18:00:49 <SlickNik> #startmeeting trove
18:00:50 <openstack> Meeting started Wed Apr  9 18:00:49 2014 UTC and is due to finish in 60 minutes.  The chair is SlickNik. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:00:51 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
18:00:53 <cweid> o/
18:00:55 <openstack> The meeting name has been set to 'trove'
18:00:58 <esp> o/
18:01:00 <juice> o/
18:01:02 <robertmyers> o/
18:01:02 <mattgriffin> o/
18:01:23 <pdmars> hello
18:01:25 <SlickNik> #link https://wiki.openstack.org/wiki/Meetings/TroveMeeting
18:01:28 <cweid> robertmyers: your name is robotmyers.
18:01:30 <cp16net> o./
18:01:31 <esmute> \o/
18:01:36 <amrith> \\o//
18:01:40 <k-pom> \o
18:01:41 <grapex> o/
18:01:41 <esp> cweid: lol
18:01:44 <kevinconway> o/
18:01:44 <dougshelley66> o/
18:01:44 <robertmyers> cweid: correct
18:01:50 <mattgriffin> o/
18:01:50 <denis_makogon> o/
18:01:51 <esmute> amrith: that was a stretch :-)
18:01:55 <amcrn> o/
18:02:08 <amrith> esmute: i need a lot of stretching ;)
18:02:12 <glucas> \o
18:02:22 <SlickNik> Notes from last meeting:
18:02:25 <SlickNik> #link http://eavesdrop.openstack.org/meetings/trove/2014/trove.2014-04-02-18.01.html
18:02:58 <SlickNik> Let's get started
18:03:11 <SlickNik> #topic Refactoring backup/restore strategies
18:03:21 <denis_makogon> it's mine
18:03:36 <denis_makogon> the point is to collect all strategies in one place
18:03:36 <vgnbkr> o/
18:03:49 <denis_makogon> it reduces amount of the confi attributes
18:03:59 <denis_makogon> and keeps code compact
18:04:15 <amrith> is this a reaction to a specific bug-fix?
18:04:16 <esp> demorris: got a link?
18:04:23 <esp> opps
18:04:31 <esp> denis_makogon: got a link to the bp?
18:04:41 <denis_makogon> #link https://bugs.launchpad.net/trove/+bug/1287684
18:04:46 <esp> thx
18:05:02 <robertmyers> I'm not sure all impl's should be in the same file
18:05:06 <denis_makogon> for future possible refactoring and extracting guest code or etc
18:05:12 <amcrn> agree with robertmyers
18:05:17 <robertmyers> it makes sense that they are separate
18:05:19 <SlickNik> denis_makogon: Can you give some concrete examples why this is needed, and what it buys us?
18:05:21 <SlickNik> +1 robertmyers
18:05:30 <amrith> +1 robertmyers
18:05:47 <vipul> o/
18:05:56 <robertmyers> we should however make it easier to configure
18:05:59 <denis_makogon> robertmyers, first of all we don't need to pollute config values each time each datastore add support of some kinds of features
18:06:08 <denis_makogon> backup/restore the exact use case
18:06:24 <robertmyers> denis_makogon: yes, but consolidating is not the answer
18:06:35 <amcrn> the backup strategy conf parameters should be moved under the datastore subheaders (i.e. [mysql], etc.)
18:06:38 <SlickNik> denis_makogon: That would point to needing some sort of lazy loading of config values.
18:06:47 <amrith> it strikes me that this is a reaction to the review of 78339
18:06:49 <robertmyers> making the config options better, or easier to split up
18:07:03 <denis_makogon> if we would not store them in one place, we would need to add another one config attr per datastore that reflects the actual place where strategy implemented
18:07:33 <denis_makogon> amcrn, why they should be moved ?
18:07:43 <amcrn> denis_makogon: your comment above just agreed with me
18:07:59 <robertmyers> personally I think we should tranistion to stevedore for these
18:08:09 <kevinconway> who is steve?
