16:00:53 #startmeeting Fedora QA meeting 16:00:53 Meeting started Mon Feb 6 16:00:53 2012 UTC. The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:53 Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:01:09 #meetingname fedora-qa 16:01:09 The meeting name has been set to 'fedora-qa' 16:01:11 #topic roll call 16:01:17 happy monday everyone 16:01:25 who's here for some awesome qa meeting fun times? 16:01:34 * mkrizek is here 16:01:49 * tflink is present 16:02:17 * kparal appears 16:02:33 oh no! a level 65 kparal! 16:03:30 * kparal looks suspiciously for a hidden meaning 16:03:43 your join message reminded me of an RPG for some reason 16:03:44 or any meaning :) 16:04:29 =) 16:04:40 I see. let's slain the latecomers! 16:04:42 alright then, let's get going - not a lot on the agenda this week, happily 16:04:52 * jskladan is in 16:04:58 hi josef 16:05:09 * pschindl is here 16:05:13 #topic previous meeting follow-up 16:05:50 "kparal to file a bug for the failure of rats #2 DVD image to boot" - that got done, during the meeting in fact 16:06:10 yup 16:06:24 #info kparal did file a bug for the rats #2 DVD boo fail: https://bugzilla.redhat.com/show_bug.cgi?id=785808 16:06:36 adamw: hey, just hanging out until the fesco meeting 16:06:42 dont mind me :) 16:06:44 "tflink to test building a boot.iso with usrmove builds " 16:06:48 josef: uh, I meant *our* josef :) 16:07:05 jskladan = also a josef 16:07:14 so this one also got done 16:07:16 yay us 16:07:44 #info tflink was eventually able to build a functional boot.iso, exposing some required changes in the build process along the way 16:07:54 ah haha 16:08:17 * tflink is currently working to build another boot.iso, will post that once its done 16:08:18 alright. anything else anyone wants to pick up on from last week? I don't see anything else 16:09:27 okey dokey 16:09:37 #topic upcoming QA events 16:09:49 so we're into the f17 alpha cycle now, TC1 is out there and under test 16:10:22 feature 'freeze' is tomorrow and RC1 is scheduled for 02-16 16:10:34 so we're going to want at least one more TC, probably two 16:11:01 i think it's probably a good idea to spin a TC2 ASAP now the /usr move has landed in Rawhide, so we have a post-/usr-move TC to bash on and compare to the pre-/usr-move one 16:11:09 anyone any thoughts there? 16:11:39 current version of lorax won't work with usrmove 16:11:44 ah, okay 16:11:50 so we'll need a new one before we can do another compose? 16:11:52 * tflink submitted a patch 16:12:04 yeah, the compose will fail silently, otherwise 16:12:21 as in it will build but won't actually work 16:12:27 #info current lorax does not work correctly with post-/usr-move packages, needs a patch tflink has submitted 16:12:40 thanks for looking into that! 16:12:46 woah, what happend to koji 16:13:08 makeover! 16:13:39 #info koji is looking sleek and sophisticated this morning 16:13:51 adamw: so the usrmove RPMs hit rawhide? 16:13:56 yup 16:14:01 hot tagged over on friday 16:14:12 s/hot/got/ 16:14:12 isn't that going to bork any additional testing w/ TC1? 16:14:27 nvm, I rethought that one 16:14:30 well, depends what you mean by 'testing' 16:14:46 you'll have to follow the upgrade procedure to get a TC1 install to update now, that's about all the consequences 16:15:06 adamw: only from DVD, netinstall shouldn't require upgrade 16:15:16 so, yeah, in case you get stuck upgrading TC1 or need to help someone else, refer to: https://fedoraproject.org/wiki/Upgrading_Fedora_using_yum#Fedora_16_-.3E_Fedora_17 16:15:33 tflink: yeah, indeed, netinstall should give you a /usr-moved system now. and hopefully should work. 16:15:39 #info in case you get stuck upgrading TC1 or need to help someone else, refer to: https://fedoraproject.org/wiki/Upgrading_Fedora_using_yum#Fedora_16_-.3E_Fedora_17 16:15:42 I tried that this weekend and it worked, but there were a couple of oddities. 16:15:52 brunowolff: what were the oddities? 16:16:27 The files get moved over without copying their context type so a relabel is needed, but I didn't see a relabel in the instructions. 16:17:14 The other was that luks password asking was a bit different. Before switching the root, each needed device got asked about 16:17:14 i think one is meant to get triggered somehow as part of the instructions, i forget how precisely 16:17:38 separately and asterisks weren't echoed like normal. 16:18:10 This happened both in the initial conversion boot and when rebooting after the update. 16:18:29 I did a relabel and rebuilt the initramfs again and both issues went away. 16:18:38 brunowolff: ah, the 'fixfiles onboot' triggers an selinux relabel next boot. 16:19:03 The relabel could probably be avoided by copy over the selinix labels when copying the files. 16:19:11 probably. 16:19:42 maybe file a bug on dracut suggesting that? as that's where the script lives 16:21:32 okay, moving on! 16:21:48 oh, i note that the first test day slot is this week but we have no event 16:21:50 OK, I didn't see the fixfiles part since I was off the script by that point since I was already running rawhide. I would have likely noticed a restorecon. 