14:01:34 <andreykurilin__> #startmeeting Rally
14:01:35 <openstack> Meeting started Mon Apr 18 14:01:34 2016 UTC and is due to finish in 60 minutes.  The chair is andreykurilin__. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:01:36 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:01:39 <openstack> The meeting name has been set to 'rally'
14:01:56 <ikhudoshyn> 0/
14:02:02 <amaretskiy> hi
14:02:47 <Yanyanhu> hello
14:03:29 <andreykurilin__> \o
14:03:36 <andreykurilin__> let's wait a bit
14:03:57 <andreykurilin__> boris-42 stpierre: ping
14:04:09 * stpierre is here
14:05:08 <andreykurilin__> good
14:05:15 <andreykurilin__> ok, let's start
14:05:39 <andreykurilin__> good news - we have an agenda for today:)
14:05:42 <andreykurilin__> #link https://wiki.openstack.org/wiki/Meetings/Rally
14:05:54 <andreykurilin__> #topic [andreykurilin] Release 0.4.0
14:06:21 <andreykurilin__> This topic was planned as happy news :)
14:06:53 <andreykurilin__> but we have problems with cutting release
14:06:58 <andreykurilin__> the process was changed
14:07:12 <andreykurilin__> and now we unable to cut release easier
14:07:20 <andreykurilin__> we need to do it via openstack/releases repo
14:08:00 <ikhudoshyn> we're under big tent, so bi bro's watching
14:08:04 <ikhudoshyn> )
14:08:17 <andreykurilin__> I push a change for new release, but gates queue is too big (250+ changes)
14:08:19 <andreykurilin__> #link https://review.openstack.org/#/c/307139/
14:08:19 <patchbot> andreykurilin__: patch 307139 - releases - Release Rally 0.4.0
14:08:48 <Yanyanhu> andreykurilin__, gate is almost broken now...
14:08:54 <andreykurilin__> yeah:(
14:09:12 <andreykurilin__> so cores please do not merge anything before new release come
14:09:15 <boris-42> andreykurilin__: hi
14:09:21 <andreykurilin__> boris-42: hi
14:09:40 <andreykurilin__> PS: release notes were merged
14:09:55 <boris-42> andreykurilin__: so now we should make that patch?
14:10:05 <boris-42> andreykurilin__: instead pushing tag by hands?
14:10:11 <andreykurilin__> boris-42: yes
14:10:21 <boris-42> andreykurilin__: heh=( no freedome
14:10:27 <andreykurilin__> yeah
14:10:50 <andreykurilin__> release notes for new release
14:10:51 <andreykurilin__> #ling http://rally.readthedocs.org/en/latest/release_notes/archive/v0.4.0.html
14:11:37 <andreykurilin__> ok, let move to the next topic)
14:11:49 <andreykurilin__> [andreykurilin] Planning for next release
14:11:52 <andreykurilin__> #topic [andreykurilin] Planning for next release
14:12:04 <rvasilets> hi
14:12:22 <andreykurilin__> as we discussed several meeting ago, next release should implement wrappers/utils refactoring spec
14:12:52 <andreykurilin__> #link https://github.com/openstack/rally/blob/master/doc/specs/in-progress/refactor_scenario_utils.rst
14:13:27 <andreykurilin__> lets discuss our steps
14:13:30 <andreykurilin__> and policy
14:14:01 <amaretskiy> proposal - let's include trends report (https://review.openstack.org/#/c/276721/) to plans for next release
14:14:01 <patchbot> amaretskiy: patch 276721 - rally - [Reports][CLI][CI] Introduce Trends report
14:14:20 <andreykurilin__> I propose to not stop merging fix to partial wrappers/utils until work for this partial service is not started
14:14:51 <ikhudoshyn> amaretskiy: I believe andreykurilin__ was askin' bout wrappers stuff policy ))
14:14:57 <rvasilets> When we will start implement services and who will do this?
14:15:07 <andreykurilin__> ikhudoshyn: +1 :)
14:15:23 <andreykurilin__> rvasilets: I think boris-42 wanted to implement the base stuff:)
14:16:17 <rvasilets> Ok boris-42 will implement base and who will move glance, nova, murano... etc to service base?)
14:16:28 <amaretskiy> i can do this work but boris-42 told me that there are some tasks waiting for me (I do not know what exact tasks)
14:16:46 <rvasilets> Will we split this work to the whole team?
