17:00:52 <oomichi> #startmeeting qa
17:00:52 <openstack> Meeting started Thu Jan  5 17:00:52 2017 UTC and is due to finish in 60 minutes.  The chair is oomichi. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:53 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:00:56 <openstack> The meeting name has been set to 'qa'
17:01:12 <oomichi> hi who join the meeting today?
17:01:50 <jordanP> hi
17:02:20 <jordanP> mtreinish, andreaf gmann you here ?
17:02:28 <oomichi> ping mtreinish, andreaf, masayukig
17:02:59 <mtreinish> o/
17:03:08 <oomichi> ok, lets start meeting
17:03:11 <oomichi> #link https://wiki.openstack.org/wiki/Meetings/QATeamMeeting#Agenda_for_January_5th_2017_.280900_UTC.29
17:03:20 <oomichi> ^^^ is today agenda
17:03:32 <oomichi> #topic PTG
17:03:44 <jordanP> question about the PTG, should we expect design sessions or coding/handson session ?
17:04:03 <oomichi> jordanP: I feel that is design session
17:04:15 <oomichi> according to https://www.openstack.org/ptg/
17:04:32 <oomichi> "project teams will meet in person to discuss priorities for the upcoming cycle, iterate quickly on solutions for complex issues, and make fast progress on critical items."
17:04:34 <jordanP> ok, but that won't last a full day, we have only 3 proposed "sessions" now
17:05:00 <mtreinish> oomichi: well it is free form. We get a room, so it's up to us to decide the format
17:05:05 <oomichi> jordanP: yeah, I see the 3, thanks for proposing
17:05:17 <mtreinish> oomichi: have you put the etherpad on the ml?
17:05:26 <jordanP> yes, I've seen it on the ML
17:05:31 <oomichi> mtreinish: can we arrange the slots as we like
17:05:41 <oomichi> ?
17:05:44 <mtreinish> oomichi: there are no slots, we get a room for a block of time
17:05:58 <oomichi> mtreinish: yeah, I have posted it on ml
17:06:24 <oomichi> mtreinish: that means we can use a room for a single QA project 2 full days?
17:06:26 <clarkb> one thing I was thinking about was potentially setting aside time for infra/qa to work with other projects on dealign with gate failures, reading logs, using elastic-recheck etc
17:06:37 <mtreinish> ok, I must ahve just missed it because of my time off, or the holidays
17:06:42 <mtreinish> clarkb: yeah that's a good idea
17:07:00 <oomichi> clarkb: cool
17:07:01 <mtreinish> clarkb: that was part of the reason why it's horizontal teams only the first couple of days
17:07:06 <clarkb> mtreinish: yup
17:07:12 <mtreinish> to try and get more people involved in stuff :)
17:08:10 <oomichi> yes, it is nice to write ideas for involving
17:08:49 <jordanP> for Tempest, I would like to spend most of my time coding with you. But I am open for discussions too :)
17:08:51 <oomichi> mtreinish: again: we can use a room for a single QA project 2 full days?
17:09:10 <clarkb> oomichi: that is my understanding
17:09:13 <oomichi> jordanP: TBH, +1 for coding :)
17:09:26 <clarkb> (I don't know if it will be one team to one room though, but you should have space for that 2 days dedicated to qa)
17:09:56 <oomichi> clarkb: I got it, so we can manage how to use the room as we can
17:10:08 <oomichi> that is nice :)
17:10:08 <jordanP> *as we want :)
17:10:16 <oomichi> yes :)
17:10:53 <oomichi> ok, are there another topic related to PTG?
17:11:25 <oomichi> ok, lets move on
17:11:36 <mtreinish> oomichi: I think it might be worthwhile to ask on the ML about specifics for topics about getting involved
17:11:54 <mtreinish> like are there things people would like to learn about qa and/or infra
17:12:26 <mtreinish> just to collect a slightly different set of topics people might be interested in
17:12:27 <oomichi> mtreinish: ah, that seems a general session a little bit like the summit?
17:12:53 <jordanP> yeah, but smaller and more interactive
17:13:00 <mtreinish> jordanP: right
17:13:01 <oomichi> for learning qa/infra in general session, not design session
17:13:28 <mtreinish> oomichi: I was thinking along the same lines as clarkb, part of the reason for having 2 dedicated days with horizontal teams was to get more people involved
17:13:51 <mtreinish> so getting some idea of what people would like to learn might give us some more ideas on how to organize the ptg time
17:14:23 <oomichi> mtreinish: that would be productive
17:15:15 <jordanP> ok:
17:15:20 <oomichi> ok, please write some ideas on the etherpad to do that and I will send a mail to get more attention
17:15:43 <jordanP> #action oomichi to ask on the ML if people are interested in some general qa/infra sessions
17:15:48 <jordanP> something along those lines :)
17:16:19 <oomichi> jordanP: yeah, that is my homework
17:16:30 <jordanP> let's move on
17:16:40 <oomichi> is it ok to move on ?
