09:00:02 <gmann> #startmeeting qa
09:00:03 <openstack> Meeting started Thu Jun 16 09:00:02 2016 UTC and is due to finish in 60 minutes.  The chair is gmann. Information about MeetBot at http://wiki.debian.org/MeetBot.
09:00:04 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
09:00:06 <openstack> The meeting name has been set to 'qa'
09:00:15 <gmann> hi who's here for today?
09:01:08 <andreaf> o/
09:01:21 <gmann> andreaf: hi
09:01:31 <andreaf> gmann: hi
09:01:43 <gmann> let's wait some time to join more people
09:05:11 <gmann> andreaf: looks like only we two are here today :)
09:05:21 <andreaf> masayukig, mkoderer around?
09:05:39 <andreaf> yes it looks like so
09:05:50 <gmann> masayukig: is out today for health checkup he mentioned
09:06:09 <andreaf> well if no-one else joins I'm not sure we should be running the meeting
09:06:32 <gmann> yea, just wait for more 2-3 min then we can cancel today one
09:06:34 <andreaf> jordanP and afazekas don't seem to be online either
09:06:44 <gmann> yea
09:07:40 <dmellado> sorry I'm late
09:07:48 <dmellado> Hi andreaf gmann
09:07:53 <gmann> dmellado: hi
09:08:01 <jlanoux_> o/
09:08:25 <gmann> jlanoux_: hi
09:08:51 <gmann> i think we have enough people now, we can start? and can finish early
09:09:18 <dmellado> +1
09:09:24 <gmann> let's start
09:09:25 <gmann> #link https://wiki.openstack.org/wiki/Meetings/QATeamMeeting#Agenda_for_April_21st_2016_.280900_UTC.29
09:09:42 <gmann> ^^ today agenda
09:10:05 <dmellado> gmann: april 21?
09:10:12 <gmann> ops
09:10:21 <gmann> #link https://wiki.openstack.org/wiki/Meetings/QATeamMeeting#Agenda_for_June_16th_2016_.280900_UTC.29
09:10:43 <gmann> #topic Specs Reviews
09:10:53 <gmann> #link https://review.openstack.org/#/q/status:open+project:openstack/qa-specs,n,z
09:11:27 <gmann> looks like all open spec are with review
09:11:39 <gmann> anyone want to bring any specific spec
09:12:53 <gmann> if no then let's move
09:12:57 <gmann> #topic Tempest
09:13:10 <gmann> we will discuss newton priority item here as listed in agenda
09:13:27 <gmann> andreaf: first one is yours - Remote client debuggability (andreaf)
09:13:57 <andreaf> gmann: well we always have the full list of priorities in the agenda, not sure we should keep it
09:14:22 <dmellado> I think we could drop it, and bring only the updates maybe
09:14:24 <mkoderer> hi
09:14:26 <gmann> andreaf: humm, or we track the current ongoing items from them
09:14:34 <dmellado> hi mkoderer
09:14:35 <gmann> mkoderer: hi
09:14:39 <gmann> dmellado: yea
09:14:54 <gmann> so just discuss the items from people here
09:15:22 <gmann> andreaf: you have anything to update from priority items?
09:15:46 <gmann> andreaf: one is client refactor i think
09:15:48 <gmann> #link https://review.openstack.org/#/q/status:open+project:openstack/tempest+branch:master+topic:bp/client-manager-refactor
09:15:48 <andreaf> gmann: I don't have updates on remote clients debuggability, it's done for API tests - but the approach is different between API and scenario tests - I'm waiting on the scenario cleanup to happen before I can do that
09:16:11 <gmann> andreaf: ok
09:16:14 <dmellado> Then that'll be my turn then I guess :D
09:16:35 <gmann> dmellado: yea we will come on your items
09:16:51 <dmellado> from my side, I've been unable to allocate *any* time upstream since I came back from Austin, so I'm on best-effort on the scenario cleanup
09:17:02 <dmellado> could we get some more volunteers in order to tackle that?
