 #info this one is deemed safe for RC3 16:51:39 ok, next selinux-policy-3.13.1-283.4.fc27 for #1483331 (FE) #1491508 #1451381 #1459081 #1485050 #1494108 16:51:53 1494108 being quite important 16:51:56 nirik: for context, we're deciding which of the FEs that are ready we still think are safe enough to include in RC3 16:52:07 yep. I read back up. 16:52:10 ok 16:52:15 also fixing mariadb and freeipa somewhat 16:52:19 .bug 149108 16:52:19 sgallagh: Bug 149108 – iiimf-le-unit works incorrectly on Greek - https://bugzilla.redhat.com/149108 16:52:22 oops 16:52:30 .bug 1491508 16:52:30 sgallagh: Bug 1491508 – [Modular Server] FreeIPA server deployment fails with SELinux in enforcing mode, despite no obvious denials - https://bugzilla.redhat.com/1491508 16:52:48 I can't read... 16:52:55 .bug 1494108 16:52:55 sgallagh: Bug 1494108 – On KDE, selinux prevents log in with newly created user - https://bugzilla.redhat.com/1494108 16:52:57 #topic selinux-policy-3.13.1-283.4.fc27 for #1483331 (FE) #1491508 #1451381 #1459081 #1485050 #1494108 16:53:00 * sgallagh copies this time 16:53:33 I'm *really* hesitant about updating SELinux between RCs. 16:53:43 but there are quite a few nice fixes in there 16:53:53 and it's only monday :) 16:53:56 kparal: And how many new denials? :) 16:54:00 so much time till thursday 16:54:19 kparal: Are you volunteering for "hero mode"? Is that what I'm hearing here? 16:54:20 well, that was proposed for blocker, but not accepted? 16:54:21 we can perhaps compose *two* RCs in that time :) 16:54:37 nirik: which one? 16:54:42 ah, thats a final ok. 16:54:44 the kde one. 16:54:47 yep 16:55:48 so are there any blockers not addressed in rc2? I guess the kde non functional one without that theme package? 16:56:00 nirik: yes, that's the only one 16:56:04 I'm going to vote 0 here. I won't block it, but I'm wary of the impact it might have. 16:56:54 so, we could do a rc3 with just that one, and a rc4 with various FE's and pick which one? but that increases load on everyone. 16:56:56 * kparal looking for a diff 16:57:17 nirik: the compose time is already unbearable 16:57:32 well, that but also 2x as many things to test, etc. 16:58:12 https://src.fedoraproject.org/rpms/selinux-policy/c/307907ae6b973c05ab009674d0feff0218a29088?branch=f27 16:58:14 I'd rather we just do a single compose with only non-risky FEs accepted 16:58:20 not as short as I hoped 16:58:36 I don't have enough insight to know how risky this SELinux change is, unfortunately. 16:59:04 But since none of the issues it addresses are blockers, I'm erring slightly on the side of not pulling it in 16:59:21 seeing this I'd perhaps avoid the selinux-policy update 17:00:15 it does seem like a lot of change. 17:01:23 ok, let's agree on avoiding it for the moment 17:01:54 +1 to avoid for rc3 17:02:16 I agree. 17:02:49 +1 to avoid 17:02:50 #info we'll avoid pulling new selinux-policy to RC3 due to it being too risky 17:03:01 and that's it 17:03:07 patch :) 17:03:13 sgallagh: feel free 17:03:35 #info Specifically, the size of the changes to the selinux-policy is significant and therefore represents an unknown degree of risk. 17:04:31 so rc3 will have the kde fix and the mate backgrounds and thats it? 17:04:51 yes 17:05:00 #topic Open Floor 17:05:19 I'll file the ticket now 17:05:24 anything else? 17:05:27 #info jkurik relies on kparal to be present on Go/No-Go meeting on Thursday as a QA representative :-) 17:05:32 yes 17:05:38 thanks in advance 17:07:42 * sgallagh feels bad. This was almost one of the shortest blocker meetings ever, until he opened his mouth :) 17:08:23 :) 17:08:36 ok, if there's nothing else, I'll conclude this meeting 17:08:49 ack 17:09:16 ack 17:09:22 ack 17:09:28 thanks for joining 17:09:32 #endmeeting