18:01:08 <amrith> #startmeeting trove 18:01:09 <openstack> Meeting started Wed May 11 18:01:08 2016 UTC and is due to finish in 60 minutes. The chair is amrith. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:01:10 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 18:01:12 <openstack> The meeting name has been set to 'trove' 18:01:16 <flaper87> o/ 18:01:21 <amrith> ./ 18:01:22 <twm2016> o/ 18:01:59 <amrith> let's give folks a couple of minutes to trickle in 18:02:05 <tellesnobrega> o/ 18:02:13 <dloi> o/ 18:02:16 <vkmc> o/ 18:02:21 <amrith> #agenda https://wiki.openstack.org/wiki/Meetings/TroveMeeting 18:02:35 <peterstac> `o/o/ 18:02:41 <amrith> that's new 18:02:48 <johnma> o/ 18:02:50 <pmalik> *\o/* 18:02:54 <mvandijk> ./ 18:03:20 <amrith> ok, sounds like a quorum to me, let's get started. 18:03:31 <amrith> #topic Action items from last week's meeting 18:03:39 <amrith> This would have been the meeting a couple of weeks ago 18:03:49 <amrith> I didn't find any. does anyone remember anything as action items? 18:04:13 <amrith> this was the meeting notes from April 20th 18:04:15 <amrith> #link http://eavesdrop.openstack.org/meetings/trove/2016/trove.2016-04-20-18.01.txt 18:05:09 <amrith> I guess not, so if no one has anything from the old meeting, we'll move along. 18:05:28 <amrith> #topic Trove pulse update 18:05:36 <amrith> #link http://bit.ly/1VQyg00 18:06:36 <amrith> the backlog of reviews (queue growht) is certainly related to the summit 18:06:49 <amrith> but we do have a number of outstanding reviews, some for over a couple of weeks. 18:07:22 <amrith> like https://review.openstack.org/#/c/306430/ 18:07:31 <amrith> which I'm sure pmackinn would like to see merged 18:07:47 <amrith> and https://review.openstack.org/#/c/300009/ 18:07:56 <cp16net> howdy 18:08:00 <amrith> both of which would be good to backport to mitaka quickly 18:08:18 <flaper87> These 2 reviews seem ready to go 18:08:18 <amrith> so any available cores who can review and approve these (and the dozen or so other changes that need a +1) 18:08:38 <flaper87> amrith: quick question re 300009. Any reason why it was not approved after 2 +2s ? 18:09:01 <amrith> flaper87, the general rule we've followed in trove is that we won't +1 a change with two +2's from the same company 18:09:01 <peterstac> I count almost 20 review that are looking for a second +2/WF+1 18:09:04 <amrith> rax respected that 18:09:07 <amrith> hp respects that 18:09:11 <amrith> we (tesora) do the same 18:09:18 <amrith> in only very specific cases 18:09:18 <flaper87> amrith: gotcha, ok! Makes sense 18:09:21 <amrith> like a broken build 18:09:38 <amrith> or something (like peterstac did yesterday) to fix the tests and reduce the mock recursion depth 18:09:41 <amrith> do we break that rule. 18:09:53 <amrith> by agreement, we decided to waive that rule for some changes 18:09:59 <amrith> those proposed by the proposal bot 18:10:06 <flaper87> yeah, I failed to notice the pattern. We used to have that in Glance too. 18:10:15 <amrith> otherwise, it is a good practice to have a wider review audience 18:10:38 <amrith> I assume that there will be situations where the build is broken or something expeditious needs to be done, and a single core can do that. 18:10:44 <cp16net> +1 18:10:45 <amrith> but these changes require some review. 