09:17:26 <dmellado> (I've been promised more time, though, but don't want to overpromise...)
09:17:27 <gmann> dmellado: which one? scenario cleanup?
09:17:35 <dmellado> gmann: scenario cleanup, yep
09:17:53 <dmellado> test class hierarchy, basically
09:18:04 <gmann> dmellado: we have spec approved for that right or any etherpad about what all needed
09:18:08 <dmellado> I don't want to stop anyone, althought I'll try my best it might not depend on me
09:18:35 <dmellado> gmann: there was an document on the methods that would need check-up
09:18:37 <andreaf> dmellado: perhaps you could go to the ML asking for volunteers?
09:18:44 <dmellado> andreaf: will do
09:18:50 <dmellado> andreaf: wouldn't you have the link around?
09:18:52 <dmellado> can't find it now
09:18:57 <gmann> yea that will be nice to get more people
09:19:24 <gmann> #link https://etherpad.openstack.org/p/newton-tempest-squash-class-hierarchy
09:19:25 <dmellado> gmann: I'll send an email to the ML with the links and formatting
09:19:35 <gmann> dmellado: Thanks
09:19:41 <andreaf> #link https://docs.google.com/spreadsheets/d/1Y241ELbiM0djdFeII8mAiL8ZEQZwpBQlsV_2xbJUMMQ/edit?usp=sharing
09:20:03 <dmellado> btw, and offtopic (sorry) is there any final time/date for the mid cycle
09:20:10 <dmellado> I'd make it there and will try to make up the most of it
09:20:45 <gmann> dmellado: mkoderer updated wiki for that
09:20:47 <gmann> #link https://wiki.openstack.org/wiki/Sprints/QAInfraNewtonSprint
09:20:55 <gmann> mkoderer: that is final dates and place right?
09:21:08 <andreaf> https://wiki.openstack.org/wiki/Sprints/QAInfraNewtonSprint
09:21:11 <mkoderer> gmann: yes
09:21:12 <dmellado> oh, nice
09:21:18 <dmellado> I was waiting for an email on the ML
09:21:23 <dmellado> but will sign myself there
09:21:25 <dmellado> thanks gmann mkoderer
09:21:31 <gmann> mkoderer: cool
09:21:35 <mkoderer> so I have booked two meeting rooms
09:21:45 <mkoderer> and tried to put all needed information in the wiki
09:21:54 <mkoderer> just let me know if somethine is missing
09:22:00 <gmann> mkoderer: Thanks may be oomichi will send on ML
09:22:31 <gmann> mkoderer: sure, Thanks for that, happy to go Germany first time
09:22:46 <mkoderer> gmann: will be fun
09:23:02 <gmann> yea
09:23:25 <gmann> ok, let's move on
09:23:31 <gmann> not sure about Neutron OO wrappers
09:23:44 <andreaf> that merged yesterday https://review.openstack.org/#/c/320495/
09:23:48 <gmann> jlwhite : hockeynut started anything on this
09:24:19 <gmann> andreaf: oh nice, i did not notice those :)
09:25:07 <gmann> next tempest-lib/plugin
09:25:09 <andreaf> regarding documentation I did not have time to do a complete check yet, it will be an ongoing effort through the cycle for me
09:25:38 <gmann> andreaf: ok, any specific thing we need to do on priority on doc
09:25:59 <andreaf> gmann: no, the most urgent are done
09:26:16 <andreaf> gmann: on tempest-lib I've been working on migrating the client manager to lib
09:26:28 <gmann> andreaf: ok, Thanks.  and yea we can keep updating those on finding
09:26:46 <gmann> andreaf: yea, i did not get time today to look into those. i will try tomorrow
09:26:55 <andreaf> #link https://review.openstack.org/#/q/status:open+project:openstack/tempest+branch:master+topic:bp/client-manager-refactor+is:mergeable
