17:01:14 #startmeeting qa 17:01:15 Meeting started Thu Oct 8 17:01:14 2015 UTC and is due to finish in 60 minutes. The chair is mtreinish. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:01:16 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:01:20 The meeting name has been set to 'qa' 17:01:21 ok, without the typo this time 17:01:27 o/ 17:01:29 o/ 17:01:31 who's here for today's meeting 17:01:37 o/ 17:01:39 man, you guys are faster than me :) 17:01:45 #link https://wiki.openstack.org/wiki/Meetings/QATeamMeeting#Proposed_Agenda_for_October_8th_2015_.281700_UTC.29 17:01:52 ^^^ Today's agenda 17:02:22 sdague, dtroyer: around? 17:03:07 well, let's get started 17:03:20 #topic Mitaka Summit Prep 17:03:31 #link https://etherpad.openstack.org/p/mitaka-qa-summit-topics 17:03:35 o/ 17:03:50 just a reminder to add topics to the brainstorming etherpad 17:04:10 we've got 8 slots and 9 proposals so if you add one your odds are pretty good :) 17:04:32 that list reminds me of a question: is the health dashboard running somewhere for public consumption somewhere? 17:04:36 I think we'll have to do the deciding soonish 17:04:46 cdent: not quite yet, there is still a bit of work on that front 17:04:55 you can spin it up locally using real data though 17:05:00 cool thanks 17:05:14 cdent: if you want to test it locally I can give you a hand with that post-meeting 17:05:43 I'll give it a go some time later (after this I'm off to dinner) 17:05:44 * mtreinish had to learn some dark secrets of js packaging to work on it :( 17:05:55 ok cool 17:06:08 ok does anyone have anything to discuss on this topic? 17:06:16 * andreaf lurks around for now 17:07:05 0/ 17:07:25 ok, I guess not 17:07:27 let's move on 17:07:29 #topic Specs Reviews 17:07:37 #link https://review.openstack.org/#/q/status:open+project:openstack/qa-specs,n,z 17:07:53 we actually have a new spec proposal this week from SpamapS: 17:08:03 #link https://review.openstack.org/230183 17:08:18 o/ 17:08:20 the first one in a while 17:08:29 * mtreinish lets SpamapS sell his spec :) 17:08:36 Lets measure things! 17:09:39 Please have a quick read if you're interested. I'm pushing hard to give us a picture of how hard we're hammering our backend services. 17:10:11 sounds interesting. I'll have a peek 17:10:18 The goal is to have some data about this as change ocurrs, and then start doing some large scale changes to reduce (or even in some cases _remove_) use of backend services through various means and by adding some new ones. 17:10:39 For instance, keystone is ripe for more caching. 17:10:54 But we don't have much of a picture of how hard it actually hits the database versus the hit/miss ratio on memcached. 17:11:42 And even more fun.. we're abusing rabbitmq *and* mysql for locking in Cinder and Heat, but using a DLM might have a significant positive impact on how much we utilize the mq and database. 17:12:01 Also, who doesn't love more graphs? 17:12:04 yeah, and the nice thing is this will also provide a framework and template for how to expand this for other metrics in the future when new questions arise 17:12:51 heh, everyone likes colors 17:13:09 ok, does anyone have anything else on specs today? (either SpamapS's or in general) 17:14:11 #topic Tempest 17:14:24 ok does anyone have anything to discuss about tempest this week? 17:14:55 mtreinish, have a question about devstack/tempest regarding flavors, not sure if can use this time :) 17:15:02 rfolco: go for it 17:15:13 rfolco: it's not like anyone else spoke up :) 17:15:32 I have a question about flavors on devstack/tempest. Since the memory requirements changed for ppc64, I sent a patch to stop creating nano/micro flavors and to start using DEFAULT_INSTANCE_TYPE=m1.tiny, 512MB - https://review.openstack.org/#/c/226905. However, the flavor_ref_alt would be m1.small - 2G, which works fine, but too large� waste of RAM. 17:15:32 Would be reasonable to introduce CUSTOM_FLAVOR_ALT= so I can have a smaller favor_alt for resize scenarios, as 768MB ? 