18:10:49 <flaper87> +1 18:10:58 <amrith> hence you'll see many where peter and I have +2'ed but someone else must +1 18:11:11 <amrith> which is what peterstac was alluding to 18:11:14 <amrith> as 20+ 18:11:16 <amrith> #link https://review.openstack.org/#/q/NOT+label:Workflow%253E%253D1+NOT+label:Workflow%253C%253D-1+label:Verified%253E%253D1%252Cjenkins+NOT+owner:self+label:Code-Review%253E%253D2+NOT+label:Code-Review-1+is:mergeable+status:open++%28project:openstack/trove+OR+project:openstack/trove-integration+OR+project:openstack/trove-specs+OR+project:openstack/python-troveclient+OR+project:openstack/trove-dashboard%29 18:11:39 <amrith> for example, I think https://review.openstack.org/#/c/313648/ should be quick :) 18:12:08 <amrith> anyway, this is back to the same issue we've dealt with for some time 18:12:22 <amrith> #link https://gist.github.com/amrith/70b8913df8109c2ef40b0d08338f6206 18:12:28 <amrith> that's the review counts 18:12:39 <amrith> and we should try and pick up the pace ... 18:12:48 <amrith> I have nothing further to add on this 18:12:57 <amrith> hi cp16net, sorry didn't see you come in 18:13:22 <cp16net> howdy 18:13:32 <flaper87> yup, noticed the review rate! Happy to help there as much as possible 18:13:37 <amrith> anything further to add re: trove pulse ... 18:13:42 * flaper87 has nothing 18:14:12 <amrith> #topic Announcements 18:14:23 <amrith> we've discussed these things already 18:14:29 <amrith> I've updated the newton release calendar 18:14:36 <amrith> #link http://releases.openstack.org/newton/schedule.html 18:14:40 <amrith> with trove specific dates 18:15:12 <amrith> The next deadline we have is: Deadline for proposing specs for Newton is R-17 (Week of June 13 to 17) 18:15:25 * flaper87 writes that down in his calendar 18:15:30 <amrith> and the spec freeze comes after that Deadline for specs for Newton to be merged is R-12 (Week of July 11 to 15) 18:16:06 <amrith> any questions or comments on this ... 18:16:30 <flaper87> none from me other that I think having a spec freeze is a good thing 18:16:36 * flaper87 shuts up 18:17:09 <amrith> anyone have any other announcements 18:17:56 <amrith> ok, moving along then ... 18:17:59 <amrith> #topic Specs that are up for review 18:18:19 <amrith> So I made a list of specs that are up for review sometime this morning 18:18:26 <amrith> I believe that some specs came in after that 18:18:48 <amrith> #link https://review.openstack.org/315141 18:19:10 <amrith> #link https://review.openstack.org/#/c/315079/ 18:20:01 <amrith> vkmc, your spec is verified failed and is tehrefore not showing up on the dashboard. I believe johnma has the reason for it in her comments. 18:20:12 <amrith> if you'd update it, it'll pop-up on people's dashbaords 18:20:15 <vkmc> amrith, yeah, amending soon 18:20:16 <amrith> (if they use the dashboard) 18:20:22 * flaper87 hasn't reviewed these last 2 specs 18:20:30 <amrith> any discussion of the specs that are up for review 18:20:45 <amrith> pmackinn, yt? 18:20:55 <pmackinn> no 18:21:05 <pmackinn> 'sup? 18:21:10 <amrith> or maybe vkmc (since pmackinn isn't here, he says), would you respond to my comments in https://review.openstack.org/#/c/315141/ 18:21:14 <tellesnobrega> im sending a new patch for review on the cinder backups this week hopefully, so that will something to review 18:21:20 <amrith> the BP is in trove-integration 18:21:30 <amrith> which I'm hoping we can kill soon(ish) 18:21:39 <amrith> do you want to register the BP in some other project? 18:22:09 <pmackinn> amrith, re-targeting to the new T-I-B 18:22:20 <pmackinn> amrith, do we have a repo? 18:22:20 <amrith> pmackinn, are you able to enter bp's there? 