16:00:50 #startmeeting qa 16:00:50 Meeting started Thu Mar 28 16:00:50 2019 UTC and is due to finish in 60 minutes. The chair is gmann. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:51 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:53 The meeting name has been set to 'qa' 16:01:04 who all here today 16:05:25 o/ 16:06:40 o/ 16:06:49 let's start 16:07:59 #topic release 16:08:24 I have pushed all the patches required for stein release on tempest/devstack/grenade etc side 16:08:42 #link https://review.openstack.org/#/q/topic:stein-release-qa+(status:open+OR+status:merged) 16:08:50 added stable jobs etc also 16:09:35 #topic ptg 16:09:48 there is etherpad of qa ptg 16:09:51 #link https://etherpad.openstack.org/p/qa-train-ptg 16:10:05 mark you availability and topics you want to discuss in ptg 16:10:13 * tosky won't attent 16:10:17 attend* 16:10:42 * afazekas is waiting for the answer 16:10:42 QA is schedule for Thursday and Friday 16:10:51 tosky: humm :( 16:11:51 afazekas: from your company about budget things ? 16:16:00 regarding migration of legacy jobs on bionic 16:16:26 except trove it is all done seems- 16:16:29 #link http://lists.openstack.org/pipermail/openstack-discuss/2019-March/004268.html 16:17:47 * afazekas reconnected .. 16:20:06 that's all i have for today 16:20:11 anything else to discuss 16:20:31 grenade! 16:21:20 tosky: yeah, go ahead 16:21:44 topic change? 16:22:26 #topic grenade 16:23:13 I've been working on the native grenade zuul v3 job from andreaf for a while 16:23:37 the status is that it seems to work, I'm fighting to find the cleanest way to capture the log files 16:23:54 because unfortunately some files are duplicated between old/ and new/ 16:24:02 and stage-output does not handle this, so I need to move files around 16:24:22 and I'm trying to do it in the cleanest way, maybe overengineering it, you can see it in the last reviews 16:24:49 apart from that, the job is complete from my point of view, so any review is more than welcome before I remove the WIP part 16:24:54 the reviews are: 16:25:22 https://review.openstack.org/#/c/637265/ - Add GRENADE_DEVSTACK_EXTERNAL to skip part of the setup 16:25:46 which allows the grenade job to skip the devstack run on old/ (now handled by the devstack roles) 16:25:56 https://review.openstack.org/#/c/548936/ - WIP Setup grenade-base, grenade, grenade-{py3,postgresql} 16:25:58 that's the main review 16:26:02 and it requires also 16:26:18 https://review.openstack.org/#/c/637523/ - Always source lib/apache when lib/tls is used 16:26:21 afazekas proposed openstack/tempest master: Unskip test_stamp_pattern https://review.openstack.org/615434 16:26:28 otherwise bad things happen (failures) 16:26:47 tosky: thanks. 16:26:52 if you want to see more complex jobs, apart from the 3 jobs (base, py3 and postgresql) in the main job, I have 16:27:01 https://review.openstack.org/#/c/639774/ heat (even if in the wrong repository) 16:27:02 stage-output thing merged ? 16:27:08 https://review.openstack.org/#/c/638390/ - sahara 16:27:21 the stage-output change is not merged yet, but I realized another issue with it (!) 16:27:32 so I may rewrite it again and conflate the changes 16:28:00 because if we use gzip --recursive, we can pass directly the logs/ directly and forget about the list of files 16:28:38 on the other hand, I found other issue: there is some code which renames the starting dot of hidden files from .file to _file to make them visible, 16:28:59 ohk 16:29:12 but this code works only if you pass the file name directly to zuul_copy_output 16:29:32 if you have a dot file in a directory, and you pass a directory to zuul_copy_output, the file is not renamed 16:30:34 but anyway, it's not a blocker for the grenade change, it can be reviewed and merged in parallel 16:30:55 ok 16:31:24 tosky: can you add the same topic on above required reviews 16:31:35 oh, sure 16:32:08 thanks. 16:32:34 I'd also mention this devstack change, which is not required by the grenade jobs, but it will help in the future because it allows us to clean useless usage of MYSQL_PASSWORD: https://review.openstack.org/#/c/637081/ 16:33:13 sure 16:34:53 i will check them sometime today or tomorrow 16:35:19 anything else ? 16:36:07 * afazekas have 5 review waiting for ayes 16:36:51 #topic open discussion 16:36:57 afazekas: can you past the link here 16:37:31 https://review.openstack.org/#/q/owner:afazekas+project:openstack/tempest+status:open 16:38:40 mostly having the machines alive when pci plug/unplug /acpi event comes 16:39:03 soft reboot test, interesting 16:39:21 afazekas: thanks, i have added them in my list 16:40:32 gmann, No API way, but if you prefer one of the alternatives let me know 16:40:54 ok 16:41:45 question was whether tempest will be able to test it or not 16:42:30 * frickler stumbles in late 16:43:27 I'd like to mention two stacks of devstack patches that I think could get reviewed and merged now 16:44:04 one is reverting the pip<10 cap at https://review.openstack.org/561597 16:44:41 gmann, The acpi event could be captured by editing acpid config files, the config differs on cirros and on another images . I f we just unskip it , some cases it will be soft another cases it will be hard , but both case expected to work 16:45:40 the other is ianw's fedora stack at https://review.openstack.org/#/c/634626/12 16:45:42 gmann, If we just ssh the thing, likely it will be soft, but the only way to guess it is: `time`, what we do not really want to do .. 16:46:22 afazekas: yeah but thing is how to verify it was soft reboot or hard otherwise it is juts reboot test 16:47:48 gmann, We can check is nova sending an acpi event, we would change the acpid, to write the event to the filesystem , if we an read it back, the soft reset button was pressed 16:48:16 frickler: pip uncap works fine now ? 16:49:02 gmann: yes, cmurphy fixed the last remaining suse issue, we just waited for stein to be cut 16:49:26 afazekas: humm, but that will be too specific with image and non consistence 16:49:48 frickler: cool, so devstack stein cut patch is merged now 16:50:06 afazekas: i will check and put these on review. 16:50:08 there may be issues for project-specific child jobs, so I'd like to uncap pip as soon as possible now to allow those to be found 16:50:13 gmann, cirros and fedora differs for sure, I did not checked how different fedora,suse,ubuntu yet 16:50:28 afazekas: ok 16:50:57 frickler: make sense. it will be better to put that on ML before we merge that 16:51:40 gmann: yeah, but I'd like to get the reviews ready for that as the first step 16:52:06 maybe I should just put a W-1 on the top one 16:52:39 ok 16:53:00 gmann, I can go back to patchset 1 as well ;-) 16:53:19 (again), hoping that I'll still get reviews on it 16:53:41 frickler: yeah, i will review the base patches. 16:54:25 afazekas: yeah, i am thinking more on that but let me check more and will add comment there. 16:56:38 ok, anything else or critical review ? else we close the office hour for today 16:57:24 thanks all for joining 16:57:33 #endmeeting