09:00:06 <gmann> #startmeeting qa
09:00:06 <openstack> Meeting started Thu Apr 21 09:00:06 2016 UTC and is due to finish in 60 minutes.  The chair is gmann. Information about MeetBot at http://wiki.debian.org/MeetBot.
09:00:07 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
09:00:09 <openstack> The meeting name has been set to 'qa'
09:00:24 <gmann> hi who's here for today?
09:00:27 <masayukig> \o/
09:00:33 <gmann> masayukig: hi
09:01:04 <jordanP> hi
09:01:21 <gmann> jordanP: hi
09:02:12 <gmann> Let's wait for some time for more people to join otherwise we can finish early today
09:02:36 <dmellado> hi guys ;)
09:02:43 <gmann> dmellado: hi
09:03:09 <gmann> let's start
09:03:15 <gmann> #link https://wiki.openstack.org/wiki/Meetings/QATeamMeeting#Agenda_for_April_21st_2016_.280900_UTC.29
09:03:24 <gmann> ^^ today agenda
09:03:48 <gmann> #topic Specs Reviews
09:04:05 <gmann> #link https://review.openstack.org/#/q/status:open+project:openstack/qa-specs,n,z
09:04:30 <gmann> seems like all spec are under review with author updates
09:04:47 <gmann> I do not have much on that, anyone want to discuss any spec
09:06:02 <gmann> ok, let's jump to next
09:06:09 <gmann> #topic Tempest
09:07:26 <gmann> so most of them patches are in good review
09:07:31 <jordanP> I'd like to have some eyes on https://review.openstack.org/#/c/304024/ (Replace Mox with Mock)
09:07:32 * masayukig thinks it'll be a short meeting this time
09:07:47 <gmann> jordanP: sure.
09:08:10 <gmann> i might check but tomorrow.
09:08:13 <dmellado> jordanP: I'll have a look over the long flight xD
09:08:34 <masayukig> jordanP: should we have a Hacking rule for it?
09:09:10 <jordanP> no hacking rule, let's just not reintroduce mox in test-requirements.txt
09:09:24 <masayukig> jordanP: ah, i see
09:09:32 <jordanP> btw, we should be conservative with any addition to requirements.txt
09:09:39 <gmann> yea that will avoid that
09:10:14 <jordanP> adding new requirements are costly
09:10:23 <gmann> jordanP: quick question
09:10:40 <gmann> why not use mockpatch from oslotest with fixture
09:11:03 <jordanP> I think that's what I use most of the time
09:11:04 <gmann> seems this is done already there - https://review.openstack.org/#/c/304024/2/tempest/tests/lib/base.py
09:11:30 <gmann> jordanP: you are using self.patchobject
09:12:35 <jordanP> in general, I prefer to use native library (mock is in stdlib in python3) than any 3rd party library
09:12:46 <jordanP> oslotest and fixture are 3rd party libraries
09:12:59 <jordanP> if we can do stuff using just the python standard library, it's usually better
09:13:13 <masayukig> jordanP: +1
09:13:50 <gmann> ok, ll check that later in detail.
