18:00:39 <SlickNik> #startmeeting trove
18:00:40 <openstack> Meeting started Wed Feb 11 18:00:39 2015 UTC and is due to finish in 60 minutes.  The chair is SlickNik. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:00:41 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
18:00:44 <openstack> The meeting name has been set to 'trove'
18:00:50 <SlickNik> Hi folks.
18:00:50 <amrith> o/
18:00:57 <amrith> hiya SlickNik
18:01:03 <georgelorch> o/
18:01:08 <abramley> hi
18:01:10 <vkmc> o/
18:01:27 <SlickNik> Agenda for the meeting is at:
18:01:29 <SlickNik> #link https://wiki.openstack.org/wiki/Meetings/TroveMeeting
18:03:11 <SlickNik> #topic OSLO Namespace change (https://review.openstack.org/#/c/150709/)
18:03:17 <SlickNik> #link https://review.openstack.org/#/c/150709/
18:03:25 <dougshelley66> o/
18:03:32 <edmondk> o/
18:03:51 <SlickNik> So if I recall correctly, this is one of the follow up items to the mid-cycle agenda.
18:03:57 <amrith> <paste-bomb>
18:03:57 <amrith> re: change https://review.openstack.org/#/c/150709/
18:03:57 <amrith> This change, as presented, is incomplete and should have a hacking rule to prevent a recurrence of this problem. a sample hacking rules is available in neutron.
18:03:57 <amrith> Other than that, this change may be low risk.
18:03:57 <amrith> We don't "HAVE TO" take it for kilo but will most likely be forced to take it for "L".
18:03:59 <amrith> Question for discussion: Do we want to hold it in abeyance for L or reconsider for Kilo if a proper hacking rules is implemented?
18:04:04 <amrith> </paste-bomb>
18:04:36 <annashen> o/
18:05:17 <SlickNik> My concern during the mid-cycle was how we would ensure we didn't reintroduce this problem with a patch merge — the hacking rules allays that.
18:05:51 <SlickNik> So I'm of the opinion that we should take this in Kilo, since we're going to have to bite the bullet in Liberty eventually anyway.
18:05:55 <amrith> SlickNik, yes. I haven't seen what was merged in neutron though. ihar said he wrote it.
18:06:02 <amrith> not necessarily
18:06:08 <amrith> doug would like to force this change in L
18:06:14 <amrith> but there's no guarantee that it will happen
18:06:16 <edmondk> Can we define hacking rules, I am not familiar with this term and I am sure many of us are not
18:06:53 <SlickNik> edmondk: good question — https://github.com/openstack-dev/hacking/
18:06:59 <amrith> they are rules that hackers adhere to. a code of some sort ;)
18:07:00 <nshah> o/
18:07:08 <edmondk> lol
18:07:32 <SlickNik> That's a set of style checks that get run along with pep8 checks when you run tox -epep8 on your project
18:07:35 <edmondk> Basically the openstack style guide
18:07:43 <edmondk> gotcha
18:08:04 <edmondk> didnt realize we ran more then pep8 good to know
18:08:31 <edmondk> this conversation makes sense now :)
18:08:37 <amrith> [10:31:29] <dhellmann> amrith: I would *like* to drop namespace packages next cycle
18:08:38 <amrith> [10:31:37] <dhellmann> amrith: that's not a definite plan, yet
18:08:44 <amrith> SlickNik, that was from #openstack-oslo
18:08:50 <amrith> on 02/09
18:09:19 <vkmc> this shouldn't be a dangerous change, it has been around for a while now in several projects... and adding the hack file would make it safer
18:09:21 <edmondk> this doesn't seem like a high priority for kilo
18:09:41 <edmondk> what does this gain us
18:09:59 <amrith> edmondk, yes. This remains a nice-to-have and if the change doesn't go into the L release in OSLO, then an unnecessary change.
18:10:00 <edmondk> besides dhearing to new standard
18:10:01 <peterstac> o/
18:10:15 <amrith> edmondk, there's no "new standard"
18:10:29 <amrith> there's a proposal for a new standard that is not yet finalized
18:10:30 <edmondk> or recommended standard?
18:10:36 <edmondk> gotcha
18:10:36 <vkmc> edmondk, http://blog.nemebean.com/content/whys-and-hows-oslo-namespace-change
18:10:46 <SlickNik> So in that case, I'd suggest waiting till we know that it is going to be a definite standard.
18:11:19 <SlickNik> (and a more concrete timeline for when oslo expects this change to be made to the projects)
18:12:22 <amrith> if that's a formal proposal, I'll second it.
18:13:29 <SlickNik> Anyone opposed?
18:14:05 <amrith> .
18:14:11 <vkmc> sounds good to me
18:14:33 <SlickNik> Okay — let's move on
18:14:33 <vkmc> maybe we should ping oslo folks to know what is the deprecation status of the previous namespaces
18:14:35 <vkmc> and work with that
18:14:57 <amrith> vkmc, that's something that has been (regularly) on the ml
18:15:30 <SlickNik> #topic datastore and strategy classification
18:15:44 <SlickNik> #link https://review.openstack.org/#/c/154119/
18:15:48 <amrith> also action item from mid-cycle.
18:15:48 <vkmc> amrith, cool, I'll check it out
18:16:26 <amrith> <paste-bomb>
18:16:26 <amrith> https://review.openstack.org/#/c/154119/
18:16:26 <amrith> This came up in the mid-cycle and in normal course, this would get reviewed as part of the spec review and action taken based on that. However there is an urgency to get this process accepted and implemented for the Kilo cycle so I am raising this for consideration at the meeting.
18:16:26 <amrith> I'd like to know if people have questions or comments, and answer any questions now. I'd encourage you to read the spec and provide comments through review as well.