09:27:45 <andreaf> on tempest CLI, the workspace manager landed as well as initial support for un
09:27:48 <andreaf> run
09:27:56 <gmann> yea, that is nice
09:28:01 <andreaf> mtreinish has a couple of patches up for run extensions
09:28:25 <andreaf> #link https://review.openstack.org/#/q/status:open+project:openstack/tempest+branch:master+topic:bp/tempest-run-cmd
09:29:04 <gmann> yea, after that i think we can adopt new workflow in gate and have our cmd tested on gate
09:30:31 <gmann> on service client migration, oomichi   is working on image one and me on identity one
09:30:34 <gmann> #link https://review.openstack.org/#/q/status:open+project:openstack/tempest+branch:master+topic:bp/consistent-service-method-names
09:31:03 <dmellado> gmann: long work on that ;)
09:31:07 <gmann> one issue(may not be) i found on identity is most of the list APIs does not support query of limit,marker
09:31:15 <gmann> except list_tenants
09:31:28 <gmann> i ran on my local env but doc state about that
09:32:01 <gmann> so i will not add query param to those list method in service clients and raise bug on doc
09:32:57 <gmann> and if they support that then we can make change to service client which i think would not cause backward incompatibity (which adds kwargs)
09:33:39 <gmann> which i think should be ok for all?
09:34:02 <dmellado> +1
09:34:07 <andreaf> gmann: I'm not sure having kwargs in those already would be an issue
09:34:31 <gmann> andreaf: yea there is no issue, and that is another option
09:34:54 <gmann> if we send limit query keystone just ignore as of now
09:35:17 <gmann> so should we go with that, m ok for both way
09:35:31 <andreaf> having kwargs allows test to try and send unexpected content to the API :)
09:36:24 <gmann> andreaf: yea, that is one benefit. ok lets go with kwargs which doc also states that
09:37:31 <gmann> next item - microversions testing (compute)
09:38:10 <gmann> this one i need to start and cover as many as we can in Newton as nova has around 30 version up till now
09:38:33 <gmann> and issue here is schema versioning we need to do if higher microversion tests are being added
09:39:07 <gmann> if we have all lower version tests we can avoid such situation
09:40:39 <gmann> ok, thats all on those.
09:40:41 <gmann> anything on Tempest or we move to next topic
09:41:42 <andreaf> I'm trying to get pre-prov creds used in the gate
09:41:42 <gmann> #topic DevStack + Grenade
09:41:55 <gmann> andreaf: yah yea, saw mail. +1 on that
09:42:11 <gmann> andreaf: you will try on neutron job first or layer4 ?
09:42:22 <andreaf> gmann: after discussing with mtreinish we decided to start with layer4 because that runs on tempest only
09:42:40 <andreaf> once we further prove stability we can move to other jobs
09:42:47 <gmann> andreaf: yea nice point
09:43:15 <andreaf> so I tested pre-prov creds with layer4 tests and I found one role missing for heat tests
09:43:20 <andreaf> fix is in #link https://review.openstack.org/#/c/330252/
09:43:32 <andreaf> it's a very simple change, reviews are welcome :)
09:43:49 <gmann> nice, ll check right after meeting
09:43:52 <gmann> andreaf: and at end we will be having around half using dynamic and half using pre-provisioned one right?