18:08:17 <SlickNik> robertmyers: +1
18:08:20 <grapex> kevinconway: It's like Steve Holt I think
18:08:22 <robertmyers> that way you could add your own impl with setup.cfg
18:08:35 <amcrn> grapex: +100 for AD reference
18:08:36 <denis_makogon> amcrn, i think we need to store all strategies in impl.py files, thats all
18:08:59 <SlickNik> robertmyers: Yes, we need to do some work to transition to installing the guest using setup.cfg before we can do that though.
18:09:22 <amcrn> robertmyers: in lieu of that now, i think we're all agreed it should not be in the same file, and given that, the impl should be defined under each optgroup, no?
18:09:27 <robertmyers> SlickNik: true
18:09:48 <robertmyers> amcrn: I think so
18:10:31 <SlickNik> amcrn: agreed.
18:10:54 <denis_makogon> so, the the answer is - implement the it another way, just add strategy name space per datastore ?
18:11:07 <esmute> The thing about making the same file is that it would be hard to be shared with other part of the code. I put a patch to move the storage strategy from the guest to trove common so that the TM can use it for deleting/copy backups
18:11:22 <esmute> #link https://review.openstack.org/#/c/86242/
18:11:31 <kevinconway> i understand that the config parameters were moving into datastore specific sections right?
18:11:37 <kevinconway> so [mongodb]
18:11:54 <SlickNik> esmute: I don't think we're moving it into the same file.
18:12:02 <denis_makogon> kevinconway, i guess yes
18:12:19 <robertmyers> esmute: we want to split the guest out eventually
18:12:32 <robertmyers> so less shared code
18:12:44 <SlickNik> Okay, so that's approaching 15 minutes.
18:12:45 <esmute> SlickNik: i thought that was what denis_makogon was proposing
18:12:58 <esmute> robertmyers: nice :-)
18:13:55 <SlickNik> I think we're all agreed that we shouldn't have it all in 1 impl file, and that we should have the config opts per datastore. Eventually move to something like stevedore for loading the strategies.
18:13:57 <denis_makogon> so, what's the actual answer to this topic ?
18:14:13 <cweid> SlickNik: +1
18:14:14 <kevinconway> where can we read more on steve dore?
18:14:15 <denis_makogon> get it
18:14:15 <SlickNik> denis_makogon: What's the question?
18:14:37 <SlickNik> If there are any other clarifications / questions let's bring it up in the channel.
18:14:39 <grapex> kevinconway: Your local library!
18:14:53 <grapex> To quote an afterschool special
18:15:06 <denis_makogon> i heard what i want, so we can move on
18:15:08 <SlickNik> #topic Moving the docs
18:15:09 <cweid> kevinconway: http://stevedore.readthedocs.org/en/latest/
18:15:55 <SlickNik> grapex: yer up
18:17:23 <grapex> Thanks SlickNik
18:17:36 <grapex> so at the midcycle meetup we talked of moving the docs from the apidocs repo into the trove repo
18:18:13 <grapex> So moving this: https://github.com/openstack/database-api to this: https://github.com/openstack/trove
18:19:16 <grapex> Basically, we'd take the apidocs directory of the "database-api" repo and move it into the root of Trove
18:19:35 <grapex> Additionally, maybe we should kill the "openstack-database-api" directory
18:19:39 <grapex> https://github.com/openstack/database-api/tree/master/openstack-database-api
18:19:43 <grapex> there's very little in it
18:19:47 <grapex> So I guess my questions are
18:20:01 <grapex> 1. since the midcycle meetup, has anyone decided this is a bad idea
18:20:12 <grapex> and 2. does removing the openstack-database-api directory seem like a bad idea
18:21:03 <dougshelley66> would this info ultimately end up in "openstack-manuals"?
18:21:19 <esp> grapex: would you leave  openstack-database-api around for a bit during the transition or just link it to the new location?
18:21:43 <grapex> dougshelley66: Yes
18:22:00 <dougshelley66> so would that process be owned by Anne G?
