19:00:25 <catherineD> #startmeeting refstack 19:00:27 <openstack> Meeting started Tue Jun 14 19:00:25 2016 UTC and is due to finish in 60 minutes. The chair is catherineD. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:00:28 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:00:30 <openstack> The meeting name has been set to 'refstack' 19:00:43 <sslypushenko> o/ 19:00:59 <pvaneck> o/ 19:01:21 <catherineD> #link meeting agenda and notes, please feel free to add items https://etherpad.openstack.org/p/refstack-meeting-16-06-14 19:01:33 <andrey-mp> hi 19:01:36 <andrey-mp> o/ 19:02:34 <catherineD> andrey-mp: hi 19:03:02 <catherineD> agenda https://etherpad.openstack.org/p/refstack-meeting-16-06-14 19:03:20 <catherineD> #topic Feature branch feature/vendor 19:03:25 <luzC> o/ 19:03:41 <catherineD> luzC: hello welcome 19:03:58 <catherineD> so last week we agreed that we will create a feature branch 19:04:22 <catherineD> and we did .. pvaneck: and andrey-mp: start to submit patches to the branch 19:04:30 <catherineD> thx 19:04:57 <catherineD> This branch is for non-production commits only 19:05:32 <catherineD> #link first commit to this branch --> Add vendor UI ( https://review.openstack.org/#/c/328780/ ) 19:05:50 <andrey-mp> btw, I didn't read your last meeting log - do you have some document with branches agreement? 19:05:53 <catherineD> Please review ... 19:06:04 <andrey-mp> I'm reviewing it 19:06:08 <catherineD> andrey-mp: yes we did .. 19:07:01 <catherineD> * AGREED: Use feature branch for development of new features and master branch for deployment (business as usual) ... (this supersedes the earlier agreement) (catherineD|2, 19:56:28) 19:07:30 <andrey-mp> :) 19:08:02 <catherineD> any other thought about this topic ? 19:08:13 <andrey-mp> I mean agreement how and when it will be merged 19:08:44 <catherineD> andrey-mp: you mean merger from feature branch to master? 19:08:52 <andrey-mp> yeah 19:08:58 <catherineD> we did not discuss when ... 19:09:24 <andrey-mp> its strange... 19:09:32 <catherineD> for the how ... I understand that there is exisiting method and procedure to merge feature branch to master 19:09:45 <catherineD> pvaneck: you take a look right? 19:10:02 <andrey-mp> you've made some desicion but don't how to apply it ) 19:10:53 <pvaneck> we will merge the feature vendor into master once we feel it is ready for production. WE already have the gerrit permissions to do it 19:10:59 <pvaneck> i think it is fine 19:11:06 <andrey-mp> ok 19:11:15 <andrey-mp> its good ) 19:11:48 <catherineD> andrey-mp: according to Doug this is the best way ... 19:12:09 <Rockyg> o/ 19:12:27 <luzC> I just take a look 19:12:41 <catherineD> Rockyg: agenda https://etherpad.openstack.org/p/refstack-meeting-16-06-14 19:12:50 <luzC> branch is only at refstack but not in refstackclient 19:13:07 <sslypushenko> andrey-mp: not may be the best, but it will definitely work) 19:13:19 <andrey-mp> it's common git method but not common to base OpenStack projects 19:13:27 <catherineD> luzC: correct ... branch is only needed for refstack not refstack-client 19:13:29 <andrey-mp> I'm ok with it ) 19:13:57 <luzC> ok, sounds good 19:14:09 <catherineD> moving on ... 19:14:21 <andrey-mp> sslypushenko: I mentioned this way two meetings ago but you preferred OpenStack way ) 19:14:51 <catherineD> #topic Comments from Xuchao 19:15:15 <catherineD> #link http://eavesdrop.openstack.org/irclogs/%23refstack/%23refstack.2016-06-14.log.html 19:16:05 <catherineD> I am kind of concern about the comments --> " You don't pay much attention to the problem ? " 19:17:21 <pvaneck> well the timezone differences make it tough to reply to irc messages in a timely matter 19:17:34 <catherineD> We pretty much answer every IRC question and comments to the bugs .... is there any thing else we should do here? 19:17:44 <catherineD> pvaneck: that is true 19:18:20 <catherineD> Rockyg: any thoughts on how to handle the situation? 