16:22:01 that's fairly usual early in the cycle as people usually want t ogo later 16:22:15 And with restorecon the relabel could have been limited to a few directories and taken less time. 16:23:50 we don't have a j_dulaney to give us a test day update, but he seems to be on top of the tickets. 16:24:08 and we have the next blocker review meeting coming up friday, of course. 16:24:16 yay! 16:24:20 #action adamw to co-ordinate with dgilmore on getting lorax updated and a new compose done 16:24:26 fun abounds! 16:24:55 sounds like there's going to be a new anaconda build today, too 16:25:00 goody. 16:26:41 okay, anything else on upcoming events? 16:28:12 alrighty then 16:28:15 #topic autoqa update 16:28:20 #chair tflink kparal 16:28:20 Current chairs: adamw kparal tflink 16:28:32 take it away, autoqa wielders 16:28:49 * kparal waiting for zimbra... 16:29:07 anyway, the biggest change from last week is that rats_install is finally merged into master 16:29:16 so kudos hongqing 16:29:52 after that we found out about that root= anaconda option that is required right now, so it will need one minor patch to get it fully working for the composes 16:30:08 and of course we need to have the repository branched, but that should happen tomorrow 16:30:20 and probably the most visible AutoQA update: http://autoqa-stg.fedoraproject.org/resultsdb/frontend we finally have the ResultsDB frontend up'n'running since last friday (even though just collecting data from staging). Looks to be working all right so far ;) (kudos to pschindl!) 16:30:38 and yes, resultsdb being the second biggest change :) 16:31:17 #info rats_install is merged into autoqa master, needs minor fixes tomorrow 16:31:29 #info resultsdb is now up and running at http://autoqa-stg.fedoraproject.org/resultsdb/frontend 16:31:35 thanks petr and hongqing! 16:31:52 we also learned that we were not sending Bodhi emails for some time due to expired/locked account. that is now also fixed 16:31:58 so what's the status of rats_install exactly (assuming the fixes tomorrow)? 16:32:10 is it actually at the point now where we can truly have it running entirely automated install tests? 16:32:30 #info bodhi mailshots were broken due to expired/locked account, now fixed 16:32:48 I need to verify that, because I still don't have the branched repo. I need to check the watchers etc. but basically it should be well prepared 16:33:14 I also need to get two bare-metal machines and add them to our staging server so that we can test it our properly 16:33:38 I still haven't considered where I will take these machines from 16:33:53 whatever you do, just don't get caught 16:33:56 kparal: you could take them from production 16:34:16 tflink: yes, that is what I will probably try 16:34:32 ie lock 2 bare metal hosts in production, add them to staging and hope it doesn't explode 16:34:50 * tflink wonders if autotest will be confused with 2 controllers 16:34:52 adamw: does this sentence make you an accomplice? 16:35:10 kparal: not as long as irc is an insecure protocol it doesn't 16:35:21 :) 16:35:40 tflink: you're testing! you're testing! 16:36:08 okay, sounds like good stuff, congratulations 16:36:16 tflink: I'll ask lmr, as long as they are locked at one controller, it shouldn't be a problem I guess 16:37:03 kparal: otherwise, I don't think it's too much of a problem to reinstall those hosts but jlaska would know more 16:37:59 I'll work on that 16:38:51 cool. anything else for autoqa? 16:39:09 that's it 16:40:51 okelydokely 16:41:07 #topic open floor 16:41:13 okay, any other business anyone? 16:42:57 I just want to ask what the current status of remote_installation is 16:43:10 * tflink is waiting for management approval of the demo 16:43:21 everything is set up and ready to go otherwise 16:43:24 on which topic, has anyone heard from benl lately? 16:43:27 great 16:43:32 i'm waiting for a reply from him too 16:43:47 he said something about traveling until tuesday, but I'm not sure 16:44:17 I can harass him in person if you want 16:44:39 he can't escape, he's at the same building, if not traveling 16:44:48 he doesn't seem to be on IRC right now 16:45:45 yeah, hasn't been for a while when i've looked 16:45:50 * tflink is wondering if he should re-setup for F17 alpha instead of F16 for remote_install, though 16:46:03 tflink: that seems reasonable, given that it's what we want to test :) 16:46:29 adamw: my only argument against it is that the demo is for remote_install, not for testing fedora 16:46:38 but that isn't a very strong argument 16:46:56 either way, I'm waiting for the OK before I do much more 16:47:35 okay 16:47:48 kparal: if you could look out for ben it'd be appreciated 16:48:03 or anyone who might know where he's gotten to 16:48:33 adamw: I'll ask around tomorrow 16:48:44 oh, hey, he modified my 1x1 slot for tomorrow. 16:48:50 so he's aliiiiiive anyway, wherever he is. 16:50:47 #info remote_install status: mostly ready to go into public testing, tflink waiting for go-ahead from benl 16:52:14 alright, i guess that's about it 16:52:30 oh, we also need to fill out tc1 testing, of course 16:52:38 looks like robatino and hongqing have knocked out quite a bit of it so far 16:54:34 alrighty then, meeting's over, time for heavy drinking everyone! thanks for coming 16:54:42 #endmeeting