00:03:55 #startmeeting Ansible_azure_working_group 00:03:55 Meeting started Thu May 10 00:03:55 2018 UTC. 00:03:55 This meeting is logged and archived in a public location. 00:03:55 The chair is Kylie_. Information about MeetBot at http://wiki.debian.org/MeetBot. 00:03:55 Useful Commands: #action #agreed #halp #info #idea #link #topic. 00:03:55 The meeting name has been set to 'ansible_azure_working_group' 00:04:03 #topic plan for 2.6 00:04:18 Jordan, we have below plan for 2.6. 00:04:54 New modules: 00:04:54 - Azure VM facts (committed by external contributor) 00:04:54 - Rest API 00:04:54 - Application Gateway 00:04:54 - AKS 00:04:54 - App Service 00:04:54 - Azure storage facts (committed by external contributor https://github.com/ansible/ansible/pull/37333) 00:05:10 #chair jborean93 00:05:10 Current chairs: Kylie_ jborean93 00:05:21 Azure VM and VMSS facts 00:06:01 can we start with this? then? 00:06:06 Bug fixes: 00:06:06 - VMSS stability 00:06:06 - ADAL 00:06:06 - UA for inventory 00:06:06 - MSI 00:06:06 - SDK version 00:06:06 - Bug fixes for networking/LB 00:06:07 - Bug fixes for inventory 00:06:07 - https://github.com/ansible/ansible/projects/22 00:06:15 Yes. Zim. Please. 00:06:34 I will leave time to Zim, Yuwei and Catherine to discuss opens for items in 2.6 plan. 00:06:40 yes, so already discussed with jborean93 00:07:19 i made changes to format option of vmss facts. currently format can have two values "raw" and "curated" 00:07:23 I haven't seen the facts one in depth but I know the VM one is being held up because there are no tests and we want the curated return values first 00:08:00 zikalino: vmss facts is an existing module right? 00:08:23 yes, i asked Gustavo if i could help, but he said he wants to do it himself 00:08:29 yes, it's an existing module 00:09:02 i just added "curated" and test for curated, so if you could review it as soon as possible, it would be great 00:09:24 ok, can you send through a link, I'll do that once I'm finished with what I'm working on now 00:09:35 i still have a few minor issues there to fix.... but i hope it will be ready to merge tomorrow morning 00:09:36 if it is an existing we can't change the default but that's ok for now 00:09:44 for the vm facts we want to make sure we have the default as curated 00:10:30 https://github.com/ansible/ansible/pull/39722 00:10:52 thanks 00:10:59 i will tell gustavo that we want to release early next week and ask him if he can finish by the end of the week 00:11:17 so next rest api? 00:11:54 yes, here: https://github.com/ansible/ansible/pull/39274 00:12:16 i added idempotence option already and tests for indepotence 00:12:41 and it would be good if nitzmahone_ could find some time to review.... 00:12:59 yea I know he wanted to have a look through that 00:13:12 oh... but back to vmss facts... could we agree on the format option names? 00:13:30 isn't it curated vs raw? 00:13:32 I'm happy with raw/curated 00:13:40 i think we all here (on msft side)agreed on raw/curated 00:14:13 raw being the raw API results and curated being the more user friendly results 00:14:41 #action Jborean93 review https://github.com/ansible/ansible/pull/39722 00:14:46 more specifically -> curated will fit main module input 00:15:02 I thought that was going to be a third input? 00:15:10 Honestly I don't know how that is going to really work though 00:15:24 because you can't do `azure_rm_virtualmachine: '{{variable_dict}}'` 00:15:43 it works for vmss 00:15:46 you could hack around with yaml anchors but that defeats the point 00:16:08 really? I thought we explicitly don't allow that for security reasons 00:16:26 #action @nitzmahone review https://github.com/ansible/ansible/pull/39274 00:17:35 ok, let's just stay with raw/curated for now :-) i think it works well for vmss. i will prepare a document about facts 00:17:57 and discuss more when nitzmahone_ is available 00:17:57 agree 00:17:59 e.g. this does work but is set to be removed in a future version https://www.irccloud.com/pastebin/HcyoHQ2p/ 00:18:12 i am happy with current agreement 00:19:35 next is application gateway, aks and appservice , those are new modules 00:19:49 i'll send out app service pr today 00:19:50 aks https://github.com/ansible/ansible/issues/36677 00:20:05 contributed by other people 00:20:54 sorry wrong link 00:20:55 Application gateway, Zim, do you have PR? 00:20:59 app gateway will be ready by the end of tomorrow / end of weekend depending on my time 00:21:18 i can create pr even today, but it will be WIP 00:21:20 yuwei: is there a PR for that ask issue? 00:21:28 s/ask/aks/ 00:22:13 i think there's some customer asking for aks, @Kylie_ ? 00:22:53 https://github.com/ansible/ansible/pull/35888 00:23:05 jborean93 00:23:15 ok thanks 00:23:28 this is held up for a long time 00:23:56 Yes for AKS module. 00:24:10 And BUILD will announce more news about AKS. 00:25:24 #action @jborean93, review https://github.com/ansible/ansible/pull/35888. Very important. 00:25:45 I'll do vmss facts and that one today 00:26:19 Another one - https://github.com/ansible/ansible/pull/37333. It seems you assigned it to yourself. Please also help take a look. It could be after AKS. Thank you. 00:27:47 As for bug fixes, Yuwei, you have urgent ones for UA and MSI. Right? 