17:00:43 <mtreinish> #startmeeting qa
17:00:44 <openstack> Meeting started Thu Dec 18 17:00:43 2014 UTC and is due to finish in 60 minutes.  The chair is mtreinish. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:45 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:00:47 <openstack> The meeting name has been set to 'qa'
17:01:01 <mtreinish> hi, who's here today?
17:01:10 <jlanoux> hi
17:01:16 <sdague> o/
17:01:20 <mtreinish> #link https://wiki.openstack.org/wiki/Meetings/QATeamMeeting#Proposed_Agenda_for_December_18th_2014_.281700_UTC.29
17:01:23 <mtreinish> ^^^ Today's agenda
17:02:20 <mtreinish> andreaf, dkranz, mkoderer, dtroyer: around?
17:02:49 <dtroyer> o/
17:03:11 <mtreinish> ok, let's get started
17:03:21 <mtreinish> #topic No meeting next 2 weeks (mtreinish)
17:03:45 <dkranz> o/
17:03:51 <mtreinish> I just though I should mention because of the holidays and most people taking off that we will not be having the meeting for the next 2 weeks
17:04:03 <andreaf> o/
17:04:03 <mtreinish> it will pick up again on Jan. 9th
17:04:17 <mtreinish> err no 8th
17:04:26 <mtreinish> (the 9th is a friday)
17:04:39 <sdague> wfm
17:05:10 <dkranz> sure
17:05:20 <mtreinish> the other thing is I won't be able to lead the meetings in Jan. because I'll be travelling for LCA and some other stuff
17:05:37 <dkranz> mtreinish: LCA ?
17:05:47 <mtreinish> dkranz: http://linux.conf.au/
17:05:59 <mtreinish> so if someone could volunteer to run the meetings while I'm on a plane that would awesome
17:06:00 <dkranz> mtreinish: ah, warm
17:06:08 <mtreinish> dkranz: heh, yep :)
17:06:38 <dkranz> mtreinish: I can do Jan 8/15
17:07:00 <mtreinish> dkranz: ok cool
17:07:17 <mtreinish> #info dkranz to run meeting on the 8th and 15th
17:07:21 <dkranz> mtreinish: I'll be some place warm the two after that
17:07:46 <ekhugen_alt> I can host Jan 22/29, if needed
17:08:01 <dkranz> Is any one else going to be working Dec 29/30/31 ?
17:08:06 <mtreinish> dkranz: ok, I'll figure something out for the other 2 days
17:08:18 <sdague> dkranz: not I, today is my last day for the year
17:08:49 <andreaf> dkranz: probably not
17:08:53 <mtreinish> dkranz: yeah neither am I
17:08:56 <andreaf> Jun 8th will be 5pm or 10pm UTC?
17:08:56 <dkranz> Just wondering if it is worth pushing anything
17:09:19 <andreaf> s/Jun/Jan
17:09:24 <dkranz> andreaf: It is three from now so that would be 10pm
17:09:28 <mtreinish> ekhugen_alt: cool thanks, I was hoping for a member of a qa core team to do it. But, if no one steps up that'll work. I'll let you know
17:09:47 <ekhugen_alt> okay, thanks mtreinish
17:10:02 <mtreinish> andreaf: yeah it'll be 2200 UTC
17:10:43 <mtreinish> it's easier to keep the cadence on off weeks so I don't have to try an update the ical
17:10:59 <mtreinish> ok does anyone else have anything on this topic?
17:11:51 <mtreinish> #topic Specs Reviews
17:12:10 <mtreinish> does anyone have an open spec review to bring up?
17:12:43 <andreaf> mtreinish: I'm working on your comments on the ssh-auth one
17:12:51 <mtreinish> andreaf: cool, thanks
17:13:01 <sdague> people should take a look at - https://review.openstack.org/#/c/137054/ - I'm actually working on a POC on it
17:13:17 <mtreinish> #link https://review.openstack.org/#/c/137054/
17:13:18 <andreaf> mtreinish: I hope to have it done tomorrow, and then move faster, I'm proposing k2 as milestone
17:13:31 <sdague> (that's the devstack remote plugins one)
17:14:07 <andreaf> sdague: sure I'll take a look at it
17:14:09 <mtreinish> oh I guess today is k-1...
