15:00:14 <d0ugal> #startmeeting Mistral
15:00:15 <openstack> Meeting started Mon Dec  4 15:00:14 2017 UTC and is due to finish in 60 minutes.  The chair is d0ugal. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:16 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:18 <openstack> The meeting name has been set to 'mistral'
15:00:28 <d0ugal> Hey - anyone around for the Mistral meeting this week?
15:04:31 <bobh> o/
15:04:40 <d0ugal> Hey bobh
15:04:56 <d0ugal> bobh: Nobody else has joined, so I was just about to cancel the meeting :)
15:05:08 <bobh> dobson: lol - ok - guess I didn't miss anything
15:05:18 <bobh> d0ugal: sorry
15:05:22 <d0ugal> bobh: no, nothing :)
15:05:28 <bobh> d0ugal: one quick question
15:05:30 <d0ugal> bobh: do you have anything you want to discuss?
15:05:32 <d0ugal> Sure
15:05:54 <bobh> d0ugal: I ran into an issue with adhoc actions that need an action_context in the base action - appears to be broken
15:06:14 <bobh> there was a bug opened last year that claimed to be fixed but I added a test that fails so looks like it's not
15:06:21 <bobh> just wondering if it's a known issue?
15:06:31 <d0ugal> bobh: Do you have a link to the bug?
15:06:48 <bobh> d0ugal: one sec
15:08:51 <bobh> https://bugs.launchpad.net/mistral/+bug/1568909
15:08:51 <openstack> Launchpad bug 1568909 in Mistral "execution of ad-hoc actions based on std.mistral_http always fails" [Medium,Fix released] - Assigned to Renat Akhmerov (rakhmerov)
15:10:35 <apetrich> o/
15:10:41 <d0ugal> apetrich: Hey
15:10:50 <apetrich> sorry my calendar is still with the wrong time for this meeting
15:11:02 <d0ugal> bobh: reading :)
15:11:07 <apetrich> me 2
15:11:20 <bobh> d0ugal: np - it took a bit of digging
15:11:20 <thrash> o/
15:11:43 <d0ugal> apetrich: in Google calendar you can set it to GMT without daylight savings.
15:12:59 <d0ugal> bobh: right, I understand now
15:13:16 <d0ugal> bobh: that action is weird, it has a special case, so I am not that surprised it is broken with ad-hoc actions :)
15:14:02 <bobh> d0ugal: that figures
15:14:40 <bobh> d0ugal: I tried injecting the action_context in the prepare_inputs of AdHocAction but then validation failed with an unexpected input
15:14:52 <d0ugal> bobh: the patch that claims to have fixed it is a huge refactor. so kinda hard to know what it did.
15:15:22 <d0ugal> bobh: fwiw, I have a patch that changes how this action works, I think it will be much nicer: https://review.openstack.org/520348
15:15:39 <bobh> d0ugal: right - and there is no existing test so no way to verify that it works
15:15:42 <d0ugal> but I have been blocked landing it by tripleo (I need to find a solution for this)
15:16:37 <d0ugal> oh! There has been a promotion of the tripleo packages, so it may be able to land now
15:16:39 * d0ugal rechecks
15:18:08 <bobh> d0ugal: that patch would definitely fix my problem
15:18:40 <d0ugal> Yeah, I hope so
15:19:10 <d0ugal> bobh: I guess it would be good if we could add a test case to show the problem and then we can verify it gets fixed more easily.
15:20:01 <bobh> d0ugal: your patch changes the signature of mistral_http to not require the action_context, so we would need another base action that requires the action_context I think
15:21:51 <d0ugal> bobh: what is action_context actually needed for?
15:22:29 <bobh> d0ugal: I don't need it - I was just trying to use mistral_http as a base for an adhoc action and that's where I ran into the problem with action_context
15:22:30 <d0ugal> bobh: the new version in my patch uses the new action conctext that is passed to the run method
15:22:37 <d0ugal> Right
15:22:59 <d0ugal> So why would we need another base action that requires the action_context?
15:23:05 <d0ugal> That is the bit I don't follow
15:23:47 <d0ugal> essentially MistralHTTPAction just adds some extra headers, I've moved that from the __init__ to the run method.
15:23:50 <bobh> d0ugal: to reproduce the issue with adhoc actions from a (now custom) base action that requires the action_context param - unless that is going away completely and it always comes out of the run context, which is a better solution anyway
15:24:02 <d0ugal> aha
15:24:05 <d0ugal> I see
15:24:12 <d0ugal> I think it is better to just remove it completely :)
15:24:17 <bobh> if the action_context param is deprecated there is no issue
15:24:49 <d0ugal> great
15:24:57 <d0ugal> bobh: if you could test/review that patch it would be very useful.
15:25:05 <bobh> d0ugal: will do - and thanks!
15:25:21 <d0ugal> There is one -1, but I am mostly ignoring it :)
15:25:29 <bobh> d0ugal: lol
15:25:55 <d0ugal> (I didn't think it was a helpful review)
15:26:00 <bobh> d0ugal: I don't suppose this patch would be a candidate for stable/pike
15:26:51 <d0ugal> bobh: good question, I hadn't planned to backport it but we could discuss with rakhmerov
15:27:01 <d0ugal> The related mistral-lib patches have already been backported to Pike
15:27:25 <d0ugal> so it *should* be fairly easy.
15:28:21 <d0ugal> bobh: Anything else you'd like to discuss?
15:28:23 <d0ugal> or anyone else?
15:28:42 <bobh> d0ugal: I'm good - thanks
15:29:15 <d0ugal> apetrich: anything you want discussed or reviewed for that YAQL function?
15:30:14 <apetrich> d0ugal, I don't think much at this point. I'm doing the minimum there and probably when the reviews come in I will need some more discussion
15:30:21 <apetrich> minimum functional
15:30:23 <d0ugal> I wanted to discuss CI, but I guess I need rakhmerov for that. https://review.openstack.org/522880
15:31:16 <d0ugal> apetrich: sounds good. I would encourage adding the smallest possible useful function. It is always easier to expand it later - but it sounds like you are already doing this :)
15:32:36 <d0ugal> thrash: btw I just reviewed https://review.openstack.org/#/c/522023
15:32:58 <d0ugal> I think we can probably end the meeting - mostly just open discussion today, but that is fine.
15:33:20 <d0ugal> I think we should consider doing the "office hours" with Mistral - other projects do this in place of the weekly meeting
15:33:55 * d0ugal waits 1 min before ending the meeting
15:35:45 <d0ugal> Okay, thanks everyone!
15:35:48 <d0ugal> #endmeeting