09:14:13 <gmann> we might need to do more cleanup on that as we use oslotest in many places too
09:14:22 <dmellado> yep, +1 on that, maybe we should change gmann link
09:15:10 <gmann> dmellado: sorry  did not get :)
09:15:38 <dmellado> gmann: so I was referring to change what is on base.py to use also standard library (again, cleanup ;))
09:16:04 <gmann> dmellado: ohk
09:16:13 <gmann> ok
09:16:33 <gmann> i have another which fixing the false pass from expected_success method
09:16:49 <gmann> #link https://review.openstack.org/#/c/306919/3
09:17:09 <gmann> but before this need to fix Ironic plugin and tempest baremetal tests
09:17:18 <gmann> which i just loaded
09:17:54 <gmann> #link https://review.openstack.org/#/c/308855/
09:17:59 <gmann> it will be very quick
09:19:05 <gmann> anything else
09:19:23 <gmann> dmellado: you are ok with venev script on your CI now
09:19:53 <dmellado> gmann: yes, I'm fine now, we'll refactor the wrapper code and reapply, so no worries
09:19:57 <gmann> dmellado: actually that patches waited for around 4 month because of same concern in rally as you have
09:20:03 <gmann> dmellado: cool
09:20:07 <dmellado> gmann: tbh, I missed it
09:20:21 <dmellado> but I guess that it would've been fine to mark it as 'deprecated'
09:20:23 <dmellado> before removing it
09:20:30 <gmann> dmellado: yea, thats normal thing :) for all of us
09:20:52 <gmann> dmellado: actually those were oslo script, we just had copy of those
09:20:59 <tosky> moreover, as we are mainly testing rpms, we probably don't even need the venv
09:20:59 <gmann> same like other projects
09:21:20 <dmellado> gmann: I see, I mean, for other tools, there was a deprecation mark for a couple of months
09:21:31 <dmellado> so maybe a deprecation mark would've been good
09:21:36 <gmann> tosky: yea but we need in many places depends on user usage
09:21:43 <dmellado> tosky: yep, but we do also use git, not only rpm
09:21:46 <dmellado> so some jobs got affected
09:21:56 <gmann> directly git install need those
09:21:59 <gmann> dmellado: yea
09:22:19 <gmann> dmellado: for tempest own interfaces we always use deprecation
09:22:27 <jordanP> (I didn"t follow this subject closely, but ironic code in tempest should go away into a plugin)
09:22:43 <gmann> jordanP: yea i fixed there too master and mitaka
09:23:05 <dmellado> gmann: jordanP in terms of plugins too, do we have a track of plugin-enabled projects?
09:23:17 <gmann> #link https://review.openstack.org/#/q/Ic4b0af5701df99621e3abb693644c4824c92dc4c,n,z
09:23:27 <jordanP> yeah but we (the tempest core team) have no way to easily test baremetal tests, so we can"t maintain them
09:23:50 <gmann> we need to fix same tests in tempest as ironic L and K branch does not have plugin and run tempest tests
09:24:10 <gmann> jordanP: totally true, its in process of removal
09:24:17 <tosky> or backport the plugin?
09:24:30 <jordanP> gmann, thanks for doing this for Ironic. But you won"t be able to fix tempest-plugin for ever. Ironic people should step up if they are interested in running functional tests
09:24:33 <dmellado> tosky: well, that would depend on ironic, not only tempest
09:24:44 <gmann> jordanP: bu before that mtreinish and jroll doing isloated plugin install on gate
09:25:04 <gmann> so that stable branches also can run plugin and then only we can remove those tests from tempest
09:25:12 <jordanP> ok
09:25:29 <dmellado> sounds fine
09:25:40 <gmann> jordanP: +2, i do not want, had to do ceilometer and congress also for removal of ceilometer tests
09:25:49 <gmann> not to be very hard on them
09:26:03 <jordanP> that's nice of you
09:26:05 <gmann> tosky: yea thats all what project want
09:26:29 <gmann> and to avoid impression that Tempest break them :)
09:26:39 <dmellado> lol
09:26:48 <gmann> which always false but nobody understand
09:27:26 <gmann> we have few projects tests which are in progress of removal
09:27:34 <tosky> that should be the role.of QA liason for each project
09:27:44 <gmann> once all those are gone in plugin we will be little bit in relive state
09:28:01 <tosky> explain the changes, I mean
09:28:16 <gmann> tosky: officially yes and we have on wifi too but that does not seems to work :)
09:29:07 <gmann> anyways that might be last fix on plugins and ask each project to do in future
09:29:29 <dmellado> gmann: about the plugins itself, is there a timeframe for that?
09:29:34 <gmann> other thing is we should have all common interface in lib ASAP to avoid such changes break plugin
09:30:02 <gmann> dmellado: you mean for projects (other than 6 core ) which tempest has tests
09:30:41 <gmann> for those we try to remove tests from tempest ASAP, but we do not have strict deadline
09:31:12 <gmann> and for other projects its upto them when they want to implement plugin for tempest coverage
09:31:53 <dmellado> gmann: I see, thanks!
09:32:19 <tosky_> for Sahara, the tests have been copied to sahara-tests, and we run the tests with the Sahara CI, but for the main gates I'm still investigating
09:32:33 <gmann> tosky_: great.