19:18:47 <andrey-mp> but is this question related to RefStack? 19:19:03 <catherineD> andrey-mp: IMO it is not 19:19:14 <Rockyg> I'm looking at the bug report now. I think this may be filed by a Huawei tester trying to pass the tests......give me a sec 19:19:14 <andrey-mp> sure 19:19:15 <catherineD> that is why I marked all bugs invalid 19:19:43 <andrey-mp> so why should care about questions that is not related to us? ) 19:19:53 <andrey-mp> also - there is a mailing list 19:19:55 <catherineD> Rockyg: ChinaCloud according to the email address .. 19:20:16 <andrey-mp> anyone can ask any question through the mailing list - I read it every day 19:20:47 <Rockyg> Hmmm. Still might be us ;-) 19:21:18 <andrey-mp> we can add some note to refstack channel - "if no one answers you then try to ask via openstack-dev mailing list" 19:21:24 <Rockyg> But, good reply. Although I think maybe also a link to the description og "flagged test" would help. 19:21:31 <Rockyg> in the bug. 19:21:52 <catherineD> andrey-mp: after all it is RefStack tests that give them the results .... like in any development env ... QA and tests are the last places to catch failures... 19:22:37 <catherineD> Rockyg: I think I did give a link in one of the bug .... the issues are most bugs are for the same issues 19:23:52 <catherineD> Rockyg: the comments in this one https://bugs.launchpad.net/refstack/+bug/1592239 gives the link about DefCore flagged tests 19:23:52 <openstack> Launchpad bug 1592239 in refstack "The test run didn't actually run any tests" [Undecided,Invalid] - Assigned to xuchao (xuchao) 19:24:05 <Rockyg> just mark them dupicates, then. and point the to the tempest original bug 19:24:26 <catherineD> This is the first time that the little "flag" in-front of the test name does not mean much to a user 19:24:45 <catherineD> Rockyg: would you be able to find out? 19:24:50 <pvaneck> Maybe the refstack site needs an FAQ or more information outlining the separation between defcore, refstack, and tempest 19:25:24 <catherineD> pvaneck: This is the reason I bring the topic here ..... Is there something that we can do? 19:25:57 <Rockyg> Yeah. agree with pvaneck 19:26:01 <catherineD> however, this is the first time we encountered such questions for being on line for about one year now 19:26:18 <sslypushenko> catherineD: we need some feedback from DefCore 19:26:22 <Rockyg> Maybe we can link flag to the bug, or the reason the flag is there? 19:27:02 <catherineD> sslypushenko: good idea .. I will add this as a topic to the DefCore working group ... 19:27:23 <sslypushenko> flagged tests - it is DefCore idea.... so from refstack point of view we need just provide some explanation 19:27:24 <Rockyg> More folks are trying to follow the steps, and china people are afraid to engage Chris. Even our people at Huawei. They need to be introduced to him before they will ask him questions 19:27:49 <sslypushenko> it is definitely looks like not a RefStack bug 19:28:24 <sslypushenko> Rockyg: heh) 19:28:26 <Rockyg> For every flag defcore applies, we should have link to the reason or the review, or the bug or whatever. Could save us and chris a lot of effort 19:29:07 <catherineD> Rockyg: I think that would complicate the UI but I will ask for pvaneck: andrey-mp: opinions 19:29:22 <Rockyg> Thanks. 