17:15:43 rfolco: although I'm not sure I want to here about power and how you need giant guests to work on it 17:16:20 rfolco: sure I think that's fine 17:17:03 mtreinish, thanks 17:17:13 rfolco: heh, was that it? :) 17:17:44 mtreinish, will submit the patch for review on this path then.. that's what I meant 17:18:05 rfolco: heh, nm 17:18:14 ok is there anything else on tempest for this week? 17:18:58 I guess one thing to note is that the tempest-lib service client migrations are in full swing 17:18:58 mtreinish: just a heads up I started working on the migration of cred providers to tempest-lib 17:19:20 so if people could keep an eye on tempest-lib that'll help that move quickly 17:19:32 andreaf: ok cool, that'll be good to have 17:19:32 mtreinish: will do 17:20:04 mtreinish: the only bit that is blocked is the tenant isolation provider, which requires the identity client to be in tempest-lib 17:20:06 andreaf: on that note, I started an etherpad somewhere with things we need to migrate to tempest-lib (when I was working on the ec2 tempest plugin) 17:20:11 mtreinish: but that should happen soon 17:20:12 I'll have to dig that up and put it on the ML 17:20:26 mtreinish: ok thanks 17:20:35 andreaf: can't we pass it the client object? 17:21:07 mtreinish: sure that's the way it is now, but it would be a bit odd to have something in tempest-lib which depends on an object in tempest 17:21:30 andreaf: heh, that's how the rest client was for a long time until we migrated auth 17:21:34 but I agree it is weird 17:22:24 ok, is there anything else to discuss on tempest this week? 17:22:25 mtreinish: anyways the identity client should be in tempest-lib relatively soon 17:22:39 andreaf: ok cool 17:23:30 ok, lets move on 17:23:32 #topic Devstack 17:23:45 does anyone have something to discuss about devstack this week? 17:23:50 dtroyer, sdague: ? 17:25:36 ok, I guess not. Let's move on 17:25:45 #topic Grenade 17:26:01 anyone have anything to discuss on grenade this week? 17:26:30 I guess the big news here (and for devstack too) is that we've branched eveything for liberty 17:26:44 so now on master changes we're running liberty->master 17:26:55 which sdague drove forward, there were some headaches IIRC 17:27:45 ok, is there anything else on grenade? 17:28:46 #topic OpenStack Health 17:29:16 ok so does anyone have anything to discuss on the openstack-health dashboard 17:29:31 I can run down a quick status if anyone wants 17:29:45 mtreinish: that would be nice 17:29:45 mtreinish: it would be great 17:29:55 So I was trying to follow the dev setup and somewhere between npm and gulp, I'm failing :) 17:30:20 dtroyer: yeah the js install stuff is awful 17:30:42 dwalleck: improves the docs: https://review.openstack.org/#/c/231746/4 17:30:47 because those extra steps are so obvious 17:31:05 ahh cool, I'll take a look 17:31:18 dwalleck: there is a follow on step for mac too :) 17:31:25 as for the status right now we're still working on getting the 3 main pages working with real data 17:31:44 the first page (an overview of all runs with per project groups) is in a good place 17:32:06 mtreinish: maybe we need some ui/ux improvements only. 17:32:08 mtreinish: hehe :) 17:32:08 the second page has the angular pieces but still needs api query support 17:32:21 ccarrara: sure, this is all iterative 17:32:30 we need to get a starting point running 17:32:38 then we can make lots of improvements easily 17:32:40 mtreinish: We think the second page already has api support. 17:32:43 I don't know if there's any place you guys need specific help, but this something I absolutely need and would like to throw some cycles at 17:33:00 mtreinish: https://review.openstack.org/#/c/231668/ 17:33:05 #link https://review.openstack.org/#/c/231668/ 17:33:10 yes that's the patch in progress 17:33:15 I was discussing what's already landed 17:33:20 nice 17:33:32 dwalleck: sure, that would definitely help 17:33:44 dwalleck: sure I think help is always welcome 17:33:47 dwalleck: the more hands on deck we can get things running for real 17:34:41 the lack of information about the tasks in progress and the target to achieve in the project can be a obstacle to income helps 17:34:47 cool. I'll be lurking around more, so I'll start poking around 17:35:05 ccarrara: well we should be using an etherpad, but that kinda fizzled away 17:35:05 I don't know if we can improve this in any