18:16:27 <amrith> </paste-bomb>
18:18:19 <SlickNik> #link https://etherpad.openstack.org/p/trove-kilo-sprint-state-of-ci
18:19:07 <SlickNik> The main idea here is to be able to communicate to deployers / users what state of CI and testing the various guest managers have undergone.
18:21:26 <SlickNik> amrith: I've given the spec a quick look over, and it generally looks reasonable to me — will update it with some more detailed feedback after the meeting.
18:21:35 <edmondk> I would think technical preview is already a superset of experimental so it's not necessary for Stable to say it needs to meet the requirements of technical Preview and experimental?
18:21:35 <amrith> SlickNik, thanks.
18:21:49 <vkmc> amrith, it looks really good
18:22:56 <dougshelley66> amrith, I think this spec accurately reflects the discussions last week
18:23:07 <dougshelley66> Can you elaborate on "Testing"?
18:23:50 <amrith> "T E S T I N G"
18:23:52 <amrith> sorry
18:23:54 <amrith> let me look
18:23:56 <amrith> one second
18:24:22 <amrith> So I was thinking that with this spec, what changes (materially) is imports
18:24:30 <amrith> and config files
18:24:41 <amrith> So in trying this out
18:24:56 <amrith> I learned that there are places where programmatically one could construct an import statement
18:25:02 <dougshelley66> ok so you were referring to specifically testing the dir structure changes, ok
18:25:13 <amrith> and finding those is harder than searching for 'path names'
18:25:19 <amrith> yes, that's basically what I was looking for.
18:25:24 <amrith> can I still build a guest image
18:25:30 <amrith> are there dependencies into trove-integration
18:25:37 <amrith> once built, does that guest image work
18:25:45 <amrith> what exactly do I have to do in order to upgrade
18:25:51 <amrith> verify those upgrade processes
18:25:55 <amrith> that's what I had in mind
18:26:02 <amrith> now, I may be missing a couple of things.
18:26:06 <amrith> so input there would be valuable.
18:27:42 <SlickNik> So one thing that comes to mind is that there are some unit tests for the extensions that we have.
18:28:30 <SlickNik> Perhaps something equivalent to test the dynamic load of the manager modules would be a good test to add.
18:28:53 <SlickNik> Still working through how exactly that would help in the upgrade scenario though.
18:28:57 <edmondk> To move out of experimental does the community vote that a specific datastore is ready, or does someone just send out a code review moving it from the experimental directory
18:29:15 <edmondk> and switch the CI from non voting to voting
18:29:21 <amrith> edmondk, I'd say the second is a better approach
18:29:25 <SlickNik> edmondk: The latter I would think
18:30:04 <SlickNik> Once we are convinced that the CI is stable enough and we have the tests that we need that are running in the gate.
18:30:05 <edmondk> Ok so moving the code from experimental dir and making the CI voting would be the official sanction
18:30:14 <edmondk> k
18:30:37 <dougshelley66> right because if anyone has any issues with the change they can comment during review
18:31:16 <SlickNik> Any other questions or clarifications regarding this?
18:31:59 <SlickNik> .
18:32:07 <SlickNik> #topic Open Discussion
18:32:37 <amrith> thanks for hosting the mid-cycle Nikhil!
18:32:40 <SlickNik> Thanks to those who could make it to the mid-cycle last week!
18:33:28 <edmondk> SlickNik, Is there an aggregate list of all the tasks from mid cycle?
18:33:40 <amrith> edmondk, funny you should ask
18:33:41 <SlickNik> For those who could not make it, and who weren't able to view the hangout sessions live — we do have recorded links to the sessions if you're interested in following the discussions that took place.
18:33:55 <SlickNik> #link https://wiki.openstack.org/wiki/Sprints/TroveKiloSprint
18:34:10 <amrith> edmondk, I added action items to https://wiki.openstack.org/wiki/Sprints/TroveKiloSprint
18:34:12 <edmondk> perfect thanks I see the table
18:34:20 <SlickNik> edmondk: There's an aggregate list of action items on that page that amrith added too!
18:34:23 <SlickNik> Thanks amrith!
18:34:29 <edmondk> thanks for doing that
18:34:35 <amrith> edmondk, so I assume you are now going to tell me that yu haven't yet done #16, ...
18:34:39 <amrith> convenient
18:34:53 <edmondk> working on it
18:34:59 <amrith> ;)
18:35:13 <edmondk> dan ritchie is looking into an issue with the configuration group tests being idempotent
18:36:01 <SlickNik> I'm making some headway into my action items — will update that list in the wiki as I go along.
18:36:09 <dougshelley66> SlickNik, is now a good time to start with #3 :)
18:36:32 <amrith> I made the list so I'm all set.
18:36:37 <amrith> just waiting for y'all to catch up.
18:37:02 <SlickNik> dougshelley66: It's funny you should mention that — I started looking at some stats over the weekend last couple of days.
18:37:27 <SlickNik> I should have something by the meeting next week to insert as a regular part of the agenda.
18:37:32 <amrith> And I did some of #5
18:37:51 <amrith> don't know how to share and host as a public resource quite yet
18:37:57 <amrith> but I have some of that ...
18:38:57 <edmondk> Added another item for myself, 32 Update python-troveclient documentation
18:39:44 <SlickNik> Thanks edmondk
18:40:24 <SlickNik> If there are any items that you feel we discussed at the mid cycle but have been left out, please add them to the list
18:40:51 <SlickNik> This is just a dump from the action items in the various etherpads — in one place.
18:42:08 <SlickNik> Anything else for Open Discussion?
18:43:15 <SlickNik> .
18:43:22 <SlickNik> #endmeeting