19:29:54 <pvaneck> right now, flag reasons are given as title text when you hover over the flag icon. But i can see how that may not be clear 19:30:07 <Rockyg> we also might consider a training class at Barcelona for newbies trying to certify 19:30:11 <catherineD> so the person does not seem to understand the concept ofDefCore using flag, aliases 19:30:31 <Rockyg> Yup. 19:30:37 <andrey-mp> Rockyg: +1 19:30:42 <Rockyg> a lot of folks don't read the docs;) 19:30:48 <andrey-mp> +100 19:31:09 <catherineD> Rockyg: we did at Austin but the turn out seems to be low ... but I beleave a training course would help ... definitely thinking about it 19:31:10 <luzC> +1 19:31:23 <catherineD> Rockyg: +++ about not reading doc 19:31:29 <catherineD> :-) 19:32:02 <Rockyg> Might consider some youtube videos. then you only have to do it once. We should talk with hogepodge and foundation about what would be useful 19:32:33 <sslypushenko> Rockyg: nobody read the docs) 19:32:50 <Rockyg> You know, it could also be a good business for a devops/tester type to consult on certs.... 19:33:03 <catherineD> so seems like action plan would be 1) consider adding FAQ on RefStack 2) Discuss this with DefCore 3) RefStack training class at summit 19:33:20 <andrey-mp> agreed 19:33:38 <catherineD> Rockyg: I did refer people to the video of Austin summmit .. 19:33:50 <Rockyg> ++ 19:34:11 <catherineD> but this time in Barcelona may be we will have a hands on class and not a speaker session? 19:34:29 <catherineD> but hands on class won't have video 19:35:34 <catherineD> #agreed 1) consider adding FAQ on RefStack 2) Discuss this with DefCore 3) RefStack training class/speaker session at Barcelona summit 19:35:45 <Rockyg> Could record it. And most of the hands on happens before the summit, so could get more time. 19:36:13 <Rockyg> and agree with all of what you just wrote... 19:36:27 <catherineD> great .. good discussion ... any thing else? 19:36:51 <catherineD> hear nothing ,,, moving on 19:36:58 <catherineD> #topic Blueprint 19:37:14 <andrey-mp> about asking in empty RefStack channel? 19:37:40 <andrey-mp> will we try to let them know about openstack-dev mailing list? 19:37:49 <Rockyg> Maybe we should have a title that says email to ....? 19:38:03 <andrey-mp> or is there any other mailing list that I can monitor? 19:38:17 <catherineD> andrey-mp: good idea ... no we use openstack-dev 19:38:41 <Rockyg> We don't use the dev mailing list. Maybe we should use the community ml instead? 19:38:47 <catherineD> BTW could everyone also monitor refstack IRC and give answer ... 19:39:01 <Rockyg> I think it's user-committee 19:39:09 <catherineD> andrey-mp: sslypushenko: your time zone may match China's better 19:39:24 <Rockyg> Or we could say to use openstack It's the general one... 19:40:00 <andrey-mp> I can check IRC channel once at hour - but it's very rare to answering... 19:40:14 <Rockyg> Should ask hogepodge which one he thinks would be more likely for a certifier to be subscribed to 19:40:17 <sslypushenko> catherineD: I'll try to help 19:40:20 <catherineD> andrey-mp: yea I know 19:40:36 <catherineD> sslypushenko: andrey-mp: thx 19:40:42 * hogepodge reads backlog 19:40:42 <Rockyg> there are bots that will send you an email if someone types in a channel... 19:40:59 <andrey-mp> People can ask in mailing list about time of discussion - and we will meet ) 19:41:03 <sslypushenko> andrey-mp: try irccloud.com... it is really helpful tool 19:41:20 <andrey-mp> sslypushenko: thanks, will try 19:41:31 <catherineD> sslypushenko: me too will try 19:41:55 <catherineD> moving on ? 19:42:00 <andrey-mp> yes 19:42:04 <hogepodge> Rockyg: I hope people aren't afraid to approach me with questions. Please please let them know that I strive to be an equal representative for developers, users, and vendors. I truly want to help 19:42:19 <catherineD> #link Implement ostestr on refstack client ( https://blueprints.launchpad.net/refstack/+spec/implement-ostestr-refstackclient ) Thanks luzC: 19:42:26 <Rockyg> I know that. It's a culture thing. 19:43:03 <catherineD> luzC: we will wait for your patch ... thanks for joining us .. 19:44:01 <luzC> yes, I'm submitting it later today :) 19:45:26 <catherineD> hogepodge: any other thoughts on how to help this user Xuchao ? 