14:16:48 <ikhudoshyn> I'm pretty busy as well
14:16:56 <boris-42> andreykurilin__: +2
14:17:07 <rvasilets> Or this will be done by one person7
14:17:07 <stpierre> iirc we had discussed boris-42 doing the base and keystone, and just getting that into the next release
14:17:11 <boris-42> andreykurilin__: I want to stop merging those things as well and just start working on the base
14:17:23 <boris-42> stpierre: that's like minial goal
14:17:24 <stpierre> and waiting for the subsequent release to try to get other services done
14:17:53 <boris-42> stpierre: nope I was talking about having at least keystone done in next release
14:18:01 <rvasilets> ok boris will move keystone I can help with some other part
14:18:15 <andreykurilin__> boris-42: why we can't merge small fixes to glance wrapper for example if we don't start work on it?
14:18:49 <stpierre> andreykurilin__++
14:19:09 <andreykurilin__> let's block patches only for keystone stuff  now
14:19:11 <rvasilets> because they are unneeded
14:19:18 <andreykurilin__> rvasilets: why?
14:19:25 <boris-42> stpierre: andreykurilin__ is it so hard to do the services base?
14:19:35 <stpierre> it makes sense to declare each wrapper deprecated once we have an actual WIP to replace it, but given how busy everyone has just declared themselves i think it's awfully optimistic to pretend that we'll get this all done quickly enough
14:19:50 <rvasilets> andreykurilin__, because we will rewrite then in the nearest month
14:19:55 <stpierre> boris-42: i dunno, the spec took well over 6 months to get merged, i don't expect anything else to be any quicker
14:20:09 <stpierre> rvasilets: and in the mean time there are real actual bugs that people are dealing with
14:20:15 <boris-42> stpierre: the spec wasn't not on priority
14:20:16 <andreykurilin__> rvasilets: we will use existing code as a base for new services
14:20:27 <boris-42> stpierre: and nobody reviewed it
14:20:33 <andreykurilin__> rvasilets: so we can improve it and fix issues
14:20:45 <rvasilets> we have a lot of bugs and they are not critical as others
14:21:03 <stpierre> we can't just declare the wrappers dead code and ignore any issues with them without having a replacement
14:21:12 <andreykurilin__> stpierre: +1
14:21:52 <rvasilets> why not?)
14:21:54 <stpierre> once we even have a replacement WIP i absolutely think we should block changes to the wrapper, but we haven't even started work on the base class, let alone the individual services
14:22:05 <stpierre> because there are real actual bugs that currently exist and are affecting people
14:22:07 <boris-42> stpierre: agree on this
14:22:09 <stpierre> there are scenarios that we can't run
14:22:27 <stpierre> that's not an abstract "we", that's us, cisco -- we have important internal work that is blocked on actual rally bugs
14:22:39 <stpierre> that we have a fix in review for
14:23:05 <andreykurilin__> ok, it looks like we have an agreement
14:23:08 <andreykurilin__> :)
14:23:16 <boris-42> andreykurilin__: stpierre ok let's merge small fixes
14:23:21 <stpierre> let's be sure to communicate well when work on a new service starts
14:23:23 <andreykurilin__> nice
14:23:28 <stpierre> so we can block further changes to that wrapper
14:23:32 <boris-42> andreykurilin__: stpierre however what I don't like is porting old scenarios to wrapper and adding more work on this
14:24:22 <stpierre> honestly, i think that should serve as a motivator to get us to write the services faster
14:24:31 <andreykurilin__> :)
14:25:24 <stpierre> but in general i concur, let's keep any fixes as small as is practical
14:26:22 <andreykurilin__> Can we move to the next topic?
14:26:32 <andreykurilin__> since we have an agreement
14:26:54 <andreykurilin__> ok
14:26:59 <andreykurilin__> #topic [amaretskiy] Let's revive DB refactoring spec and give it high priority, since we can not effectively switch to new task format without this spec
14:27:08 <andreykurilin__> #link https://review.openstack.org/#/c/167060/
14:27:08 <patchbot> andreykurilin__: patch 167060 - rally - [spec] Making DB scalable and flexible enough
14:28:21 <amaretskiy> we have to do a lot of changes in database schema - adopt it for new task results format, add extra data about workloads and cloud and so on
14:28:34 <amaretskiy> #link https://github.com/openstack/rally/blob/master/doc/specs/in-progress/new_rally_input_task_format.rst
14:28:50 <amaretskiy> it's time to revive spec https://review.openstack.org/#/c/167060/
14:28:50 <patchbot> amaretskiy: patch 167060 - rally - [spec] Making DB scalable and flexible enough
14:29:05 <amaretskiy> we can do all these changes well without spec
14:29:07 <ikhudoshyn> I'm working on it
14:29:19 <amaretskiy> I propose to plan this patch for next rlease
14:29:33 <amaretskiy> ikhudoshyn: great!
