16:00:10 <xarses> #startmeeting fuel
16:00:10 <xarses> #chair xarses
16:00:10 <xarses> Todays Agenda:
16:00:10 <xarses> #link https://etherpad.openstack.org/p/fuel-weekly-meeting-agenda
16:00:10 <xarses> Who's here?
16:00:11 <openstack> Meeting started Thu Mar 31 16:00:10 2016 UTC and is due to finish in 60 minutes.  The chair is xarses. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:13 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:15 <romcheg> \o
16:00:15 <mwhahaha> hi
16:00:16 <openstack> The meeting name has been set to 'fuel'
16:00:17 <openstack> Current chairs: xarses
16:00:20 <evgenyl> hi
16:00:20 <maximov> hi
16:00:28 <kozhukalov> hi
16:00:33 <dmellado> hi everyone
16:00:34 <akislitsky_> hi
16:00:45 <fzhadaev> Hi!
16:00:46 <dpyzhov> hi
16:00:50 <srwilkers> hi
16:01:00 <EZPZ> hello
16:01:11 <angdraug> \o
16:01:26 <ashtokolov> hi
16:01:35 <holser_> \o
16:01:42 <xarses> #topic action items from last week
16:01:45 <zynzel> greetings from wrong timezone ;)
16:01:49 <alex_didenko> hi
16:01:52 <ikalnitsky> o/
16:01:55 <xarses> angdraug holser_ will raise UX tzar ML
16:02:23 <yottatsa> o/
16:02:41 <holser_> xarses: ok
16:02:46 <askb_> 0/
16:02:56 <mihgen> hi
16:03:03 <xarses> holser_: any update?
16:03:16 <holser_> no updates from my side
16:03:27 <xarses> thanks
16:03:30 <xarses> holser_ angdraug with collaborate on [asutin schedule]
16:03:37 <xarses> Draft: https://etherpad.openstack.org/p/fuel-austin-agenda
16:03:43 <xarses> Schedule: https://etherpad.openstack.org/p/fuel-newton-summit-planning
16:03:45 <holser_> xarses:
16:03:51 <holser_> I will provide an update
16:03:58 <holser_> https://etherpad.openstack.org/p/fuel-newton-summit-planning
16:04:07 <holser_> Here is a planning with schedule
16:04:18 <holser_> We picked the most hottest topics
16:04:41 <holser_> However, schedule may be adjusted
16:04:42 <kozhukalov> in fact most of topics, not only hottest
16:05:00 <holser_> we assigned leaders and mandatory participants
16:05:08 <holser_> and started doing a draft of agenda
16:05:27 <holser_> so Leader of every topic should work on details of agenda
16:05:46 <holser_> I will follow up on that personally with Topic Leaders
16:05:52 <holser_> that’s all from my side
16:05:58 <xarses> thanks
16:06:07 <holser_> yw
16:06:27 <xarses> #topic Deploy RDO using Fuel (kozhukalov, dmellado)
16:06:45 <kozhukalov> i added this topic
16:07:00 <kozhukalov> the idea is to make it possible to deploy RDO
16:07:07 <mattymo> kozhukalov, wow^ I applaud you. Sounds like we can build on top  of what we started with the UCA feature
16:07:18 <kozhukalov> here is the BP https://blueprints.launchpad.net/fuel/+spec/deploy-rdo-using-fuel
16:07:34 <kozhukalov> in a nutshell there are two major work items
16:07:46 <kozhukalov> 1) put some of fuel packages to RDO
16:07:52 <kozhukalov> ideally all of them
16:08:01 <kozhukalov> but mcagents at least
16:08:26 <kozhukalov> 2) implement rpm image build in fuel-agent
16:08:39 <kozhukalov> and looks like we have volunteer
16:08:55 <mattymo> kozhukalov, who is our volunteer?
16:09:00 <dmellado> I'll try to help on that
16:09:08 <kozhukalov> we just need to help him, answer questions, and help with design
16:09:08 <dmellado> on a best-effort volunteer based, though  ;)
16:09:22 <dmellado> so we'll need more people on that
16:09:47 <xarses> by all means
16:09:54 <kozhukalov> we have already talked and discussed preliminary design
16:10:12 <kozhukalov> first step will be PoC
16:10:20 <kozhukalov> deploy RDO over base Centos
16:10:25 <mattymo> my first question is do we intend to do just CentOS? Or also Fedora?