17:14:56 <mtreinish> sdague: sure, I'll put it on my list too
17:15:06 <mtreinish> it seems like a good idea to have that
17:15:32 <andreaf> mtreinish: also I'd like to close on https://review.openstack.org/#/c/139578/ if possible so I can start working on it
17:15:48 <sdague> also - https://review.openstack.org/#/c/140843/ - should be approved unless there are major gripes. That's basically just a write up of what was talked about at summit by dean
17:16:03 <sdague> dtroyer have you started hacking on it yet?
17:16:30 <mtreinish> andreaf: sure, I gave it a +2 earlier today, if gets another +2 I'll push it through
17:16:39 <mtreinish> andreaf: I can help out with pushing bits on that one too
17:16:56 <mtreinish> sdague: sure, I'll take a look after the meeting and approve it if it looks good
17:16:56 <andreaf> mtreinish: great, thanks
17:17:06 <sdague> I guess a question would be what code effort are stalled on specs now that we should unstall?
17:17:13 <dtroyer> sdague: yup...
17:17:23 <dtroyer> https://review.openstack.org/140565
17:17:46 <sdague> dtroyer ok cool
17:18:14 <mtreinish> sdague: yeah that's a good question :)
17:19:18 <mtreinish> honestly there aren't that many open spec reviews right now though
17:19:36 <mtreinish> or at least ones without negative reviews
17:19:47 <mtreinish> one on devstack that sticks out is from adam_g: https://review.openstack.org/136101
17:20:39 <mtreinish> ok if there aren't any other specs to discuss we can move on
17:20:47 <sdague> oh, yeh, I should look at that one
17:21:00 <dtroyer> ditto, I had missed it
17:21:52 <mtreinish> #topic Blueprints
17:22:04 <mtreinish> ok are there any open blueprints that we need to discuss
17:22:08 <mtreinish> or any status updates?
17:22:55 <andreaf> mtreinish: on the resource cleanup one
17:23:14 <mtreinish> andreaf: https://blueprints.launchpad.net/tempest/+spec/resource-cleanup ?
17:23:15 <andreaf> mtreinish: ekhugen_alt has started working on it with a few patches already (thanks!)
17:23:57 <andreaf> mtreinish: yes - there is an etherpad - if someone wants to contribute put your name in there next to a test set you want to migrate
17:24:01 <mtreinish> cool, that's doing the split of the setup and teardown into finer grained pieces right?
17:24:05 <mtreinish> andreaf: sure, link?
17:24:16 <andreaf> #link https://etherpad.openstack.org/p/resource-cleanup-tempest-2
17:24:37 <andreaf> mtreinish: yes that's correct
17:25:40 <mtreinish> cool, thanks
17:25:54 <mtreinish> are there any other bps that anyone would like to bring up
17:26:28 * andreaf has to leave the meeting early today
17:26:41 <sdague> nope, though I have some devstack bits when we get to that
17:27:03 <mtreinish> ok, well then let's move onto that. (since it seems there aren't any other bps to discuss today)
17:27:12 <mtreinish> #topic Devstack
17:27:36 <mtreinish> sdague, dtroyer: go ahead
17:27:42 <sdague> ok, so I've been chugging through the nounset patches in the background
17:27:44 <dtroyer> sdague mentioed the plugin work that he is starting, he can fill in more there…
17:27:52 <sdague> https://review.openstack.org/#/q/status:open+project:openstack-dev/devstack+branch:master+topic:unset,n,z
17:28:16 <sdague> this is going to be a long chug, but the bottom patch is a set of cleanups that is probably good to go in
17:28:22 <dtroyer> I started the logging spec and have the devstack jobs working,  but Grenade needs to learn how to swing both ways before that can proceed
17:28:47 <mtreinish> sdague: https://review.openstack.org/#/c/139287/ ?
