16:01:03 <rakhmerov> #startmeeting Mistral
16:01:03 <openstack> Meeting started Mon Feb 13 16:01:03 2017 UTC and is due to finish in 60 minutes.  The chair is rakhmerov. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:01:05 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:01:07 <openstack> The meeting name has been set to 'mistral'
16:01:09 <d0ugal> Hey!
16:01:12 <mgershen1> o/
16:01:13 <ddeja> o/
16:01:22 <rakhmerov> hi everyone )
16:01:52 <rakhmerov> as you may guess we won't have a meeting next week
16:01:59 <rakhmerov> and the week after
16:02:21 <rakhmerov> ok, let's start
16:02:30 <rakhmerov> #topic Review Action Items
16:02:33 <rakhmerov> no action items
16:02:42 <rakhmerov> #topic Current Status
16:03:59 <rakhmerov> my status: completed refactoring of RPC serialization together with Dawid (https://review.openstack.org/#/c/429643/), fixed one simple bug with workflow context internal data, almost finished working on PTG etherpad
16:04:17 <rakhmerov> share your updates, if any
16:04:27 <d0ugal> Why wont there be a meeting after the PTG?
16:04:44 <rbrady> no update from me.  just preparing for PTG
16:04:47 <ddeja> Not much from my side, helped Renat with RPC serialization
16:05:01 <d0ugal> No update from me, other than I have tired legs from skiing.
16:05:07 <rakhmerov> d0ugal: well, you can have it, but I won't be available
16:05:18 <d0ugal> rakhmerov: okay, that's fine - we can skip it. Just checking.
16:05:21 <rakhmerov> I'll take a couple of days off
16:05:28 <d0ugal> Good idea :)
16:05:49 <mgershen> working on documenting the tasks function. https://review.openstack.org/#/c/433096/
16:05:54 <rakhmerov> d0ugal: let's get back to it in the end of this week. If you'll have what to discuss, why not have it
16:06:05 <rakhmerov> mgershen: is it ready for review?
16:06:24 <mgershen> yes
16:06:31 <rakhmerov> ok, great
16:06:46 <rakhmerov> d0ugal: still waiting for your review on https://review.openstack.org/#/c/429643/
16:06:57 <d0ugal> mgershen: nice! I'll look check it out.
16:07:00 <rakhmerov> i'd like to push it into rc2
16:07:09 <mgershen> and I really want input on how to document it, because I fear we might want to change what exactly should return for each task in the list
16:07:37 <mgershen> d0ugal: thanks, please do
16:07:46 <d0ugal> rakhmerov: I actually did review it, I just didn't vote yet. I wasn't sure I was awake enough to fully understand it
16:07:50 <d0ugal> rakhmerov: I think it looks good.
16:07:50 <rakhmerov> mgershen: do not fear, may the OpenStack force be with you
16:07:52 <rakhmerov> :)
16:08:27 <d0ugal> rakhmerov: I guess either you or ddeja needs to review it? who else would be the second reviewer?
16:08:29 <mgershen> haha :)
16:08:32 <rakhmerov> d0ugal: did you get an idea and what problem I was trying to solve?
16:08:55 <rakhmerov> d0ugal: mostly Dawid
16:08:57 <d0ugal> rakhmerov: no, not fully - it would be good it the problem was stated in the commit message :)
16:09:02 <rakhmerov> well, I don't know, we both can
16:09:26 <rakhmerov> d0ugal: how about talking tomorrow in IRC?
16:09:34 <d0ugal> rakhmerov: sure, that sounds good.
16:09:46 <rakhmerov> I listed some major points, maybe still need more details, dunno
16:09:50 <rakhmerov> I can explain
16:09:56 <rakhmerov> ok
16:10:01 <rakhmerov> let's move on
16:10:13 <rakhmerov> #topic Ocata RC-2 readiness
16:10:31 <mgershen> if anyone else wants to review here is the docs preview (gate-tripleo-ci-centos-7-nonha-multinode test is stuck for a while): http://docs-draft.openstack.org/96/433096/1/check/gate-mistral-docs-ubuntu-xenial/00c127f//doc/build/html/dsl/dsl_v2.html#tasks-function
16:10:34 <d0ugal> rakhmerov: the points in the commit message say what you did, but I don't think they really explain "why" - but maybe I need to read it again
16:10:38 <rakhmerov> this we're releasing RC2 which is supposed to be the final Ocata release
16:11:09 <rakhmerov> d0ugal: I'll look at the message again and see if I can improve it
16:11:25 <d0ugal> thanks
16:11:28 <rakhmerov> mgershen: thanks, usually TripleO gate takes long, that's normal
16:12:16 <rakhmerov> d0ugal: briefly, our serialization system didn't work for custom types at all. Although we had, for example, ResultSerializer. It wasn't just used
16:12:38 <rakhmerov> d0ugal: moving forward we'll need more custom classes sent over RPC
16:13:10 <rakhmerov> I looked at OVO (oslo versioned project) which was a good candidate to use, but I thought it was kind of an overkill for what I wanted to achieve
16:13:25 <rakhmerov> ok, so RC2 this week
16:13:52 <rakhmerov> I'd like to ask all of you to check whether you want to squeeze anything else into RC2
16:14:09 <rakhmerov> and please be ready to review stuff in case of some urgent bugs
16:14:15 <rakhmerov> d0ugal and ddeja especially
16:14:18 <ddeja> rakhmerov: I'll have one thing
16:14:27 <rakhmerov> test coverage?
