14:01:26 <boris-42> #startmeeting Rally
14:01:27 <openstack> Meeting started Mon Jul 13 14:01:26 2015 UTC and is due to finish in 60 minutes.  The chair is boris-42. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:01:29 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:01:31 <openstack> The meeting name has been set to 'rally'
14:01:36 <andreykurilin> aloha!
14:02:00 <e0ne> hi
14:02:07 <rvasilets> \o
14:02:08 <boris-42> Todays agenda https://wiki.openstack.org/wiki/Meetings/Rally
14:02:12 <yingjun> hi!
14:02:16 <andreykurilin> #link https://wiki.openstack.org/wiki/Meetings/Rally#Agenda
14:02:30 <kun_huang> :)
14:02:46 <e0ne> oops. i forgot about agenda:(
14:03:03 <boris-42> e0ne: you have few minutes
14:03:06 <boris-42> e0ne: to fix it=)
14:03:10 <e0ne> :)
14:03:24 <e0ne> i've got 2 annoucements only
14:04:00 <boris-42> e0ne: just mention them in Agenda
14:04:33 <boris-42> okay let's start and hope that other will come to meeting as well=)
14:05:02 <boris-42> stpierre: are you around?
14:05:05 <stpierre> yes
14:05:13 <boris-42> stpierre: nice =)
14:06:02 <boris-42> #topic Supporting Solaris
14:06:10 <boris-42> albertw: ^
14:06:18 <boris-42> stpierre: so?)
14:06:20 <albertw> I'm here
14:06:36 <stpierre> i'm not sure why i'm leading this topic tbh, but i'll give it a go i guess :)
14:07:23 <stpierre> it sounds like there'd be a lot of value to be gained (both for us and for people who want to run rally) by supporting solaris. but we don't have a realistic idea of the costs. we need CI infrastructure for solaris if we're going to support it, and we need to make sure that our docs and supporting shell scripts work on solaris
14:08:32 <boris-42> #info stpierre: so I will work on CI as soon as possible
14:08:39 <boris-42> it's really not hard task
14:09:26 <kun_huang> actually it is not 100% necessary to maintain Solaris CI
14:09:51 <kun_huang> I was in Swift, and there are also many operating systems need support
14:10:05 <stpierre> there's no need to repeat the mistakes of other projects, though :)
14:10:11 <andreykurilin> :)
14:10:15 <andreykurilin> albertw: small question. What version of python is delivered with latest Solaris?
14:10:22 <albertw> let me know if you need any help. I'd already suggest some changes to get the install script to work on solaris, but as stpierre mentioned in a bug there are other changes needed to really support it (e.g. check the required packages are installed)
14:10:30 <kun_huang> just keep the users submit patch theirself
14:10:41 <albertw> python 2.7.9
14:10:41 <boris-42> albertw: btw does solaris suport python 2.7?
14:10:46 <andreykurilin> nice
14:10:46 <boris-42> albertw: great
14:10:52 <stpierre> our CI coverage is already limited enough -- e.g., we don't test Python 2.6 although we support it -- so if we're going to expand to other platforms i think it's reasonable to expect to expand CI coverage along with it
14:11:02 <boris-42> #info Solaris support 2.7.9 (shoudn't be hard to support)
14:11:30 <boris-42> stpierre: as far as I know Oracle is ready to give us resources for testing Solaris
14:11:32 <stpierre> albertw: IMO the best course of action will be to get Solaris CI going, then expand our test coverage of install_rally.sh and fix whatever we find
14:11:36 <stpierre> boris-42: awesome
14:11:44 <boris-42> stpierre: so hardware shoulnd't be issue
14:12:13 <andreykurilin> stpierre: what do you mean by "we don't test Python 2.6 "? we have gate-rally-python26
14:12:33 <stpierre> hrm, is that new?
