22:00:38 #startmeeting qa 22:00:39 Meeting started Thu Sep 18 22:00:38 2014 UTC and is due to finish in 60 minutes. The chair is mtreinish. Information about MeetBot at http://wiki.debian.org/MeetBot. 22:00:40 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 22:00:42 The meeting name has been set to 'qa' 22:00:47 hi who's here today? 22:00:50 hi 22:00:53 Hey 22:00:55 #link https://wiki.openstack.org/wiki/Meetings/QATeamMeeting#Proposed_Agenda_for_September_18th_2014_.282200_UTC.29 22:00:56 o/ 22:00:58 hi 22:00:59 hi 22:01:01 ^^^ Today's agenda 22:01:13 o/ 22:01:30 o/ 22:01:54 ok let's get started 22:02:03 #topic Summit topic brainstorming (mtreinish) 22:02:29 so I figured I'd just remind everyone about the etherpad for brainstorming summit topics 22:02:36 #link https://etherpad.openstack.org/p/kilo-qa-summit-topics 22:02:53 I couldn't remember if I did it for this meeting timeslot before or not 22:03:03 although the agenda logs show that I did 22:03:30 but, anyway if you've got ideas for a topic you'd like to discuss at summit feel free to add it to the agenda 22:03:54 that's really all I had for this 22:04:38 ok, lets move on 22:04:50 #topic tempest-lib commit history tracking (mtreinish) 22:05:10 so tempest-lib is almost ready for the first release 22:05:37 but as part of trying to do the second migration I realized that I was mistaken and we could migrate git history from tempest a second time 22:05:52 so the history we have there is what we'll have moving forward 22:06:18 mtreinish: what about andreaf's change to the base test class around resource cleanup? 22:06:24 so to keep a record of what work is being migrated I changed the script to generate commits like: https://review.openstack.org/121994 22:06:54 dkranz: so right now the base test class I migrated to tempest-lib is very minimal and doesn't really include anything 22:07:09 and it'll just be for the cli tests 22:07:27 why we need to move history also? 22:07:49 we can check sync by seeing the code. 22:08:02 between tempest and tempest-lib 22:08:04 oomichi: well it would have been nice to keep the history so things like git blame, etc would work. 22:08:42 ah, ok. nice for me. 22:08:59 mtreinish: thanks 22:09:04 andreaf, dkranz: I'm not sure migrating the base test class over is a good idea at this point 22:09:30 mtreinish, because of the resource-cleanup bp? 22:09:33 or moving forward either the resource cleanup bp shows how coupled that is to tempest, and if we moved that out of tree fixing things will be more difficult in the future 22:09:39 andreaf: well for the short term yeah 22:09:40 mtreinish: I'm not either. I am just concerned about the stable api issue 22:10:37 mtreinish: I am not sure what the rush is for tempest-lib if its interface is not stable 22:10:57 dkranz: well that's why I'm concentrating on things with a stable api so we can lock them down 22:11:05 mtreinish: Because it only has real value if other projects use it 22:11:18 the cli clients, subunit-trace, the rest clients and auth layer are next 22:11:21 mtreinish: right, but that is almost nothing 22:11:38 mtreinish: ok 22:12:01 mtreinish: all of the rest clients? 22:12:03 dkranz: well, having the cli framework will let the clients do there own functional testing so we can rip them out of tempest 22:12:14 andreaf: just the core, common/rest_client.py 22:12:23 mtreinish: ok 22:13:13 anyway this discussion is why I think an in person talk about tempest-lib direction at summit will be useful 22:13:23 so we can have a more concrete direction on it's future 22:13:28 yu[ 22:13:34 yup 22:13:57 ok let's move on, unless anyone else has something to add 22:14:45 mtreinish: whats about updates going on on files moved to tempest-lib and not yet removed from tempest 22:15:17 gmann: if they're in tempest-lib and the reviews are still in progress we should probably tell them to respin the commit against tempest-lib 22:15:21 will we force those patch to do for tempest-lib or sync again while removing them from tempest 22:15:25 gmann: do you have an example 22:15:31 ok 22:15:44 gmann: I think we should probably try to avoid a resync 22:16:10 not exactly but some CLI patches are there but need to check if they change base class which is moved to tempest=lib 22:16:14 I'm planning to push the first release after: https://review.openstack.org/121995 and https://review.openstack.org/121988 merge 22:16:42 gmann: they'll definitely have to be rebased if they don't land before: https://review.openstack.org/117649 22:16:59 mtreinish: ok, sounds good 22:17:16 ok, let's move on 22:17:19 #topic Specs Review 22:17:39 so does anyone have an open qa-specs that need discussion or reviews 22:18:04 although at this point in the cycle I don't see us taking on too much more work until after summit 22:18:09 mtreinish, my usual one - asking for reviews on the ssh auth spec 22:18:20 mtreinish, https://review.openstack.org/#/c/94741/ 22:18:36 #link https://review.openstack.org/#/c/94741/ 