14:00:54 <vgridnev> #startmeeting sahara
14:00:54 <openstack> Meeting started Thu Jun 16 14:00:54 2016 UTC and is due to finish in 60 minutes.  The chair is vgridnev. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:56 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:58 <openstack> The meeting name has been set to 'sahara'
14:01:02 <elmiko> hi
14:01:05 <vgridnev> #link https://wiki.openstack.org/wiki/Meetings/SaharaAgenda
14:01:24 <tosky> hi
14:01:28 <vgridnev> hi
14:01:35 <egafford> o/
14:02:11 <mionkin> hi!
14:02:38 <vgridnev> #topic News / updates
14:03:04 <egafford> o/ tellesnobrega!
14:03:06 <tellesnobrega> o;
14:03:10 <tellesnobrega> o/
14:03:10 <elmiko> sadly, not much from me, i am working on a presentation for an upcoming conference
14:03:20 <mionkin> I'm working on adding CDH 5.7 in sahara-elements and in sahara
14:04:06 <egafford> I have a validation script for Ambari that's in pretty decent shape; working on getting the image generation script and remote working now that folks seem to be reasonably happy with the spec.
14:04:42 * egafford is probably doing the usual trying-to-get-too-much-of-the-feature-done-before-posting-any-of-the-feature thing. There will be code by early next week, even if it's a little drafty.
14:04:45 <tosky> refining jobs for saha-ra-tests
14:05:19 <elmiko> saha-ra, sounds like a character from one of the mummie movies
14:05:20 <vgridnev> not so much from me too, I was unable start testing my impl for plugins-api, so now I'm squashing our bugs and making improvements on our gate jobs.
14:05:46 <vgridnev> so, we have then <= 70 open bugs
14:06:01 <egafford> vgridnev: That's better than >= 70 open bugs for sure.
14:06:04 <tellesnobrega> i'm starting to take a look on storm 1.0.0 so I can bring the newer version to sahara and I will also start a spec to allow python jobs to be run in sahara+storm
14:06:11 <egafford> Should we have a bug day/week?
14:06:27 <egafford> tellesnobrega: \o/
14:06:42 <vgridnev> egafford, I guess that we can do that after around of end of N3 milestone
14:07:35 <tosky> elmiko: Saha-ra, Princess of Power
14:07:50 <elmiko> hehe, nice!
14:07:59 <vgridnev> esikachev, are you around?
14:08:07 <egafford> vgridnev: Sure; maybe if we have a ridiculous number it'd be good to do one earlier, but that sounds good.
14:08:10 <tosky> vgridnev: good thing; are those 70 all bugs including wishlists?
14:08:43 <vgridnev> tosky, that includes incomplete's and wishlists
14:10:07 <vgridnev> do we have other news?
14:10:54 <vgridnev> #topic Review priorities
14:11:01 <vgridnev> #link https://etherpad.openstack.org/p/sahara-review-priorities
14:11:26 <vgridnev> so, I think that there is a good change to discuss
14:11:29 <vgridnev> #link https://review.openstack.org/#/c/326582/
14:11:38 <egafford> Yeah, agreed with that.
14:13:01 <vgridnev> so, egafford, seems like it's ok to put Apache 2.0 license on MIT licensed code, and required copyright was added.
14:13:05 <elmiko> theoretically, MIT is compatible with APL
14:13:44 <nanhai> hello?
14:13:50 <elmiko> hi
14:13:54 <egafford> elmiko, vgridnev: Ack; if so, my blocking objection is withdrawn.
14:13:54 <egafford> Still, I suppose one question I'd ask is: do we truly need to use this functionality? If other projects aren't using sphinx-argparse and they're generating doc successfully, and sphinx-argparse is broken, is there a better way to address this problem?
14:14:08 <tosky> vgridnev: so basically, will this change reverted if the new requirement lands in time for Newton?
14:14:18 <elmiko> egafford: i think that's an excellent question
14:14:40 <elmiko> should we add a DocImpact to that review and try to get some guidance from the doc team?
14:15:02 <vgridnev> tosky, I guess that we it should be reverted if we will get that in globals
14:16:02 <egafford> I'm okay with the change as it is, mind you: I don't think it's the end of the world to copy a module we can legally copy. elmiko: I like bringing in a doc team eyeball or two a bit, yeah. If they don't have time soon enough we can push this change.
14:16:09 <vgridnev> egafford, no other projects are using this lib. We are using to bring simple docs auto-generation
14:17:57 <vgridnev> other projects not using that for generation docs, seems like them are mostly hand-made
14:18:12 <nanhai> sphinx-argparse?
14:18:17 <nanhai> auto generate?
14:18:45 <egafford> vgridnev: Hm; well, I'll admit that I don't like handmade and I do like auto-generation. :)
14:18:53 <vgridnev> https://raw.githubusercontent.com/openstack/python-openstackclient/master/doc/source/command-objects/address-scope.rst
14:19:00 <vgridnev> http://docs.openstack.org/developer/python-openstackclient/command-objects/address-scope.html#address-scope-create
14:19:09 <vgridnev> just an example
14:19:33 <egafford> Yeah, we don't want that for sure.
14:20:26 <vgridnev> nanhai, do you a specific questions?
14:20:54 <nanhai> no, i just listenning
14:20:56 <egafford> vgridnev: Changed my -1 to a +1 with a note that we might want to ask the doc team if they had specific suggestions about doc autogeneration.
14:21:03 <nanhai> i am listenning
14:21:41 <vgridnev> egafford, I ll put DocImpact now
14:22:06 <egafford> Cool; if we don't hear back at all by next meeting let's just do this and ask later.
