16:00:01 <kozhukalov> #startmeeting Fuel
16:00:02 <openstack> Meeting started Thu Dec 11 16:00:01 2014 UTC and is due to finish in 60 minutes.  The chair is kozhukalov. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:03 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:06 <openstack> The meeting name has been set to 'fuel'
16:00:12 <aglarendil> hi everyone
16:00:14 <kozhukalov> hey everyone
16:00:16 <izinovik> hello
16:00:21 <kozhukalov> #chair kozhukalov
16:00:22 <openstack> Current chairs: kozhukalov
16:00:31 <kozhukalov> agenda as usual
16:00:41 <kozhukalov> #link https://etherpad.openstack.org/p/fuel-weekly-meeting-agenda
16:00:43 <evgeniyl__> hi
16:00:50 <kozhukalov> let's start
16:01:09 <kozhukalov> #topic Announcements (6.0 HCF and status)
16:01:21 <aglarendil> hi, fuelers
16:01:34 <aglarendil> so far we called Hard Code Freeze for 6.0 release of Fuel
16:01:44 <aglarendil> thanks to all of you who participated in this effort
16:02:11 <aglarendil> our preliminary plan for release is Dec 18 Pacific time
16:02:17 <christopheraedo> howdy
16:02:31 <aglarendil> currently, we have several bugs targeted for 6.0 which are not resolved yet
16:02:41 <aglarendil> #link https://bugs.launchpad.net/fuel/+bug/1401152
16:02:44 <uvirtbot> Launchpad bug 1401152 in fuel "[Upgrade] Versions of packages in 5.1.1 are older than in 6.0" [Critical,Triaged]
16:03:00 <aglarendil> this bug should be updated as fixed after we merge updated packages into stable mirror
16:03:14 <aglarendil> #link https://bugs.launchpad.net/mos/6.0.x/+bug/1401261
16:03:15 <uvirtbot> Launchpad bug 1401261 in mos/6.0.x "qemu-img convert intermittently corrupts output images" [Critical,New]
16:03:46 <aglarendil> AFAIK, fix is not yet ready, we are waiting for our MOS-Linux team to apply it and update qemu package
16:03:53 <aglarendil> #link https://bugs.launchpad.net/mos/6.0.x/+bug/1401503
16:03:55 <uvirtbot> Launchpad bug 1401503 in mos/6.0.x "Redeployment of controller failed when murano is enabled" [Critical,In progress]
16:04:23 <aglarendil> Root cause is identified - fix is being tested by developers right now and should be pushed on review shortly
16:04:47 <aglarendil> #link https://bugs.launchpad.net/mos/+bug/1401118
16:04:49 <uvirtbot> Launchpad bug 1401118 in mos "Glance can't upload image by link from horizon dashboard" [Critical,Confirmed]
16:05:08 <aglarendil> this bug is still in the progress of Root Cause Analysis
16:05:21 <kozhukalov> any estimate for RC iso?
16:05:38 <aglarendil> according to the latest info it is present in current upstream code
16:05:46 <aglarendil> #link https://bugs.launchpad.net/mos/+bug/1369524
16:05:48 <uvirtbot> Launchpad bug 1369524 in mos/6.0.x "lvremove can't delete volume with --force on Ubuntu" [High,In progress]
16:05:58 <aglarendil> fix for Ubuntu LVM packages and corresponding components is ready
16:06:03 <aglarendil> and is in progress of merge
16:06:30 <aglarendil> there are also two bugs related to stat collection feature
16:06:33 <aglarendil> #link https://bugs.launchpad.net/fuel/+bug/1401572
16:06:35 <uvirtbot> Launchpad bug 1401572 in fuel "Public ip addresses are collected by stats in action_logs" [High,Confirmed]
16:06:45 <aglarendil> #link https://bugs.launchpad.net/fuel/+bug/1401521
16:06:46 <uvirtbot> Launchpad bug 1401521 in fuel/6.1.x "Usage statistics isn't sent if postgres was restarted" [High,New]
16:07:07 <aglarendil> I do not have any info on them. can anyone from fuel-web team comment on this ?