22:18:48 ok, yeah that one is on me, I keep forgetting to review it :) 22:19:13 I'll try to get to it tomorrow 22:19:15 mtreinish: well feedback from anyone is welcome 22:19:40 yes very true, everyone should review it 22:20:18 does anyone else have a spec, otherwise lets move on 22:20:49 #topic Blueprints 22:20:57 #link https://blueprints.launchpad.net/tempest/juno 22:21:16 so at this point we still have a fair number of open bps 22:21:27 does anyone have a status update on a bp? 22:21:36 mtreinish: resource-cleanup 22:21:52 about tempest-client-scenarios, The status is that: Total: 18 test files (MERGED:13, APPROVED:1, SUBMITTED:4, NOT YET:0) 22:21:54 mtreinish: I'm try to push the first part out as fast as possible 22:22:05 masayukig, sorry go ahead 22:22:19 #link https://etherpad.openstack.org/p/tempest-client-scenarios 22:22:35 almost there 22:22:35 masayukig: oh, cool we have all of them submitted now! 22:22:51 mtreinish: yeah! 22:23:16 once those last 4 land we can start removing all the official client handling code 22:23:23 which will be awesome, and clean a lot of things up 22:23:28 mtreinish, masayukig: yes so I migrated test_stamp https://review.openstack.org/122196 22:23:48 mtreinish, masayukig which fails but it fails for the same reason why the test is skipped 22:24:00 I have a patch on top to unskip to see the test running 22:24:09 otherwise I think the code is ok 22:24:49 andreaf: thanks, I'll have a look later. 22:24:51 andreaf: ok, yeah I'll take a look at that. I'll ask jgriffith about that test too 22:25:01 because it's been skipped for a very long time 22:25:21 yes basically vdb doesn't show up even if attach is ok according to API 22:26:12 andreaf: ok, yeah thats what I remember happening, although I thought it was a nondeterministic failure. But whatever 22:26:27 andreaf: ok, was there anything else on resource-cleanup? 22:26:53 I just changed my patches according to your and afazekas' comments 22:27:05 so I believe the base patch is ready to go now 22:27:15 andreaf: ok cool 22:27:25 so I need reviews for ithttps://review.openstack.org/121537 22:27:31 #link https://review.openstack.org/121537 22:27:43 the rest should be easy to review is pretty mechanical 22:28:13 so I'd really love to see that going through quickly, we can do the split of resource_setup with more time 22:28:48 andreaf: ok, I'll put it on my list for tomorrow. Unless someone beats me to it 22:29:01 does anyone else have an update on a bp in progress? 22:29:04 regarding test-accounts, there are only two patches left, which are both in good shape 22:29:24 however we should wait for the infra change to be able to test non-isolated run on demand 22:29:42 https://review.openstack.org/120861 22:29:58 then we can have all the code merged 22:30:04 andreaf: ok I'll bug clarkb, jeblair, and fungi about that 22:30:13 mtreinish, thanks 22:30:44 are there any other bps to discuss? 22:30:57 mtreinish: https://review.openstack.org/112581 was just pushed 22:31:17 mtreinish: that is the resource cleanup from dpatterson 22:31:18 dkranz: oh, good. I'll take a look at that soon 22:31:31 mtreinish: I played with it a bit today. Pretty nifty. 22:32:15 dkranz: do you know which timezone dpaterson is in? 22:32:31 mtreinish: Boston 22:33:14 ah, ok that's convenient. 22:33:32 from a quick glance I wanted to work through some of the docs stuff on that patch. It'll probably be easier to do interactively 22:33:58 another call for open bps to discuss, otherwise we'll move on 22:35:00 #topic DevStack 22:35:10 so dtroyer said he couldn't make it today 22:35:25 but there were 3 patches he wanted to discuss if he could make it 22:35:33 #link https://review.openstack.org/105785 22:36:16 which was more a discussion on timeline for switching the default over to neutron 22:36:31 a lot of +1s :) 22:36:57 mtreinish: asap if you ask me 22:36:58 I think that really deserves a ML thread, because it involves a lot of people 22:37:22 oomichi: oh, wow I completely missed that the first time :) 22:37:36 dkranz: yeah I tend to agree, but it might be a bit political 22:37:51 mtreinish: I'm not so sure. But we should find out. 22:38:02 the second patch for discussion was: 22:38:04 #link https://review.openstack.org/98854/ 22:38:19 that switches every endpoint to use ssl 22:38:36 which was I think a priority discussion 22:38:51 and the 3rd was: 22:38:53 #link https://review.openstack.org/#/c/111734/ 22:39:19 which I think was a call for reviews 22:39:44 re: 1st one: I agree that we need to discuss it on -dev ml. 22:40:11 that should depend on nova-net vs. neutron. 