way 17:35:11 I'll see if I can fix that soon 17:35:13 agree 17:35:38 the other big thing is getting the real deployment stuff working 17:35:47 there are a couple patches in progress for that: 17:35:56 #link https://review.openstack.org/232255 17:36:02 mtreinish: I think once we have the fist pages up and running we could start using specs, in the meanwhile the etherpad should be good to see what's going on 17:36:03 for real config support on the frontend 17:36:32 oh it looks like the wsgi container support landed when I wasn't looking 17:36:46 #link https://review.openstack.org/#/c/227849/9/ 17:36:57 currently working on the improvements suggested by mtreinish 17:37:33 glauco: yep, that'll be how we deploy it with puppet 17:37:43 there will be a corresponding system config patch to use it when it's ready 17:37:58 andreaf: oh, yeah specs might be a good review mechanism for adding pages 17:38:27 ok, so my summary was kinda a mess :) 17:38:35 but was there anything else to discuss on this topic? 17:38:54 mtreinish: maybe we couldn't help so much :) 17:38:55 mtreinish: was there any news on ci-watch? 17:39:00 to interject, here's a running snapshot of the current master branch: http://208.94.241.194:8080/#/ 17:39:08 for those looking for an example 17:39:11 #link http://208.94.241.194:8080/#/ 17:39:21 timothyb89: heh, that's actually really helpful I think 17:39:22 cool! 17:39:22 thanks 17:39:36 andreaf: I think they're still working on that on the infra side 17:39:39 timothyb89: cool thanks 17:39:47 andreaf: or did you mean the ci-watch spec? 17:40:02 mtreinish: I meant the actual service 17:40:10 mtreinish: from the infra folks 17:40:13 (it's not at all confusing that the spec for this work was labelled ci-watch and the 3rd party dashboard service is called ci-watch) 17:40:22 andreaf: I haven't been following it too closely 17:40:29 but I think they're making progress there 17:40:46 mtreinish: I guess we should have a new spec, reflecting the initial 4 pages, just for sake of tracking it? 17:40:59 mtreinish: a spec in the qa-specs repo 17:41:07 andreaf: is that you volunteering to write it? 17:41:14 mtreinish: yep, indeed 17:41:16 because that would probably help 17:41:29 #action andreaf to write qa spec for the work in progress on openstack-health 17:41:36 it's all sorts of official now :) 17:42:07 ok, is there anything else on this topic? 17:43:13 andreaf: I would like to help on this spec as well, if possible 17:43:47 ok, lets move on 17:43:49 #topic Critical Reviews 17:43:50 glauco: sure I'll put up an initial draft 17:43:56 glauco: and we can take it from there 17:44:04 does anyone have any reviews they'd like to get extra eyes on? 17:44:16 yes 17:44:25 https://review.openstack.org/#/c/224338/ 17:44:39 #link https://review.openstack.org/#/c/224338/ 17:44:44 oh, that's an important one 17:44:51 yes. If we could get some reviews on https://review.openstack.org/#/c/231668/ 17:45:04 we get at least 1 question a week because the docs recommended method of running tempest fails 17:45:08 mtreinish: no so critical, but this is the series for preparing the cred provider migration, at least the part I have up until now: # link https://review.openstack.org/#/q/status:open+project:openstack/tempest+branch:master+topic:bp/tempest-library,n,z 17:45:20 so if people could review dpaterson's patch that would be good 17:45:29 #link https://review.openstack.org/#/c/231668/ 17:45:38 #link https://review.openstack.org/#/q/status:open+project:openstack/tempest+branch:master+topic:bp/tempest-library,n,z 17:45:58 andreaf: I think you messed up the formatting, i'm not sure it was gonna get picked up 17:46:17 mtreinish: thanks 17:46:20 and I have 1 this week too: 17:46:21 #link https://review.openstack.org/232592 17:46:40 which is kinda simple, but something I hit in testing the tempest_ec2 plugin I've been playing with 17:47:21 ok, are there any other reviews to bring up? 17:47:52 mtreinish: +2 17:48:14 andreaf: thanks 17:48:25 ok, lets open the floor 17:48:28 #topic Open Discussion 17:48:38 Does anyone have anything to discuss which wasn't on the agenda? 17:51:05 ok, I guess let's end it here today 17:51:08 thanks everyone 17:51:13 #endmeeting