20:07:55 #startmeeting deb_packaging 20:07:56 Meeting started Mon Oct 10 20:07:55 2016 UTC and is due to finish in 60 minutes. The chair is zigo. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:07:57 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 20:07:59 The meeting name has been set to 'deb_packaging' 20:08:28 #topic Project status: upload to unstable 20:08:45 I uploaded everything to Unstable this week. 20:08:57 #link https://qa.debian.org/developer.php?login=openstack-devel@lists.alioth.debian.org 20:09:01 mistral-dashboard too ? 20:09:18 Only networking-scf and neutron-taas aren't migrated. 20:09:22 They wouldn't build. 20:09:27 mistra-dashboard too yeah. 20:09:30 cool! 20:09:46 I made the package today, it's in the NEW queue now. 20:10:07 yeah, i've seen ITP 20:10:08 senlin-dashboard was fixed for Django 1.10 20:10:09 I also uploaded it. 20:10:09 from you 20:10:27 Only trove-dashboard is missing. 20:10:37 i'm still working on it 20:10:40 tlbr: You saw the patch from rob, right? 20:10:41 sec 20:10:49 to trove-dashboard ? 20:10:54 zigo, ^^^ 20:11:10 #link https://review.openstack.org/#/c/381999/ 20:11:12 This one ! 20:11:17 sec 20:11:25 gosh! 20:11:29 thanks! 20:11:43 It's still not a 100% with it. 20:11:50 There's 3/4 failures still. 20:12:02 which one ? 20:12:07 I can't upload without the fixes. 20:12:11 Hang on, let me show you. 20:12:18 "NoReverseMatch: u'en-us' is not a registered namespace" 20:12:24 this is what i'm curently checking 20:12:56 FAIL: test_add_parameter_invalid_value (trove_dashboard.content.database_configurations.tests.DatabaseConfigurationsTests) 20:12:56 ---------------------------------------------------------------------- 20:12:56 Traceback (most recent call last): 20:12:56 File "/usr/lib/python2.7/dist-packages/openstack_dashboard/test/helpers.py", line 115, in instance_stub_out 20:12:56 return fn(self, *args, **kwargs) 20:12:57 File "/home/zigo/sources/openstack/newton/trove-dashboard/build-area/trove-dashboard-7.0.0~rc2/trove_dashboard/content/database_configurations/tests.py", line 341, in test_add_parameter_invalid_value 20:12:59 ['Value must be a number.']) 20:13:01 File "/usr/lib/python2.7/dist-packages/django/test/testcases.py", line 406, in assertFormError 20:13:03 self.fail(msg_prefix + "Response did not use any contexts to render the response") 20:13:05 AssertionError: Response did not use any contexts to render the response 20:13:10 Thats' the only one remaining. 20:13:16 hm 20:13:31 Then we got to actually TRY it, and make sure there's no exceptions. 20:13:43 Just in case some problems aren't caught by the unit tests ... 20:14:02 Same with the other dashboards, we need to check for them (at least browse in Horizon). 20:14:06 have you uploaded the changes to gerrit ? 20:14:17 For trove-dashboard, no. 20:14:19 i want to pick your changes and try to fix it 20:14:28 The rest of, yes. 20:14:33 ok 20:14:38 Though mistral-dashboard isn't in Gerrit yet, just in Alioth. 20:14:43 ok 20:15:01 Oh, we also need to do python-glareclient... 20:15:12 yes :) Michael told me about it 20:15:12 And move both glare and glareclient to Gerrit. 20:15:38 yeah, he will also release a new version soon, with nice improvements 20:15:52 i'm talking about Glare 20:15:57 #topic Project status: Still heat_integration problem with tempest 20:16:05 That one, I'm really not sure how to fix. 20:16:28 Currently, keystonemiddleware doesn't produce an auth_type = password kind of directives. 20:16:30 is it the same issue you showed a week earlier ? 20:16:39 Yup. 20:16:53 Same issue, I guess we don't need to discuss it again. 20:17:02 I need to chat with some Keystone specialists. 20:17:16 stevemar for example. 20:17:22 ok 20:17:31 IMO, it's ok to delay the work on it. 20:17:52 #topic Project status: BTS clean-up in progress 20:18:05 I've cleaned-up about 15 bugs today! :) 20:18:11 :) 20:18:26 #link https://bugs.debian.org/cgi-bin/pkgreport.cgi?which=maint&data=openstack-devel%40lists.alioth.debian.org&archive=no&raw=yes&bug-rev=yes&pend-exc=fixed&pend-exc=done 20:18:41 There's more to come. 20:18:55 I'm satisfied when we reach a level where no action can be made. 20:19:05 That should be around 10 bugs, no more. 20:19:31 I've been slowed down a lot by "Hash mismatch" issues on the CI though. 20:19:51 Hopefully, we'll have a "real" debian mirror some day to avoid these. 20:20:00 is it still happening ? 20:20:05 A LOT !!! 20:20:10 =\ 20:20:10 ok 20:20:12 Some patches when up to 8 rechecks, still in error ... :( 20:20:27 =( 20:20:47 #link https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=837238 20:20:47 Debian bug 837238 in src:trove-dashboard "trove-dashboard: FTBFS: compressor.exceptions.OfflineGenerationError: No template loaders defined. You must set TEMPLATE_LOADERS in your settings or set 'loaders' in your TEMPLATES dictionary." [Serious,Open] 20:20:56 shall we patch settings.py ? 