22:00:33 <mtreinish> #startmeeting qa
22:00:34 <openstack> Meeting started Thu May 29 22:00:33 2014 UTC and is due to finish in 60 minutes.  The chair is mtreinish. Information about MeetBot at http://wiki.debian.org/MeetBot.
22:00:35 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
22:00:38 <openstack> The meeting name has been set to 'qa'
22:00:42 <mtreinish> hi who's here today?
22:00:46 <mlavalle> hi
22:00:49 <vrovachev> hi
22:00:53 <dpaterson> hi
22:00:54 <masayukig> o/
22:00:55 * clarkb lurks
22:00:59 <mtreinish> #link https://wiki.openstack.org/wiki/Meetings/QATeamMeeting#Proposed_Agenda_for_May_29_2014_.282200_UTC.29
22:01:05 <mtreinish> ^^^ Today's agenda
22:01:15 <mtreinish> it's just the boilerplate one today...
22:01:39 * sarob lurks
22:02:12 <mtreinish> ok, well let's get started
22:02:24 <mtreinish> #topic Specs Review
22:02:38 <mtreinish> so does anyone have a spec review they'd like to bring up
22:02:41 <mtreinish> or to discuss
22:02:49 <sdague> o/
22:03:00 <mtreinish> sdague: go ahead
22:03:22 <sdague> oh, that was mostly just (I'm here) but while we're at it - https://review.openstack.org/#/c/96445/
22:03:34 <sdague> is the write up of the javelin2 overview
22:03:52 <mtreinish> heh, ok well that works out :)
22:04:37 <sdague> so right now it mostly demonstrates the cli and the yaml that's being used. Comments welcomed
22:04:46 <mtreinish> #link https://review.openstack.org/#/c/96445/
22:04:50 <sdague> also breaks down the work chunks expected for the cycle
22:05:16 <sdague> easier read here - http://docs-draft.openstack.org/45/96445/2/check/gate-qa-specs-docs/57ce662/doc/build/html/specs/javelin2.html
22:05:39 <mtreinish> #link http://docs-draft.openstack.org/45/96445/2/check/gate-qa-specs-docs/57ce662/doc/build/html/specs/javelin2.html
22:05:49 <sdague> that's it unless there are questions
22:06:03 <mtreinish> I need to look into having a place to publish the specs after they're merged at some point
22:06:23 <mtreinish> sdague: it looks sane to me, but I'll do a detailed review once it gets a +2 from someone else
22:08:05 <mtreinish> which reminds me we need more reviews on qa-specs in general. Things have felt pretty stagnent on that review queue
22:08:20 <mtreinish> but I'll bug people about that after the meeting :)
22:08:40 <mtreinish> ok are there any other spec reviews that people want to bring up?
22:08:43 <mlavalle> mtreinish: I want the team to know I am working on https://review.openstack.org/#/c/95600/. It is for the Neutron scenarios we will be implementing in Juno. It is an early version, but please review it and give feedback
22:09:08 <mtreinish> #link https://review.openstack.org/#/c/95600/
22:09:18 <mlavalle> I already got a -1 from yfried. That's what I need at this point, though
22:09:46 <sdague> mtreinish: yeh, I was hoping to get a new unified dashboard with all of that in place. Still haven't quite figured out a way to land project specific dashboards without bugging clarkb
22:10:29 <mtreinish> mlavalle: it looks like a good start though. That's about what I was expecting for that spec.
22:10:41 <mtreinish> sdague: yeah having that dashboard might help prioritize things
22:10:41 <mlavalle> coll, thanks!
22:11:23 <mtreinish> ok, if there aren't any other specs to discuss we'll move onto the next topic
22:11:28 * afazekas WIP: https://review.openstack.org/#/c/95794/2/specs/network-debug.rst,cm
22:12:11 <mtreinish> #link https://review.openstack.org/#/c/95794/2/specs/network-debug.rst,cm
22:12:11 <masayukig> WIP: https://review.openstack.org/#/c/96163/
22:12:45 <masayukig> ^^^ about tempest server/client/GUI spec
22:12:54 <masayukig> #link https://review.openstack.org/#/c/96163/
22:12:57 <sdague> masayukig: cool
22:13:11 <masayukig> sdague: thanks :)
22:13:28 <mlavalle> afazekas: I'll review it by this weekend
22:13:45 <mtreinish> masayukig: oh, I bet that'll be a contentious one :)
22:13:53 <mtreinish> don't let jogo see it...
22:14:49 <masayukig> mtreinish: yeah...
22:15:06 <mtreinish> afazekas: I'll do a dive down into that at some point
22:15:58 <mtreinish> I have some concerns about doing out of band debug stuff in tree, like what we have now with the sudo stuff
22:16:30 <mtreinish> ok is that it for specs, or are there any others?
