16:00:35 <rakhmerov> #startmeeting Mistral
16:00:36 <openstack> Meeting started Mon Sep  8 16:00:35 2014 UTC and is due to finish in 60 minutes.  The chair is rakhmerov. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:37 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:39 <openstack> The meeting name has been set to 'mistral'
16:00:41 <rakhmerov> hi
16:01:26 <bhavenst> hi
16:01:28 <akuznetsova> hello)
16:01:50 <rakhmerov> let's wait for others a little bit
16:02:11 <nmakhotkin_> hi!
16:02:38 <rakhmerov> hi
16:02:44 <rakhmerov> ok, let's start
16:03:36 <rakhmerov> #topic Action Items
16:04:15 <rakhmerov> we didn't actually have action items for the last week except the same ongoing thing related to discussion different workflow types
16:04:40 <rakhmerov> one thing that Fuel folks want us to do is implement a workflow type based on task priorities
16:05:09 <bhavenst> Hard priorities or weighting?
16:05:12 <bhavenst> Just curious
16:05:13 <rakhmerov> it's when basically we specify a bunch of tasks with priorities and Mistral figures out the order of execution
16:05:21 <dzimine> hi all.
16:05:46 <rakhmerov> bhavenst, it's just a general idea
16:05:50 <bhavenst> Sounds pretty non-deterministic
16:05:52 <bhavenst> cool
16:05:56 <rakhmerov> details are to be discussed :)
16:06:16 <rakhmerov> most likely next week I'll be able to tell more exactly
16:06:28 <rakhmerov> when I get familiar with what they do in Fuel
16:06:32 <rakhmerov> hi Dmitri )
16:06:40 <rakhmerov> ok, so let's move on
16:06:57 <rakhmerov> #topic Current status (progress, issues, roadblocks, further plans)
16:07:18 <rakhmerov> let's quickly report everyone's status
16:08:16 <bhavenst> Started looking into our (Ericsson) blueprints for mistral.  Haven't managed to discuss w/ author, but found a little more doc.  Have a few questions and plan to discuss a bit today.
16:08:26 <rakhmerov> my status: designed how policies and engine instructions (like 'pause', 'fail') should work, made a bunch of changes related with policies, a lot of bug fixes and reviews
16:08:50 <akuznetsova> I have written suite of integration tests for CLI v2, i'm planning to stark working on test for API v2
16:08:57 <rakhmerov> bhavenst, cool, I included a topic about that, we'll get to it
16:09:02 <bhavenst> great
16:09:34 <rakhmerov> akuznetsova, were you able to find out what the problem was with v2 tests?
16:09:46 <rakhmerov> 'no reply on topic..' thing
16:10:05 <_nmakhotkin> last week I finished work on client v2 and CLI
16:10:24 <akuznetsova> rakhmerov, me not, Nikolay can tell more about it
16:10:41 <_nmakhotkin> also I implemented policies: wait-before, wait-after and retry
16:10:57 <rakhmerov> ok
16:11:21 <akuznetsova> rakhmerov, I asked _nmakhotkin  to help me and he said that he knows what the problem is
16:11:25 <rakhmerov> _nmakhotkin, can you briefly tell about that? Do we know exactly what happens?
16:11:53 <_nmakhotkin> fake rpc_backend doesn't work as expected
16:11:59 <_nmakhotkin> only in engine v2
16:12:05 <rakhmerov> do we know how to fix it?
16:13:42 <rakhmerov> _nmakhotkin?
16:13:55 <_nmakhotkin> not yet, but we'll find a way to fix it
16:14:10 <rakhmerov> ok
16:14:20 <rakhmerov> I hope there's nothing serious
16:15:11 <rakhmerov> dzimine, all client related patches have been merged in
16:15:23 <dzimine> good.
16:15:24 <rakhmerov> can we borrow Kirill for a couple of days now? :)
16:15:43 <dzimine> later this week, yes.