14:12:45 <stpierre> i know i've submitted at least one patch that had a python 2.7-ism, and it wasn't caught by unit tests
14:12:50 <boris-42> stpierre: nope it's not new=)
14:13:05 <boris-42> stpierre: not everything is covered by unit test (and that is second issue)
14:13:18 <boris-42> stpierre: but we should add as well python2.6 dsvm-cli job
14:13:25 <stpierre> ah, that may be it
14:13:29 <boris-42> stpierre: I will try to understand how to do that in Mirantis Rally CI
14:13:38 <boris-42> stpierre: so we will increase the coverage *
14:13:41 <stpierre> yay!
14:13:51 <boris-42> stpierre: as well it's spearated topic from Solaris
14:13:57 <stpierre> yeah
14:14:02 <boris-42> as far as it supports 2.7 things are good
14:14:06 <stpierre> yep
14:14:17 <stpierre> and if we're going to get solaris CI, i think that's the best way to start
14:15:01 <boris-42> #agreed We will support Solaris and we will start from CI for it
14:15:11 <boris-42> okay let's move to next topic
14:15:22 <boris-42> #topic Should we support py26?
14:16:22 <e0ne> boris-42: yes. if rally is a lib/framework/service/etc and supports old openstack releases, it should support python2.6 too:(
14:16:50 <boris-42> e0ne: we are installing always latest clients
14:17:09 <andreykurilin> and latest clients will drop py26 in near future
14:17:25 <boris-42> e0ne: so actually there is only one problem if you are trying to run Rally against system without py2.7 or py3
14:17:46 <kun_huang> only centos 6.6 use python 2.6 by default in my mind
14:17:52 <e0ne> boris-42, andreykurilin: imo, rally should follow the same practiece as clients
14:17:58 <stpierre> as a developer, i'd really love for the answer to be 'no.' as a former sysadmin, i feel like the answer should probably still be 'yes'. RHEL 6 is still supported for 5.5 more years, after all.
14:18:06 <andreykurilin> If we stop support of py26, we will be able to remove https://github.com/openstack/rally/blob/master/rally/common/costilius.py module
14:18:25 <andreykurilin> :)
14:18:26 <stpierre> and while we don't necessarily need to follow RHEL's support model (after all, they've got SCLs now), py2.6 is still obviously a going concern for lots of people
14:19:10 <boris-42> stpierre: andreykurilin so let's support it until it is possible
14:19:11 <andreykurilin> we have docker image, which should help to launch rally on distros without py27
14:19:34 <stpierre> that's true
14:20:20 <andreykurilin> boris-42: what about release 0.1.0 and say that this is the last release with support of 2.6 . We can mark it with some tag
14:20:31 <andreykurilin> ?
14:20:43 <boris-42> andreykurilin: why do you think that 0.1.0 will be the last release?)
14:20:59 <boris-42> andreykurilin: btw we should learn how to cup our reqruirements
14:21:13 <boris-42> andreykurilin: otherwise during time we will get our release broken (
14:21:28 <andreykurilin> boris-42: we have not ability to do it now
14:22:11 <andreykurilin> maybe dhellmann or lifeless can help us with such ability
14:22:58 <boris-42> andreykurilin: we can remove rally from requriments checker job (or make it non voting)
14:23:08 <boris-42> andreykurilin: cup releases
14:23:21 <boris-42> andreykurilin: cut release
14:23:26 <andreykurilin> boris-42: hm...
14:23:33 <boris-42> andreykurilin: rever cup and make job voting back
14:23:54 <boris-42> not the best ever workflow but...
14:26:21 <boris-42> okay what we need to do is
14:26:42 <andreykurilin> boris-42: move on?:)
14:26:56 <boris-42> #action Talk without dhellmann about cuping release versions in requriments
14:27:01 <boris-42> #topic stucked specs
14:27:07 <boris-42> kun_huang: ^ your idea?)
14:27:51 <kun_huang> "Enhance Rally API" ?
14:28:08 <kun_huang> I and yingjun would like to work on it:)
14:28:17 <kun_huang> yingjun: ping
14:28:28 <yingjun> i’m here
14:29:40 <kun_huang> boris-42:  yingjun and I talked about this privately, and we two could push this work further
14:29:45 <andreykurilin> #link https://review.openstack.org/#/c/182245/
14:30:03 <boris-42> kun_huang: yingjun okay sounds nice to me I will try to find time to re-review it today
14:30:24 <boris-42> kun_huang: yingjun so I believe that topic should be more preciese next time
14:30:30 <kun_huang> but is Alexander Maretskiy here?
