13:01:26 <HenryG> #startmeeting neutron_db
13:01:27 <openstack> Meeting started Mon Sep 15 13:01:26 2014 UTC and is due to finish in 60 minutes.  The chair is HenryG. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:01:28 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
13:01:30 <openstack> The meeting name has been set to 'neutron_db'
13:01:59 <salv-orlando> I need to go afk in 15 minutes. If you have any item for me cna you please move it forward?
13:02:08 <HenryG> Not much left to discuss
13:02:23 <HenryG> Reorganize migrations is merged
13:02:38 <HenryG> Migration unit tests are working
13:02:58 <akamyshnikova_> salv-orlando, I have question how can I recheck vmware?
13:03:57 <salv-orlando> akamyshnikova_: vmware-recheck, but we suspended neutron jobs last weeks.
13:04:02 <salv-orlando> I resumed them this morning.
13:04:27 <akamyshnikova_> salv-orlando, ok, thanks!
13:04:32 <salv-orlando> jlibosva: I was on PTO - this merged https://review.openstack.org/#/c/117913/ and people said they had no problem updating from havana?
13:05:00 <salv-orlando> I think they should have had problems… according to your comments?
13:05:12 <salv-orlando> or did someone else fix the patch for me?
13:05:15 <HenryG> salv-orlando: I/we fixed while you were away
13:05:17 <jlibosva> salv-orlando: HenryG used the conditions for operations
13:05:47 <salv-orlando> jlibosva: thanks guys, this is what I was planning to do
13:05:56 <jlibosva> salv-orlando: all credit goes to HenryG :)
13:06:02 <salv-orlando> but hardly found any time to sit in front of a keyboard last week
13:06:29 <HenryG> salv-orlando: We are here to serve :)
13:07:37 <jlibosva> HenryG: salv-orlando does it mean in the end we don't support offline migrations now, right?
13:07:56 <HenryG> jlibosva: Only for Juno
13:08:04 <jlibosva> Is it documented somewhere?
13:08:16 <HenryG> Not yet
13:08:34 <HenryG> We need to update the deployer/operator docs
13:09:00 <HenryG> From kilo we will once again support offline migrations
13:09:37 <HenryG> I.e. Juno -> Kilo -> L___ -> etc.
13:10:26 <HenryG> It's only Icehouse -> Juno that needs to be online
13:11:28 <HenryG> #action HenryG to ensure operator docs get updated
13:11:33 <salv-orlando> correct. It is a suspension in offline migration support.
13:11:46 <salv-orlando> we already agreed to this when we did the healing script
13:13:11 <HenryG> Yup
13:13:27 <HenryG> Anything other questions for salv-orlando ?
13:14:33 <HenryG> salv-orlando: You may be excused. :)
13:14:50 <HenryG> #topic Migration Unit Tests
13:15:09 <HenryG> akamyshnikova_: These now work great!
13:15:13 <amotoki> We need to mention Juno migration does not support offline mode in the release note. I believe Kyle will cover this :-)
13:16:04 <HenryG> amotoki: thanks, I will get in touch with mestery about that
13:16:31 <mestery> amotoki: Thanks for the reminder here, I'll make sure to cover that as well.
13:16:44 <amotoki> HenryG: mestery: we check the release note as a community. it is wiki-based :-)
13:16:58 <mestery> amotoki: Exactly, please feel free to update that yourself! :)
13:17:14 <akamyshnikova_> akamyshnikova_, this working ok now :) so it needs review and we are ready to merge :)
13:17:37 <salv-orlando> akamyshnikova_: I will do a review of those patches later on today.
13:17:49 <akamyshnikova_> salv-orlando, thanks!
13:18:06 <salv-orlando> I have to leave now. Enjoy the rest of meeting. If there is any note for me I’ll read the log when I’ll be back.
13:18:16 <HenryG> akamyshnikova_: I see a comment from Maru about moving it to functional job. I will talk to him about that.
13:18:33 <HenryG> salv-orlando: Thanks for your time.
13:19:27 <HenryG> akamyshnikova_: Since no root or devstack env is required, it does not need to be in the functional job.
13:20:11 <jlibosva> HenryG: akamyshnikova_ I don't think the reason is root or devstack but the system interaction
13:20:27 <amotoki> HenryG: doesn't it depend on db back-ends like mysql or postgres?
13:20:52 <HenryG> Yes, it does
13:21:03 <akamyshnikova_> HenryG, yes, in fact these tests was created in oslo to be implemented in unittests
13:21:03 <HenryG> But it will skip if they are not found.