22:17:08 <mtreinish> ok then let's move on
22:17:11 <mtreinish> #topic Blueprints
22:17:24 <mtreinish> ok, does anyone have a blueprint status update
22:17:31 <mtreinish> I still need to do the blueprint purge
22:17:40 <mtreinish> but we have a handful of open bps with specs approved
22:18:12 <mtreinish> sdague: we hit some more branchless tempest edge cases this week right?
22:18:38 <sdague> yeh, ceilometer has a set of tests that are impossibly long in icehouse
22:18:48 <sdague> but work under the juno code
22:18:59 <vrovachev> yeees :(
22:19:32 <sdague> so we need to do some kind of signalling for those in the gate
22:20:03 <sdague> vrovachev: I was actually thinking about this on my bike ride today, would there be a way that we could have some ceilo extension that would give us bogomips of the system?
22:20:18 <sdague> basically an idea of what our event processing rate might be
22:20:40 <sdague> then you could actually ask ceilometer, and know programatically if the tests are viable
22:20:47 <mtreinish> sdague: just to be a bit harsh we could always rip ceilo out of the icehouse gate to get around this too...
22:21:08 <mtreinish> but that should be a last resort, and not something we want to do
22:21:09 <sdague> mtreinish: we could, there aren't that many tests anyway, but it makes for a good test case
22:21:56 <vrovachev> sdague, I think it is unlikely.
22:22:30 <sdague> vrovachev: the issue is I'm not very thrilled about the idea that some magic piece of information is needed to know if this would work in a cloud
22:22:39 <sdague> because it's not discoverable
22:22:59 <sdague> and the thing we were planning to have to use this for was new extensions, which are discoverable
22:23:09 <sdague> so this is a very weird thing
22:23:27 <mtreinish> sdague: we do have other non-discoverable feature flags in the config now
22:23:32 <mtreinish> not many though
22:23:47 <sdague> mtreinish: not in the feature matrix though
22:23:59 <mtreinish> that's fair because most of them don't work in the gate anyway
22:24:08 <mtreinish> and if they do they default on
22:24:35 <sdague> anyway, can probably take it offline
22:24:48 <mtreinish> sdague: you're right a prereq for doing this should be having ceilo have some kind of discovery that sql actually works
22:24:57 <mtreinish> sdague: yeah let's take this offline
22:25:10 <mtreinish> ok are there any other bp status updates?
22:25:29 <vrovachev> I do not know how you can work around this problem in tempest. because problem in the ceilo api in icehouce
22:26:26 <mtreinish> ok if there aren't any other bps let's move on
22:26:39 <mtreinish> #topic Neutron testing
22:26:51 <mtreinish> mlavalle: this is your topic right?
22:26:55 <mlavalle> mtreinish: yeah
22:27:10 <mlavalle> another couple of api tests merged since last meeting
22:27:21 <mlavalle> we are at 25 only another 3 to go
22:27:30 <mlavalle> to complete the set we defined in January
22:27:41 <mtreinish> mlavalle: we're getting there...
22:27:49 <mtreinish> how many new apis were added since that list though?
22:27:55 <mlavalle> also made progress this week with fwaas and vpaas scenario tests
22:28:14 <mlavalle> we will be able to implement in one devstack
22:28:23 <mlavalle> no need of multinode for those tests
22:29:06 <mlavalle> and finally I am tracking what markmcclain is doing as far as nova network parity, in case help is needed
22:29:25 <mlavalle> that's all I have this week
22:29:30 <mtreinish> mlavalle: that's too bad I was actually hoping that would be a motivator to actually get someone to set up multinode soon
22:29:33 <mtreinish> mlavalle: ok, thanks
22:29:45 <mtreinish> does anyone have anything else to add on neutron testing?
22:29:56 <mtreinish> mlavalle: do you know where we stand with a parallel full run
22:29:58 <mlavalle> multinode will be needed for other scenarios, so I still see it as my cause
22:30:08 <mlavalle> :-)
22:30:12 <mtreinish> because I think we need to get that in before we start ramping up more tests?
22:30:22 <mlavalle> agree
22:30:54 <mtreinish> I can bug rossella_s and salv_orlando about that later though
22:31:01 <mlavalle> ok
22:31:20 <mtreinish> ok, if no one else has anything to discuss about neutron testing let's move on
22:31:55 <mtreinish> #topic Bugs
22:32:21 <mtreinish> so I haven't had a chance to look at the bug list lately, I imagine it's grown since our last bug day
22:32:33 <mtreinish> but does anyone have any bugs they'd like to bring up
22:32:42 <mtreinish> or anything else to talk about open bugs
22:33:36 <sdague> not here, I just want to bug masayukig to take a look at my javelin patches - https://review.openstack.org/#/q/status:open+project:openstack/tempest+branch:master+topic:bp/javelin2,n,z :)
22:33:54 <mtreinish> heh, ok well that's a good segway into the next topic
22:34:02 <mtreinish> #topic Critical Reviews
22:34:20 <mtreinish> #link https://review.openstack.org/#/q/status:open+project:openstack/tempest+branch:master+topic:bp/javelin2,n,z
22:34:23 <sdague> note, we discovered at summit that javelin wasn't doing what we thought
22:34:25 <masayukig> sdague: sure
22:34:29 <mtreinish> yes someone look at those
22:34:37 <mtreinish> just so sdague stops bugging me about it :)
22:34:48 <sdague> so this is an attempt to do a rebuild that will live in tempest that we can call from grenade
22:34:50 <sdague> :P
22:35:11 <mtreinish> does anyone else have any reviews they'd like to get extra eyes on?