14:22:23 <elmiko> it may also be prudent to reach out to the doc team, just a heads up
14:22:36 <egafford> elmiko: Ack.
14:22:56 <vgridnev> ok, we can put that to ML and ask OpenStack client team
14:22:58 <vgridnev> too
14:23:34 <egafford> vgridnev: Yeah, I can't imagine other teams are uninterested in autogenerating docs.
14:24:00 <vgridnev> oops
14:24:06 <egafford> vgridnev: :)
14:24:39 * egafford wrote 'vgridnev: Yeah, I can't imagine other teams are uninterested in autogenerating docs.' in your extremely brief absence.
14:25:51 <vgridnev> ok
14:26:39 <egafford> I think the other high-prio reviews just need a few more eyes; they look pretty good.
14:26:39 <vgridnev> #topic API v2 progress
14:26:50 <elmiko> not much to report since last week
14:26:57 <elmiko> i don't think there has been any movement
14:27:32 <vgridnev> ok then
14:27:35 <elmiko> it's possible, that i might have a little time after this conference to get back into things. we'll see
14:27:46 <elmiko> i'll definitely keep the team up to date
14:27:56 <vgridnev> ok, thanks elmiko
14:28:01 <vgridnev> #topic Specs / features status
14:28:23 <vgridnev> so, there are two specs on review now
14:28:37 <vgridnev> #topic Initial kerberos integration
14:28:37 <egafford> I'll read the kerb spec today.
14:28:51 <elmiko> yea, me too. i need to read that spec
14:29:06 <vgridnev> ok, super short topic then
14:29:23 <vgridnev> #link https://review.openstack.org/#/c/327616/
14:29:27 <tellesnobrega> i started looking into it, but still haven't been through, will finish it today
14:29:37 <vgridnev> #topic CLI for Plugin-Declared Image Declaration
14:29:45 <vgridnev> #link https://review.openstack.org/#/c/327111/
14:30:04 <vgridnev> egafford, seems like my questions was addressed
14:30:23 <egafford> Responded to your thoughts and comments (which were good ones), esp. vgridnev's question about deps. I think externalizing the libguestfs dep in a separate module for packagers is definitely a good idea.
14:30:34 <egafford> It gives them a lot of options.
14:31:10 <egafford> vgridnev: Yeah, happily I designed the last feature while mostly thinking about this feature. :)
14:31:21 <egafford> So the spec mostly wrote itself.
14:32:16 <vgridnev> I did not got that: Yeah, happily I designed the last feature while mostly thinking about this feature. :)
14:32:47 <egafford> Oh, just that when I was writing the validation spec last cycle, I was mainly thinking about how to implement this on top of it.
14:33:16 <vgridnev> ok, thanks
14:33:21 <egafford> So it seems like that worked and people think the spec fits pretty well into the overall effort, which was the goal. That's all.
14:33:56 <vgridnev> ok, I'm pushing that
14:34:01 <egafford> vgridnev: \o/
14:34:19 <vgridnev> cool
14:34:30 <vgridnev> #topic Open discussion
14:34:35 <egafford> (Some) code incoming by next meeting.
14:35:39 <elmiko> vgridnev: added a few questions to the kerberos spec
14:36:51 <tellesnobrega> I discussed with vgridnev this week about storm upgrade, we came to an agreement that no spec is needed for that, does that make sense to all? how about python jobs for storm, spec is needed for that?
14:37:03 <esikachev> saharaclient functional tests moved to sahara-tests
14:37:14 <esikachev> job migration - in progress
14:37:16 <egafford> tellesnobrega: Python jobs seems worth a spec probably.
14:37:27 <vgridnev> tellesnobrega, have you create blueprints for that?
14:37:42 <tellesnobrega> sure egafford
14:37:50 <tellesnobrega> vgridnev, not yet, will do it soon
14:37:50 <egafford> We haven't dealt with storing binaries for an uncompiled, uncompressed language, for one thing.
14:38:06 <egafford> It can probably be a short spec.
14:39:07 <vgridnev> tellesnobrega, I think if spec will be there till the next meeting, we can review that quickly and push that at the meeting
14:39:40 <tellesnobrega> vgridnev, awesome, i will try to put that up to review asap
14:42:13 <vgridnev> elmiko, answered your questions on review
14:42:39 <vgridnev> will put short note about first question in spec
14:42:48 <elmiko> cool, thanks
14:44:02 <egafford> vgridnev: Both Ambari and CDH have built-in KDCs that we can use then?
14:44:16 <egafford> (Without installing any additional packages in SIE?)
14:44:53 <elmiko> vgridnev: makes sense, i think we are going to have issues with the proxy users feature and using an existing kdc
14:45:03 <elmiko> or any kdc
14:45:11 <elmiko> so, something to keep in mind
14:45:23 <vgridnev> additional packages are needed, as I remember
14:45:40 <egafford> vgridnev: Makes sense.
14:46:13 <vgridnev> but we will put that on images, I think. If not presented (I don't know, because of old openstack) we can install that
14:47:49 <vgridnev> elmiko, I hope no, proxy users are needed to access swift, not for hadoop services
14:48:35 <elmiko> ok, let's hope not ;)
14:49:01 <vgridnev> for example, YARN or Oozie will his own key tabs to access other services like HDFS
14:49:27 <elmiko> yeah, it makes sense
14:50:36 <vgridnev> ok, any other topics to discuss?
14:51:25 <tellesnobrega> i got nothing
14:51:35 <egafford> Nothing here.
14:51:45 <vgridnev> #endmeeting