16:02:22 <rakhmerov> #startmeeting Mistral
16:02:23 <openstack> Meeting started Mon Jan 23 16:02:22 2017 UTC and is due to finish in 60 minutes.  The chair is rakhmerov. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:02:24 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:02:27 <openstack> The meeting name has been set to 'mistral'
16:02:35 <rakhmerov> hi
16:02:39 <ddeja> hello
16:02:56 <rakhmerov> ddeja: everything's fine?
16:03:38 <ddeja> rakhmerov: yeah, sure. How about you
16:04:10 <rakhmerov> aah, more or less ok too
16:04:28 <rakhmerov> a lot of stuff going on
16:04:37 <rakhmerov> ok, is there anybody else here?
16:04:40 <rakhmerov> d0ugal: ^
16:04:57 <rakhmerov> mgershen1: ^
16:05:06 <rakhmerov> rbrady: ^
16:05:19 <rbrady> hey
16:05:23 <rakhmerov> hi Ryan
16:05:36 <rakhmerov> I reviewed your WIP, let's discuss it in a few
16:05:40 <rbrady> ack
16:05:46 <rakhmerov> ok, let's start
16:06:26 <rakhmerov> #topic Review action items
16:06:31 <rakhmerov> 1. review https://review.openstack.org/#/c/411412/
16:06:40 <rakhmerov> done (will discuss it later)
16:06:46 <d0ugal> Hey all
16:06:49 <rakhmerov> 2. rakhmerov: find Hardik's proposal on testing OpenStack actions
16:06:49 <d0ugal> Sorry I'm late.
16:06:50 <rakhmerov> done
16:06:53 <rakhmerov> d0ugal: np
16:06:58 <rakhmerov> thanks for joining
16:07:04 <rakhmerov> #topic Current status
16:07:41 <rakhmerov> my status: sent a few relatively simple patches including fixing a bug with task() function w/o a task name, made a bunch of reviews
16:08:08 <rakhmerov> keep investigating big input size issue reported by tuan__ and colleagues
16:08:35 <tuan__> Hi Renat
16:08:42 <tuan__> let;s talk about it tomorrow
16:08:47 <ddeja> my status: finished refactoring of getting the rpc-related flag and finishing the multi-rabbit support in kombu driver. I'm hoping to get it merged before the freeze
16:08:50 <tuan__> i will try to provide more information
16:09:33 <rakhmerov> tuan__: yes, ok. I don't have any new updates on it now, I should have more tomorrow
16:10:17 <rakhmerov> ddeja: can you explain a bit why multi-rabbit support is needed?
16:10:18 <d0ugal> I fixed an issue with name validation and posted this patch: https://review.openstack.org/412433
16:10:26 <rakhmerov> what use case does it cover?
16:10:30 <d0ugal> Looks like I have a CI issue to resolve, but that aims to improve action loading in Mistral.
16:10:45 <rakhmerov> d0ugal: ok, cool
16:11:01 <rakhmerov> ddeja: and how does it actually work?
16:11:11 <ddeja> rakhmerov: when you have more than one rabbit host
16:11:28 <ddeja> and want to automatically switch beetwen them when one of them dies
16:11:30 <rakhmerov> ooh, you mean for rabbit deployed in HA?
16:11:33 <ddeja> yes
16:11:37 <rakhmerov> I see
16:11:48 <rakhmerov> but it's still one Rabbit cluster?
16:11:51 <rakhmerov> right?
16:11:53 <ddeja> yes
16:11:57 <rakhmerov> ok, now clear
16:12:07 <rakhmerov> I thought it's for multiple clusters :)
16:12:18 <ddeja> no no, that would be magic ;)
16:12:22 <rakhmerov> I was wondering why we would ever need it..
16:12:27 <rakhmerov> yes
16:12:46 <rakhmerov> #topic Ocata-3 state
16:13:00 <rakhmerov> small announcement
16:13:09 <rakhmerov> this week we're releasing O-3
16:13:30 <rakhmerov> this cycle was very short comparing to the previous ones
16:14:03 <d0ugal> yeah, scary short :-D
16:14:20 <rakhmerov> so, before this Fri we need to merge all the new features that we want to have in Ocata
16:14:55 <rakhmerov> I don't think we have many of those but please check whether you need something urgently
16:15:24 <rakhmerov> so, starting this Friday no new features in Ocata
16:15:59 <rakhmerov> we can use FFE if we need to but only if we really need something very much
16:17:14 <rakhmerov> I would also ask you to check all of your patches for everything to see if what is still possible to merge before Fri
16:17:24 <rakhmerov> maybe some small improvements, bugs etc.