22:35:15 <sdague> well, I'm antsy to start actually testing resources living across upgrades again...
22:35:47 <afazekas> #link https://review.openstack.org/#/c/94203/
22:37:11 <mtreinish> afazekas: yeah that's probably a good fix
22:37:40 <sdague> afazekas: can we simplify that code a bit, map filter, and 2 lambdas are kind of dense
22:38:26 <afazekas> sdague: basically it is copied from the same api test
22:38:27 <sdague> what about computes = [x for x in hosts if x.service == 'compute']
22:38:42 <sdague> and then computes[0].host_name
22:38:53 <sdague> afazekas: it's still pretty dense
22:38:59 <afazekas> ok
22:39:15 <mtreinish> sdague: heh, I was fine with it, but that'll probably be a bit easier to read
22:39:23 <mtreinish> ok are there any other reviews?
22:39:37 <sdague> yeh, as someone that spent much of this week reading odd parts of tempest, I'm all for future readability
22:39:48 <mtreinish> #link https://review.openstack.org/#/c/92606/
22:40:13 <mtreinish> #link https://review.openstack.org/92519
22:40:28 <mtreinish> the first should fix a periodic failure
22:40:29 <sdague> +2 to the first one
22:40:38 <mtreinish> and the second has just been sitting for a little bit
22:41:12 <mtreinish> ok well if there aren't any other review, let's open the floor
22:41:12 <sdague> the -1 on that one is probably fair, to respin the readme
22:41:53 <mtreinish> sdague: oh I completely missed the -1
22:41:59 <mtreinish> I only saw the +2 sorry
22:42:14 <mtreinish> #topic Open Discussion
22:42:29 <mtreinish> ok, does anyone have a topic they'd like to bring up that wasn't on the agenda
22:42:40 <vrovachev> guys, please see sahara api tests https://review.openstack.org/#/c/90101/
22:42:56 <mtreinish> #link https://review.openstack.org/#/c/90101/
22:43:18 <vrovachev> thanks :)
22:43:54 <sdague> mtreinish: what about a preview of the mid cycle event?
22:44:05 <mtreinish> oh yeah that's a good idea :)
22:44:11 <sdague> what's your thinking for both bootstrapping days and the QA dedicated day?
22:44:21 <ylobankov> vrovachev: :)
22:44:35 <mtreinish> so for the bootstrapping days it's going to be an overview of both how infra works
22:44:41 <mtreinish> and an explanation of tempest and grenade
22:44:48 <sdague> #info QA / Infra Mid Cycle July 14 - 18
22:44:49 <mtreinish> how they work etc...
22:45:08 <mtreinish> for the dedicated qa day, I was thinking we either could have a day of f2f discussion
22:45:25 <mtreinish> or if there was a topic we needed to work through we could have a hacking day
22:45:26 <sdague> ok, would be interesting to get a rough agenda, even if it's just broken up by "morning / afternoon" topics across the days.
22:45:45 <mtreinish> sdague: yeah I need to work with jeblair to get more details ironed out
22:45:49 <sdague> coolio
22:46:09 <mtreinish> I should also say the schedule I posted to the ML is tenantive
22:46:14 <mtreinish> we might move things around a bit
22:46:17 <sdague> ok
22:46:28 <mtreinish> and steal a bootstrapping day for more qa stuff if we need it
22:46:39 <sdague> well good to get that nailed down soon, as I expect people will be starting to sort out travel soon
22:47:01 <mtreinish> for those who missed the post:
22:47:03 <mtreinish> #link http://lists.openstack.org/pipermail/openstack-dev/2014-May/036262.html
22:47:09 <masayukig> The mid-cycle meet-up is interesting! But I need to talk my boss(es) to get the budget.
22:47:20 <masayukig> Germany is very far from Japan...
22:47:27 <mtreinish> mlavalle: you're a boss now though :)
22:47:48 <masayukig> mtreinish: heh :)
22:48:28 <mtreinish> also if anyone has suggestions on something they'd like to see for the mid-cycle, please ping me
22:49:00 <mtreinish> ok are there any other topics to discuss? otherwise we'll end a little early today
22:50:03 <vrovachev> heh, go to sleep :) in Russia 2.49 AM
22:50:23 <mtreinish> ok, well I guess I'll call it for today
22:50:26 <mtreinish> thanks everyone
22:50:33 <mtreinish> #endmeeting