17:28:59 <sdague> the approach involves https://review.openstack.org/#/c/139287/11/stack.sh,cm setting it not to be on by default
17:29:01 <sdague> mtreinish: yeh
17:29:20 <mtreinish> #link https://review.openstack.org/#/c/139287/
17:29:21 <sdague> that's enough to pass on my local config
17:29:39 <sdague> stack.sh, unstack.sh, and clean.sh all work with nounset locally with that
17:29:53 <dtroyer> nice
17:30:03 <sdague> but, there are a lot more configs in the gate
17:30:23 <sdague> so the top of the patch series sets things to true by default, and then I'm piling up fixes under it
17:31:00 <sdague> also, I introduced a unittest framework in 139287 - https://review.openstack.org/#/c/139287/11/tests/unittest.sh,cm
17:31:15 <sdague> because, all the test scripts were very one off
17:31:27 <sdague> only the new tests for trueorfalse use that
17:31:40 <sdague> but I think that would be a good cleanup to move to that
17:31:57 <sdague> the other thing, is the devstack plugins
17:32:07 <mtreinish> sdague: is there a reason you didn't use something like shunit?
17:32:20 <sdague> mtreinish: mostly, not pulling in another large dependency
17:32:54 <mtreinish> ok, just curious
17:32:56 <sdague> this is something we've got over in devstack-gate in a similar form
17:33:28 <sdague> I looked at shunit at some point, and ... this was just simpler
17:33:43 <sdague> ok, devstack plugins
17:34:16 <mtreinish> ok that's fine. The only reason I asked is I think I remember shunit having subunit support so we could get subunit streams from the unit tests. But that's not really important
17:34:17 <sdague> based on the glusterfs patch - https://review.openstack.org/#/c/133102/ - I felt like this needed to happen a little sooner
17:34:45 <sdague> because I'm not convinced we should merge the glusterfs patch, as it requires 3rd party epel repos
17:35:00 <dtroyer> ++
17:35:07 <sdague> so chmouel wrote the spec
17:35:14 <sdague> see above ^^^
17:35:30 <mtreinish> sdague: +1 yeah 3rd party repos are a no go
17:35:32 <sdague> here is a first pass - https://review.openstack.org/#/c/142805/
17:35:48 <sdague> honestly, given the library work, it turned out to only be 40 lines of code
17:36:07 <sdague> it's slightly different from the spec right now
17:36:26 <sdague> because this was more straight forward, but it's close enough (will update the spec if people like this direction)
17:37:14 <sdague> a "more real" version is needed with some error checking, but I'm testing that works in a follow on patch
17:37:54 <chmouel> sdague: ah i was just about to comment on the review about that
17:38:31 <sdague> one concern, for the gate, ERROR_ON_CLONE is on, so we can only use this in the gate for stuff that's in the project list I think
17:38:48 <sdague> anyway, that limitation will need to be sorted, I'm hacking around it on the test patch
17:39:08 <chmouel> or if we really needed it we could maybe have infra mirror the git repo?
17:39:11 <sdague> chmouel: yeh, honestly, I started writing parsing code and said ... you know what, lets just let the user specify
17:39:17 <sdague> chmouel: yep
17:39:18 <mtreinish> sdague: yeah I don't think it'll be an issue in the gate. Most people have been doing this manually with gate hooks
17:40:07 <chmouel> sdague: i think that look pretty good, i mean at least syntax wise
17:40:08 <mtreinish> so the repos are already there for the plugin
17:40:40 <sdague> also... by using the same GIT* arrays this actually lets you do interesting things like intentionally override pre defined repos. Which I think might be a useful side effect.
17:41:28 <sdague> that's it from me
17:41:38 <mtreinish> ok, thanks
17:41:39 <chmouel> except for forks i am not sure how this could be used but well :)
17:41:40 <dtroyer> sdague: thanks…
17:41:45 <mtreinish> does anyone have anything else on devstack?