16:14:32 <ddeja> worse
16:15:00 <d0ugal> I don't have anything.
16:15:00 <rakhmerov> ok
16:15:06 <ddeja> today I looked on test coverage, why it may fail and why on earth kombu driver have coverage 0%
16:15:06 <rakhmerov> ddeja: what's that?
16:15:15 <rakhmerov> so?
16:15:19 <ddeja> so I did this: https://review.openstack.org/#/c/433074/
16:15:36 <ddeja> so, from some point (I have no idea when it stared)
16:15:54 <ddeja> tests from unit/engine/rpc_backend/kombu/* didn't get run at all
16:15:56 <rakhmerov> wooooooow!!
16:16:05 <d0ugal> :(
16:16:06 <mgershen> X.X
16:16:07 <rakhmerov> epic fail :))
16:16:14 <ddeja> I remember when it was failing
16:16:24 <rakhmerov> goosh..
16:16:34 <ddeja> becouse I was making changes that was making tox to fail and I was fixing unit tests
16:16:59 <ddeja> but when I was making kombu driver work with rabbit in HA, I haven't fixed any unit tests
16:17:03 <d0ugal> ddeja: the __init__ was deleted here: https://github.com/openstack/mistral/commit/5341fb2b5bc255410a04726ef24de2d95607c1f6
16:17:03 <rakhmerov> ok, do you think it's that serious?
16:17:13 <rakhmerov> is there a lot of stuff that really needs to be fixed?
16:17:15 <rakhmerov> I can help
16:17:21 <ddeja> rakhmerov: not a lot
16:17:25 <ddeja> I'll do it tommorow
16:17:30 <rakhmerov> I need this 100% ready
16:17:42 <ddeja> It's only 15 tests after all
16:17:46 <ddeja> oh, thanks d0ugal
16:18:17 <rakhmerov> ok, it was done by mistake I guess. Happens sometimes
16:18:19 <rakhmerov> no worries
16:18:23 <d0ugal> yup
16:18:25 <rakhmerov> let's just fix it
16:18:27 <ddeja> hm, pycharm
16:18:39 <rakhmerov> ddeja: I'll be available to help
16:18:42 <ddeja> when I deleted the only file in dir, it also deleted __init__
16:18:49 <d0ugal> :(
16:18:50 <rakhmerov> any time of the day and night
16:18:56 <d0ugal> This is why I use vim :-D
16:19:25 <ddeja> rakhmerov: thanks, but from the logs it's nothing hard to fix
16:19:27 <rakhmerov> d0ugal: I'm thinking to learn it back (I used to use it a lot ~8 years ago)
16:19:32 <rakhmerov> ok
16:19:50 <ddeja> I should have it by tommorow 12 UTC
16:20:02 <ddeja> if no, I'll let you guys know :)
16:20:08 <rakhmerov> d0ugal: any chance we can get https://bugs.launchpad.net/mistral/+bug/1627689 fixed?
16:20:08 <openstack> Launchpad bug 1627689 in Mistral "Openstack service client caching breaks security and backward compatibility" [Medium,In progress] - Assigned to Dougal Matthews (d0ugal)
16:20:20 <rakhmerov> I guess unlikely, right?
16:20:30 <d0ugal> rakhmerov: good question. I think it is quite complicated so it might be a bit late
16:20:41 <d0ugal> I had forgotten about it
16:21:02 <rakhmerov> :) Ok, in Pike I'll do a better job reminding you about it
16:21:20 <d0ugal> hah, sorry - I should do a better job of keeping up with my bugs.
16:21:28 <d0ugal> Ocata has been too short :(
16:21:47 <rakhmerov> yes, true
16:22:17 <rakhmerov> ok, then please look at your tasks and move them to Pike if you see you won't complete them
16:22:52 <rakhmerov> I'll try to finish my two tasks on doc
16:23:44 <rakhmerov> d0ugal: I'd ask you to look at https://bugs.launchpad.net/mistral/+bug/1641174
16:23:44 <openstack> Launchpad bug 1641174 in Mistral "Update the Mistral documentation to reflect the addition of Jinja2" [Medium,Confirmed] - Assigned to Dougal Matthews (d0ugal)
16:23:58 <rakhmerov> maybe you'll have time in the next 2 days to do it
16:25:09 <d0ugal> rakhmerov: I assume documentation changes can be added after RC2?