22:40:24 Second one is most controversial to me 22:41:17 dkranz: really? I'm curious why 22:41:19 If devstack is a developer tool it might waste a lot of runtime with ssl 22:41:27 On the other hand it is more realistic 22:41:47 when previous release(icehouse), nova team anounced that nova-net should be CURRENT, I'm not sure current status about that. 22:41:52 oh, that's a good point if the traffic is encrypted it's harder to intercept to figure out what's going on 22:42:30 mtreinish: So I think it is really about what is the primary mission of devstack 22:42:30 oomichi: yeah, but that was more caused by a bunch of big regressions caused by freezing the code 22:42:40 mtreinish: and we are talking just about the default anyway 22:42:57 dkranz: yeah, that might warrant a ml thread too :) 22:43:13 since not many devstack cores are around now 22:43:28 is ssl support with devstack thoroughly documented in the event a developer wants to configure it that way? 22:43:49 * andreaf 's brain is shutting down against my will - going to bed for today 22:43:57 andreaf: zzzzzzz 22:44:04 lbragstad: probably not, most of the devstack documentation is through code I think. like devstack.org is really out of dat 22:44:18 andreaf: :) 22:44:19 mtreinish: gotcha 22:44:36 mtreinish: RTFC :) 22:45:10 dkranz: heh, yeah something like that 22:45:25 it's a good way to learn how to deploy openstack too 22:45:27 anyway we've got ~15min left. so let's move on 22:45:41 #topic Bugs 22:45:59 dkranz started the weekly rotation etherpad last week 22:46:03 dkranz: do you have a link 22:46:03 ? 22:46:15 mtreinish: just a sec 22:46:50 https://etherpad.openstack.org/p/qa-bug-triage-rotation 22:46:55 #link https://etherpad.openstack.org/p/qa-bug-triage-rotation 22:47:07 gmann: ok, thanks 22:47:22 so if people could volunteer to cover a week of bug triage that would be great 22:47:24 https://etherpad.openstack.org/p/qa-bug-triage-rotation 22:47:42 once we get down to 0 new bugs, keeping on top of bug growth is a lot easier 22:47:56 because the weekly volunteer just has to make sure a few bugs get triaged 22:48:04 mtreinish: so in addition to triaging new bugs, I knocked about a dozen off the backlog. 22:48:13 I'm down if I am qualified :) 22:48:20 mtreinish: If each person does that over the next six weeks it should be gone 22:48:33 dkranz: oh, nice yeah we can widdle away that will be good too 22:48:48 dpaterson: yeah, anyone can take a week 22:48:59 K 22:48:59 you just need to join the tempest-bugs lp group, which has open membership 22:49:03 mtreinish: It's a little unfortunate that hardly any of them end up actually being tempest bugs 22:49:13 dkranz: that's always been the norm 22:49:20 yeah 22:49:26 I think it comes from a general lack of knowledge about how the gate works 22:49:41 and what the role of tempest and the other pieces are 22:49:54 Yes, and I desire to move on when you know the problem is not with your commit, and probably not even with your project 22:49:59 the desire 22:50:19 yep 22:50:30 ok is there anything else to discuss about bugs? 22:50:37 mtreinish: dkranz: I wrote up brief summary about numbers- https://etherpad.openstack.org/p/Tempest-bug-report 22:50:49 gmann: thanks! 22:50:55 gmann: awesome 22:50:57 which i am thinking of keep updating weekly bases 22:51:10 gmann: yeah that'll be great 22:51:21 gmann: also tracking incomplete will be useful too 22:51:27 mtreinish: I have a script that does that if you want me to add the tempest project to it 22:51:27 because those don't always autoexpire 22:51:36 sure. thanks 22:51:37 lbragstad: oh, cool sure 22:51:39 http://keystone-weekly-bug-report.tempusfrangit.org/weekly-bug-reports/ 22:52:25 gmann: that should make tracking thing easier :) 22:52:28 lbragstad: what determines if a bug is on that list? 22:52:53 dkranz: it uses the lanchpadlib api to pull the bugs for the last 7 days on a project 22:52:57 mtreinish: yes, will check that 22:53:15 lbragstad: you mean reported in the last 7 days? 22:53:22 reported and opened, sure 22:53:27 lbragstad: ok, thanks 22:53:38 lbragstad: having tempest there would be great 22:53:46 lbragstad: thanks. it will really helpful 22:53:58 sure, I can do that 22:54:16 ok, let's move on 22:54:23 #topic Critical Reviews 22:54:41 ok does anyone have any open reviews that could use a little extra attention 22:55:19 I have at least one this week: 22:55:21 #link https://review.openstack.org/113674 22:55:25 which is a doc update 22:55:44 #link https://review.openstack.org/118744 22:55:53 which should clean up the console output from a gate run a bit 22:56:16 and: 22:56:18 #link https://review.openstack.org/122527 22:56:29 which is a critical fix so we start running the cli tests again... 22:57:38 someone else must have a review. I don't want this topic to seem like me stealing meeting time to bother people with my backlog :) 22:59:15 heh, ok well there's 1min left. So is there anything else? 22:59:43 mtreinish: http://keystone-weekly-bug-report.tempusfrangit.org/weekly-bug-reports/tempest-weekly-bug-report.html 22:59:44 mtreinish: I have a question about a bug. can we talk briefly in the qa channel? 22:59:55 lbragstad: ok awesome, thanks 22:59:59 np 23:00:00 mlavalle: sure 23:00:10 ok well that's time 23:00:12 thanks everyone 23:00:12 I don't want to hold anybody here 23:00:15 :-) 23:00:16 #endmeeting