17:16:44 <oomichi> got it
17:16:48 <oomichi> #topic Specs Reviews
17:17:14 <oomichi> small number of specs on #link https://review.openstack.org/#/q/status:open+project:openstack/qa-specs,n,z
17:17:43 <oomichi> HA spec seems needed to be updated
17:18:43 <oomichi> resource one also needs to be updated, but I'd like to skip it on this Ocata cycle due to less time
17:19:14 <mtreinish> oomichi: well we can prod andreaf some more to still revise it
17:19:24 <mtreinish> having it in good shape before the ptg would be best I think
17:19:57 <oomichi> mtreinish: now the resource spec is owned by me, and yeah it is nice to prepare for PTG before
17:20:30 <mtreinish> oomichi: ah, ok. I forgot you took it over
17:20:55 <oomichi> yeah, it is easy to forget that due to different owner :)
17:21:00 <oomichi> including me
17:21:31 <oomichi> ok, lets move on if we dont have more items about spec
17:22:11 <oomichi> #topic Tempest
17:22:45 <oomichi> The bug triage seems in very good progress from https://github.com/oomichi/bug-counter#current-graph
17:23:09 <oomichi> now the bug number is less 150
17:24:06 <oomichi> This week bug triage is done by chandankumar
17:24:23 <oomichi> but he doesnt seem online today
17:24:35 <mtreinish> heh, someone was busy right before the holidays. That's quite the drop off
17:24:54 <oomichi> mtreinish: yep, that is :)
17:25:21 <oomichi> do we have topics about Tempest today?
17:26:35 <jordanP> nope, but I'd like some reviews on https://review.openstack.org/#/c/413739/
17:26:40 <jordanP> Port object_storage tests to Py3.
17:26:44 <jordanP> nothing urgent though
17:26:59 <mtreinish> #link https://review.openstack.org/#/c/413739/
17:27:05 <mtreinish> jordanP: cool, I'll take a look
17:27:17 <mtreinish> jordanP: so you're running tempest with py3 locally?
17:27:41 <oomichi> jordanP: I got it, btw do we run Tempest on the gate with both py3 and py2?
17:27:53 <mtreinish> oomichi: only in unit tests
17:28:01 <jordanP> mtreinish, yeah, I did run tempest on py3 locally
17:28:01 <mtreinish> oomichi: tempest runs are still only py2
17:28:06 <jordanP> results are encouraging
17:28:29 <mtreinish> jordanP: oh, cool. It's been a while since I last tried that, it wasn't in the best shape back then
17:28:37 <jordanP> I say we are very close to having it running fine, at least locally
17:28:39 <oomichi> I did breke some gate before due to this kind  of patches on Nova side, so I'd like to confirm that
17:29:00 <mtreinish> oomichi: dims is working on getting devstack running in the gate with py3
17:29:06 <jordanP> I am running tempest wirh py3, all other services running Py2
17:29:16 <mtreinish> although that technically isn't a blocker for running tempest under py3
17:29:23 <oomichi> jordanP: oh, awesome.
17:29:32 <mtreinish> it does make it easier from a gating perspective to just have a global flag USE_PYTHON3 in devstack
17:29:37 <oomichi> mtreinish: that is good work
17:30:54 <oomichi> ok, I will review it later anyways
17:31:53 <oomichi> lets move on if we don't have more topics related to Tempest today
17:32:14 <oomichi> #topic DevStack + Grenade
17:32:55 <oomichi> do we have disussion items about them?
17:33:25 <oomichi> there are a lot of +2ed patches on the devstack queue anyways
17:33:38 <oomichi> #link https://review.openstack.org/#/q/project:openstack-dev/devstack+status:open
17:33:44 <jordanP> yep
17:34:45 <oomichi> jordanP: ah, you are deprecating a lot of Tempest options and that would require Devstack patches also for removing the options
17:35:04 <jordanP> yeah, yeah, devstack changes have been proposed already
17:35:25 <mtreinish> oomichi: there is the py3 work dims was doing that I mentioned earlier
17:35:28 <oomichi> jordanP: cool, I will see them
17:35:33 <mtreinish> but I think all the patches he had up for that are already approved
17:36:25 <oomichi> mtreinish: https://review.openstack.org/#/c/416064/ is ?
17:36:52 <mtreinish> oomichi: that's the latest patch yes
17:36:59 <mtreinish> there were others that I think have already merged
17:37:18 <oomichi> mtreinish: are there patches for Tempest?
17:37:57 <mtreinish> not that I've seen
17:38:05 <mtreinish> well besides jordanP's from earlier :)
17:38:39 <oomichi> hehe, I got it
17:39:16 <oomichi> ok, lets move on to the next topic
17:39:20 <oomichi> #topic openstack-health
17:39:26 <jordanP> there have been a mass +2 on devstack changes
17:40:46 <oomichi> jordanP: yeah, thare is a lot
17:41:00 <mtreinish> I have nothing on openstack-health this week. I think things have been pretty quiet on it lately
17:41:27 <oomichi> mtreinish: yeah there are not so many active patches today on https://review.openstack.org/#/q/project:openstack/openstack-health+status:open
17:41:40 <oomichi> ok, lets move on
17:41:52 <oomichi> #topic Critical Reviews
17:42:10 <oomichi> do we have patches needed to be reviewed asap today?