14:29:41 <ikhudoshyn> new patchset would be available in a next couple days
14:30:50 <ikhudoshyn> yet i'd like to warn that it's implementation may take pretty much long
14:31:15 <andreykurilin__> It is sad
14:31:34 <ikhudoshyn> since predicted schema change gonna be huge I plan to implement it piece by piece
14:31:46 <amaretskiy> let's add a high priority to this patch and review new patch sets without delay
14:32:14 <amaretskiy> any comments?
14:32:14 <ikhudoshyn> at least let's set hi prio to the spec))
14:32:26 <amaretskiy> sure :)
14:32:51 <boris-42> amaretskiy: ikhudoshyn seems like that we have more priority tasks than people
14:32:52 <andreykurilin__> ok
14:33:00 <andreykurilin__> boris-42: heh
14:33:09 <boris-42> what about trying to push this release services bases and just pause everything else
14:33:19 <andreykurilin__> hm...
14:33:41 <rvasilets> agree
14:33:45 <rvasilets> big agree
14:33:46 <boris-42> I don't think that trends or new db are rising problem (it's like constant amount of work to it now or in one release)
14:33:49 <ikhudoshyn> ok I'm fine with that
14:33:56 <andreykurilin__> perfect
14:34:00 <boris-42> however we are getting every day new plugins and using more and more wrappers
14:34:05 <andreykurilin__> stpierre: you should be happy:)
14:34:31 <stpierre> heh
14:34:43 <andreykurilin__> boris-42: ok, I can work on nova service
14:35:01 <rvasilets> me on glance for example
14:35:23 <andreykurilin__> stpierre: do you want to take some service to yourself? :)
14:35:31 <ikhudoshyn> I'll try to finish the spec and would probably need some time off in the nearest future
14:35:46 <andreykurilin__> amaretskiy: and you?
14:35:55 <rvasilets> stpierre, you can take glance if you like)
14:35:57 <stpierre> seems like we should prioritize glance, cinder, keystone, and nova-net/neutron, since those are the wrappers we have
14:35:59 <stpierre> i'll do glance
14:36:09 <amaretskiy> i can take network service
14:36:17 <rvasilets> i'll take cinder then...
14:36:20 <andreykurilin__> amaretskiy: Could you take neutone/n-net ?
14:36:44 <andreykurilin__> since you was an author of old wrapper:)
14:37:04 <amaretskiy> andreykurilin__: network service? sure
14:37:58 <andreykurilin__> thanks
14:38:03 <rvasilets> and could handle heat also
14:38:24 <rvasilets> i'll try to use help of heat core
14:38:32 <andreykurilin__> rvasilets: I think redixin could do it
14:38:38 <andreykurilin__> but he is not here now
14:39:04 <rvasilets> there is vm
14:39:14 <rvasilets> maybe redixin will do vm?)
14:39:18 <andreykurilin__> ok, boris-42: we are waiting for base classes ;)
14:39:24 <boris-42> andreykurilin__: ok
14:39:27 <rvasilets> year)
14:39:35 <boris-42> andreykurilin__: I will try to manage to get this soon
14:39:38 <rvasilets> *yeah
14:39:47 <andreykurilin__> boris-42: thanks
14:40:05 <andreykurilin__> anything else for this topic?
14:40:28 <rvasilets> I think no)
14:41:15 <andreykurilin__> zuul queue contains 355 patches...
14:41:16 <andreykurilin__> ;(
14:41:25 <andreykurilin__> #topic [yanyanhu] Basic support for Senlin service in Rally:
14:41:35 <andreykurilin__> #link https://review.openstack.org/298109
14:41:38 <Yanyanhu> hello, everyone :)
14:41:44 <rvasilets> o/
14:41:46 <andreykurilin__> #link https://review.openstack.org/307170
14:41:54 <Yanyanhu> hi, I'm from Senlin team
14:41:59 <andreykurilin__> #link https://review.openstack.org/307050
14:42:05 <Yanyanhu> we hope to use Rally to benchmark Senlin service
14:42:11 <andreykurilin__> \o/
14:42:22 <Yanyanhu> so we proposed a patch to add basic support for senlin into rally
14:42:33 <Yanyanhu> https://review.openstack.org/298109
14:42:50 <Yanyanhu> this one
14:43:14 <Yanyanhu> and I just proposed patch to infra to add gate job
14:43:30 <Yanyanhu> here:  https://review.openstack.org/298109
14:43:54 <Yanyanhu> and local plugins and jobs in Senlin repo: https://review.openstack.org/307170
14:44:37 <Yanyanhu> but I think the following two ones rely on the basic senlin support patch to work :)
14:44:54 <andreykurilin__> Yanyanhu: why do you want to store all your plugin at the local repo?