17:42:07 <dtroyer> just one more thing here, I'm working on making DevStack run with a global venv
17:42:17 <sdague> dtroyer: ++
17:42:23 <mtreinish> dtroyer: awesome
17:42:48 <chmouel> dtroyer: coool :)
17:42:48 <dtroyer> first tests just completed and the results were…surprising.  one of the grenade jobs succeeded and I don't yet know why
17:42:49 <sdague> chmouel: sure, but sometimes forks are valid things to test against (especially as you can set branch)
17:42:55 <chmouel> sdague: yeah true
17:43:20 <dtroyer> I expect this will also have grenade implications
17:43:20 <mtreinish> dtroyer: it succeeded and it shouldn't have? :)
17:43:35 <chmouel> that would be very nice for very large feature to implement across all different projects (i.:e like nfv)
17:43:36 <dtroyer> mtreinish: yes, but probably because 'new' doesn't run stack.sh
17:43:54 <dtroyer> so it ran the system packages
17:44:01 <mtreinish> ah, ok
17:44:41 <dtroyer> anyway, I have a script to do the venv build and install the pypi packages that have distro package prereqs
17:44:55 <sdague> dtroyer: nice
17:45:13 <dtroyer> and its mostly activation and  assumption fixes from there
17:45:32 <dtroyer> that's it from me
17:45:40 <mtreinish> ok, thanks
17:45:46 <mtreinish> let's move onto the next topic then
17:45:48 <mtreinish> #topic Grenade
17:46:03 <mtreinish> does anyone have any updates on the world of grenade?
17:46:16 <mtreinish> dtroyer, sdague, jogo: ^^^
17:46:33 <sdague> we had a work around this week
17:46:44 <dtroyer> no details here yet beyond what I already mentioned
17:46:45 <sdague> let me find the patch
17:46:51 <mtreinish> the ceilo paste pipeline thing?
17:46:55 <sdague> https://review.openstack.org/#/c/142075/
17:46:56 <sdague> yeh
17:47:15 <sdague> I fast merged that tues morning to help unstick things
17:47:48 <sdague> turns out... for the last year... new swift was running with old ceilometer in it's pipeline
17:48:03 <sdague> and we didn't notice until the requirements capping
17:48:03 <mtreinish> heh, nice
17:48:37 <sdague> anyway, the new order address that, I also think cdent was going to talk with the swift and ceilometer teams and try to come up with a better long term plan
17:49:09 <sdague> https://review.openstack.org/#/c/142741/ - is a quick cleanup grenade review that's easy (removes a work around needed to land 142075 and another old work around)
17:49:12 <cdent> yeah, in today's ceilometer meeting I got the job of carrying a branch to swift
17:49:47 <sdague> cdent: great
17:49:57 <mtreinish> ok, cool
17:50:17 <mtreinish> sdague: +A
17:50:29 <mtreinish> ok is there anything else on grenade? Otherwise lets move on
17:50:38 <sdague> (also... on plugins ... looks like it totally works - http://logs.openstack.org/15/142815/2/check/gate-tempest-dsvm-large-ops/e5171fb//logs/devstacklog.txt.gz#_2014-12-18_17_25_32_896 )
17:51:09 <mtreinish> awesome
17:51:40 <mtreinish> #topic Critical Reviews
17:51:55 <mtreinish> does anyone have any reviews they'd like to get some extra eyes on?
17:52:11 <dkranz> mtreinish: just the one I put in the agenda
17:52:15 <mtreinish> #link https://review.openstack.org/#/c/141014/
17:52:17 <mtreinish> dkranz: ^^^
17:52:24 <mtreinish> it looks like it's WIP though
17:52:42 <dkranz> mtreinish: right. I just wanted to know if folks were comfortable with the approach
17:52:50 <dkranz> I have not been able to find ken-ichi
17:53:04 <dkranz> And am not sure why it was marked WIP
17:53:16 <mtreinish> dkranz: it seems reasonable. It's basically what I had in mind for doing that
17:53:27 <dkranz> If we are comfortable with breakout to a base class and accepting keyword args for all the conf values I might take it over
17:53:32 <mtreinish> (it's basically what we did for the CLI tests too, just in reverse order)
17:53:37 <dkranz> mtreinish: ok, cool
17:54:02 <dkranz> mtreinish: I am nursing the neutron client patch through gate failures and rebases
17:54:10 <dkranz> hopefully no more of the latter
17:54:26 <dkranz> mtreinish: those two are what is needed for the neutron team to move ahead
17:54:40 <dkranz> mtreinish: that's all for that one I think
17:54:48 <mtreinish> ok, thanks
17:54:58 <mtreinish> does anyone else have any reviews to bring up?
17:56:22 <mtreinish> ok, if there aren't any other reviews, I guess we'll end here
17:56:26 <mtreinish> thanks everyone
17:56:45 <mtreinish> #endmeeting