20:00:08 #startmeeting Ansible Windows Working Group 20:00:08 Meeting started Tue Nov 13 20:00:08 2018 UTC. 20:00:08 This meeting is logged and archived in a public location. 20:00:08 The chair is nitzmahone. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:00:08 Useful Commands: #action #agreed #halp #info #idea #link #topic. 20:00:08 The meeting name has been set to 'ansible_windows_working_group' 20:00:33 yo 20:00:35 jborean93: I hope it's as simple as pulling the wrong file 20:01:05 #chair jborean93 20:01:05 Current chairs: jborean93 nitzmahone 20:02:20 Nothing new on the agenda today 20:02:26 #topic open floor 20:02:30 I probably had one question to bring up 20:02:42 fire away 20:03:47 https://github.com/ansible/ansible/pull/46516#issuecomment-437179660 20:04:01 pretty much what we would want to name the `win_psmodule` source/repo module 20:04:21 yeah, win_psrepository works for me 20:04:33 (though it sounds a lot like "suppository" ;) ) 20:04:37 cool, simple to me 20:04:58 slightly related, I had a giggle when I came across https://docs.microsoft.com/en-us/windows/desktop/api/shtypes/ns-shtypes-_shitemid 20:05:19 lol 20:05:26 hey 20:05:29 hey 20:05:34 #chair jhawkesworth_ 20:05:34 Current chairs: jborean93 jhawkesworth_ nitzmahone 20:05:40 hey Jon! 20:06:09 hey. 20:06:14 just catching up 20:06:18 shite 20:06:42 I bet it seemed like a great idea at the time 20:07:02 jborean93 you ready for https://github.com/ansible/ansible/pull/48404 to get merged? 20:07:11 Looks like it's just failing on the docker mess 20:07:18 yep, merge away if you're good with it 20:07:20 It's holding up testing on ... other things ;) 20:07:38 cool, will do 20:07:44 I know dag's win_wait_for_process has held up by it as well 20:07:58 and someone just posted a comment earlier saying win_psexec was failing 20:08:04 is shippable not happy atm? 20:08:05 when run with async in devel 20:08:26 shippable seemed to have downgraded the Docker version on our nodes causing some issues 20:08:35 it should be back to normal now AFAIK 20:09:48 ah ok I bounced someone azure pr earlier to see if it was just flakiness 20:10:42 Any other fun for today? 20:11:00 I'm good, should be nearing a win_mapped_drive revamp sometime soon but coming across some interesting issues 20:11:30 I have a question. Is psrp ready to go now... I've been a bit out of touch recently but was wondering whether to try it for my next round of playbook upgrades 20:12:00 I'd still call it "experimental" personally; is there any reason you want to use it? 20:12:03 * jhawkesworth_ sees its documented connection type now so I guess its considered ready to use 20:12:22 I'm hoping for any speed enhancements I can get 20:12:39 I think in general the underlying stuff is a lot cleaner. I'd be surprised if you get performance boost from that. 20:12:54 got some feedback from users that some automation stuff is getting skipped because its slower in some cases than manual upgrades. 20:13:09 dag claims it is faster, local testing does say it is faster as well but only minimal gains 20:13:32 yes, manual work is always faster than automated runs.... 20:13:37 I'll take any gains I can get. 20:13:50 ... but then you get snowflakes and headaches 20:14:06 exactly 20:14:18 sounds like people just want to continue running stuff manually 20:14:20 For most stuff it should be stable; you should definitely try it out 20:14:46 ok I'll set it up when I come to test the playbooks and capture some timings. 20:14:51 yep, there were a few minor bugs we found when it came to unicode characters but that should mostly be fixed now 20:14:56 There are probably some dusty corners still though where things may break 20:15:23 Any feedback is more than welcome, especially if you come across any of those dusty corners 20:15:51 sure 20:16:36 curious if playing with max_envelope_size has any effect, so I might try upping that 20:16:59 Probably not much; we hardcode internally to ~200k for things like file transfers 20:17:45 I wanted to make it totally dynamic, but querying the config costs roundtrips 20:18:07 (and most people don't mess with it anyway) 20:18:29 and having it set too high is actually counterproductive anyway 20:18:40 oh ok.. file transfers being most likely sort of calls for any gains. 20:18:49 sounds like one to leave alone then. 20:20:22 any other burning issues this week? 20:20:54 not from me. Thanks for everyone's help with my epic refactor of `win_say` 20:21:10 :D 20:21:28 if only VBox allowed me to run it :) 20:22:00 * nitzmahone has still never actually seen it work; my one "real" Win10 host is silent as can be 20:22:02 I have literally heard of 1 other person who has used it successfully... to prank to his manager 20:22:49 yeah I was hoping to prove that it does work at 'fest, but your talk was scheduled same time as mine... and my demo failed anyway :-( 20:23:31 nice to have a low impact module to get head round using Ansible.Basic though 20:23:41 jborean93: shows how slurp is the same across multiple plays https://paste.fedoraproject.org/paste/EW6A3uOhS41394cte3cqPQ 20:24:48 that just tells me the keytab is always the same 20:25:33 Well, until next week then! 20:25:35 #endmeeting