18:22:29 <amrith> if not I may have to fix lp permissions 18:22:32 <amrith> pl lmk 18:22:40 <amrith> lp project is trove-image-builder 18:22:48 <amrith> re: repo, there was this whole flap about the name 18:22:49 <pmackinn> https://blueprints.launchpad.net/trove-image-builder/+spec/trove-image-builder 18:22:52 <vkmc> amrith, I'll review last update on 315141 soon 18:22:58 <vkmc> and reply to your comments 18:23:12 <amrith> if you are fine with the name I'll add the bp link to my infra changes and WF+0 them 18:23:42 <amrith> https://review.openstack.org/#/c/312805/ and https://review.openstack.org/#/c/312806/ 18:23:46 <amrith> vkmc, thx 18:23:52 <vkmc> np 18:23:54 <pmackinn> amrith, re: comments we still have agreement in principle right? no need to get on the merry-go-round more at this point i would hope 18:24:29 <amrith> yes we have agreement in principle re: new repo, trove-integration dib elements find a new home, a wrapper to make it easier to use them, a new element to install the guest agent, ... 18:24:34 <amrith> I don't think I've missed anything 18:24:56 <amrith> I think the broader community will want to get involved when we start thinking about other tool chains. 18:25:10 <pmackinn> \o/ 18:25:37 <amrith> I think that is still a point of some concern in the community at large, and I would like to make sure we are congisent of that. 18:25:59 <amrith> I see your new BP 18:26:06 <amrith> I will tag that and move forward for the repo 18:26:11 <pmackinn> not sure what concerns there would be at this point in the discussion 18:26:20 <pmackinn> and, hey community now 18:26:58 <amrith> one caution re: redstack build-image which I put in the review 18:27:08 <amrith> we can talk about that offline (#openstack-trove) 18:27:15 <amrith> any other discussion re: specs 18:27:16 <pmackinn> noted 18:27:18 <amrith> thanks tellesnobrega 18:27:21 <amrith> will look for that 18:27:24 <tellesnobrega> np 18:27:42 <amrith> tellesnobrega, you also had a q: re the old spec, I'd say abandon it for now. we can restore when appropriate 18:28:00 <tellesnobrega> amrith, sounds good, will do that 18:28:11 <flaper87> +1 for abandoning it 18:28:19 <amrith> #action [amrith] update infra specs to point to new BP and WF +0 18:28:26 <amrith> any other action items here? 18:28:58 <amrith> #action [tellesnobrega] abandon spec on CEPH (don't have a number handy) 18:29:17 <amrith> anything else? 18:29:29 <tellesnobrega> https://review.openstack.org/#/c/256057/ 18:30:00 <amrith> #action [tellesnobrega] abandon spec on CEPH (https://review.openstack.org/#/c/256057/) 18:30:44 <amrith> if everyone would take a look at the specs that are up for review and post feedback that'd help us hit our deadline. 18:30:53 <amrith> anything else ... 18:31:24 <amrith> I'm going to skip the next two sections in the agenda since no one added anything in the agenda. 18:31:25 <amrith> but ... 18:31:43 <amrith> if you want to talk about proposals for review or projects for newton ... now's the chance 18:32:40 <amrith> ok, let's move along 18:32:43 <amrith> #topic Updating the API documentation 18:33:33 <amrith> I added this to the agenda after stubbing my toe on http://developer.openstack.org/api-ref-database-v1.html 18:33:47 <amrith> actually someone else did and pointed me at it. 18:33:54 <amrith> This is rather old 18:34:04 <amrith> not updated in a while 18:34:08 <amrith> inaccurate in some places 18:34:19 <amrith> so I asked other projects how they handled their API documentation 18:34:30 <amrith> and the answer is that they have it as part of the source tree 18:34:43 <amrith> and whenever a change updates the API, the document must be updated in the same commit 18:35:01 <amrith> does anyone have experience with this in another project? 18:35:10 <amrith> I'm thinking first of zaqar, glance, ... 18:35:24 <amrith> but others too ... 18:35:35 <flaper87> Yeah, we do that in both Zaqar and Glance. We're just moving Glance's WADL support to RST: https://review.openstack.org/#/c/312259/ 18:35:46 <flaper87> so, if we do this in Trove, we should probably start off with RST directly 18:36:01 <amrith> Yes, that was the recommendation from a couple of other teams 18:36:18 <amrith> how did it work out to have to update the spec as part of the same commit 18:36:22 <flaper87> As a reviewer/contributor, I appreciate having these docs in-tree since it makes it easier to keep things updated 18:36:25 <amrith> was there a lot of conflict on merges? 