18:22:06 <grapex> esp: We could leave it in the openstack/database-api repo
18:22:11 <grapex> but I don't think it's being used
18:22:15 <grapex> the last update was 8 months ago
18:22:41 <grapex> dougshelley66: I'd email everyone involved in the docs. She heard this plan at the meetup and was ok with the notion.
18:22:50 <esp> grapex: k
18:23:05 <dougshelley66> grapex: i presume someone would have to do that conversion?
18:23:08 <robertmyers> +1 on consolidating docs
18:23:14 <grapex> dougshelley66: Sure, I can do it.
18:23:19 <SlickNik> grapex: ajaeger seems to be the most active committer to that database-api repo. I'd make sure to doublecheck with him as well.
18:23:26 <kevinconway> shouldn't every doc get it's own directory?
18:23:27 <grapex> SlickNik: I have the email ready to send right now to Andreas
18:23:44 <dougshelley66> grapex: i meant the conversion to the format of the rest of the openstack-manuals content?
18:24:58 <grapex> dougshelley66: Hmm.. I don't know. My guess is we'll see if moving it interferes with anything the doc team wants to do, and if it does we'll hold off or figure out a better plan
18:25:18 <dougshelley66> i have no issue with moving it - was just asking the longer term question
18:26:22 <SlickNik> grapex: I also think it would be a good idea to check with the doc team and figure out if there's a way for them to consume this so that we don't end up having to duplicate this info in 2 places when they need to produce a docbook.
18:26:41 <grapex> SlickNik: def
18:27:52 <SlickNik> Sweet, any other comments on this?
18:28:31 <SlickNik> Let's move on.
18:28:35 <SlickNik> #topic Design Summit Session Proposals
18:28:47 <SlickNik> This is everyone. :)
18:29:08 <SlickNik> Please get your proposals in for the design summit in ATL.
18:29:20 <SlickNik> #link http://summit.openstack.org/
18:29:23 <amrith> don't everybody step forward at the same time
18:29:24 <denis_makogon> SlickNik, when the deadline for the submitting ?
18:29:45 <mattgriffin> denis_makogon, the 20th
18:30:01 <denis_makogon> mattgriffin, hah, my birthday =)
18:30:06 <denis_makogon> mattgriffin, thanjs
18:30:13 <mattgriffin> :)
18:30:58 <denis_makogon> i have some ideas for the design session related to replication use case B
18:31:05 <SlickNik> hub_cap, the rest of the cores and I are planning on doing a pass to review these early next week (for scheduling) so please try and get them in by end of this week.
18:31:08 <hub_cap> SlickNik: and i are going to add sessions to address the defecits we have in the integration process
18:32:27 <annegentle> hey I can circle back on that docs question if you want grapex SlickNik (sorry so late) :)
18:32:51 <hub_cap> ++ w annegentle
18:32:56 <SlickNik> That's pretty much all I had for the summit sessions, so yes we can get back to docs. :)
18:33:10 <SlickNik> #topic Moving the docs
18:33:16 <annegentle> so I was thinking you'd bring database-api into your trove/doc repo
18:33:27 <annegentle> Andreas has nice JSON validation tests and so on
18:33:29 <grapex> annegentle: Me too
18:33:36 <annegentle> go for it! :)
18:33:41 <grapex> I think we can merge the tox file and add the new targets to the Trove tox file
18:33:47 <annegentle> grapex: cool
18:33:56 <grapex> annegentle: I'll be sure to email Andreas still
18:33:59 <annegentle> that's ideally, really, keep api docs near code
18:34:07 <grapex> since he hs def. put in the miles on it right now
18:34:30 <grapex> annegentle: I agree- it will help us tremendously in keeping them up to date
18:34:31 <annegentle> now the install doc I'm still seeking help on, Laurel is working on it, but the apt package isn't available yet, etc. So it's hard to get going
18:34:54 <annegentle> the bug is https://bugs.launchpad.net/openstack-manuals/+bug/1304584
18:35:16 <annegentle> if you can offer any notes or suggestions or anything, please put it in that bug.
18:35:39 <grapex> annegentle: Will do
18:36:05 <annegentle> such as: how to make the VM, what platforms you know it works on, all that good info
18:36:08 <annegentle> thanks!
