17:02:38 <mtreinish> #startmeeting qa
17:02:39 <openstack> Meeting started Thu Apr 23 17:02:38 2015 UTC and is due to finish in 60 minutes.  The chair is mtreinish. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:02:41 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:02:43 <openstack> The meeting name has been set to 'qa'
17:02:49 <mtreinish> Hi, who's here today?
17:02:51 <dstanek> hi
17:02:57 <kir> hi
17:02:58 <jlanoux> hi
17:03:02 <Yogi> hello
17:03:04 <nfedotov> hi
17:03:20 <mtreinish> #link https://wiki.openstack.org/wiki/Meetings/QATeamMeeting#Proposed_Agenda_for_April_23rd_2015_.281700_UTC.29
17:03:23 <dtroyer> o/
17:03:25 <mtreinish> ^^^ Today's agenda
17:04:00 <mtreinish> dkranz, afazekas, jogo, andreaf, mkoderer, sdague: around?
17:04:29 <mtreinish> ok, let's get started
17:04:35 <sdague> o/
17:04:45 <mtreinish> #topic Summit Session Brainstorm
17:04:49 <jogo> o/
17:04:53 <mtreinish> #link https://etherpad.openstack.org/p/liberty-qa-summit-topics
17:05:16 <mtreinish> so I just wanted to continue reminding people about the brainstorming etherpad
17:05:46 <mtreinish> I think we said next thurs was the deadline for proposing a summit session
17:06:09 <mtreinish> so if you've got a session idea feel free to put it on the wiki page
17:06:21 <mtreinish> we've got a few slots this time
17:06:31 <mtreinish> with 5 fishbowl sessions, 4 work room sessions and a full day meetup shared w/ infra and rel-mgt on Friday
17:07:09 <mtreinish> that was all I really had for this topic, does anyone else have something else?
17:08:01 <dkranz> echo
17:08:23 <mtreinish> #topic Tempest-4 tag (mtreinish)
17:08:56 <mtreinish> so with the kilo release quickly approaching next week I quickly wanted to discuss the next tempest tag to mark it
17:09:07 <mtreinish> we have a bit of flexability in when we push the tag
17:09:30 <mtreinish> but I just wanted to keep track of any things we should be targetting for inclusion in it
17:10:02 <mtreinish> they should mostly be things that don't make deep functional changes
17:10:20 <mtreinish> I put a few on the agenda
17:10:35 <mtreinish> although the last one I think we're fine holding off on until after the tag
17:11:19 <mtreinish> does anyone else have anything that would be something else to look out for?
17:11:33 <mtreinish> I guess I really should make an etherpad to track these
17:11:47 <sdague> yeh, an etherpad for such issues would be good.
17:12:50 <mtreinish> #link https://etherpad.openstack.org/p/tempest-tag-4
17:12:57 <sdague> I was just looking back at the qa topics, in unwinding some of these requirements issues, and timing issues in devstack / grenade building tempest venvs, I'd really like to figure out how to make installable tempest with a cli a priority in liberty
17:13:27 <sdague> especially as there become more and more jobs using the REGEX in project-config as well
17:13:53 <mtreinish> sdague: do you want to put a session for that on the summit session etherpad?
17:14:05 <sdague> well, I don't want to lead it :)
17:14:09 <mtreinish> I think that would be a good work room topic to just get a poc out there
17:14:20 <mtreinish> just leave leader blank and say work session
17:14:45 <dkranz> sdague: I think it is in bounds to propose a session but not volunteer to lead it
17:14:59 <mtreinish> dpaterson did push his wip code up for the cli using cliff
17:15:07 <sdague> yeh
17:15:15 <mtreinish> https://review.openstack.org/#/c/175168/
17:15:39 <mtreinish> it definitely needs to be split up before it can land
17:16:56 <mtreinish> ok is there anything else on the next tempest tag, or design summit sessions?
