17:00:24 <alaski> #startmeeting nova_cells
17:00:25 <openstack> Meeting started Wed Dec  2 17:00:24 2015 UTC and is due to finish in 60 minutes.  The chair is alaski. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:26 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:00:29 <openstack> The meeting name has been set to 'nova_cells'
17:00:45 <alaski> anyone around for cells today?
17:01:21 <melwitt> o/
17:01:27 * dansmith gurgles
17:01:58 <alaski> heh.  are your insulin levels back to reasonable levels?
17:02:11 <alaski> #topic Cells testing
17:02:12 <dansmith> yes, belly fat, not so much
17:02:13 <vineetmenon> o/
17:02:21 <alaski> dansmith: lol
17:02:39 <alaski> melwitt: anything concerning these days?
17:02:51 <melwitt> alaski: not that I'm aware of
17:03:04 <alaski> great.  no news is good news
17:03:18 <alaski> #topic Specs
17:03:40 <alaski> anyone have outstanding specs for review?
17:03:48 <alaski> according to https://etherpad.openstack.org/p/mitaka-nova-spec-review-tracking we're all merged
17:04:50 <alaski> cool
17:05:00 <alaski> #topic Open Reviews
17:05:36 <alaski> dansmith: I wanted to get back to your concern about sqlalchemy in this series https://review.openstack.org/#/c/201606/
17:06:00 <alaski> and it being needed on computes
17:06:43 <dansmith> alaski: is there an update for me to look at?
17:07:23 <alaski> well, I never saw any comments from you on the reviews
17:07:37 <dansmith> oh snap, sure enough
17:07:44 * dansmith slaps his own wrist
17:08:05 <dansmith> so, thinking about this again,
17:08:30 <dansmith> I'm wondering there is really anything to be gained by not having that code in the object directly
17:08:42 <dansmith> if we have no db url configured, do we never load the actual sqlalchemy module?
17:08:48 <dansmith> or do we get all the way to that point anyway?
17:09:03 <dansmith> if so, then we're not adding any new dependency, or rather, we aren't really free from that one like I thought we were
17:09:41 <alaski> yeah, I think it would still be imported either way
17:10:04 <alaski> unless there's a lazy import somewhere in db/api.py
17:10:51 <dansmith> if there's not, this prevents us being able to add one later I guess
17:10:58 <melwitt> I was thinking about it yesterday and I while I don't like adding the dependency in a new place other than the db abstraction layer, I can see the reasoning since no one else would be calling the db api
17:11:15 <dansmith> I was thinking more about something that loads sqlalchemy/api or rax/foobarapi instead
17:11:44 <dansmith> melwitt: well, I like having the sqla code in the object because db_api and sqlalchemy_api are both layers that don't buy us much
17:11:50 <dansmith> and they're more work every time we do a thing
17:12:09 <dansmith> I just don't like forcing us to have very direct sqlalchemy code that high up from a dep point of view
17:12:25 <melwitt> yeah
17:12:29 <dansmith> alaski: I suppose we could make a fake sqlalchemy module that we load on the computes if we importerror on importing it,
17:12:38 <dansmith> which we already do with the db-errors-out wedge
17:12:41 <dansmith> so maybe that's the thing to do
17:12:52 <alaski> I was just considering that as well
17:13:35 <dansmith> so yeah, I guess I'll withdraw my concern
17:13:36 <alaski> I think I'm on board with that
17:14:42 <alaski> okay, cool
17:15:21 <alaski> we'll want to get a blueprint together for that work and then get it in place before this pattern is copied too many more times
17:15:34 <dansmith> for the wedge?
17:15:59 <dansmith> I think we can always do that later, right? and it's just a class with a getattr handler that blows up I think
17:16:12 <dansmith> or do you mean the pattern for the import handler?
17:16:25 <alaski> the import handler
17:16:31 <dansmith> okay
17:17:25 <alaski> anything else for open reviews?
17:17:49 <alaski> #topic Open Discussion
17:18:08 <alaski> anyone have a topic for today?
17:18:08 <belmoreira> sorry for being late
17:18:37 <belmoreira> do you talked about flavors reviews?
17:18:38 <dansmith> we should discuss that
17:18:40 <dansmith> heh
17:18:51 <alaski> :)
17:18:59 <alaski> belmoreira: we did talk about them
17:19:01 <belmoreira> dansmith :)
17:19:20 <alaski> specifically about the sqlalchemy import, but for now nothing needs to be done differently
17:19:43 <belmoreira> ok I will read the logs then
17:20:06 <lalitd> alaski: in patch https://review.openstack.org/#/c/213041 day before jenkin was +1 but now it is merge conflict
17:20:42 <alaski> lalitd: you'll need to rebase the change
17:20:47 <belmoreira> I will take care of it
17:21:17 <alaski> really the series needs to be rebased when that happens
17:21:41 <lalitd> alaski: ok
17:22:17 <alaski> if nothing else then we'll end here
17:22:27 <alaski> thanks everyone
17:22:31 <alaski> #endmeeting