16:17:32 <ddeja> sure
16:17:52 <rakhmerov> I believe we have lots of relatively small things that have been waiting their turn for a while
16:18:16 <rakhmerov> and please also review stuff actively this week
16:18:39 <rakhmerov> in general, this Ocata cycle turned to be pretty messy
16:18:44 <rakhmerov> mostly my fault, of course
16:19:07 <rakhmerov> although we discussed a lot of stuff in Barcelona, not much of that was really done
16:19:12 <rakhmerov> for various reasons
16:19:21 <ddeja> like in other projects, I guess
16:19:33 <rakhmerov> including changing priorities during development (internal influence)
16:19:58 <rakhmerov> ddeja: yeah, but usually we did slightly better in this regard
16:20:08 <rakhmerov> anyway, I think that's generally ok
16:20:23 <rakhmerov> that fact that priorities are changing is good
16:20:40 <rakhmerov> people are using Mistral and coming with new urgent requests
16:20:44 <d0ugal> I think everyone was shocked how fast the time went
16:20:51 <rakhmerov> d0ugal: yeah
16:21:00 <rakhmerov> typically we would have 1.5 months more
16:21:15 <rakhmerov> ok
16:21:24 <rakhmerov> that's all I wanted to say
16:21:26 <rakhmerov> on this
16:21:45 <rakhmerov> we still have about a month for bugfixing etc.
16:22:39 <d0ugal> Sounds good, there are lots of bugs and documentation improvements we can make :)
16:23:02 <rakhmerov> true
16:23:11 <rakhmerov> on documentation, the light is green
16:23:23 <rakhmerov> we can be improving it all the way till the PTG
16:23:37 <d0ugal> (and beyond!)
16:23:43 <rakhmerov> yes )
16:23:56 <mgershen> +1
16:23:59 <rakhmerov> and btw, I think we made a really noticeable progress on docs
16:24:11 <rakhmerov> mgershen: ooh, I again didn't realize you were here :)
16:24:13 <rakhmerov> hi!
16:24:31 <mgershen> Hi!
16:24:41 <rakhmerov> mgershen: Michal, any updates from your side?
16:24:53 <rakhmerov> mgershen: thanks for reviewing patches actively!
16:25:08 <rakhmerov> you're making good progress
16:25:20 <mgershen> I should update the documentation with the new function I added a few weeks back.
16:25:41 <mgershen> rakhmerov: thanks :)
16:25:42 <rakhmerov> mgershen: ok, good
16:26:00 <rakhmerov> #topic Open discussion
16:26:20 <rakhmerov> ok, any topics you'd like to discuss?
16:26:30 <rakhmerov> we put even more topics for the PTG
16:26:45 <rakhmerov> I think we now have VERY MUCH to discuss there
16:27:00 <rakhmerov> I'll be working on filling them with more details
16:27:24 <rakhmerov> not too much time left till then
16:28:03 <rakhmerov> rbrady, d0ugal: I reviewed a WIP for keystone utils, exceptions etc. in mistral-lib
16:28:18 <rakhmerov> I left some comments, please take a look
16:28:23 <rbrady> rakhmerov: thanks.  I've been reading
16:28:37 <d0ugal> I'll take a look later.
16:28:59 <rakhmerov> my big concern with this whole thing is decomposition of source code into multiple modules
16:29:29 <rakhmerov> maybe it's not something that we can find a good vision for, but I'm just bringing it up so we're more prepared before the PTG
16:29:54 <rakhmerov> for example, do we need to move all of our exceptions from 'mistral' to 'mistral-lib'?
16:30:09 <rbrady> rakhmerov: I can explain a little bit that might set some context
16:30:10 <rakhmerov> same about context thing, keystone, utils etc.
16:30:15 <rakhmerov> yes
16:30:55 <rakhmerov> my fear is that it won't be convenient during server-side development because 'mistral' will have another dependency that should be changing all the time
16:31:06 <rakhmerov> rbrady: yes, go ahead
16:31:43 <rakhmerov> so, we need to reasonable about what we keep in 'mistral' and what we move into 'mistral-lib'
16:32:01 <rbrady> rakhmerov: when I started moving code into the mistral_lib repo, I was under the assumption that mistral would take a dependency on mistral_lib
16:32:14 <rakhmerov> right
16:32:18 <rbrady> mistral_lib would then contain the minimum code needed for the actions
16:32:34 <rakhmerov> yes
16:32:46 <rbrady> I knew from past experience the keystone_utils were important if you were doing anything with openstack lients
16:32:52 <rbrady> so I started there
16:33:08 <rbrady> to move over the keystone utils, I needed the context
16:33:11 <rakhmerov> and it sucked in the rest?
16:33:22 <rbrady> I couldn't make a circular dependency, so the context was ported
16:33:27 <rbrady> yes!
16:33:32 <rakhmerov> aah...
16:33:35 <rakhmerov> I see now
16:33:50 <rakhmerov> I didn't realize it when I was reviewing the patch actually
16:34:26 <rakhmerov> hah..
16:34:59 <rakhmerov> so somehow we should keep it small
16:35:19 <rbrady> so the conventions issue - I assumed the non-public classes and methods that would be subject to change would follow the underscore naming convention
16:35:21 <rakhmerov> maybe not all of the keystone stuff is needed that we have on the server-side?
16:35:57 <rbrady> that's a question I can look into, but I don't have an answer to that yet
16:35:58 <rakhmerov> rbrady: you mean even classes should contain _ if they are private to the project?