16:15:48 <rakhmerov> ok
16:16:04 <rakhmerov> #topic Release 0.1 progress (go through the list of what's left)
16:16:34 <rakhmerov> https://launchpad.net/mistral/+milestone/0.1
16:16:53 <rakhmerov> this is a page with information about the release we're now working on
16:17:21 <rakhmerov> basically, the only thing we haven't touched so far is https://blueprints.launchpad.net/mistral/+spec/mistral-engine-instructions
16:17:33 <rakhmerov> which is on me, I'm planning to start working on it tomorrow
16:17:48 <rakhmerov> it should take about two days
16:18:15 <_nmakhotkin> I forgot to say - guys who work on UI for actions/tasks ask me today and they need endpoint for actions
16:18:35 <rakhmerov> and the second one which we've only scratched a little bit is https://blueprints.launchpad.net/mistral/+spec/mistral-ui
16:18:51 <rakhmerov> hopefully, it won't take too long as well
16:19:41 <rakhmerov> _nmakhotkin, I've already made some changes related to this. May be I'll do it tomorrow, it should take about an hour or less
16:20:13 <_nmakhotkin> yes :)
16:20:26 <_nmakhotkin> just for information
16:20:37 <rakhmerov> anyways, we can spread it between two of us so that everything is ready by the moment Kirill starts working on UI changes
16:20:43 <akuznetsova> will I have enough time to test it? cause i've never seen our ui )
16:20:56 <rakhmerov> :)))
16:20:58 <rakhmerov> sure
16:21:30 <akuznetsova> it's good
16:21:39 <rakhmerov> so the overall picture now is that we're mostly on target with the release
16:22:21 <rakhmerov> but I assume we may not be able to finish everything up
16:22:41 <rakhmerov> by the end of the week
16:22:59 <rakhmerov> mostly I mean some docs, all the testing, maybe small bugs
16:23:18 <rakhmerov> so it may take 2-3 days longer
16:23:21 <_nmakhotkin> .. and demo :)
16:23:51 <rakhmerov> yeah
16:24:30 <rakhmerov> on REST API/Client/CLI I guess the only thing that left is actions, correct?
16:24:40 <rakhmerov> _nmakhotkin, akuznetsova?
16:24:52 <_nmakhotkin> yes, correct
16:25:00 <akuznetsova> rakhmerov, "maybe small bugs"  - you are optimistic)
16:25:12 <rakhmerov> :) I am
16:26:00 <rakhmerov> #action rakhmerov, nmakhotkin, enykeev: Add action into REST API/Client/CLI
16:26:33 <rakhmerov> Nikolay, did you have time to look at action params etc.?
16:26:40 <rakhmerov> you know what I mean..
16:27:19 <rakhmerov> so that we could see some info on actions when we type 'mistral action-list' or more specifically 'mistral action-get nova.servers_get' for example
16:27:57 <_nmakhotkin> yes, I try to do it
16:28:02 <rakhmerov> info about their parameters and, if possible, short description taken from docstrings
16:28:04 <_nmakhotkin> maybe tomorrow
16:28:07 <rakhmerov> ok
16:28:34 <_nmakhotkin> Merlin guys also asked me about it )
16:28:52 <rakhmerov> #action nmakhotkin, we should see action parameters and description (if possible) when calling API
16:29:02 <rakhmerov> yeah, Timur asked me about that too
16:29:23 <rakhmerov> otherwise, it will be a nightmare to work with all this huge number of actions
16:29:34 <rakhmerov> we should be able to help somehow
16:29:40 <rakhmerov> do what we can do
16:29:58 <rakhmerov> ok, workflow policies are mostly done, right?
16:30:13 <rakhmerov> btw, folks from Fuel asked us about one more policy
16:30:17 <akuznetsova> _nmakhotkin, will we have filters in actions? for example mistral action-list type=nova ?
16:30:21 <rakhmerov> which should be pretty simple to implement
16:30:25 <rakhmerov> timeout
16:30:40 <rakhmerov> akuznetsova, yes, good call
16:30:47 <rakhmerov> currently they're not implemented
16:30:50 <dzimine> oh yes, timeout as a task policy, especially on async tasks.
