20:03:55 #startmeeting Ansible Windows Working Group 20:03:55 Meeting started Tue Apr 24 20:03:55 2018 UTC. The chair is nitzmahone. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:03:55 Useful Commands: #action #agreed #halp #info #idea #link #topic. 20:03:55 The meeting name has been set to 'ansible_windows_working_group' 20:04:04 hello 20:04:18 Good $greeting_in_your_timezone 20:04:35 you too 20:04:43 #chair jhawkesworth_ 20:04:43 Current chairs: jhawkesworth_ nitzmahone 20:05:13 so uh, how does this work? 20:05:35 * nitzmahone looks at agenda 20:05:40 * chopraaa is confused. 20:06:01 https://github.com/ansible/community/issues/294 20:06:26 Looks like the only new item is Jordan asking about some changes to Convert-ToSID 20:06:32 chopraaa: link above is the agenda for the meeting 20:06:56 It's a public holiday in Australia today, so Jordan won't be here, but the idea sounds good to me 20:07:01 I think I kept jordan up late discussiong win_package paths 20:07:26 So I guess: 20:07:29 #topic open floor 20:07:39 i posted again 20:08:06 I have no opinion on having the SID magic. 20:08:25 #topic https://github.com/ansible/community/issues/294#issuecomment-384062697 20:08:34 (splitting up win_disk_X modules) 20:08:46 I'm all for it 20:09:06 there's a win_disk_mgmt module PR going on 20:09:23 That's one of the reasons win_disk_management has been sitting out there 20:09:25 made me throw up a bit 20:09:36 IMO it's trying to do too much 20:09:57 I think there hasn't been a consensus that it's something we want to ship in its current state. 20:10:18 im all up for splitting it into 2 or 3 modules 20:10:27 The disk facts I could go either way on, since it's already out there (and IIRC it doesn't take significantly longer to collect all that stuff than to have separate modules for it) 20:10:32 same for the disk_facts module 20:10:52 The difference there being that it's already been shipped 20:10:52 I haven't had a need to manage windows disks, but prefer smaller modules that do 1 thing well 20:10:58 +1000 20:11:17 yeah no harm in having the existing disk facts module. 20:12:10 this is what i did that works for me https://gist.githubusercontent.com/chopraaa/18062f052ca3001a807ec038c2f2c0fc/raw/a98df39901a6c2f9e1aaa55c1192c7a6b9398f0d/win_vol.ps1 20:12:43 if something better can be done I'd rather either see existing disk_facts module enhanced or different facts modules created. Once they are merged we can think about deprecating others 20:13:13 the idea was to be able to use params like get-vol uses 20:13:29 and return more data, using wmi/cim 20:13:31 Yeah, I don't think we need to split the disk facts- if there are things missing, feel free to add them to what's there 20:13:46 sounds good 20:13:47 But on the management side for sure 20:14:02 it's the disk_mgmt module thats the turd in the punch bowl for me 20:14:12 Also: be careful about asserting admin rights- you don't necessarily have to be in the admins group to do those thigns 20:14:43 Yep, IIRC that PRs been out there for a long time because none of us are big fans either 20:15:22 I guess win32_Voume has been around for a while - supporting older windows versions (down to S2008 at the moment) is something to bear in mind when working on modules 20:15:23 when i spoke to dag he said something about being able to include functions like test-admin 20:16:11 yeah, the current disk_facts module uses get-volume i think 20:16:25 Sure, that's what module_utils are for, but you can do those things without being an admin 20:16:33 which uses msft_volume, for 2012+ 20:16:34 I guess we could have something in module_utils ... but it may be that its dependent on privileges, rather than users 20:16:36 So that test is overly restrictive in nearly all cases 20:16:36 gotcha 20:17:21 so whats the way to go if a module is using a combination of cmdlets that use/dont use admin privs? 20:17:33 Anyway, sounds like you've got several of us that would be willing to review and work with you on more granular disk management modules if you want to get the ball rolling 20:17:52 Easiest is to just let the underlying stuff fail if the privilege is missing 20:17:58 It's ugly, but it works. 20:18:14 Once you know what the privileges are, you can assert them yourself directly if you want 20:18:34 alright 20:18:34 But I'd rather delegate to the underlying system protections than place overly restrictive tests that are unnecessary 20:18:55 "Access denied" is usually pretty descriptive ;) 20:19:21 * jhawkesworth_ hates seeing "access is denied" so much 20:19:25 (though if you know what the more granular tests are, it's nice to fail with a prescriptive "do you have priv X?" 20:19:55 * chopraaa scratches head. 20:19:57 but yeah, please catch it and return a message that gives the user a clue how to fix the issue if you can 20:20:02 no habla engles 20:20:11 ok i got this ^ 20:21:10 OK, anything else for today? 20:21:12 so yeah, go ahead, make some PRs 20:21:33 #agreed chopraaa to hack on more granular disk management modules 20:21:35 so its cool if i switch things around right? 20:21:52 and replace the get-vol get-disk stuff with cim 20:22:03 Do whatever you like for the new stuff, but don't break existing functionality in stuff that's already shipped (eg win_disk_facts) 20:22:06 since we're talking about support for 2003+ then 20:22:21 Well, 2008+, since Ansible's not supported on anything less 20:22:25 fortunately ansible won't talk to 2003 20:22:40 i'll keep the output the same, just use cim or wmi 20:22:58 and add more facts since i use them myself 20:23:09 but yes CIM/WMI more likely to be compatible with older windows, so go ahead 20:23:12 Sounds like a plan- looking forward to seeing them 20:23:16 and i guess anyone dealing with webhosting would bump into the same issues i did 20:23:54 perhaps, I have no experience of webhosting 20:24:02 and about jordan's pypsexec module 20:24:21 * chopraaa is curious about it's status. 20:24:30 maybe best to wait till jordan is around to quiz him on that 20:24:31 I just approved for 2.6 20:24:35 Is that getting added or did i miss the merge? 20:24:35 cool 20:24:40 cool 20:25:00 well thats it for me 20:25:12 I think he was going to tweak a couple more little things, but yeah, I :thumb 20:25:30 gave it the thumbs up yesterday 20:25:40 nothing from me, book stuff is parked so started looking at issues and PRs again this week 20:25:46 yeah it's great 20:25:48 OK, short and sweet. 2.5.2 ships next week. 20:26:14 (and future 2.5 releases will be every 2 weeks for awhile) 20:26:36 Thanks all! 20:26:37 great. 20:26:40 #endmeeting