18:37:28 <grapex> annegentle: Thanks Anne!
18:37:32 <SlickNik> laurelm: If there specific areas that we can look at to help this along, let us know. I'd be happy to put together some info to fill in places where we might have gaps.
18:37:39 <cp16net> what happened to the that we had some people were working on
18:37:43 <SlickNik> Thanks annegentle!
18:37:49 <cp16net> i reviewed it but never saw an update on it
18:38:22 <SlickNik> Oh you mean this one cp16net: https://review.openstack.org/#/c/85803
18:38:34 <cp16net> no...
18:38:42 <cp16net> that one is new
18:39:03 <cp16net> SlickNik: this one https://review.openstack.org/#/c/78608/
18:39:08 <laurelm> SlickNik: will do
18:39:17 <annegentle> cp16net: yes that API doc is different from database-api slightly; api-site is a reference listing for end users of the API (think: sdk devs)
18:39:51 <annegentle> cp16net: and as for a manual installation guide, we're looking for a chapter that plugs into somethign like http://docs.openstack.org/havana/install-guide/install/apt/content/
18:39:55 <annegentle> which we work on across four distros
18:39:57 <cp16net> ok
18:40:38 <SlickNik> Anything else to add here?
18:40:44 <annegentle> cp16net: are you Craig? Cuz your review of https://review.openstack.org/#/c/78608/ is great -- just need that info for four distros in another location is all
18:40:52 <cp16net> annegentle: yup
18:41:10 <annegentle> cp16net: kk thought so :) I'm getting good at this IRC mapping
18:41:18 <cp16net> thats my -1 ;-P
18:41:42 <annegentle> cp16net: so that looks really nearly there, laurelm all you have to do is convert it to docbook and get it as a chapter in the openstack-manuals install-guide
18:41:58 <annegentle> cp16net: laurelm: I could do a first pass if we want, get it into the community faster
18:42:27 <dougshelley66> annegentle: that is what we are doing. We went thru that document, clarified some things for laurelm and she is creating the docbook.
18:42:31 <annegentle> I just hate to do it if laurelm has a head start
18:42:32 <cp16net> ok i dunno who that was that started that review but it was a good start
18:42:38 <annegentle> dougshelley66: ok, great! That's it then
18:42:54 <annegentle> sorted.
18:43:42 <SlickNik> dougshelley66 / laurelm: Super; I'm not familiar with the docbook format, but I can help with information / editing once we have something out there.
18:43:59 <dougshelley66> SlickNik - thx sounds good
18:44:26 <SlickNik> Okay, we're good with that. Moving on...
18:44:30 <annegentle> dougshelley66: yeah the main thing is get it in the repo for reviews as soon as you can, we'll get it better and better
18:44:52 <SlickNik> #topic Open Discussion
18:44:58 <kevinconway> o/
18:45:06 <kevinconway> key signing
18:45:08 <kevinconway> that is all
18:45:18 <cp16net> i just did that refresh your keys
18:45:23 <SlickNik> kevinconway: I signed the key! :)
18:45:27 <SlickNik> key(s)
18:45:50 <amrith> my key hasn't been signed ;(
18:45:57 <amrith> only Justin and Andrew did ;(
18:46:13 <kevinconway> amrith: i must have missed yours. will fix that
18:46:41 <SlickNik> amrith: I could've sworn I signed yours as well (did I forget to push it back up?) I'll double check.
18:47:10 <kevinconway> so that's four of us that are working on it. what about the other 20?
18:47:40 <SlickNik> lol kevinconway. The WoT grows slowly.
18:47:53 <SlickNik> Any other items?
18:48:07 <SlickNik> btw kevinconway: https://github.com/dreamhost/stevedore
18:48:40 <SlickNik> If there's nothing else, I'm more than happy to end this early.
18:48:47 <SlickNik> going, going....
18:48:47 <dougshelley66> +1
18:48:53 <grapex> SlickNik: do it, do it, do it...
18:48:57 <SlickNik> #endmeeting