14:45:13 <Yanyanhu> we hope to contribute them to rally's repo
14:45:24 <andreykurilin__> nice
14:45:36 <Yanyanhu> but not sure whether we should do that :) so I temporarily keep them in senlin's reop
14:45:38 <Yanyanhu> repo
14:45:49 <andreykurilin__> so I'm ok with mergin support of senlinclient to our repo now
14:46:32 <Yanyanhu> thanks andreykurilin__, with that patch being merged, more code for senlin plugin can be added :)
14:46:35 <Yanyanhu> and also jobs
14:46:59 <andreykurilin__> Yanyanhu: btw, clients entity in rally is a pluggable too, so you able to store it in local repo too ;)
14:47:04 <boris-42> andreykurilin__: yep I don't think that we should block this
14:47:16 <boris-42> andreykurilin__: I think that we need just to help to get CI job done for this
14:47:25 <Yanyanhu> I see.
14:47:30 <andreykurilin__> Yanyanhu: can you add your experimental job to our project too?
14:47:36 <rvasilets> will we create experimental job at Rally7
14:47:37 <Yanyanhu> sure
14:47:39 <Yanyanhu> I will
14:47:44 <andreykurilin__> nice
14:47:51 <rvasilets> nice
14:48:02 <rvasilets> =)
14:48:10 <Yanyanhu> thanks you guys :)
14:48:16 <andreykurilin__> so as soon as we got new release we will be able to merge your base patch
14:48:26 <Yanyanhu> got it
14:48:37 <andreykurilin__> I think it will done in a day(zuul is dead)
14:48:56 <Yanyanhu> yes... the gate is broken I think
14:49:03 <andreykurilin__> anything else?
14:49:16 <rvasilets> for this topic?)
14:49:20 <andreykurilin__> yes
14:49:25 <rvasilets> тщ
14:49:26 <rvasilets> no
14:49:29 <andreykurilin__> ok
14:49:36 <andreykurilin__> #topic [andreykurilin] Make Glance V2 the default version
14:49:54 <andreykurilin__> as far as I know, Glance V1 will deprecated soon
14:50:23 <andreykurilin__> I think we should switch to use V2 by default
14:51:16 <andreykurilin__> Last week we merged several patches for compatibility with it (thanks to stpierre ylobankov)
14:51:50 <stpierre> i'd like some extra eyes on the image creation code in rally for glance v2
14:52:17 <stpierre> AFAICT glance v2 is really and truly horrible and the code is correct, but before we make it the default i'd just like to double- or triple-check that we aren't doing anything stupid
14:52:24 <stpierre> or at least, no more stupid than we need to be to deal with glance v2
14:52:29 <andreykurilin__> :)
14:52:39 <rvasilets> ok
14:52:57 <stpierre> #link https://github.com/openstack/rally/blob/master/rally/plugins/openstack/wrappers/glance.py#L151-L196
14:53:25 <stpierre> if that checks out, then +1 to making glance v2 the default
14:53:27 <andreykurilin__> a bit of magic
14:53:28 <andreykurilin__> lol
14:53:29 <rvasilets> I could ask in glance chat will we do everything correct
14:53:44 <andreykurilin__> ok, let's switch to glance v2 as soon as stpierre propose patch for it :D
14:53:51 <rvasilets> ))
14:54:20 <stpierre> andreykurilin__: sounds good :)
14:54:29 <andreykurilin__> perfect
14:54:59 <andreykurilin__> Do we have more topics?
14:55:14 <rvasilets> yes I want to ask review to https://review.openstack.org/#/c/283180/
14:55:14 <patchbot> rvasilets: patch 283180 - rally - [Spec] Rally Task Validation refactoring
14:55:24 <andreykurilin__> nop
14:55:28 <andreykurilin__> :D
14:55:35 <rvasilets> =(((
14:55:40 <rvasilets> crying
14:55:47 <andreykurilin__> rvasilets: please, no...
14:55:58 <andreykurilin__> I'll try to review it this week
14:56:12 <andreykurilin__> anything else?
14:56:19 <rvasilets> So please review it. And after services we will be able to refactor validators
14:56:34 <rvasilets> nothing else
14:56:55 <ikhudoshyn> https://review.openstack.org/#/c/242329/ actually is waiting for boris-42 eyes
14:56:55 <patchbot> ikhudoshyn: patch 242329 - rally - Transform DB layer to return dicts, not SQLAlchemy...
14:57:09 <andreykurilin__> boris-42: ^
14:57:12 <andreykurilin__> we need you
14:57:14 <ikhudoshyn> others are welcome as well
14:57:26 <boris-42> andreykurilin__: ok
14:57:44 <andreykurilin__> ok, let's finish
14:57:47 <andreykurilin__> thank you all
14:57:54 <andreykurilin__> #endmeeting