16:10:43 <kozhukalov> i don't know
16:10:51 <mwhahaha> i'm sure there will be a bunch of issues in library around centos
16:10:52 <dmellado> mattymo: I think Centos will be easier as a first step, as there was already a base for that
16:10:58 <mwhahaha> and rdo package names
16:11:04 <dmellado> but I'm not an expert on that, by any means
16:11:07 <kozhukalov> if RDO is compatible with Fedora then yes but later
16:11:53 <kozhukalov> mwhahaha: yes, you are right
16:11:58 <xarses> kozhukalov: what do we need to do to move this along?
16:12:00 <kozhukalov> but will see
16:12:14 <mwhahaha> yea just something we'll need to keep an eye on, the first step is just being able to deploy centos again
16:12:16 <dmellado> mwhahaha: doesn't delorean works on centos?
16:12:19 <mattymo> dmellado, just curious, what time zone are you in?
16:12:24 <dmellado> I'm CEST
16:12:50 <mwhahaha> dmellado: no the issue is that we've been deving with deb packages so the various other packages that we use for things like pacemaker/haproxy may not exist
16:13:02 <mwhahaha> i know haproxy will be a problem because we use a custom version that supports conf.d
16:13:05 <dmellado> mwhahaha: I see
16:13:19 <mwhahaha> but we also have some other custom names like for our mysql packages
16:13:26 <kozhukalov> ok, I think it is just an announcement, and I'd like we to be as friendly as possible to help
16:13:28 <mattymo> haproxy, rabbitmq, and ceph will all present issues, based on what we saw with deploying on top of Ubuntu openstack packages
16:13:59 <mwhahaha> yea i'd like to see the centos deployment so just something to be aware of :)
16:14:42 <xarses> do we need any actions here?
16:14:52 <kozhukalov> xarses: not yet
16:14:55 <maximov> kozhukalov: good to hear that fuel is going to support centos deployment for slave nodes again.
16:14:56 <xarses> k, thanks
16:15:03 <xarses> #topic bit.ly/1RD6JLR - many review requests (mihgen)
16:15:25 <mihgen> hey folks there are many reviews in the queue
16:15:32 <mihgen> how do we deal with it.. ?
16:15:46 <xarses> and SCF is next week
16:15:55 <ikalnitsky> there're a lot of specs
16:16:02 <ikalnitsky> and fuel-mirror
16:16:10 <xarses> specs for 10 are fine
16:16:13 <ikalnitsky> afaiu, fuel-mirror is targeted to 10.0
16:16:29 <ikalnitsky> 'cause it's about refactoring
16:16:34 <ikalnitsky> kozhukalov: can you confirm that?
16:17:03 <kozhukalov> ikalnitsky: which one exactly?
16:17:12 <xarses> can we just merge https://review.openstack.org/#/c/297798
16:17:18 <xarses> its fixing spec markup
16:17:32 <kozhukalov> ahh, yes, it is refactoring
16:17:34 <ikalnitsky> kozhukalov: https://review.openstack.org/#/q/status:open+project:openstack/fuel-mirror+branch:master+topic:perestroika/refactoring-2
16:17:48 <kozhukalov> i will poke build team to review this
16:18:52 <ogelbukh> I still have 2 specs pending for 9
16:18:52 <xarses> #topic API guidelines http://git.openstack.org/cgit/openstack/api-wg/tree/guidelines (kozhukalov)
16:19:07 <kozhukalov> it is again from my side
16:19:09 <xarses> sorry, do we still need to talk about last
16:19:38 <xarses> #topic bit.ly/1RD6JLR - many review requests (mihgen)
16:19:52 <xarses> ogelbukh: yes, guys please review ogelbukh's specs
16:20:09 <xarses> its landed already and needs to be closed
16:20:23 <xarses> #link https://review.openstack.org/#/q/project:openstack/fuel-specs+owner:%22Oleg+Gelbukh+%253Cogelbukh%2540mirantis.com%253E%22+(topic:bp/upload-deployment-facts-to-configdb+OR+topic:bp/serialized-facts-nailgun-api)
16:20:55 <xarses> #topic API guidelines http://git.openstack.org/cgit/openstack/api-wg/tree/guidelines (kozhukalov)
16:21:05 <xarses> ok, back to you kozhukalov
16:21:10 <kozhukalov> the idea is to follow community API guidelines and actively discuss these guidelines
16:21:36 <kozhukalov> I'd like Igor to be our liasion in this group
16:21:56 <kozhukalov> ikalnitsky: what do you think?