16:25:15 <d0ugal> but yeah, I will try and look into it
16:25:22 <rakhmerov> ok
16:25:26 <mgershen> sorry for the stupid question, where can I see all the issue in mistral that are assigned to me in Ocata?
16:25:31 <d0ugal> but we really need to avoid getting into this situation again - can we require docs for features from Pike onwards?
16:25:43 <rakhmerov> mgershen: https://launchpad.net/mistral/+milestone/ocata-rc2
16:25:54 <rakhmerov> it's for Ocata RC2
16:26:46 <rakhmerov> d0ugal: yes, we touched this topic on the previous meeting. We'll need to require docs for all new features as part of "definition of done" for any task
16:26:47 <mgershen> rakhmerov: thank you
16:26:55 <d0ugal> rakhmerov: great
16:27:14 <d0ugal> I shall read the meeting log
16:27:36 <rakhmerov> d0ugal: having said that, I still think it should be OK to split into separate patches but we should look at "Implements" and "Closes-bug" more carefully and require docs
16:27:59 <d0ugal> yeah, maybe we should require at least a docs draft
16:28:01 <rakhmerov> it could be sent as dependent patches
16:28:11 <d0ugal> yup
16:28:12 <rakhmerov> yes, something like this
16:28:27 <rakhmerov> d0ugal: we can even require docs in the same patch if it's something small
16:28:29 <d0ugal> It also makes it easier to review a new feature if you have docs :)
16:28:34 <rakhmerov> I mean, common sense will help
16:28:42 <rakhmerov> true
16:29:19 <rakhmerov> d0ugal: a general comment: we're very bad at planning now
16:29:22 <mgershen> I think all suggestions are good.
16:29:27 <rakhmerov> which is, of course, mostly my fault
16:29:41 <rakhmerov> I'd appreciate if you helped me improve in this aspect
16:30:04 <d0ugal> I can try - how can we make planning more of a team effort?
16:30:44 <rakhmerov> first thing I see: be more active with that in ML
16:30:56 <rakhmerov> schedule meetings for planning
16:31:02 <rakhmerov> for but triaging
16:31:06 <rakhmerov> etc.
16:31:29 <rakhmerov> I'll come up with proposals on what we need to be doing
16:31:35 <d0ugal> sounds good
16:31:49 <rakhmerov> and that should be done no matter what :)
16:32:20 <rakhmerov> except very rare cases may be like summits, PTGs etc.
16:33:01 <rakhmerov> next topic
16:33:13 <rakhmerov> #topic PTG topics final review
16:33:38 <rakhmerov> https://etherpad.openstack.org/p/mistral-ptg-pike
16:34:16 <rakhmerov> I put some more details into it today (Dawid did too)
16:34:51 <d0ugal> Looks like a great doc.
16:34:52 <rakhmerov> I'd suggest you review the list of topic again and see if something important is missing
16:35:14 <rakhmerov> either in the topic list itself and/or some info about particular topics
16:35:21 <rakhmerov> d0ugal: thanks, trying hard )
16:35:50 <rakhmerov> A couple of things are still don't have enough details
16:35:55 <rakhmerov> like Mistral/Glare
16:36:03 <rakhmerov> which is an important thing IMO actually
16:36:12 <d0ugal> I am curious to hear about that :)
16:36:14 <rakhmerov> I'll do it tomorrow
16:36:17 <ddeja> rakhmerov: BTW, I'd like to ask for that thing
16:36:29 <rakhmerov> d0ugal: yeah, I guess I already told you a little bit about this idea
16:36:30 <ddeja> who's request is to add this support?
16:36:42 <rakhmerov> it's been my dream for a couple of years to implement something like this
16:37:00 <rakhmerov> ddeja: it's been mostly my idea
16:37:12 <ddeja> ok, so it's an inner request, not coming from some company :)
16:37:17 <rakhmerov> but people asked for something similar too
16:37:17 <ddeja> thanks
16:37:31 <rakhmerov> in a different form maybe but the reason is the same
16:37:38 <rakhmerov> coming, coming
16:38:00 <rakhmerov> not only internal
16:38:05 <ddeja> ok
16:38:14 <rakhmerov> now we just finally have resources to implement it
16:38:17 <rakhmerov> really
16:38:29 <rakhmerov> I think it will be done in the next few months
16:38:47 <rakhmerov> ddeja: for example, kong was requesting it too
16:38:54 <ddeja> OK
16:39:03 <rakhmerov> he even tried to implement a similar thing already
16:39:16 <rakhmerov> the patch is still on review somewhere, I can find it
16:39:22 <rakhmerov> the demand was the same
16:39:30 <d0ugal> I hadn't even heard of Glare before this :)
16:39:43 <rakhmerov> ok, I'll let you all know when I put more details (ideally with WF snippets)
16:40:03 <rakhmerov> d0ugal: yeah, it's making some noise now in the community
16:40:22 <rakhmerov> there's even an idea to replace most of Glance with it
16:40:24 <rakhmerov> :)
16:40:54 <rbrady> I only saw that recently in the ML thread.  Has there been other discussions about glare vs glance somewhere else?