16:36:49 <rakhmerov> how typical is it for python? If it is pretty typical then it's probably OK
16:37:04 <rbrady> rakhmerov: https://www.python.org/dev/peps/pep-0008/
16:37:12 <rakhmerov> I'm not insisting really, I just want to make it clear somehow for users of these public clases
16:37:48 <d0ugal> rakhmerov: if you look at the oslo projects, they use underscores to mark private files, variables etc.
16:38:02 <rbrady> I've been thinking that most of the other information intended to be passed from server-side to a custom action - it would be appended to the context object
16:38:32 <rakhmerov> rbrady: yeah, but what I remembered is that this doesn't tell anything about module and class names
16:38:55 <rakhmerov> d0ugal: ok, no problem
16:39:10 <rakhmerov> d0ugal: yeah, I saw some
16:39:44 <rakhmerov> d0ugal, rbrady: ok, let's proceed with underscores, no objections
16:39:54 <tuan__> rakhmerov: it should be like this in python
16:39:56 <rbrady> rakhmerov: there are some naming guidelines for classes
16:40:24 <rakhmerov> honestly, I probably just don't like this underscore thing.. :)
16:40:34 <rakhmerov> but that's my personal issue though
16:40:38 <rakhmerov> nevermind
16:40:49 <rbrady> rahkmerov: lol
16:40:54 <rakhmerov> :)
16:41:07 <rakhmerov> yeah, I'm still kind of alien in Python
16:41:17 <d0ugal> lol
16:41:20 <rakhmerov> although I've been using it for 4 years
16:41:26 <rakhmerov> ok
16:41:37 <rakhmerov> :)
16:41:37 * rbrady tries to find alien python shirt for rahkmerov
16:41:44 <rakhmerov> yeah :)))
16:42:06 <rakhmerov> that would be a really cool idea for a t-shirt design :)
16:42:49 <ddeja> since we are bringing funny stuff
16:42:53 <ddeja> http://dom.trojmiasto.pl/Mistral-i249.html
16:43:02 <rbrady> rahkmerov: I appreciate the review and will respond to the patch comments to keep the conversation going
16:43:11 <rakhmerov> "Still feel an alien in the Python world? You're not alone!"
16:43:13 <ddeja> just found out that there is a buliding developer in my city that is called mistral
16:43:22 <d0ugal> nice
16:43:38 <rakhmerov> rbrady: ok, np
16:43:52 <rakhmerov> let's just try to prepare for the PTG as much as possible
16:43:58 <rbrady> rahkmerov: ack
16:44:09 <rakhmerov> including getting familiar with SS actions design
16:44:29 <rakhmerov> ddeja: haha, cool :)
16:44:44 <d0ugal> sorry, I need to leave. I have some chores to do
16:44:53 <ddeja> that's better link :D http://www.mistral.ekolan.pl/
16:44:57 * d0ugal waves
16:45:03 <tuan__> :D
16:45:06 <rbrady> bye d0ugal
16:45:06 <rakhmerov> d0ugal: ok, see you online tomorrow
16:45:12 <tuan__> well, for the underscore stuff
16:45:17 <tuan__> https://shahriar.svbtle.com/underscores-in-python
16:45:27 <tuan__> you guys can take a look to this one
16:45:32 <tuan__> but it is necessary
16:45:35 <rakhmerov> ok, rbrady: I don't see a simple solution for that dependency problems
16:45:42 <rakhmerov> we need to brainstorm it somehow
16:45:57 <rakhmerov> I'd like that ddeja to be also involved into this
16:46:06 <rbrady> rahkmerov: I will attempt to write up some thoughts to present it better to the group at PTG
16:46:10 <rakhmerov> he usually comes with good ideas
16:46:19 <rakhmerov> yes, please
16:46:45 <rbrady> rahkmerov: I think I might be able to explain it better in person.  maybe.
16:46:48 <rakhmerov> I really want to knock it down in Pike
16:47:01 <rakhmerov> but for that we need to solve all the fundamental problems
16:47:08 <rakhmerov> yep
16:47:47 <rakhmerov> ok, if there's nothing else you'd like to discuss now we can close the meeting
16:47:56 <rakhmerov> 12 mins left
16:47:59 <ddeja> rakhmerov: I'm done
16:48:04 <rakhmerov> ok
16:48:11 <rbrady> I'm good
16:48:22 <rakhmerov> ddeja: btw, do you have a context of this Actions API thing?
16:48:32 <ddeja> rakhmerov: a little bit
16:48:41 <rakhmerov> it'd be cool if you could participate too
16:48:50 <rakhmerov> at least in thinking and discussions
16:48:55 <ddeja> I can try ;)
16:49:10 <rakhmerov> it's not as simple as we thought it would be so we probably need more brains
16:49:29 <rakhmerov> ok
16:49:36 <rbrady> rahkmerov: my intent is to try to make it as simple as possible.  I promise. :)
16:49:49 <rakhmerov> ddeja: we can discuss more it in our channel
16:49:59 <rakhmerov> rbrady: hope so!
16:50:04 <rakhmerov> ok, bye everyone
16:50:08 <rbrady> bye
16:50:12 <rakhmerov> thanks guys and have a good wekk
16:50:14 <rakhmerov> week
16:50:22 <rakhmerov> #endmeeting