20:21:14 That one was before the upload of all of Newton to Sid. 20:21:23 I didn't close the bug, because we have another one with Django 1.10. 20:21:30 got it, thanks 20:21:31 Though the topic and the reason is now completely different. 20:21:51 Maybe we should close that one and open a new bug... 20:21:57 I'd say: useless if we have the fix. 20:21:58 i agree 20:22:20 I'll continue working on closing issues tomorrow. 20:22:44 #topic Opening Ocata dev 20:23:02 I need to have infra guys to help with that. 20:23:19 pabelanger will need to add new ocata repo for example. 20:23:20 you mean make branches for projects ? 20:23:26 ah, got it 20:23:31 Branches is another thing, bug yes ! :) 20:23:32 new repo along with branches 20:23:39 ok :) 20:23:43 I don't really want to open 250+ branches by hand... :P 20:23:53 We can in theory do it in Gerrit though. 20:24:03 Do you know if we could script clicking on Gerrit? 20:24:07 sec, i'll check if we have rights 20:24:20 We do have the rights to make new branches on Gerrit, that's for sure. 20:24:27 The only question is how to automate that ... 20:24:36 zigo: the release team typically handles branch creation fwiw 20:24:43 I think they may have scripts for it 20:24:44 no, we can't do that by button 20:24:55 clarkb: Ah, thanks a lot for your input ! :) 20:25:18 clarkb, is there anyway to do it ourselves ? 20:25:34 *any way 20:25:35 clarkb: Our workflow is a little bit different though, instead of "master" (which is upstream master), we would like to open "debian/ocata" right away now. 20:25:53 zigo: I am sure that can be parameterized 20:26:07 clarkb: Great, who should I ask? Doug? 20:26:41 ya I think I would start with Doug 20:26:47 Ok, will do. 20:27:07 #action Zigo to ask Doug about opening new debian/ocata branches when we are to start 20:27:29 tlbr: I prefer finishing the BTS clean-up first though, so I don't have to backport the fixes. 20:27:54 tlbr: I'd like to start doing packaging from trunk though, which means packaging each and every release of libs, clients, etc... 20:28:06 tlbr: So it has to happen fast ... 20:28:18 zigo, got it 20:28:23 tlbr: Do you think starting next week is reasonable? 20:28:38 zigo, i guess so 20:28:40 ie: one week to clean-up everything ... 20:28:44 Ok, let's do that then. 20:28:50 do you need help with bugs ? 20:29:00 Contrib is always welcome! :) 20:29:04 :) 20:29:06 So, yes for sure ... 20:29:11 #topic install-guide contribution 20:29:26 #link https://wiki.openstack.org/wiki/Documentation/NewtonDocTesting#Testers 20:29:34 I started reviewing every bits ... 20:29:41 And a few CR 20:29:51 tlbr: Do you have time for that work as well? 20:30:05 zigo, what should I add there ? 20:30:10 annegentle: ping 20:30:15 i'll try to spend some time on it 20:30:34 tlbr: The thing is, the doc team insist that we "test" the install-guide, and make sure it works. 20:31:03 I know the packages so well that I don't need to do actual tests for all. For example, for glance, I know it works, so I marked the debconf one as passed... 20:31:17 Need to add some stuff from the non-debconf bits though. 20:31:22 This type of things ... 20:31:36 Everything is in the status though. 20:31:54 The hardest bits will be networking. 20:32:02 we have tests for services 20:32:05 The problem is that I don't have enough space available on my laptop. 20:32:12 Nova, Neutron, Glance, etc. 20:32:18 tlbr: Do we have hardware available somewhere for this type of testing? 20:32:21 I mean functional tests 20:32:28 nope =\ 20:32:42 The goal is *not* to test if packages work, but if the install-guide is accurate and works. 20:32:55 I do know packages are working ! :) 20:33:00 (thanks to tempest testing...) 20:33:13 ok, i can try to do so 20:33:31 tlbr: Do you have enough space on your computer to do the network tests? 20:33:45 That's really the hardest part ... 20:33:47 zigo, yes, i can find it 20:33:57 Because you got to have all of the other parts working. 20:34:03 let's discuss it on #openstack-pkg 20:34:07 Ok. 20:34:26 #topic Open discussion 20:34:26 i'll need some help with it 20:34:51 tlbr: On this topic, I wanted to know: who's taking care of that sync git script for MOS ? 20:35:02 A. Tsamutali ? 20:35:14 zigo, we haven't started yet 20:35:22 there are still CI preparation 20:35:23 tlbr: Yeah, but who's doing it? 20:35:32 zigo, no one at this moment 20:35:35 Ok. 20:35:43 yes, we have Sync, but it thinks to branch 20:35:46 all the changes 20:35:51 Well, it has to sync from git.openstack.org now, I guess you know that ... :P 20:36:02 yeah, I've Sasha about it today 20:36:08 we will see what we can do 20:36:14 *i've told 20:36:18 Ok. 20:36:36 Anything else? 20:36:57 I guess no. 20:37:00 Thank you! 20:37:04 :) 20:37:09 have a good night then. 20:37:17 #endmeeting