09:32:47 <tosky_> whether to wait for this: https://review.openstack.org/#/c/297446/ or to see if the way of the designate project can work
09:32:59 <tosky_> (install tests from devstack plugin)
09:33:01 <gmann> tosky_: you might need to change their job to redirect the tests to sahara plugin
09:33:15 <tosky_> gmann: Sahara CI is already fixed
09:33:42 <gmann> tosky_: yea, after that, it will be nice to install isolated plugin
09:34:21 <gmann> tosky_: can it be done before that how other projects installing now (with all-plugin tox env)
09:34:33 <gmann> that is bad way but in case.
09:35:02 <tosky_> I was hoping to go directly to the good one :)
09:35:17 <gmann> tosky_: ok, thats fair enough :)
09:35:34 <gmann> hope we get rid of all non-core projects tests by starting of Newton
09:36:27 <gmann> ok, anything on tempest from anyone
09:37:41 <gmann> seems like expected_success changes exposing more bug in tests
09:37:42 <gmann> http://logs.openstack.org/34/308834/1/check/gate-tempest-dsvm-ironic-pxe_ssh/d0ba3b3/logs/testr_results.html.gz
09:37:43 <gmann> :)
09:38:23 <gmann> ok, let's move on
09:38:49 <gmann> #topic DevStack + Grenade
09:39:05 <gmann> if nothing on those we skip this
09:39:41 <gmann> #topic  Austin Summit
09:40:02 <gmann> so we have all schedule up on site now
09:40:40 <gmann> #link https://www.openstack.org/summit/austin-2016/summit-schedule/global-search?t=QA%3A
09:40:51 <gmann> #link https://www.openstack.org/summit/austin-2016/summit-schedule/global-search?t=Infra%2FQA
09:41:00 <gmann> hope this search covered all
09:41:07 <masayukig> cool
09:41:19 <gmann> and hoping not much overlap sessions
09:41:37 <gmann> i have one with nova on Thursday but seems like cannot be changed :(
09:42:19 <gmann> we just start preparing for those sessions and have a great summit at the end
09:42:43 <gmann> anyone want to talk/discuss anything on this
09:43:13 <dmellado> I think I'm also more or less fine
09:43:20 <dmellado> will be giving a speak early Thu
09:43:25 <gmann> dmellado: oh
09:43:25 <dmellado> but should be free for the remaining
09:43:27 <gmann> what time?
09:43:32 <dmellado> gmann: let me check
09:43:52 <gmann> dmellado: Thursday we start at 11.50
09:44:01 * dmellado checking
09:44:29 <gmann> masayukig: hope our is not overlapping with general sessions?
09:44:37 <gmann> honestly i missed to check that :P
09:44:57 <masayukig> gmann: not overlapping
09:45:03 <gmann> ohk.
09:45:07 <masayukig> (fortunately :)
09:45:08 <dmellado> gmann: yep
09:45:12 <dmellado> it collides
09:45:18 <dmellado> 11:50-12:30
09:45:21 <gmann> dmellado: oh
09:45:31 <dmellado> https://www.openstack.org/summit/austin-2016/summit-schedule/events/8626/midonet-scalability-testing-with-neutron-and-open-source-tools
09:45:51 <gmann> dmellado: same time mine was overlapping with Nova
09:46:02 <gmann> oomichi: can you check this if possible :)
09:46:18 <gmann> ok let's move on and see what oomichi can do
09:46:25 <dmellado> cool
09:46:33 <gmann> #topic Critical Reviews
09:46:46 <gmann> any reviews to look on priority
09:47:56 <gmann> if nothing let's jump to open
09:47:59 <gmann> #topic Open Discussion
09:48:10 <gmann> anything else to discuss or we can close early
09:48:44 <dmellado> well, maybe I mised the mail but oomichi was speaking about having a dinner on Mon
09:49:00 <dmellado> he said he was going to send one (maybe not yet)
09:49:00 <gmann> dmellado: oh, did he send mail?
09:49:06 <gmann> ohk, may be
09:49:25 <gmann> may be we will have. :)
09:49:42 <dmellado> so let's check that later ;)
09:50:00 <dmellado> gmann: I'll ping you also on Mon to discuss the session' details
09:50:08 <gmann> dmellado: ok
09:51:05 <gmann> ok let's close the meeting.
09:51:06 <gmann> 3...
09:51:10 <gmann> 2..
09:51:15 <gmann> 1.
09:51:16 <gmann> Thanks all
09:51:28 <gmann> #endmeeting