00:02:31 #startmeeting Ansible Azure Working Group 00:02:31 Meeting started Thu Mar 22 00:02:31 2018 UTC. The chair is jborean93. Information about MeetBot at http://wiki.debian.org/MeetBot. 00:02:31 Useful Commands: #action #agreed #halp #info #idea #link #topic. 00:02:31 The meeting name has been set to 'ansible_azure_working_group' 00:02:53 hey yuwei when starting a meeting you need to set the meeting name `Ansible Azure Working Group` 00:02:59 #chair yuwei 00:02:59 Current chairs: jborean93 yuwei 00:02:59 i guess not so many people today 00:03:06 #chair zikalino 00:03:06 Current chairs: jborean93 yuwei zikalino 00:03:18 nitzmahone is out today and I believe Kylie said she won't be here 00:03:23 Yes I m looking for the sentence 00:04:17 i think my only topic is prs today. 00:04:40 Yep 00:04:42 i have one urgent pr, as seems like container instance is broken now 00:05:02 I am working through them, currently testing the networkinterface with the vnet being in a different resource group 00:05:15 there is one more networkinterface PR I am aware of and that was the end of the list I had 00:05:48 this one: https://github.com/ansible/ansible/pull/37714 00:06:02 but i think i have to disable tests as they are not stable 00:06:18 I have added myself as a reviewer for that one 00:06:24 anyway, container instance didn't have version limitation <0.4.0 00:06:40 and 0.4.0 seems to be broken 00:06:53 At some point soon we need to look into the unstable tests and get them stable again 00:07:02 so i think we should have fixed version 0.3.1 (or i can change to 0.4.0) 00:07:27 i am actually looking into all unstable / disabled tests this week 00:07:42 so there are 9 right now 00:09:10 i also have two following prs: 00:09:12 https://github.com/ansible/ansible/pull/36826 00:09:43 https://github.com/ansible/ansible/pull/36768 00:10:07 I'm just about to merge 36768, my local tests were successful 00:10:30 and less urgent one: 00:10:32 https://github.com/ansible/ansible/pull/36818 00:10:54 but i need to add some more explanation so i will ping you later about this 00:11:12 ok, for the latter one it would be great if we can get a test for it 00:11:33 Hi Jordan how about this https://github.com/ansible/ansible/pull/36109 00:12:08 I found about 7 same PR sent in the past year 00:13:52 that sounds like a good option, we would need a way to configure what variable to use for `ansible_host` though. Have a look at `ec2.py` and I would try and replicate a similar setup to how they define it in `ec2.ini` 00:14:26 36818? yes, actually tests are disabled for deployment template, i am changing tests at the moment to use different template, so i will add tests for this one as well. 00:15:50 sounds good 00:16:24 yuwei: if you can find the best implementation of that feature (or do your own) with a way to configure the variable used as the host connection, send it through and we can merge it in 00:17:21 Okay, I will add it to my backlog 00:17:33 thanks 00:17:37 Also I find a PR very out of date, can we close it https://github.com/ansible/ansible/pull/19669 00:18:01 If it is either fixed or superseeded by another PR then yes we can 00:18:06 Are you asking me to close it? 00:18:59 Seems a PR to fix a bug in 2016, but not reproduce in later version 00:19:27 Affect 2.3 00:19:37 so i think it shoudl be closed, but we can't close it, right Yuwei? 00:19:41 ok, will close saying they can re-open and rebase if it is still an issue 00:20:38 i also created following pr, to fix the doc: 00:20:38 closed 00:20:39 https://github.com/ansible/ansible/pull/37626 00:21:11 but seems like there are lot of tests disabled because of this particular problem. 00:21:22 so i was wondering if theres' any reason not to fix it 00:21:37 hmm, shouldn't we just add the default choice to the azure_rm_common arg spec? 00:21:45 or should i just go ahead, enable tests again and merge it? 00:22:13 nitzmahone may be the best to ask on that as he understands Azure stack more than I do 00:23:04 well, what puzzles me is why to keep documentation out of sync? ok, anyway i will bother him 00:23:38 Small pr about inventory https://github.com/ansible/ansible/pull/34980 00:23:40 ¯\_(ツ)_/¯ he usually does things for a reason 00:23:46 ok, and last issue from me, or actually from another developer, i am just helping him to review: 00:23:49 https://github.com/ansible/ansible/pull/37333 00:25:34 the latter may take some more time, I'm mostly focusing on bug fixes and not new modules right now but have added myself as a reviewer 00:26:04 ok, thanks! 00:26:06 the inventory pr seems like a simple fix, will look at it today 00:26:35 hi Jordan, these days I work for AKS module, when I write integration test, since the AKS need a pair of valid Service Principal as parameter, can I get them in integration test? 00:27:12 mattclay is the best person to ask about that 00:27:32 Okay I will bother him 00:27:36 Thanks 00:28:13 yuwei: Are you able to use the service principal used for the rest of the tests, or do you require a dedicated one for AKS? 00:29:18 I want to use the auth one of integration test 00:29:55 Like "{{ Azure_Client_id }}" 00:30:39 But seems when running integration playbook, the azure.py only sends in resource group 00:30:59 yuwei: You can use `{{ lookup('env', 'VAR_NAME') }}` to access env vars, including the service principal details. 00:31:03 yes, i think this would be useful for some other tests too 00:31:21 i was working on authorization module and it will also require such things 00:31:40 Oh yes thank you. I will have a try 00:32:23 yuwei: All the vars I listed here should be available: https://github.com/ansible/ansible/pull/35888#issuecomment-375136747 00:34:22 Thanks I will try 00:35:17 ok, i disabled test in container instance pr, as still unstable, but fix should be merged urgently: https://github.com/ansible/ansible/pull/37714/files 00:35:30 i will try to figure out test failures later today 00:36:02 should be also backported to 2.5 00:39:02 ok, is there anything else we want to discuss? 00:39:20 No from my side 00:40:22 sounds good, I'll close the meeting if that is it 00:40:24 yes, i think that's all from me 00:41:10 sounds good 00:41:12 #endmeeting