13:21:32 <akamyshnikova_> HenryG, yes if backend is not available tests just skip
13:22:27 <HenryG> But Maru is the test guru, so we should still listen carefully to his reasoning.
13:22:46 <akamyshnikova_> sure
13:23:10 <amotoki> it doesnt require devstack but a specific backend needs to be installed. for some case postgresql should be installed. Can we satisfy the requirement?
13:23:54 <HenryG> amotoki: The unit test requirements are currently python only.
13:24:13 <amotoki> HenryG: yes
13:24:24 <akamyshnikova_> in gate we have all necessary backend for this test available(mysql,postgres)
13:24:47 <akamyshnikova_> locally you have to install it yourself
13:25:34 <HenryG> The advantage of having the tests in unit tests is that it is easier for developers to test than setting up the environment for the functional tests.
13:25:51 <HenryG> So it will be easier for devs to test locally.
13:26:24 <akamyshnikova_> I added information in developer's documentation about this test
13:26:40 <amotoki> Okay, i got the situation. It is a kind of functional tests, but the situation around db is a bit different. thanks.
13:27:04 <jlibosva> HenryG: someone will still need to setup db, no?
13:27:38 <HenryG> jlibosva: Yes, but it takes like 30 seconds to do it manually. One time.
13:30:09 <HenryG> However, setting up the functional test environment is not hard either. And moving the tests to the functional job is not a bad thing.
13:30:32 <jlibosva> I'm more inclined to move it to functional
13:30:48 <HenryG> I would prefer it to be in unit tests, but I will talk to Maru.
13:32:09 <akamyshnikova_> HenryG, I also thinks that it should be unit tests, but of course we need to know why Maru suggest this.
13:32:51 <akamyshnikova_> I will be online next 2 hours if you will need me
13:33:08 <HenryG> #topic Open discussion
13:33:27 <HenryG> Anything else?
13:33:57 <akamyshnikova_> HenryG, I want to ask what sort of docs should I provide for test, but I think we need to get feedback from Maru first
13:35:41 <HenryG> akamyshnikova_: You already have the information in the comments. It's just a question of where to put them officially.
13:36:33 <akamyshnikova_> it won't be added in docs.openstack.org?
13:36:44 <akamyshnikova_> automatically
13:37:37 * HenryG knows too little about how documentation is done :(
13:37:54 <akamyshnikova_> http://docs-draft.openstack.org/20/76520/71/check/gate-neutron-docs/87432c0/doc/build/html/devref/db_layer.html#module-neutron.tests.unit.db.test_migration
13:38:36 <akamyshnikova_> this is result of gate-neutron-docs
13:38:53 <HenryG> Very nice, I did not know that.
13:39:41 <akamyshnikova_> :)
13:40:03 <HenryG> akamyshnikova_: We can check with emagana if that is sufficient or if some other steps are needed.
13:40:16 <akamyshnikova_> HenryG, great thanks!
13:41:26 <HenryG> akamyshnikova_: In your commit message you can put "DocImpact: developer docs"
13:41:51 <akamyshnikova_> HenryG, sure, thanks, for pointing this
13:42:03 <amotoki> http://docs.openstack.org/developer/openstack-projects.html is generated automatically when neutron repo is updated.
13:42:28 <amotoki> DocImpact for developer docs is unnecessary.
13:42:50 <akamyshnikova_> amotoki, ok
13:43:23 <amotoki> DocImpact flag is mainly targeted for documentations manually updated like admin guide or something.
13:43:40 <akamyshnikova_> amotoki, is there anything else what I should provide?
13:44:06 <amotoki> i think dev doc in neutron repo is enough
13:44:31 <amotoki> as long as it does not affect operators.
13:45:01 <akamyshnikova_> amotoki, ok, thanks!
13:45:06 <amotoki> akamyshnikova_: does it make sense?
13:46:04 <akamyshnikova_> amotoki, yes, thank you for explanation :)
13:46:45 <HenryG> amotoki: You are talking about doc/source/devref/* ?
13:47:23 <amotoki> HenryG: In my understanding, doc under neutron repo is the developer docs.
13:47:46 <HenryG> amotoki: OK, thanks
13:48:05 <HenryG> There is also TESTING.rst in the root dir
13:48:49 <HenryG> Anyway, we'll figure it out.
13:49:05 <HenryG> Anything else to discuss?
13:50:54 <amotoki> none from me. Thanks for your great work on migrations!
13:55:00 <HenryG> Thanks everyone!
13:55:05 <HenryG> #endmeeting