19:45:33 <catherineD> luzC: thx 19:46:02 <catherineD> #topic Open reviews 19:46:53 <catherineD> andrey-mp: thx for --> Add 'cancel' registration feature ( https://review.openstack.org/#/c/329329/ ) 19:47:19 <catherineD> question: Should we update the spec? So far, we use the spec as document for API.. 19:48:50 <hogepodge> catherineD: sorry, lots of side things going on. I'm trying to help him, and sent a private e-mail. My best suggestion is to resend a new patch 19:49:06 <pvaneck> I don't think it would hurt. Probably should if you want to keep an accurate record of the API for when/if API docs are created 19:49:10 <catherineD> hogepodge: Thanks 19:49:10 <hogepodge> since he abandoned the changes I made that would have fixed his problem. 19:49:45 <hogepodge> (which isn't a problem anyway, since the tests are flagged... I was trying to guide him through the technical aspects of the patch as a new contributor) 19:49:55 <catherineD> pvaneck: great .. I think we should ... I will submit a spec update 19:49:57 <andrey-mp> spec are good to discuss something before implement... but they are not intended to add something. Nova team uses additional spec but it is very ugly... 19:50:31 <sslypushenko> catherineD: andrey-mp +1 for spec update 19:50:56 <Rockyg> +1 for spec update. 19:51:03 <catherineD> andrey-mp: sslypushenko: Let discuss more next week .. I do want to keep accurate info of our API 19:51:12 <sslypushenko> we can consider specs like a documentation... 19:51:37 <catherineD> sslypushenko: yea ... 19:51:50 <andrey-mp> if we treat spec as documentation - it's ok for me to update them 19:51:50 <Rockyg> sslypushenko, ++ Less confusing to update than add a new spec for each change 19:52:52 <catherineD> andrey-mp: that is how we use the API spec so far ... but we may want to start thinking about document the API outside of spec ... will discuss next week 19:53:22 <catherineD> meanwhile please review https://review.openstack.org/#/c/328780/ and https://review.openstack.org/#/c/329329/ 19:53:31 <Rockyg> catherineD, Docs group is working on getting a process for documenting api 19:53:46 <Rockyg> s in place. It's close but not wuite there yet 19:54:01 <andrey-mp> 328780 - merged 19:54:27 <catherineD> we want to have the UI complete before the DefCore midcycle so that we can show DefCore ( especially hogepodge: ) the UI before putting it on line 19:54:48 <pvaneck> andrey-mp, thanks! 19:54:54 <catherineD> andrey-mp: super speed :-) 19:55:30 <andrey-mp> I've started before meeting ) 19:55:33 <catherineD> #topic Open discussion 19:55:54 <andrey-mp> what is the next UI change will be? 19:55:58 <hogepodge> I was wondering if refstack would be a good place to add a javascript-based configuration generator for interop testing 19:56:32 <catherineD> andrey-mp: we can add product or add more fature to vendor 19:56:33 <hogepodge> so a user could go to a page, enter information about their cloud, and get a basic tempest.conf and accounts.yaml file that they can start from based on our experiences and talks 19:56:45 <catherineD> which one do we prefer ? pvaneck: andrey-mp: 19:57:07 <andrey-mp> hogepodge: good feature for product(cloud) page 19:58:40 <hogepodge> mmm, no. Just something to get the basics for non-admin single account testing 19:58:48 <hogepodge> help with defcore tempest config 19:59:43 <Rockyg> hogepodge, do you use the clouds.yaml file at all? 19:59:57 <catherineD> hogepodge: I will add the tempest conf topic for next week ... since we are about out of time 20:00:22 <Rockyg> Some devvs/infra think that would be a good file to have sent for cloud config info... 20:00:26 <catherineD> need to end the meeting .. more discussion next week 20:00:32 <catherineD> thx everyone .. 20:00:37 <catherineD> #endmeeting