14:30:39 <amaretskiy> kun_huang: hi :)
14:30:44 <boris-42> kun_huang: he should be LOL=)
14:31:03 <kun_huang> amaretskiy: let's finish the spec together first :)
14:31:45 <amaretskiy> i left my comments on this spec
14:32:17 <amaretskiy> I'm watching this spec
14:32:43 <amaretskiy> and will make a review as soon as new patch set is ready
14:32:56 <kun_huang> amaretskiy: thanks :)
14:33:03 <stpierre> next stuck spec: "Refactoring Rally cleanup" has been stuck for a while. i propose splitting it in two: 1) consistent resource naming (which has two separate POCs, so is basically code complete); and 2) the actual cleanup.
14:33:05 <stpierre> #link https://review.openstack.org/#/c/172967/
14:33:19 <boris-42> stpierre: +2
14:33:20 <kun_huang> +1
14:33:29 <stpierre> ok, i will make it so. that was easy :)
14:33:38 <boris-42> stpierre: put agreed tag*
14:33:57 <amaretskiy> but who will submit patch sets on spec Enhance Rally API ?
14:34:15 <kun_huang> amaretskiy: I will
14:34:18 <amaretskiy> ok
14:34:27 <stpierre> #agreed Split "Refactoring Rally cleanup": 1) consistent resource naming; 2) the actual cleanup.
14:34:34 <boris-42> amaretskiy: yep kun_huang will work on it
14:34:43 <boris-42> amaretskiy: you have a lot of work on graphs part
14:34:50 <amaretskiy> boris-42: yes :)
14:34:56 <boris-42> kun_huang: but please don't start working on patches until I review the spec
14:35:14 <kun_huang> hah, no problem
14:35:29 <boris-42> kun_huang: that spec is part of very very big refactoring of Rally
14:35:45 <boris-42> kun_huang: so it can be make in different ways but only one is right=)
14:36:10 <boris-42> Okay le'ts move to next topic?
14:37:14 <rvasilets> yes)
14:37:26 <boris-42> #topic sucked patches (pain)
14:37:31 <boris-42> I will start this topic
14:37:34 <andreykurilin> guys, sorry, I need to go:(
14:37:45 <boris-42> We have set of pboldin patches
14:37:58 <boris-42> https://review.openstack.org/#/c/177014/25
14:38:04 <boris-42> #link https://review.openstack.org/#/c/177014/25
14:38:17 <boris-42> ^ that are ready for merge (almost all branch)
14:38:25 <boris-42> so please cores take a look at it
14:39:56 <boris-42> any other extra stuck patches?
14:40:39 <rvasilets> I have no clue
14:41:21 <boris-42> stpierre: kun_huang e0ne  ^
14:42:46 <boris-42> guys?
14:43:32 <amaretskiy> https://review.openstack.org/#/c/194745/9
14:43:45 <e0ne> boris-42: nothing from me
14:43:47 <boris-42> amaretskiy: how about next one?
14:43:58 <boris-42> amaretskiy: when it will be ready?
14:44:20 <amaretskiy> i'm working on streaming algorithms (percentile, median)
14:44:42 <boris-42> amaretskiy: but there are already streaming alogrithms for that
14:44:43 <amaretskiy> they are already working, I'm going to push patch set tomorrow
14:44:58 <stpierre> nothing from me. i've got a few older patches, but it's not causing me any pain
14:45:10 <amaretskiy> we do not have streaming median and percentile
14:45:14 <boris-42> amaretskiy: https://github.com/openstack/rally/blob/master/rally/common/streaming_algorithms.py#L39-L82
14:45:24 <boris-42> amaretskiy: https://github.com/openstack/rally/blob/master/rally/common/streaming_algorithms.py#L38
14:45:32 <amaretskiy> mean is not median
14:45:37 <boris-42> amaretskiy: ah yep sure
14:45:48 <boris-42> amaretskiy: ok
14:45:51 <boris-42> amaretskiy: sorry *
14:45:57 <amaretskiy> actually median is 50 percent percentile
14:46:03 <boris-42> amaretskiy: yep
14:46:06 <amaretskiy> so the only algorithm is percentile
14:46:17 <boris-42> amaretskiy: okay let's move to net topic
14:46:17 <amaretskiy> got it almost working
14:46:31 <amaretskiy> just making more accurate _ tests
14:46:35 <boris-42> #topic Who is in charge for moderating 2 next meetings?