16:22:11 <kozhukalov> is it going to be useful?
16:22:28 <ikalnitsky> kozhukalov: taking into account not good shape of nailgun api (and not only nailgun), i think it might be a good idea
16:23:18 <kozhukalov> ok, then are interested to be Fuel liasion in API-wg?
16:23:30 <ikalnitsky> yep
16:23:43 <ikalnitsky> i'll propose myself
16:23:50 <kozhukalov> ok, nice )
16:23:59 <kozhukalov> that is it on this topic
16:24:03 <xarses> #topic Mixed team status (zynzel)
16:24:14 <zynzel> Fixed team is working on multipath, idempotency and bugfixing/reviewing process.
16:24:17 <zynzel> Multipath - one issue left (https://bugs.launchpad.net/fuel/+bug/1562836), but for now it is not reproductive.
16:24:18 <openstack> Launchpad bug 1562836 in Fuel for OpenStack "Provision of node with multipath device fails on boot" [Medium,In progress] - Assigned to Aleksey Zvyagintsev (azvyagintsev)
16:24:20 <zynzel> Idempotency - one issue found today (https://bugs.launchpad.net/fuel/+bug/1564358).
16:24:21 <openstack> Launchpad bug 1564358 in Fuel for OpenStack "cgroups task is not idempotent" [High,In progress] - Assigned to Bartosz Kupidura (zynzel)
16:24:23 <zynzel> QA/CI tests for idempotency are in review, tests for ensurability in progress.
16:24:26 <zynzel> QA/CI patch for collecting fixtures was merged. Those fixtures will be used by fuel-noop-fixtures framework.
16:24:31 <zynzel> questions?
16:24:50 <holser_> zynzel:
16:25:01 <holser_> Did you verify idempotency for all tasks
16:25:12 <holser_> not just run one .pp in one task
16:25:26 <holser_> one task may configyre something but another configure in different way
16:25:30 <zynzel> we use 2 methods to find not idempotent task
16:25:40 <zynzel> 1) run 2 times in a row each granular
16:26:01 <zynzel> 2) run whole deployment against deployed env
16:26:16 <zynzel> so we find all not idempotent tasks afaik
16:26:32 <xarses> now that we have classes for each granular, should we start using beaker to enforce this?
16:27:37 <zynzel> i dont think this question is for me
16:27:50 <alex_didenko> why?
16:28:02 <alex_didenko> we already have integration tests
16:28:16 <alex_didenko> based on fuel-qa
16:28:30 <mwhahaha> we should look at being able to test the classes leveraging beaker to do simple validation
16:28:41 <angdraug> alex_didenko: we can't rely on just integration tests for validating _everything_
16:28:43 <mwhahaha> rather than having to spin up the multi-node stuff we already do
16:28:44 <xarses> to test for idempotency?
16:29:04 <mwhahaha> it's much easier to test idempotency via beaker by running the classes again then trying to elverage fuel to do it
16:29:19 <alex_didenko> how about multi-tasks idempotency?
16:29:33 <alex_didenko> what if one task changes the same resoruces/files as the other
16:29:36 <mwhahaha> we managed to get some reporting items into astute to allow us to report on idempotency but it's basically a double run of deploments
16:29:41 <mwhahaha> we tested both
16:29:44 <alex_didenko> so if you run one task multiple times - it's idempotent
16:29:53 <alex_didenko> but if you run both - the won't be idempotent
16:29:57 <mwhahaha> 1) double run the tasks serially and 2) double run deployments
16:30:15 <alex_didenko> so double deployments with fuel-qa and beaker
16:30:16 <mwhahaha> we identified only a few places where the tasks are not idempotent
16:30:25 <xarses> sounds like we need to continue to visit idempotency testing in newton
16:30:30 <mwhahaha> well the first one could be done via beaker
16:30:36 <mwhahaha> and the double deployment via fuel-qa
16:31:00 <mwhahaha> we should continue this conversation as the framework we have is very time and labor intensive at the moment
16:31:21 <alex_didenko> but we can't replace it with beaker
16:31:24 <mwhahaha> and would be a swarm check and not CI
16:31:29 <mwhahaha> no we should augment with beaker
16:31:33 <alex_didenko> we can add some additional test coverage with beaker
16:31:51 <alex_didenko> but we should research if it brings any value
16:32:01 <xarses> some one want an action item here?