09:45:08 <andreaf> gmann: that's my hope. pre-prov creds can help prevent regressions on resource cleanup, but we need to start running the cleanup script in dry mode in the gate to make that easier to troubleshoot in case it happens
09:45:58 <gmann> andreaf: ok, dry mode will help in that
09:46:48 <gmann> andreaf: one issue i faced on pre-provisioned account in Austin summit when we tried to run tempest with init workspace
09:47:31 <gmann> i think i talked to you may be, there was some issue on releasing the accounts
09:47:44 <gmann> but i forgot totally, i will try to reproduce that if i can
09:48:06 <andreaf> gmann: ok let me know
09:48:11 <gmann> sure
09:49:01 <andreaf> time's flying, we should probably move on
09:49:05 <gmann> yea
09:49:26 <gmann> we will skip devstack one if nobody has anything on those
09:49:38 <dmellado> yep
09:49:51 <gmann> openstack-health too? as masayukig mtreinish are not here
09:49:56 <gmann> or andreaf you have anything on this
09:50:16 <andreaf> nothing from my side
09:50:19 <gmann> ok
09:50:24 <gmann> #topic Critical Reviews
09:50:48 <gmann> most of priority review we talked, apart from them any other review anyone want to bring up
09:52:00 <gmann> #topic QA mid-cycle meetup
09:52:09 <gmann> i think we already talked on this
09:52:11 <gmann> #link https://wiki.openstack.org/wiki/Sprints/QAInfraNewtonSprin
09:52:21 <gmann> thanks mkoderer again
09:52:39 <gmann> we will jump to open
09:52:44 <gmann> #topic Open Discussion
09:53:03 <andreaf> ok just one note on the client manager from my side
09:53:14 <gmann> andreaf: sure
09:53:36 <andreaf> since not all clients are in tempest.lib yet, the initial client manager for lib will only contain those migrated already
09:54:09 <andreaf> because I don't want to have something in lib that depends on clients in tempest - changing that later would break backward compatibility
09:54:16 <gmann> andreaf: yea that make sense.
09:54:49 <gmann> and gradually we can keep adding migrated one
09:55:06 <andreaf> also I started preparing the stable client manager in lib_manager.py on tempest side because I want to make sure the interface looks good before it's moved to tempest.lib
09:55:57 <andreaf> for now it only includes clients in tempest, I will add an interface to register extra clients from plugin later in the patch series
09:55:59 <dmellado> sounds good, also if that affects another tempest plugins, they can migrate that later too
09:56:54 <gmann> andreaf: and do we want any plugin to adopt that before migrating this to lib ?
09:57:01 <andreaf> dmellado: yes I'm maintaining manager.Manager same as before for plugins that already use that for now - so they are not broken by my changes
09:57:05 <gmann> to avoid any interface change in lib
09:57:33 <andreaf> once the stable manager is in lib plugins can start to migrate to it and we can then drop manager.Manager
09:57:50 <gmann> can we ask any plugin to adopt new one and test new interface throughly  while in tempest
09:58:07 <andreaf> basically I'm having to maintain backward compatibility on non-stable classes for a bit to be nice to plugins - since we didn't have a stable client manager ready in time
09:58:08 <gmann> andreaf: i was thinking before migraion to lib
09:58:30 <andreaf> gmann: well but then they will have to change twice
09:58:55 <gmann> andreaf: yea thats true but that will make sure all our interface works well
09:59:01 <andreaf> gmann: and until it's on lib side it can change so they may be broken by changes
09:59:08 <gmann> may be we can volunteer on those to update twice
09:59:37 <andreaf> gmann: it would be good to have a set of changes making the switch in the plugins for testing purposes
10:00:13 <gmann> yea, once we are ready to migrate then first test with 1 plugin then migrate
10:00:34 <gmann> i can talk to mistral or congress one, my team mates work in those
10:00:44 <dmellado> gmann: yep, also I was speaking with the mistral guys
10:00:47 <dmellado> to modify the plugin
10:00:54 <dmellado> as it didn't have the skip_checks there
10:00:57 <andreaf> gmann: thanks
10:01:01 <dmellado> and is quite small
10:01:06 <andreaf> well, I guess we are at time
10:01:08 <dmellado> so it'd be great to have it for testing
10:01:10 <gmann> dmellado: sure let me know which one i can ask
10:01:16 <gmann> ok time up, sorry guys
10:01:16 <andreaf> we can continue in the qa room
10:01:22 <gmann> let's switch to qa
10:01:28 <dmellado> sure
10:01:31 <gmann> yea
10:01:33 <gmann> thanks all
10:01:34 <gmann> #endmeeting