14:46:35 <amaretskiy> * + tests
14:46:48 <boris-42> amaretskiy: kun_huang stpierre andreykurilin ^
14:47:36 <stpierre> i'll take them
14:47:48 <stpierre> i'm going to be out of town for most of august, so i'll get it out of the way now
14:48:04 <boris-42> #agreed stpierre is going to be in charge for next two meethings
14:48:07 <boris-42> stpierre: thanks
14:48:23 <boris-42> so basically it's not hard task a bit boring=) but okay
14:48:59 <boris-42> #topic Murano benchmarks
14:49:01 <boris-42> rvasilets: ping
14:49:18 <rvasilets> I'm here
14:49:21 <rvasilets> typing
14:50:46 <rvasilets> So, #link https://review.openstack.org/#/c/195677/ Murano bench is already finished. And there functional is working. I want you to review it. Alsoo please take a look is all temp files is deleted
14:51:17 <rvasilets> For me they all are delete but its not proper behavior
14:51:23 <rvasilets> So please review it
14:51:28 <rvasilets> eom
14:51:56 <boris-42> rvasilets: ok thanks for update
14:52:02 <boris-42> let's move to next
14:52:12 <boris-42> #topic Cinder benchmarks
14:52:14 <boris-42> e0ne: ping
14:52:34 <e0ne> few anouncements from me
14:53:00 <e0ne> in cinder, we're going to make rally job voting
14:53:08 <boris-42> e0ne: WOOT
14:53:11 <e0ne> after merging this patch https://review.openstack.org/#/c/194180/
14:53:19 <e0ne> i'll create patch for infra:)
14:53:29 <boris-42> e0ne: so i need to finish asap support of V2 version=)
14:53:37 <e0ne> boris-42: yep:)
14:53:46 <boris-42> e0ne: there is not to much to do
14:53:50 <e0ne> boris-42: and i'll send this mo openstack-dev ml
14:54:19 <e0ne> boris-42: and, of cource, i'm working on list of cinder api rally coverage
14:54:33 <e0ne> boris-42: i hope, will finish it tonight
14:55:07 <e0ne> so, that's all from my side:)
14:55:12 <boris-42> e0ne: great
14:55:19 <boris-42> thank you for update
14:55:36 <boris-42> so let's finish meeting?
14:55:43 <boris-42> we discussed everything
14:55:59 <rvasilets> no)
14:56:02 <rvasilets> ironic
14:56:06 <boris-42> rvasilets: oh
14:56:09 <boris-42> #topic ironic
14:56:11 <boris-42> rvasilets: okay
14:56:17 <rvasilets> #link https://review.openstack.org/#/c/186064/
14:56:27 <rvasilets> They are already finished
14:56:36 <rvasilets> and pboldin already review it
14:56:49 <rvasilets> I have fixed his comments
14:56:49 <boris-42> rvasilets: what about ironic guys?
14:57:13 <rvasilets> I don't no/ I have not ping them
14:57:28 <rvasilets> I asked to see only one Ironic guy
14:57:35 <rvasilets> but he is not core
14:57:49 <boris-42> rvasilets: core in ironic?
14:57:55 <rvasilets> yes
14:58:07 <boris-42> rvasilets: so please ask as well somebody from core team
14:58:13 <boris-42> I will review it as well
14:58:19 <boris-42> okay now we have to finish meeting
14:58:23 <rvasilets> They a vote +1 previously)
14:58:25 <boris-42> see you guys in rally chat
14:58:30 <boris-42> #endmeeting