18:36:32 <flaper87> Not really 18:36:42 <amrith> should we structure it as a single file or multiple files for the API? 18:36:47 <amrith> (to reduce conflict) 18:36:51 <flaper87> I can't think of any blocking/annoying issue related to this 18:37:00 <flaper87> multiple files is better, IMHO 18:37:01 <amrith> cp16net, how does this tie in with the example generator? 18:37:26 <cp16net> it should reference the output examples and inject them into the doc 18:37:31 * amrith rubs the magic lamp three times and says grapex grapex grapex 18:38:04 <cp16net> since the examples are statically named they should be able to be referenced by their file name 18:38:06 <amrith> cp16net, my concern was always whether these samples were complete 18:38:32 <cp16net> yeah they probably are not complete with the new api features that were added 18:38:46 <cp16net> the existing ones should be valid tho because they are run as part of the tests 18:38:57 <cp16net> unless that was sneakily removed 18:39:09 <amrith> well, they are complete insofar as the tests exercise the API, no? 18:39:28 <amrith> i.e. it is based on test coverage not on actually parsing the code, right? 18:39:30 <cp16net> i know they were used in the wadl doc generation that was there when you built with mvn or something 18:39:37 <cp16net> but its been a long time since i saw that 18:40:11 <cp16net> right they are bascially another set of tests that run to verify the example output 18:40:36 <cp16net> and that would need to be updated for the new api calls i.e. logs, etc... 18:40:48 <peterstac> cp16net, but they're run in the same vein as the fake tests, right? 18:40:54 <cp16net> possibly even clusterings 18:40:59 <amrith> cp16net, you refer to #link http://git.openstack.org/cgit/openstack/trove/tree/generate_examples.py 18:40:59 <cp16net> yeah 18:41:13 <cp16net> they are all run with fakemode so they are faster and verify api output 18:41:21 <peterstac> so those are getting pretty stale since everything new is done in the scenario realm now 18:41:51 <cp16net> yeah and fake mode is being ignored 18:42:02 <cp16net> and not run with the scenario tests 18:42:14 <cp16net> fakemode was using the int-tests suite 18:42:18 <amrith> so, now that the API doc's are (along with other docs) coming back to projects, how do we handle this? 18:42:21 <cp16net> or rather based on it 18:42:31 <peterstac> well, the scenario tests also use the int-tests suite 18:42:57 <peterstac> do you need to do something special to get the fakes to run? 18:43:16 <peterstac> (i.e. are they hard-coded to the api ones ...) 18:43:17 <cp16net> you would need to make sure ot import the scenarios when runnign fake mode 18:43:18 <amrith> peterstac, not that I know of 18:43:28 <cp16net> and i suspect that it will take a bit of work 18:44:02 <amrith> so, one thing I will point out is that other projects don't generate examples in an automated way. 18:44:08 <amrith> they do it the 'old way' 18:44:12 <amrith> I did it once 18:44:16 <amrith> go look at schema.py 18:44:23 <amrith> then look at each entry point (mapping) 18:44:31 <amrith> and see what other undeclared parameters it uses 18:44:35 <amrith> and write them all down. 18:44:54 <amrith> once we get that in place, we'll be in a better place to require updates on commit 18:44:55 <cp16net> https://github.com/openstack/trove/blob/master/run_tests.py#L229 18:45:05 <cp16net> i believe this is where they get loaded 18:45:47 <amrith> given that the completeness of these automated samples is dependent on the test coverage, do we want to continue this path? 18:46:37 <flaper87> mmh, sounds like quite some work and based on the current planned specs/work, it might be a bit unrealistic to also add this item 18:47:49 <flaper87> it doesn't feel like a critical item to me but others may disagree 18:47:51 <amrith> so what's the priority of this activity? 