17:18:07 <sdague> ok, added to the etherpad here
17:18:12 <mtreinish> sdague: ok cool, thanks
17:18:19 <mtreinish> #topic Specs Reviews
17:18:27 <mtreinish> so does anyone have any open specs to discuss
17:18:36 <mtreinish> (probably not given the point in the cycle)
17:19:10 <rockyg> o/
17:19:36 <rockyg> Dross project session Sean proposed for API/functional tests?
17:20:11 <mtreinish> rockyg: that's a open spec review?
17:20:26 <rockyg> I'd like to make sure we get Refstack (CatherineD) involved and try to get a direction for following projects from what Nova and others have learned so far
17:20:26 <mtreinish> or were you talking about the summit session etherpad?
17:20:33 <rockyg> Ooops.
17:20:44 <rockyg> No.  Open discussion. not open spec
17:20:47 <rockyg> Nevermind.\
17:20:55 <mtreinish> heh, no worries :)
17:21:11 <mtreinish> ok let's move on to the next topic
17:21:14 <mtreinish> #topic Blueprints
17:21:28 <mtreinish> are there any in progress bps to discuss?
17:22:07 <mtreinish> jlanoux: is there any update on the ssh auth one?
17:22:50 <jlanoux> mtreinish: we still need some eyes to look at it. Only andreaf and afazekas started to review
17:23:04 <mtreinish> jlanoux: ok, I'll try to take a look soon
17:23:05 <jlanoux> first patch was merged on Monday
17:23:29 <jlanoux> mtreinish: thanks!
17:23:33 <mtreinish> jlanoux: one thing I did tell andreaf about was having a patch up to add a section to the config guide about all the new options and how to use them would be good
17:23:42 <mtreinish> obviously at the end of the series
17:23:53 <jlanoux> okidoki
17:24:15 <mtreinish> ok are there any other bps?
17:25:18 <mtreinish> ok then let's move on
17:25:22 <mtreinish> #topic Devstack
17:25:29 <dtroyer> heyo
17:25:34 <mtreinish> dtroyer: anything exciting this week from the world of devstack?
17:25:48 <dtroyer> the big news this week is the stable/kilo branch was cut
17:26:23 <mtreinish> cool
17:26:42 <dtroyer> that always leads to a bit of cleaning up, this time most of it was elsewhere
17:27:22 <mtreinish> yeah I guess that I need to go through and generate a hardcoded extension list for kilo now
17:27:25 <sdague> yeh, that triggered discovery of some issues with oslo code
17:27:42 <mtreinish> sdague: the service shutdown thing? or something else
17:27:44 <sdague> stop in kilo is less reliable with a lot of projects
17:28:07 <dtroyer> in screen or out?
17:28:29 <sdague> dtroyer: out
17:28:41 <sdague> https://review.openstack.org/#/c/176910/ is a tactical work around
17:28:54 <mtreinish> #link https://review.openstack.org/#/c/176910/
17:29:05 <sdague> https://bugs.launchpad.net/oslo-incubator/+bug/1446583 is the bug
17:29:05 <openstack> Launchpad bug 1446583 in oslo-incubator "services no longer reliably stop in stable/kilo" [Critical,Fix committed] - Assigned to Julien Danjou (jdanjou)
17:29:11 <dtroyer> I wonder what changed, I thought the process group thing was working pretty well
17:29:29 <mtreinish> sdague: heh, just kill it twice
17:29:29 <sdague> dtroyer: it's not devstack
17:29:59 <dtroyer> ok, the pgroup approach is sensitive to the way a process daemonizes though
17:30:00 <sdague> it's services that are reading from a socket apparently are dropping the term signal on the ground
17:30:11 <dtroyer> oh, that's different
17:30:40 <mtreinish> sdague: we just never caught it because we never have to stop anything for grenade on master?
17:30:44 <sdague> right, it's actually bugs in the code, but we only discovered it after having a stable/kilo branch to test against
17:31:41 <sdague> it's worth thinking about how to catch this stuff earlier
17:31:48 <dtroyer> so if we did a grenade master -> master (actually just start/stop/start) we'd essentially have the long wanted restack.sh
17:31:59 <sdague> yes... sort of?