16:41:14 <rakhmerov> btw, Mike (Glare PTL) will be at the PTG too, we'll be able to discuss this topic with him
16:41:45 <rakhmerov> rbrady: I guess not yet, but there will be definitely some
16:42:10 <rakhmerov> stay tuned
16:43:27 <rakhmerov> just a couple of words about how I'd like to approach PTG discussions: I formally divided all topics into three categories (functional reqs, non functional, architecture)
16:43:45 <rakhmerov> I'd like to discuss at least top 3 in each categories
16:44:14 <rakhmerov> so we need to make sure that top 3 are really that we think have higher priority
16:45:15 <rakhmerov> by "discuss" I mean to have all necessary work items be able to proceed with it, all conceptional roadblocks should be clarified and a solution is found
16:45:32 <rakhmerov> that's my goal
16:45:44 <ddeja> sounds good
16:45:55 <rakhmerov> yeah
16:46:10 <mgershen> how many hours do we have?
16:46:10 <rakhmerov> let's see how many people will come though
16:46:17 <rakhmerov> mgershen: 3 days
16:46:37 <rakhmerov> practically, we'll have 2.5 days
16:46:42 <d0ugal> rakhmerov: I think it would be a good idea to try and give some days/times for the different topics
16:46:44 <rakhmerov> I don't count Fri afternoon
16:47:04 <d0ugal> Otherwise if somebody can't attend every day, how will they know when to come?
16:47:16 <rakhmerov> d0ugal: yes, reading my thoughts, I was gonna say I'd plan this for tomorrow
16:47:23 <d0ugal> cool
16:47:39 <rakhmerov> d0ugal: agree
16:47:43 <d0ugal> way ahead of me then :)
16:48:23 <rakhmerov> rbrady, d0ugal: I'd ask you to look at Actions API topic and add all things that you're worried about
16:48:36 <rbrady> rakhmerov: ack
16:48:37 <rakhmerov> so that we can discuss productively
16:48:44 <d0ugal> +1
16:48:49 <rakhmerov> I put all that I was able to recall today
16:48:52 <rakhmerov> :)
16:49:08 <rakhmerov> can't get more from my head I guess )
16:49:17 <d0ugal> I will look at this tomorrow
16:49:24 <rakhmerov> yep, thx
16:50:21 <rakhmerov> so, just summarizing: add whatever info you think is needed, I'll be looking at it again and fixing/beautifying before the event
16:50:33 <rakhmerov> it's better to have more than less
16:50:38 <rakhmerov> for this kind of work
16:50:54 <rakhmerov> that's probably all from my side
16:51:23 <d0ugal> Sounds good
16:51:35 <rakhmerov> any other comments/remarks?
16:51:57 <d0ugal> Nothing from me
16:52:03 <rakhmerov> ok
16:52:07 <rakhmerov> #topic Open Discussion
16:52:30 <rakhmerov> anything else may be on other things?
16:52:37 <rakhmerov> mgershen, ddeja, rbrady
16:52:46 <ddeja> nothing from my side
16:52:50 <rakhmerov> ok
16:52:59 <mgershen> I really want feedback about the docs, but it can be done there.
16:53:22 <rakhmerov> mgershen: no worries, we'll review your patch and give a feedback
16:53:32 <rakhmerov> or you want it now?
16:53:33 <rbrady> nothing from me.  I'll look at updating the etherpad
16:53:40 <rakhmerov> ok
16:54:12 <rakhmerov> mgershen: ^
16:54:15 <mgershen> so see you next week?
16:54:20 <rakhmerov> ok :)
16:54:21 <rakhmerov> yes
16:54:24 <mgershen> and in the review:-D
16:54:27 <d0ugal> yup, next week or in #openstack-mistral
16:54:29 <rakhmerov> right )
16:54:30 <d0ugal> ... or in gerrit!
16:54:44 <rakhmerov> ok, thanks guys
16:54:51 <d0ugal> cya
16:54:52 <rakhmerov> pleasure to work with you
16:54:53 <rakhmerov> bye
16:54:58 <mgershen> bye
16:55:03 <rakhmerov> #endmeeting