23:59:26 <nitzmahone> #startmeeting Ansible Azure Working Group
23:59:26 <zodbot> Meeting started Wed Jan 23 23:59:26 2019 UTC.
23:59:26 <zodbot> This meeting is logged and archived in a public location.
23:59:26 <zodbot> The chair is nitzmahone. Information about MeetBot at http://wiki.debian.org/MeetBot.
23:59:26 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
23:59:26 <zodbot> The meeting name has been set to 'ansible_azure_working_group'
23:59:46 <nitzmahone> #chair yungezz zikalino82 yuwei jborean93
23:59:46 <zodbot> Current chairs: jborean93 nitzmahone yungezz yuwei zikalino82
23:59:54 <nitzmahone> Hey all
00:00:13 <yungezz> Hello
00:00:56 <zikalino82> Hi All :-)
00:00:56 <yungezz> I haven’t special topic today except prs, let me find links
00:01:22 <nitzmahone> I just merged the devtestlabs and vmss_extension ones
00:01:31 <jborean93> Hey
00:01:32 <zikalino82> great :-)
00:01:38 <nitzmahone> Had a question about the VM module auto-deletion stuff that I figured we'd just cover here
00:02:42 <zikalino82> yes?
00:03:36 <nitzmahone> At a glance, it seems like the new auto-delete-stuff-I-created change is a much bigger change to the deletion logic than is necessary
00:04:00 <nitzmahone> and adds a new arg that is planned for deprecation... We should probably figure out a different way to handle that
00:04:47 <nitzmahone> Creating a new arg that's planned to be deprecated is a pain for users
00:05:18 <zikalino82> yes, we were wondering how to handle that, and we finally came up with that additional argument....
00:05:19 <nitzmahone> It changed a lot from what we talked about a couple weeks ago
00:08:14 <zikalino82> so basically we should have "all_autocreated" choice which will become default when remove_on_absent is obsoleted?
00:08:41 <nitzmahone> Do we even need to deprecate `remove_on_absent`?
00:08:42 <zikalino82> and just remove this new option?
00:09:10 <nitzmahone> Seems better to me
00:09:29 <zikalino82> ok, so i will change that...
00:09:33 <nitzmahone> That way someone who wants to opt into the new behavior doesn't have to again change their playbook later to remove reference to the temporary option
00:09:41 <zikalino82> anyway. talking about "remove_on_absent"
00:09:54 <zikalino82> vmss also have this option, but it does nothing.
00:10:33 <zikalino82> i was wondering whether we should remove it.
00:10:56 <nitzmahone> Oh, because the module doesn't actually manage the lifetime of the VMs?
00:11:35 <nitzmahone> Yeah, either just remove it or if you want to be very conservative, deprecate it first, but since it's a no-op, I'm guessing not a lot of people specified it :D
00:12:11 <zikalino82> no, when you delete vmss, vms are deleted. so actually vmss doesnt' create anything that can be deleted
00:12:33 <zikalino82> we were thinking of just silently deleting it, as deprecating seems too heavy process here.
00:12:51 <zikalino82> is it ok?
00:12:59 <nitzmahone> OK by me- that's part of the joy of status: "preview" ;)
00:13:15 <zikalino82> ok :-)
00:13:24 <nitzmahone> If it were something people were actually using, I'd be more inclined to say deprecate first, but since it's always been nonfunctional...
00:14:25 <zikalino82> ok, so about vm's remove_on_absent i will make the ux change and ask yungezz to merge then :-)
00:14:52 <nitzmahone> k
00:15:34 <yungezz> Another Pr on web app slot https://github.com/ansible/ansible/pull/48862
00:15:46 <yungezz> It’s reviewed before
00:16:12 <nitzmahone> Cool- I'll do final review and merge
00:16:23 <yungezz> Thanks Matt
00:16:26 <zikalino82> and we have this pr from 3rd party: https://github.com/ansible/ansible/pull/42400
00:17:11 <zikalino82> i have access to his fork so i can modify it. i have rebased already. but seems like there are some windows related features there which we should discuss.
00:17:43 <yungezz> Yes about win_rm
00:17:46 <zikalino82> i think we should decide how to proceed
00:17:47 <nitzmahone> I thought the win_rm stuff was kinda deprecated in favor of using an extension to configure it, but maybe I'm thinking wrong...
00:18:55 <jborean93> I've always used an extension for this
00:18:58 <nitzmahone> Maybe not, can't remember where I saw that, but https://docs.microsoft.com/en-us/azure/marketplace/cloud-partner-portal/virtual-machine/cpp-winrm-over-https seems to imply that it's still the preferred way
00:19:27 <nitzmahone> Oh
00:19:34 <jborean93> yea I think it has some good integration into Azure to store the cert for the https listener
00:19:40 <nitzmahone> No, it does seem to imply that extensions are the right way
00:20:48 <nitzmahone> I dunno, that doc is all over the place. It says to use ARM templates and "customized scripts", which I read as script extension
00:23:29 <nitzmahone> I'm not opposed to it if that's still the "recommended" way to enable WinRM
00:24:40 <nitzmahone> The VM identity seems incorrect as well- `choices` will only allow the system assigned identity to be used, so it'd have to be removed to support a user-assigned identity
00:25:41 <nitzmahone> I can add 42400 to my review queue and really poke at it I guess
00:26:22 <nitzmahone> Any other items for today?
00:27:23 <yungezz> Thanks, we also will look at if win_rm really deprecated
00:27:35 <yungezz> No from me
00:28:03 <nitzmahone> Yeah, maybe just comment on the PR with whatever you find out
00:28:16 <yungezz> Ok
00:28:20 <nitzmahone> OK, if nothing else for today, we can wrap up early! Closing in 5...
00:28:23 <zikalino82> yes, let's try to finish it in a week
00:28:30 <nitzmahone> 4..
00:28:33 <zikalino82> the contributor is not very happy about the delay :-(
00:28:42 <nitzmahone> 3..
00:28:46 <nitzmahone> 2..
00:28:51 <nitzmahone> 1..
00:28:59 <nitzmahone> Thanks all- see you next week!
00:29:01 <yungezz> Thanks all!
00:29:06 <yungezz> Bye!
00:29:06 <nitzmahone> #endmeeting