23:59:17 <nitzmahone> #startmeeting Ansible Azure Working Group
23:59:17 <zodbot> Meeting started Wed Apr 17 23:59:17 2019 UTC.
23:59:17 <zodbot> This meeting is logged and archived in a public location.
23:59:17 <zodbot> The chair is nitzmahone. Information about MeetBot at http://wiki.debian.org/MeetBot.
23:59:17 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
23:59:17 <zodbot> The meeting name has been set to 'ansible_azure_working_group'
23:59:28 <yungezz> Hello
23:59:28 <nitzmahone> #chair jborean93 yungezz
23:59:28 <zodbot> Current chairs: jborean93 nitzmahone yungezz
23:59:33 <nitzmahone> Hiya
23:59:50 <yungezz> Hello Matt and Jordan
00:00:29 <yungezz> Hi zim
00:00:35 <jborean93> hey
00:00:38 <zikalino826544> good morning :-)
00:00:48 <nitzmahone> #chair zikalino826544
00:00:48 <zodbot> Current chairs: jborean93 nitzmahone yungezz zikalino826544
00:01:04 <nitzmahone> Jake from our cloud team is lurking today ;)
00:01:30 <yungezz> Welcome
00:01:57 <thedoubl3j> :wave:
00:02:05 <nitzmahone> #chair yuwei
00:02:05 <zodbot> Current chairs: jborean93 nitzmahone yungezz yuwei zikalino826544
00:02:08 <yungezz> Hi zim, do you want to talk about the subs id Pr?
00:02:36 <zikalino826544> ah, yes, just let me find it
00:03:00 <zikalino826544> here: https://github.com/ansible/ansible/pull/55203
00:03:43 <zikalino826544> so apparently subcription_id is masked in returned facts when credentials are passed as parameters
00:04:06 <zikalino826544> what makes all returned resource ids useless (as in linked issue)
00:04:14 <nitzmahone> ah, yeah, it's the stupid heuristic masking
00:04:19 <zikalino826544> we were just wondering why it was masked in first place
00:04:47 <zikalino826544> also the issue submitter suggests client_id and tenant should be not masked either
00:04:48 <nitzmahone> The heuristic masker tries to "help" by finding return values named the same as a no_log value (eg in invocation block)
00:05:01 <zikalino826544> but i don't think these are important
00:05:14 <nitzmahone> Actually this is a bug in the underlying Ansible module_util
00:05:33 <zikalino826544> so is there anything we didn't consider, or should we just merge this fix?
00:05:56 <nitzmahone> It's better to keep things working if we merge the fix
00:06:32 <nitzmahone> But I'll make an issue to revisit the heuristic masking, or at least to have an explicit way to "opt out" where we say "this return value *needs* to be returned unmasked.
00:06:54 <nitzmahone> (it's an Ansible-wide issue, but most things don't run into it)
00:07:36 <zikalino826544> actually, i understood so far that no_log means it shoudlnt' be displayed, but it's completely removed from returned value
00:07:54 <jborean93> it scans all return values and masks it if the value is found
00:08:10 * nitzmahone hates that behavior, as it's overly blunt and causes dumb problems like this
00:08:13 <jborean93> yep
00:08:50 <nitzmahone> So yeah,  fine to merge the change, and we'll revisit how best to solve it longer-term
00:09:09 <zikalino826544> ok, thx!
00:09:11 <yungezz> Great
00:09:19 <nitzmahone> It's probably not necessary for the others, as you say I'm not aware of any modules returning those values right now
00:10:24 <nitzmahone> Anything else from your side today? We have something here if not...
00:10:34 <yungezz> We found 2.8 release delayed
00:10:45 <yungezz> I don’t have any other topic
00:10:46 <zikalino826544> well, i think not from me...
00:11:18 <nitzmahone> Yeah, apparently there have been some issues with packaging and late-breaking security things, and the 2.8 release manager just went on PTO for a few days...
00:11:39 <nitzmahone> He just decided to delay it a few days instead of having someone else take it over
00:12:02 <yungezz> Ok
00:12:14 <nitzmahone> So Jake is involved a bit in the certification program
00:12:42 <nitzmahone> He's pointed out that a few of the modules that are on the list for certification currently still have missing or disabled tests
00:13:00 <nitzmahone> This was one of the bigger problem ones: https://github.com/ansible/ansible/issues/54155
00:13:58 <nitzmahone> We also probably need tests to exercise azure_rm_resourcegroup, but they'd have to be tagged off for CI due to privileges (similar to what happened with rediscache and other long-running ones, but for a different reason)
00:14:45 <nitzmahone> mattclay: are you lurking?
00:14:48 <yungezz> You mean delicated resource group test?
00:14:51 <nitzmahone> yes
00:15:03 <yungezz> Ok
00:15:32 <nitzmahone> Jake was just finishing at a meetup- not sure if he's back in here yet. Jake- anything you want to add/clarify there?
00:15:36 <zikalino826544> in general we could assume that resource group is tested by the framework, right?
00:15:49 <nitzmahone> I don't think it uses the module
00:16:22 <yungezz> If we add more test or fix test for these modules in devel, should they backport to stable 2.8?
00:16:30 <nitzmahone> Ideally, yes
00:16:55 <nitzmahone> so long as they're test additions/cleanups or bugfixes (not new features of course)
00:17:04 <yungezz> Ok.
00:17:18 <zikalino826544> maybe if some tests are too long, we coudl at least have check mode tested...
00:17:28 <jborean93> we try and backport test fixes to all supported versions unless that isn't feasible
00:17:39 <nitzmahone> That would be ideal- at least something that's actually running the module, even if most of it is skipped in CI
00:17:40 <jborean93> i.e. it adds a new module arg or is a big code change
00:18:06 <thedoubl3j> sorry, I am juggling some things here. my apologies. resourcegroup and managed_disk were the only things that came up that need coverage either fixed or added
00:18:41 <yungezz> Ok
00:18:43 <thedoubl3j> _catches up_ you hit everything there Matt
00:19:15 <yungezz> We will look at them and add/fix tests
00:19:19 <nitzmahone> Cool, thanks!
00:19:28 <thedoubl3j> awesome! thank you!
00:19:32 <nitzmahone> I think that's all from our side this week- anything else for today?
00:19:52 <yungezz> I m ok
00:21:31 <nitzmahone> OK, if nothing else, I'll wrap it up in 5..
00:21:35 <nitzmahone> 4..
00:21:39 <nitzmahone> 3..
00:21:42 <nitzmahone> 2..
00:21:44 <nitzmahone> 1..
00:21:49 <nitzmahone> Thanks all- until next week!
00:21:50 <yungezz> Thanks all!
00:21:53 <nitzmahone> #endmeeting