17:43:03 <oomichi> from my side, I don't have
17:43:17 <jordanP> there was https://review.openstack.org/#/c/417009/ earlier
17:43:27 <jordanP> which got abandonned, but asks a question
17:43:52 <jordanP> what should we do if we break a project because they improperly imported/used non tempest/lib code ?
17:44:15 <oomichi> jordanP: Ideally, that should be fixed on project side, IMO
17:44:24 <oomichi> like this case
17:44:44 <mtreinish> jordanP: normally there is a quick fix in the consuming project because there is a supported way of doing it
17:44:53 <mtreinish> but if there isn't a quick fix I think a revert is appropriate
17:45:04 <oomichi> but if the effect is so big, it is possible to revert it case by case
17:45:16 <jordanP> ok, we are on the same page
17:45:20 <jordanP> oomichi, https://review.openstack.org/#/c/415532/
17:45:34 <jordanP> and https://review.openstack.org/#/c/405691/
17:45:38 <jordanP> they are duplicates
17:46:21 <jordanP> but yours depend on a patch that is not going to be merged soon :)
17:46:38 <jordanP> anyway, it's not critical, we can discuss this in #qa
17:46:44 <oomichi> jordanP: I win the race, but yeah that contains problem as you said ;)
17:47:00 <mtreinish> oomichi: heh, I like how sdague +2'd both :)
17:47:04 <oomichi> jordanP: I drop it now :)
17:47:13 <oomichi> haha, that is not good
17:47:24 <jordanP> sdague, was like 'let's see who wins this !'
17:48:05 <oomichi> jordanP: yeah, then you win now
17:48:16 <jordanP> bottom line is, we should all do more devstack reviews :)
17:48:56 <oomichi> jordanP: btw you seem to check unnecessary options of Tempest, the check is done already ?
17:49:08 <jordanP> I think I have it all covered yes
17:49:20 <jordanP> but a second pair of eyes always help
17:49:22 <oomichi> jordanP: cool, thanks so much for that
17:49:50 <jordanP> I've only done half of the job, the rest is to actually delete the options
17:49:55 <jordanP> now they are just deprecated
17:50:43 <oomichi> yeah, they are nice steps
17:51:36 <oomichi> do we have more patches to be reviewed today?
17:51:53 <oomichi> #topic Open Discussion
17:51:58 <mmedvede> hi, I'd like to request permission for voting on DevStack patches for IBM PowerKVM CI, is there a quorum here for a decision?
17:52:51 <oomichi> mmedvede: is that stable now?
17:52:55 <mmedvede> it is mostly stable, 7 day history: http://ci-watch.tintri.com/project?project=devstack&time=7+days
17:53:29 <mmedvede> I'd say it sometimes gives a false negative, but it is not much worse than jenkins at that, and we try to recheck when it happens
17:54:47 <jordanP> I am not found of the idea
17:55:01 <jordanP> we already have a hard time we the community gate
17:55:20 <jordanP> mmedvede, why do you want to make it voting ?
17:55:40 <jordanP> I mean, if we broke you and you are known here we can quickly work together
17:55:44 <jordanP> to come up with a fix
17:55:47 <mmedvede> It does catch valid problems sometimes for the platform (KVM on Power)
17:56:13 <mtreinish> jordanP: fwiw, voting doesn't really mean much. It just means it can leave a -1 on the patch, but it doesn't block anything
17:56:15 <mmedvede> voting gives  a bit more visibility, that is what voting for, right? :)
17:56:22 <jordanP> your CI should give you confidence about the stuff you deliver to your customers, why involve the community ?
17:56:36 <jordanP> mtreinish, I didnt know that
17:56:40 <jordanP> I though it would blocl
17:56:42 <jordanP> block
17:56:53 <mmedvede> jordanP: yes, it would not block the gate
17:57:17 <jordanP> then, I am not against having it voting
17:57:29 <jordanP> it's something :)
17:57:48 <oomichi> if that would be voting and give -1 on unrelated patches, that makes the review process slow
17:58:10 <oomichi> because reviewers tend to avoid -1 voting patches
17:58:49 <oomichi> and patch owners start getting angry to the CI which gives -1
17:59:13 <mmedvede> that is correct, but our CI is stable enough
17:59:46 <jordanP> we are out of time, sorry...
17:59:49 <mmedvede> and we also disable reporting when we catch our ci going haywire
17:59:50 <oomichi> mmedvede ok, let's discuss it more on qa channel due to the time
17:59:54 <mmedvede> ok
17:59:59 <oomichi> #endmeeting