16:31:00 <rakhmerov> dzimine, yep )
16:31:21 <rakhmerov> it actually escaped from my head somehow even though it's very important
16:31:44 <_nmakhotkin> if we want it, it will be
16:31:49 <akuznetsova> _nmakhotkin, is it the same like tags in workflow/workbook?
16:32:07 <akuznetsova> i mean implementation
16:32:14 <rakhmerov> #action nmakhotkin, enykeev: Implement filters on REST API endpoints which work with multiple items (such as /tasks, /workbooks etc.)
16:32:33 <rakhmerov> no, it's not really the same
16:32:57 <rakhmerov> tags are not used anywhere now
16:33:17 <_nmakhotkin> we can use tags in our UI
16:33:24 <rakhmerov> but I see what you're going to akuznetsova
16:33:30 <rakhmerov> ...where...
16:33:34 <rakhmerov> yep
16:33:48 <rakhmerov> and potentially we can do filtering based on them too
16:34:00 <rakhmerov> a little bit different type of filtering
16:34:07 <akuznetsova> rakhmerov, should i create a blueprint for filters or it is part of some existing?
16:34:27 <rakhmerov> well, I guess it should be a part of existing BPs
16:34:29 <dzimine> I suggest a blueprint - so we can prioritize it.
16:34:33 <rakhmerov> but let's check real quick..
16:34:54 <rakhmerov> no, it's not a part of BPs
16:35:05 <rakhmerov> akuznetsova, can you create it for us pls?
16:35:18 <rakhmerov> dzimine, agree
16:35:39 <akuznetsova> yes, sure
16:35:44 <rakhmerov> thanks a lot!
16:36:17 <rakhmerov> so other than that I think we're good
16:36:46 <rakhmerov> except Nikolay found some bugs with action/workflow parameters
16:36:51 <rakhmerov> not even bugs
16:37:06 <rakhmerov> we just need to do some enhancements
16:38:07 <rakhmerov> #action nmakhotkin, rakhmerov: make it possible to use different data types in action/workflow params (None, False/True, numbers, expressions)
16:38:19 <rakhmerov> ok, we have about 20 mins left
16:38:38 <rakhmerov> guys, I included an item to discuss Metrics Collector BPs
16:38:45 <rakhmerov> let's switch to it now
16:39:28 <rakhmerov> #topic Metrics collector BPs
16:39:53 <rakhmerov> so basically we have two BPs:
16:39:57 <rakhmerov> 1. https://blueprints.launchpad.net/mistral/+spec/mistral-metrics-collector
16:40:06 <rakhmerov> 2. https://blueprints.launchpad.net/mistral/+spec/mistral-metrics-collector-api
16:40:36 <rakhmerov> submitted by folks from Ericsson
16:41:31 <rakhmerov> The idea seems to be pretty interesting to me and I think at this point we could let Bryan (or anyone else) start working on them
16:42:08 <rakhmerov> so I would ask all of you to take a look at them, leave your comments/remarks/questions
16:42:18 <bhavenst> Yeah, would definitely appreciate that
16:42:56 <rakhmerov> the main thing here is to stay consistent with the rest of the system in terms of design, terminology etc.
16:43:06 <bhavenst> The way it's written, it would persist nothing in Mistral db, but use Ceilometer for that.  So that's one question I have.
16:43:11 <bhavenst> if that's best
16:43:13 <rakhmerov> I got the general idea but I think we just need to polish details
16:43:28 <rakhmerov> ok
16:43:33 <rakhmerov> sounds good to me
16:43:36 <bhavenst> And then architecture of either a) fixed set of measurements or b) some way to create custom measurement
16:43:56 <bhavenst> and finally for API, should there be a new endpoint or put it on top of the existing endpoints
16:44:38 <rakhmerov> yes
16:44:51 <rakhmerov> at this point I see it as a new endpoint
16:44:56 <bhavenst> OK
16:45:25 <rakhmerov> I think it makes sense to separate this because it's sort of special functionality
16:45:33 <rakhmerov> reporting
16:47:02 <rakhmerov> which may not be even related with individual objects but rather providing averages and other integral (not sure it's the right word :) ) information
16:47:15 <rakhmerov> so I'm 99% sure it should be separate
16:47:48 <rakhmerov> as far as set of measurements it's a little hard to tell now
16:48:35 <rakhmerov> I would suggest that we start digging into that first and may be create some more detailed spec (a simple etherpad would be fine)
16:48:54 <rakhmerov> with description of what exactly we're going to implement
16:49:23 <bhavenst> OK will do
16:49:45 <rakhmerov> and then let our imagination do a job of improving, polishing and making it more flexible etc etc
16:49:50 <bhavenst> Like I said, I'll talk to the author as well.  It's mostly obvious but maybe he can give some more details regarding intended purpose etc
16:49:54 <bhavenst> Sounds good
16:50:01 <akuznetsova> btw, https://blueprints.launchpad.net/mistral/+spec/mistral-items-filtering
16:50:12 <rakhmerov> ok, thanks :)
16:50:26 <rakhmerov> bhavenst, you mean Sergey?
