13:00:29 <HenryG> #startmeeting neutron_db
13:00:30 <openstack> Meeting started Mon Sep  8 13:00:29 2014 UTC and is due to finish in 60 minutes.  The chair is HenryG. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:00:31 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
13:00:33 <openstack> The meeting name has been set to 'neutron_db'
13:00:54 <HenryG> #topic Migration Unit Tests
13:01:07 <HenryG> akamyshnikova: can you give an update?
13:02:46 <akamyshnikova> yes, my update is that I talked with two oslo.db guys rpodolyaka and victors. They try to help, but we still didn't come to anything.
13:03:18 <HenryG> OK
13:04:06 <HenryG> So I have some thoughts on the tests ...
13:04:40 <HenryG> I am not sure if there is much value any more in testing all the plugins.
13:05:40 <HenryG> It may be more value at this point to select just ML2 and have a migration test in jenkins
13:06:18 <HenryG> Something that will catch errors in migration scripts proposed in gerrit
13:06:53 <HenryG> Migrations in the past are not that interesting
13:07:00 <HenryG> What do you think?
13:08:52 <akamyshnikova> HenryG, I was thinking about something like that..
13:09:57 <akamyshnikova> I think if I with oslo.db guys don't come to any other solutions to the end of this week, this will be the only variant.
13:10:38 <HenryG> Thanks akamyshnikova
13:10:51 <akamyshnikova> HenryG, is this sound ok? or we should hurry up?
13:12:20 <HenryG> I think it is high priority to get the script test in Juno, so perhaps just create a second code review using the same bug.
13:14:18 <akamyshnikova> HenryG, ok
13:14:31 <HenryG> That will allow you to keep working on the all-plugins tests if there is new progress to made there.
13:15:18 <akamyshnikova> HenryG, yeah, thanks for this idea!
13:15:46 <HenryG> Another thought I had was: is it possible to check migrations against models in unit tests?
13:16:29 <HenryG> Maybe it can be done with some HACKING rules?
13:17:09 <HenryG> It gets kind of boring to manually review people's migrations and models.
13:17:57 <HenryG> Let's give that some thought and discuss at next weeks
13:18:01 <HenryG> meeting.
13:18:42 <akamyshnikova> HenryG, hm... I'm a bit confused. My test is checking that models and migrations are the same. About what extra checks are you talking?
13:20:07 <HenryG> akamyshnikova: Sorry, I meant without needing the openstack_citest user/database, which most developers don't set up locally.
13:21:06 <HenryG> Sort of a lower barrier to checking.
13:22:02 <akamyshnikova> HenryG, I see, I don't think about that in this way. I think after this test will be merged there was not such problem as checking that they are the same manually, as it will tests but unit tests.
13:22:48 <HenryG> Yes, you are right. They will get -1 from Jenkins.
13:23:31 <HenryG> And we can educate on how easy it is to set up the openstack_citest account/db for local testing.
13:24:21 <akamyshnikova> I think we test will be merged I will write an email with detailed description what and how now is checked.
13:24:39 <akamyshnikova> *when the test will be merged
13:26:04 <HenryG> akamyshnikova: Great! I will help with that. There is also developer docs, README for testing, etc.
13:26:37 <akamyshnikova> HenryG, thanks!
13:26:55 <akamyshnikova> HenryG, I appreciate your help :)
13:27:21 <HenryG> akamyshnikova: Any time!
13:27:27 <HenryG> #topic Open discussion
13:28:08 <jlibosva> https://review.openstack.org/#/c/117913/
13:28:28 <jlibosva> I think we should do the same as we did with the migrations after icehouse, don't we?
13:29:12 <HenryG> Can you elaborate?
13:29:56 * jlibosva searching for a link
13:31:25 <jlibosva> I saw a patch from Salvatore implementing decorators that would skip modification of table if the table doesn't exist
13:33:24 <HenryG> That might work. We should discuss that with Salvatore though.
13:34:39 <HenryG> jlibosva: Can you put that in a comment in the review?
13:34:43 <akamyshnikova> jlibosva, HenryG  this one link https://review.openstack.org/#/c/109768/
13:35:08 <jlibosva> akamyshnikova: ooh, yes. Thanks!
13:35:10 <jlibosva> HenryG: will do
13:37:04 <akamyshnikova> jlibosva, np :)
13:38:42 <jlibosva> Just one note: I'm gonna abandon https://review.openstack.org/#/c/112928/ per our discussion on previous meetings
13:39:00 <jlibosva> that we will go with head.py for Juno and hopefully for Kilo I'll come up with something usable.
13:39:23 <HenryG> Yes, agreed. Thanks jlibosva
13:40:11 <HenryG> Anything else to discuss?
13:41:34 <jlibosva> nope
13:41:40 <akamyshnikova> nope
13:42:05 <HenryG> Thanks!
13:42:14 <HenryG> #endmeeting