17:32:11 <dtroyer> but that might have found this problem
17:32:14 <mtreinish> sdague: we also have the unstack unit test we could add a ps sanity check too
17:32:17 <sdague> yeh, I think it would have
17:32:44 <sdague> mtreinish: yep, the other option would be putting unstack.sh into the normal devstack-gate flow
17:33:01 <sdague> anyway, a thing to think about
17:33:20 <sdague> not this week, because my brain is burnt out from the requirements unwind
17:33:52 <mtreinish> ok, then is there anything else to discuss on devstack?
17:34:00 <dtroyer> a couple of other interesting things that are in progress or soon will be is the dual-stack IP v4/v6 work from other Sean
17:34:14 <dtroyer> and there was a CirrOS release so we can expect attempts to upgrade that too
17:34:16 <kir> are there plan to support devstack on ipv6 only?
17:34:57 <dtroyer> I saw that in my inbox this morning, so someone is thinking about it.  we don't have it on the short list now
17:35:11 <sdague> dtroyer: I also added to the summit etherpad that we should discuss things to move out of devstack
17:35:18 <mtreinish> dtroyer: yeah I know the cirros pre-release
17:35:26 <dtroyer> I have the neutron re-write first which might be helpful
17:35:29 <mtreinish> should include some ipv6 stuff too
17:35:38 <sdague> dtroyer: yes, that would be helpful
17:35:46 <dtroyer> Cirros was just a pre-release?  I read that email too fast then
17:35:54 <mtreinish> kir: I think sc68cal was working on that
17:36:25 <dtroyer> I wasn't sure if he had v6-only included
17:36:27 <mtreinish> dtroyer: yeah I thought so, let me dig it up again
17:36:52 <dtroyer> otherwise, that's all I had to mention
17:37:00 <mtreinish> #link http://lists.openstack.org/pipermail/openstack-dev/2015-April/062403.html
17:37:22 <mtreinish> dtroyer: I just assumed sc68cal was working on a v6 only setup too
17:37:30 <dtroyer> oh, quickly, morgan is working on a keystone v3-only setup too
17:38:04 <mtreinish> cool, that'll be interesting to see
17:38:29 <mtreinish> ok then does anyone have anything else on devstack?
17:39:03 <sdague> not I
17:39:10 <mtreinish> ok then let's move on
17:39:15 <mtreinish> #topic Grenade
17:39:27 <mtreinish> sdague: I'm sure you've got something to say this week about grenade :)
17:39:33 <sdague> except, I guess the fact that OSC takes 4.5 minutes of aggregate time :)
17:39:50 <sdague> yeh, the grenade refactor is done, landed, and branched
17:39:58 <mtreinish> yay
17:40:14 <sdague> there is probably still some useful docs cleanup, and an project template
17:40:28 <mtreinish> yeah, that's what I was just going to ask
17:40:35 <sdague> and then at some point, probably on the plane to vancouver, figuring out enable_plugin for grenace
17:40:37 <sdague> grenade
17:40:53 <mtreinish> do you think adding some grenade bits to devstack cookiecutter repo will be useful too?
17:41:06 <sdague> yeh
17:41:27 <sdague> we should probably find a good project that wants to work through this with us together
17:41:40 <sdague> maybe sahara?
17:41:46 <mtreinish> sdague: I think adam_g and ironic were looking at it before too
17:41:47 <sdague> as they don't currently have upgrade testing
17:41:52 <sdague> oh, right, ironic
17:42:03 <mtreinish> although sahara is probably a good choice too
17:43:01 <sdague> oh, that's why we have no test nodes, mtreinish pushed a giant tempest series that fails tests :P
17:43:15 <mtreinish> sdague: heh, I figured that would happen
17:43:23 <mtreinish> a 9 patch tempest series has that effect
17:43:29 <mtreinish> sdague, dtroyer: oh, unrelated but before I forget, if I can trouble you to update https://wiki.openstack.org/wiki/QA/releases with everything you guys had to do for branching devstack and grenade
17:43:39 <mtreinish> so we have the steps written down
17:43:43 <sdague> it was ... ask clark
17:44:01 <mtreinish> sdague: there weren't any pre or post patches?