18:48:27 <amrith> given that not having documentation is a problem, shouldn't it be a higher priority. maybe not critical but not ignoreable either 18:48:35 <amrith> what (if anything) do we want to commit to doing in Newton? 18:49:50 <flaper87> I agree with the sentiment it should probably be a higher priority. Not sure how much of it we would be able to do in Newton, though. I'd need to spend some extra time looking into what's required in Trove's case to get this done. 18:49:59 * flaper87 learned something from the above discussion 18:50:30 <amrith> cp16net, peterstac anything to add? 18:50:57 <amrith> I'm going to guess not. let's revisit in a couple of weeks. 18:51:01 <cp16net> nothing here 18:51:03 <amrith> #topic Open Discussion 18:51:12 <amrith> anybody have anything ... 18:51:14 <cp16net> i have one thing i'd like to bring up 18:51:15 <pmalik> This has been sitting in review with +2 from peterstac. 18:51:16 <pmalik> It does not change how and what the tests run, but it's a needed first step if we ever want 18:51:16 <pmalik> to see the new tests as voting jobs. 18:51:16 <pmalik> It only affects the test code and the tests pass... Thx. 18:51:16 <pmalik> https://review.openstack.org/#/c/309190/ 18:51:31 <amrith> cp16net, please go ahead 18:51:37 <flaper87> Is there a Trove midcycle planned for Newton? 18:51:41 <cp16net> i put up for a bug patch for mariadb clustering https://review.openstack.org/#/c/306638/ 18:51:55 <amrith> cp16net, saw that 18:52:00 <cp16net> What was eluded to in comments may be a larger issue across instance create as well for neutron networks. If someone with more knowledge of neutron would like to take a look that would be helpful. 18:52:34 <amrith> cp16net, we could start with this fix 18:52:43 <amrith> after all, there are many ways to specify a neutron nic 18:52:48 <amrith> and this would fix the common case 18:52:51 <amrith> but not all cases 18:53:05 <amrith> I'm more inclined to go ahead with this change and leave the bigger case for the future 18:53:18 <amrith> i.e. don't let the perfect be the enemy of the good 18:53:25 <cp16net> ok i'll work on fixing the unit failure 18:53:48 <amrith> dougshelley66 may not be here, I'll speak with him about this; he and I had talked when this happened originally. 18:53:48 <cp16net> just didnt want what was mentioned to fall through the cracks 18:54:11 <amrith> thx cp16net 18:54:14 <cp16net> np 18:54:18 <amrith> pmalik, will look at that review 18:54:41 <amrith> flaper87, the current plan is for August 24, 25, 26 or thereabouts 18:54:41 <cp16net> its small enough to fix the issue that was raised 18:54:42 <amrith> in NYC 18:54:57 <amrith> was in meeting minutes two weeks ago 18:55:02 <amrith> and for some weeks before that. 18:55:08 <amrith> flaper87 ^^ 18:57:26 <amrith> pmackinn, I've updated https://review.openstack.org/#/c/312806/ and https://review.openstack.org/#/c/312805/ 18:57:36 <amrith> please review. vkmc, pmackinn ^^ 18:58:00 <dougshelley66> amrith i'm here now..just catching up 18:58:06 <dougshelley66> i certainly don't want to be the enemy of the good 18:58:13 <dougshelley66> i'll take my -1 off that review 18:58:19 <pmackinn> amrith, thx 18:59:05 <amrith> anything else ... 18:59:09 <amrith> 50s to go ... 18:59:30 <amrith> #endmeeting