00:27:57 yes 00:28:05 are they simple PRs? 00:28:20 https://github.com/ansible/ansible/pull/39785 00:28:36 I know that's subjective but if they are smaller fixes and have been reviewed by more than 1 person you should be able to merge yourself 00:28:48 same with azure rm common 00:28:55 adfs fixing is ready also, https://github.com/ansible/ansible/pull/37909, we talked this last week, matt was aware of the fixing 00:29:03 for inventory 00:31:02 ok, then i think i can take care of these and merge. is it ok? 00:31:20 Yes. Catherine and Zim could review azure_rm_common related fix if it is simple and we have much confidence in it. 00:31:30 ok 00:31:41 i added it to my list now 00:32:21 as long as multiple people merge simple bugfixes are fine 00:32:53 As for https://github.com/ansible/ansible/pull/39785, it is for inventory to add UA and MSI. @yungezz and @zikalino could you please see whether you could take care of it. I see changes are for one file azure_rm.py. 00:33:45 ok 00:34:05 how are you going with the test stability fixes 00:34:23 Jordan, https://github.com/ansible/ansible/pull/37909 still need your review. Catherine address Matt's comments. 00:35:05 not as well as i would like to.... most of the problems are some random internal issues on azure side 00:36:28 I believe Matt is doing the ADFS stuff, he understands that a lot more than I do 00:36:43 zikalino: so what's the proposed solution going forward? 00:38:52 i will put more effort into this at some point next week when all the planned modules are merged. i think i should be able to reenable some tests already next week. 00:40:00 about some issues i am trying to chase appropriate teams internally (responsible for particular service) 00:40:57 ok sounds good 00:42:19 aha I have a backport of 00:42:25 a bugfix 00:42:40 https://github.com/ansible/ansible/pull/39266 00:43:37 yuwei: can you add the changelog fragments in the backport PR 00:43:53 ok i ll add it 00:44:04 do you know the format? 00:44:11 once that's in, I'll merge it 00:44:54 I can find it in other pr 00:45:09 ok, let me know if you need help 00:45:39 or it is better you send me a link 00:45:45 thx 00:45:50 here is one https://github.com/ansible/ansible/blob/stable-2.5/changelogs/fragments/39818-loop_control_task_vars.yaml 00:46:02 basically create a file in `changelogs/fragments`/.yaml` 00:46:38 https://www.irccloud.com/pastebin/It5yGzPM/ 00:46:50 thx 00:46:53 you can have multiple entries if there ar emultiple fixes 00:47:25 Nice. Any other open 00:47:26 And i have 2 feature request 00:47:26 ? 00:47:32 Go ahead. 00:47:40 https://github.com/ansible/ansible/pull/37627 00:48:19 This is addressed the agreement we arrived last week 00:48:32 Refer resources in other e.g. 00:48:34 Rg 00:49:01 https://github.com/ansible/ansible/pull/38643 00:49:24 This is support loadblancer in mic 00:49:24 Nic 00:50:03 just briefly looking at it, it would be nice to see an example of the dict format to show users how to do it 00:51:22 Those are new features because they added parameters for existing modules. Right? 00:51:30 yep 00:51:39 Okay I will move some test to example 00:51:52 it would be up to Matt whether it could be considered a bugfix but usually they are considered features 00:52:21 but because the whole option name was changed (with an alias) I would be surprised if it could be backported 00:53:56 Since 2.6 is upcoming, no worry for backported. But would like to see what Catherine and Zim could approve. 00:54:03 New modules, definitely need your approval. 00:54:29 Change the name and add alias is suggested by Matt 00:54:42 yep, I agree with the what was done :) 00:54:51 Small bug fixes, Catherine and Zim could review and approve if ok. 00:55:45 Bug fixes involving in new parameter, how to handle those PRs in future? 00:55:47 Any one which got comment from Jordan/Matt or any other core committer, still need to get their approval. 00:55:56 yep small bug fixes are fine with me 00:56:16 honestly it depends on how large the change, new modules we definitely want to look at it 00:56:24 bug fixes should be fine with you to review/merge 00:57:09 I'll probably err and get a review from us on features that add new parameters 00:58:00 we can always revert the changes if something bad get's through, just not after a release so might be better to get a review the closer we get to a freeze/release or backporting 00:58:47 Understand. Let us move forward with current process and discuss it if needed in the future. - new module, need approval from Jordan/Matt; fixes adding new parameters, need approval from Jordan/Matt; small bug fixes - Catherine/Zim could make the call. 00:59:00 sounds good to me :) 00:59:22 #action @Yuwei, update #37627, #38643 and #39266 ASAP and ask for review to move forward. 00:59:47 Yes 01:00:04 Any other open? Yuwei, Zim and Catherine 01:00:36 No 01:01:02 for me I'm about to start reviewing 39722, 35888, 37333 01:01:04 Zim's network stop working 01:01:18 39266 is waiting for the fragment but that's pretty simple 01:03:37 Yes. Thank you. And we will ping you when the contributor updated https://github.com/ansible/ansible/pull/38279 for VM facts. 01:03:46 sounds good 01:03:48 Then that's all. Thank you Jordan. 01:03:59 have a good day all 01:04:08 You too. 01:04:10 #endmeeting