14:00:13 <SergeyLukjanov> #startmeeting sahara 14:00:13 <huichun> hello 14:00:13 <openstack> Meeting started Thu Feb 11 14:00:13 2016 UTC and is due to finish in 60 minutes. The chair is SergeyLukjanov. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:14 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:17 <openstack> The meeting name has been set to 'sahara' 14:00:17 <esikachev> hi! 14:00:19 <SergeyLukjanov> #chair elmiko 14:00:21 <openstack> Current chairs: SergeyLukjanov elmiko 14:00:24 <elmiko> o/ 14:00:32 <SergeyLukjanov> (backup) 14:00:36 <SergeyLukjanov> #link https://wiki.openstack.org/wiki/Meetings/SaharaAgenda 14:00:36 <tosky> hi 14:00:38 <elmiko> good idea ;) 14:00:47 <NikitaKonovalov> o/ 14:00:52 <mionkin> hello 14:00:54 <vgridnev> o/ 14:00:58 <AndreyPavlov> hi 14:01:04 <SergeyLukjanov> #topic News / updates 14:01:42 <_crobertsrh> Still looking for reviews on the UI reorg. https://review.openstack.org/#/c/270478/ will get you all the changes to try out. 14:01:49 <elmiko> i've been working on api v2, and looking at more security bugs and improvements on our bandit gate 14:02:00 <AndreyPavlov> i was working on periodic tasks 14:02:10 <_crobertsrh> Ideally, I'd like to get reviews soon because rebasing them every 20 min for 3 months is not my idea of a fun springtime. 14:02:31 <vgridnev> Working on health mostly, there are few changes on review already: https://review.openstack.org/#/q/status:open++branch:master+topic:bp/cluster-verification 14:02:55 <_crobertsrh> vgridnev: cool, I can't wait to see that in action 14:03:55 <tosky> and the feature freeze is in... 18 days? 14:03:59 <sreshetn1ak> o/ 14:04:18 <SergeyLukjanov> NikitaKonovalov could you please review the UI reorg to land it faster? 14:04:21 <tmckay> hi folks! 14:04:27 <NikitaKonovalov> SergeyLukjanov: yes sure 14:04:43 <tmckay> https://review.openstack.org/#/q/owner:tmckay%2540redhat.com+status:open 14:04:53 <tmckay> please review :) ^^ some are old 14:05:35 <tmckay> also, we turned off the mapr gate test in sahara-ci, I have built a mapr image and plan today to figure out why it fails (something wrong in apt-get for oozie) 14:06:13 <SergeyLukjanov> tmckay ack, thx 14:06:15 <tmckay> 2 more patches for regex searching (job_execution, which has special stuff, and job_binary_internal, which hardly matters I think :) ) 14:06:31 <tmckay> also, in open discussion, I have an idea for a regex search tweak 14:07:11 <tmckay> And, I fixed #1453953 for job binary internal upload from horizon 14:07:23 <tmckay> https://review.openstack.org/#/c/278027/ 14:07:41 <tmckay> This was tried and abandoned a few months ago, but I hit it, so I decided to try to fix it :) 14:07:58 <crobertsrh> wtg tmckay 14:07:59 <tmckay> simple fix, and only the one case that was breaking for me 14:08:36 <tmckay> those are all my updates :) nothing left to say 14:09:43 <SergeyLukjanov> #topic API v2 progress 14:09:49 <SergeyLukjanov> #link https://review.openstack.org/#/c/273316/ 14:09:49 <SergeyLukjanov> #link https://wiki.openstack.org/wiki/Sahara/api-v2 14:09:55 <elmiko> thanks SergeyLukjanov 14:10:04 <elmiko> just a quick update, the initial commit is ready for reviews 14:10:12 <elmiko> and i am adding more content to the wiki page 14:10:26 <elmiko> once the initial commit is merged, i'll have more to follow. 14:10:43 <elmiko> and hopefully there will be a clear workload that others can join in on =) 14:11:13 <elmiko> that's all from me, unless there are questions about api v2 14:12:49 <SergeyLukjanov> #topic Open discussion 14:13:02 <SergeyLukjanov> we have a short agenda for today :) 14:13:09 <SergeyLukjanov> elmiko thx for the updates on v2 14:13:49 <NikitaKonovalov> good news, voting for talks is now open 14:13:53 <elmiko> \o/ 14:14:00 <elmiko> should we all throw up links now? ;) 14:14:12 <NikitaKonovalov> why not 14:14:18 <elmiko> https://www.openstack.org/summit/austin-2016/vote-for-speakers/presentation/7797 14:14:19 <crobertsrh> heh 14:14:25 <elmiko> =D 14:14:32 <huichun> vgridnev: I have updated suspend EDP job patch according to your reviews https://review.openstack.org/#/c/201448/ 14:14:58 <vgridnev> ok will do review later 14:15:05 <huichun> vgridnev: add engine support and cluster check before running job 14:15:28 <SergeyLukjanov> Deploy OpenStack as containers using Marathon and Mesos -- https://www.openstack.org/summit/austin-2016/vote-for-speakers/Presentation/8113 14:15:29 <SergeyLukjanov> Data Processing in OpenStack: The State of the Art in Sahara -- https://www.openstack.org/summit/austin-2016/vote-for-speakers/Presentation/8228 14:15:29 <SergeyLukjanov> Architecture and Best Practices to Deploy Hadoop and Spark Clusters with Sahara -- https://www.openstack.org/summit/austin-2016/vote-for-speakers/Presentation/8683 14:16:05 <SergeyLukjanov> would appreciate votes :) 14:16:30 <elmiko> SergeyLukjanov: did you see this one, https://www.openstack.org/summit/austin-2016/vote-for-speakers/presentation/6826 14:16:44 <elmiko> curious if they are using sahara at all 14:16:59 <NikitaKonovalov> I also participate in https://www.openstack.org/summit/austin-2016/vote-for-speakers/Presentation/8683 and https://www.openstack.org/summit/austin-2016/vote-for-speakers/Presentation/8687 14:17:13 <elmiko> i'm guessing that they are not, but i'm curious none-the-less 14:17:20 <SergeyLukjanov> elmiko no, thx for the link 14:18:34 <tmckay> okay, so I have an additional idea for regex searching, not sure if it's good or not 14:19:22 <tmckay> As you know, we have some fields with regex (name, description, url for example) and some exact match (uuids, version numbers, job types, basically ids and things with defined values) 14:19:29 <SergeyLukjanov> elmiko I think they are doing bare metal provisioning of mesos or k8s 14:19:39 <SergeyLukjanov> most probably k8s 14:20:06 <tmckay> but, I was playing yesterday and thought that maybe an OR function on something exact might be nice. Like job types "pig|shell" 14:20:06 <tosky> I have two questions related to sahara-scenario: a) I saw there is a pending request to add a the project sahara-tests to "Sahara Official", is it sahara-scenario that needs to be renamed? 14:20:12 <elmiko> SergeyLukjanov: ack, thanks 14:20:18 <tmckay> What if you want exact match on more than one thing? 14:20:36 <tmckay> So, what do you think about allowing "|" in exact match cases? 14:21:05 <elmiko> tmckay: would we be passing the "|" through the rest api? 14:21:22 <tmckay> yeah. I tried it yesterday from horizon 14:22:02 <elmiko> i'm only asking because there has been work in the api-wg on creating a guideline for more complex search queries, i'm trying to find the review now 14:22:23 <tmckay> the logic would be "if an exact match field contains an "|", then allow regex but treat the stuff between the |'s as exact matches" 14:22:34 <tmckay> elmiko, ok. Maybe defer then 14:23:07 <tmckay> The current logic is that in general, you do not want to do regex on things like version number (but you might want to search for more than one) 14:23:44 <tmckay> of course, if the spec covers comparators, then you could treat version numbers as actually numbers and not strings, and do ranges like in requirements 14:23:56 <elmiko> i think the prevailing wisdom from the api gurus, was to separate all search terms and pass them using specific args params in the rest 14:24:37 <SergeyLukjanov> tosky rename will be tomorrow 14:24:38 <elmiko> tmckay: my only concern would be us building this network of regex-yes/regex-no search fields. it could make it tough for end-users to know when they can use which types of search. 14:24:41 <tmckay> ah, I see. so you may have a field appear multiple times 14:24:44 <SergeyLukjanov> esikachev ^^ (rename) 14:24:51 <tosky> SergeyLukjanov: ah, good to know, thanks 14:25:38 <tmckay> elmiko, yes. so far what we have done matchines nova. some fields are exact, others are regex. Fields will be listed in user docs somewhere (haven't done that yet) 14:25:44 <tosky> now, second question (input appreciated from you all): as scenario-tests is branchless like tempest, I was thinking that we could move API tempest tests there, outside tempest repository 14:25:48 <tosky> what do you think? 