16:50:32 <bhavenst> yeah exactly
16:50:36 <rakhmerov> is he from Ericsson too?
16:50:39 <bhavenst> Yes
16:50:45 <rakhmerov> ok
16:51:00 <bhavenst> He might have had an @conceptwave.com mail address or something earlier, but Ericsson now.
16:51:40 <bhavenst> Oh I see was @gmail
16:51:45 <bhavenst> yeah, he's in our project
16:51:47 <rakhmerov> #action bhavenst: start working on Metrics Collector, the first step is to make the requirements more detailed and capture them in a specification document (e.g. etherpad)
16:52:31 <rakhmerov> btw, there are other folks who asked us about similar thing
16:52:38 <rakhmerov> from Rackspace
16:53:01 <rakhmerov> it was one of their major requirements for some reason )
16:53:22 <rakhmerov> but it was around 10 months back when we just started the project
16:53:29 <bhavenst> ok good to know
16:53:33 <rakhmerov> yeah
16:53:55 <rakhmerov> btw, I can actually share a document they've written with you
16:53:58 <bhavenst> I guess they didn't formulate any
16:54:00 <bhavenst> oh great
16:54:07 <rakhmerov> it was a detailed whitepaper
16:54:23 <bhavenst> Then I can try to merge the two as part of the action pt
16:54:26 <rakhmerov> but I have to ask their permission first though
16:54:28 <bhavenst> sure
16:54:31 <bhavenst> no problem
16:54:42 <rakhmerov> yes, merging I think is a good idea
16:56:03 <rakhmerov> from my perspective, I don't clearly see how it should be integrated with Ceilometer
16:56:47 <rakhmerov> but I'm not a Ceilometer expert, I just may need to digg some info on that
16:57:03 <rakhmerov> or we could discuss it either personally or in IRC
16:57:15 <bhavenst> I have to dig into it as well :)
16:57:21 <rakhmerov> :)
16:57:24 <rakhmerov> it's normal
16:57:29 <rakhmerov> OpenStack is huge
16:57:32 <rakhmerov> almost endless
16:57:36 <bhavenst> We have one of the core ceilometer folks here in E/// so can talk to her too
16:57:45 <rakhmerov> great
16:58:16 <rakhmerov> ok, bhavenst, please keep us posted on this work and feel free to ask anything you want
16:58:18 <bhavenst> will do
16:58:19 <bhavenst> thanks
16:58:31 <akuznetsova> we have a couple of 'ceilometer' guys in our location (me and _nmakhotkin )
16:58:42 <bhavenst> ok, good to know!
16:58:53 <rakhmerov> ooh, you know Ceilometer well?
16:59:10 <rakhmerov> ok, we've almost done here hah? ) Have to free the room
16:59:10 <akuznetsova> me - no) i know who knows
16:59:18 <rakhmerov> ooh, alright
16:59:21 <rakhmerov> ok, guys
16:59:28 <rakhmerov> thanks for coming!
16:59:30 <_nmakhotkin> bye!
16:59:33 <bhavenst> bye
16:59:33 <akuznetsova> bye
16:59:34 <rakhmerov> have a good one )
16:59:37 <bhavenst> same to you
16:59:39 <rakhmerov> bye
16:59:45 <rakhmerov> #endmeeting