16:32:02 <mwhahaha> we also didn't test fuel master idempotency either
16:32:22 <mwhahaha> but that should be covered as part of an effort to upstream a fuel module
16:32:40 <mwhahaha> https://blueprints.launchpad.net/fuel/+spec/split-fuel-puppet-module
16:33:38 <xarses> #action mwhahaha alex_didenko zynzel will document remaining gaps in idempotency coverage
16:33:42 <xarses> #topic UI Team status (vkramskikh)
16:33:46 <vkramskikh> Hi! Here is our status for 9.0 release. We've finished working on features and switched to bugfixing. Since the last week we had 8 outgoing and 6 incoming bugs. Currently we have 36 bugs, 7 of them are High. Currently we're fixing both High and Medium bugs - we want to fix some Mediums before SCF.
16:34:48 <xarses> vkramskikh: do these medium bugs already have patches to review?
16:35:02 <vkramskikh> xarses: some of them
16:35:19 <vkramskikh> like https://review.openstack.org/#/c/295401/ and https://review.openstack.org/#/c/299456/
16:35:33 <xarses> SCF is next week, if we don't have CR's up before monday, don't expect to them to land
16:35:48 <vkramskikh> ok
16:36:17 <xarses> #topic Bugfix status (dpyzhov)
16:36:28 <dpyzhov> hi guys
16:36:34 <dpyzhov> We have a SCF next week and we are in pretty good shape with it. We’ve reviewed all medium/low/wishlist priority bugs and already moved some of them to the Newton release.
16:36:43 <dpyzhov> With QA help we’ve marked some of bugs with keep-in-9.0 and we’ll try to fix them in 9.0. However we cannot commit on that.
16:36:50 <dpyzhov> I guess we’ll fix the rest of them at the begging of Newton release.
16:37:16 <dpyzhov> I don't expect any other unassigned bug to be fixed in 9.0
16:37:52 <dpyzhov> As for numbers we have 127 medium/low/wishlist bugs in total and it's smaller than at the begging of the release
16:38:23 <xarses> dpyzhov: other than working on them early in newton, is there something we can do to help reduce this number?
16:38:24 <dpyzhov> It will be good to get some help from SME guys
16:38:48 <dpyzhov> in order to review bugs with 'feature', 'covered-by-bp', 'tech-debt', 'need-bp' tags
16:39:05 <dpyzhov> we need to make sure that there are no risks hidden there
16:39:34 <dpyzhov> xarses: I guess we are already doing the best
16:39:49 <xarses> dpyzhov: thanks
16:39:52 <maximov> folks, I would like to pay attention on bug https://bugs.launchpad.net/fuel/+bug/1548271, it keeps reproducing again and again.. it was reported on Feb 22 first time.
16:39:53 <openstack> Launchpad bug 1548271 in Fuel for OpenStack "Access denied for user 'root'@'localhost'/Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock errors during cluster deployment" [Critical,Confirmed] - Assigned to Alex Schultz (alex-schultz)
16:40:22 <dpyzhov> xarses: we can ask people to take more bugs now. But we don't have time to fix and merge them
16:40:40 <xarses> ok
16:40:54 <dpyzhov> That's all for medium. Also we have 76 bugs in high priority
16:41:04 <dpyzhov> 17 in library, 59 in python
16:41:18 <dpyzhov> 23 in new features, 12 in networking and 24 in all other areas
16:41:40 <dpyzhov> maximov: thanks, I'll look at the bug
16:41:56 <dpyzhov> I guess the number of bugs is ok
16:42:03 <xarses> ok, moving?
16:42:12 <dpyzhov> that's all from my side
16:42:15 <holser_> maximov: I am working on 1548271
16:42:22 <xarses> #topic Telco Team Status (fzhadaev)
16:42:28 <fzhadaev> My update will be extremely short.
16:42:28 <fzhadaev> Our main activity during this week was fixing bugs:
16:42:28 <fzhadaev> 7 done
16:42:28 <fzhadaev> 6 in progress
16:42:28 <fzhadaev> Any questions?
16:42:49 <xarses> fzhadaev: are you in good shape for SCF next week?
16:42:54 <ogelbukh> short update is good
16:43:14 <fzhadaev> yep. most of our open bugs are high
16:43:58 <xarses> thanks
16:44:02 <xarses> #topic open discussion
16:44:13 <xarses> does any one have anything to raise?
16:44:44 <xarses> Otherwise I will close the meeting early (gasp)
16:45:51 <xarses> ok guys, thanks for playing, see you next time
16:45:57 <ogelbukh> o/
16:45:58 <xarses> #endmeeting