14:26:12 <tosky> if you think it's worth of, I guess I would need a spec 14:26:14 <tmckay> oh, by the way, there are pep8 errors on tempest tests in sahara on master right now 14:26:26 <tmckay> I keep seeing them when I check changes. Why is that? 14:26:41 <elmiko> tosky: not a bad idea, imo. i'm curious how we will layer in the v2 tests, is there a separate directory or config for each api version? 14:27:00 <tosky> elmiko: I don't know, I will check the other projects with multiple APIs 14:27:13 <tosky> but I suspect it's a solved problem somewhere else 14:27:15 <elmiko> tmckay: on the question of the expanding the regex, i think it's a cool idea but maybe we should generate a spec to fight it out over? 14:27:46 <elmiko> tosky: yea, most likely it is solved. i'm just curious. in general though, i don't have an objection to moving those tests into the scenario repo. 14:27:53 <tmckay> elmiko, yeah, just looking for initial feedback. If everyone said "No, terrible" I wouldn't bother :) 14:28:00 <elmiko> tmckay: ack 14:28:31 <tmckay> tosky, ++ I am all for as much local control in the sahara repos as possible 14:28:36 <tosky> esikachev: what would you think about moving tempest API tests into sahara-tests (using the tempest plugin mechanism of course)? 14:29:40 <SergeyLukjanov> tosky I'm not sure that it'll be working correctly with current way how jobs done 14:29:59 <tosky> SergeyLukjanov: do you mean the jenkins jobs? 14:31:18 <esikachev> tosky: tempest tests of sahara is a plugin for tempest not good move them to new repo 14:31:48 <tmckay> oh, crobertsrh, I had a thought. I think the nova panels have a "*" next to the search items that are regex, and no "*" next to the ones that are exact 14:31:53 <tosky> esikachev: why not? The alternative is move it into sahara repository, but then it's going to be not branchless 14:31:59 <tmckay> what do you think of that? 14:32:15 <crobertsrh> I'm all for being consistent with other services 14:32:16 <tmckay> So it looks like "Name *" on the filter list 14:32:23 <tosky> esikachev: the new repo is perfect, as it follows the same life of tempest; if you don't need to run the tests, you just don't run them 14:32:31 <SergeyLukjanov> tosky yup, jenkins job 14:32:34 <tmckay> I will double check that that's actually true, but I'm pretty sure I saw it 14:32:46 <elmiko> tmckay, crobertsrh, does horizon have the ability to create those little question mark helper links? (with a hover over explaining what is going on) 14:32:57 <tosky> SergeyLukjanov: they can be easily adapted, we already do it for python-saharaclient tempest (scenario) tests, which are in python-saharaclient repo 14:33:17 <tmckay> that would be cool 14:33:22 <elmiko> regardless of what nova is doing, i think having a hover-over with an explanation would be better 14:33:37 <tmckay> elmiko, ++ 14:33:56 <crobertsrh> Might be useful as well 14:34:02 <SergeyLukjanov> tosky if it could be fixed, I would say, I like idea of having tempest in sahara-tests 14:34:29 <crobertsrh> elmiko: We can do pretty much anything you can imagine. 14:34:42 <crobertsrh> Whether or not it "fits" is another issue. 14:35:18 <tosky> SergeyLukjanov: as vgridnev remembers from the last review, even tempest core suggested moving API tests outside tempest core 14:35:38 <esikachev> SergeyLukjanov: tosky let's do it? 14:35:44 <elmiko> crobertsrh: i thought maybe the horizon folks had already implemented something like this 14:36:07 <tosky> esikachev, SergeyLukjanov: do we need a spec, right? 14:36:14 <tmckay> crobertsrh, I wouldn't pioneer tooltips unless they are already there. Or, maybe, if you want to be a trailblazer :) 14:36:15 <crobertsrh> Oh, they have "?" with hover that appear in places. Not sure I've seen it on a table page yet though. 14:36:18 <tosky> and would it be a candidate for a feature exception? 14:36:29 <tosky> (as time is running) 14:36:46 <SergeyLukjanov> tosky spec is good 14:36:53 <tosky> okidoki 14:39:55 <SergeyLukjanov> anything else to discuss today? 14:40:44 <elmiko> nothing from me 14:40:46 <tmckay> nothing from me 14:44:53 <elmiko> 3 14:44:56 <elmiko> 2 14:45:00 <elmiko> 1 14:45:08 <elmiko> #endmeeting