17:44:01 <sdague> clarkb please tag stable/kilo at HASH
17:44:19 <sdague> just the normal .gitreview update
17:44:35 <sdague> I guess I changed 2 names in grenaderc as well
17:44:42 <mtreinish> sdague: sure, I just thought it would be good to just have the steps written down somewhere even if it's ask clarkb and update a couple files
17:45:28 <mtreinish> ok is there anything else on grenade?
17:46:26 <sdague> well, we're breaking on neutron right now
17:46:41 <sdague> https://review.openstack.org/#/c/176723/ fixes it
17:46:47 <sdague> that was being blocked by the nova not stopping issue
17:47:12 <mtreinish> sdague: ok that's simple enough +2
17:47:23 <sdague> which we have as a work around in stable/kilo devstack
17:47:42 <sdague> waiting for that to get tests back to figure out if it's good enough to fix things
17:48:10 <mtreinish> ok, sure
17:48:32 <mtreinish> let's move on we're at ~10min left
17:48:37 <mtreinish> #topic Critical Reviews
17:48:56 <mtreinish> does anyone have an open reviews that they'd like to get some extra eyes on?
17:49:10 <kir> https://review.openstack.org/#/c/157816/
17:49:24 <kir> feature is already in neutron
17:49:30 <dkranz> mtreinish: this is waiting for non-red-hat +A: https://review.openstack.org/#/c/161370/
17:50:30 <mtreinish> dkranz: oh, I remember that one, tosky asked me about a while ago
17:50:37 <mtreinish> I'll take a look later on today
17:50:49 <mtreinish> kir: if it's a feature add, does it also work on the stable branches?
17:50:53 <dkranz> mtreinish: thanks
17:51:19 <kir> don't know, it works with neutron master
17:51:58 <mtreinish> #link https://review.openstack.org/#/c/172507/
17:52:10 <mtreinish> ^^ just a doc patch but it's at the bottom of a bunch of doc cleanups
17:52:34 <mtreinish> kir: ok, I'll take a look in a bit
17:52:42 <mtreinish> it at least looks like it's passing on the stable branches too
17:53:13 <mtreinish> ok are there any other reviews?
17:54:23 <mtreinish> ok I guess not
17:54:36 <mtreinish> #topic Tempest with python 3 (mtreinish)
17:54:48 <mtreinish> probably the last topic for this week
17:55:16 <mtreinish> this morning I started doing some cleanups needed to work towards getting the unit tests running under py34
17:55:23 <mtreinish> #link https://review.openstack.org/#/q/status:open+project:openstack/tempest+branch:master+topic:python3-support,n,z
17:55:46 <mtreinish> I figured it would probably be good if we started working towards gating on 34 and 27 (like we do for tempest-lib)
17:56:18 <mtreinish> that series gets us to being able to run testr list-tests and run the majority of the unit tests
17:56:52 <mtreinish> well, except for a tempest-lib release which is needed to remove py2 code from the auth module
17:57:29 <mtreinish> that was all I had on this, I just thought I should bring up that I was starting work on it
17:57:32 <mtreinish> in case anyone cared
17:57:47 <mtreinish> does anyone have anything else, on this topic or any other?
17:58:39 <rockyg> just hoping sean will extend the api/func test session at the summit to include lessons learned, how others can do the same, to add consistency across projects
17:58:57 <rockyg> Horizontal consistency across projects for in projects testing
17:59:08 <rockyg> sorry.  sdague
17:59:17 <mtreinish> rockyg: I assumed that was part of it. The session is named show and tell iirc
17:59:38 <rockyg> Kewl.  Getting the refstack folks involved, too.
17:59:47 <rockyg> They'll be there
18:00:02 <rockyg> And, thanks!
18:00:12 <mtreinish> ok, then I guess we'll end here today
18:00:15 <mtreinish> thanks everyone
18:00:17 <mtreinish> #endmeeting