16:07:29 <kozhukalov> akislitsky: around?
16:08:13 <evgeniyl__> it doesn't look like high
16:08:24 <evgeniyl__> akasatkin: could you please comment?
16:08:56 <akasatkin> yes, i'm looking into them
16:09:00 <aglarendil> kozhukalov: there is already RC1 and corresponding community ISO is
16:09:04 <aglarendil> #link http://seed.fuel-infra.org/fuelweb-iso/fuel-community-6.0-34-2014-12-11_03-35-45.iso.torrent
16:09:18 <kozhukalov> aglarendil: great, thanx
16:09:28 <aglarendil> so far I am done with 6.0 announcements
16:09:32 <akasatkin> 1401521 requires some investigation, estimate 1 day
16:10:03 <aglarendil> regarding 6.1 we are starting design and planning
16:10:18 <aglarendil> everyone feel free to propose their blueprints for 6.1 milestone
16:10:22 <akasatkin> 1401572  is simple , i'll make fix now
16:10:22 <kozhukalov> aglarendil: it is the next topic
16:10:40 <kozhukalov> akasatkin: great
16:11:14 <kozhukalov> ok, looks like we are in time for 6.0
16:11:31 <kozhukalov> #topic preliminary plans for 6.1
16:11:57 <aglarendil> so I will start with fuel-library plans
16:12:02 <kozhukalov> there is still no official list of features which are supposed to be included into 6.1
16:12:26 <kozhukalov> aglarendil: yes please, share your team's plans
16:12:37 <aglarendil> kozhukalov: thx
16:12:52 <aglarendil> we are going to work on following:
16:13:04 <aglarendil> Superfeature called Granular Deployment
16:13:27 <aglarendil> which should allow us to split deployment into a lot of small parts and allow user
16:13:32 <aglarendil> to inject his own pieces
16:13:43 <aglarendil> related-blueprints:
16:13:45 <aglarendil> #link https://blueprints.launchpad.net/fuel/+spec/granular-deployment-based-on-tasks
16:14:00 <aglarendil> #link https://blueprints.launchpad.net/fuel/+spec/fuel-library-modularization
16:14:24 <aglarendil> this should allow us to implement seamless patching and upgrades for OpenStack
16:14:31 <aglarendil> and underlying system components
16:14:45 <aglarendil> along with ability to improve our plugin engine
16:14:56 <aglarendil> and CI infrastructure
16:15:08 <kozhukalov> afaiu, zero step implementation for granular deployment is not supposed to be based on Mistral, am I right?
16:15:15 <aglarendil> kozhukalov: right
16:15:38 <aglarendil> but we are going to start using Mistral as an orchestration engine starting from 7.0
16:15:56 <aglarendil> and this feature should knock everyone dead :-)
16:16:14 <kozhukalov> great, this gonna be huge progress in our task management
16:16:28 <aglarendil> next feature group is what we call 'Power management'
16:16:44 <aglarendil> base blueprint is
16:16:46 <aglarendil> #link https://blueprints.launchpad.net/fuel/+spec/ha-fencing
16:17:02 <aglarendil> we are eager to start using advanced power management techniques
16:17:13 <aglarendil> such as IPMI/UPS/PDU
16:17:21 <aglarendil> so that we can manage power of slave nodes
16:17:31 <kozhukalov> this BP is kind of adjacent to this one https://blueprints.launchpad.net/fuel/+spec/cobbler-ironic-substitute
16:17:31 <aglarendil> and send corresponding data to pacemaker during deployment
16:17:33 <aglarendil> of controllers
16:17:59 <aglarendil> kozhukalov: correct, but we are not tied to ironic as cobbler also can work with IPMI
16:18:00 <kozhukalov> as far as we are going to switch to ironic (if it is possible)
16:18:14 <aglarendil> our initial plan is to at least provide IPMI support in 6.1
16:18:28 <aglarendil> and we hope Ironic with Fuel Agent will make its way to 6.1 also
16:18:33 <kozhukalov> aglarendil: you are right, but ironic is a priority, i think
16:19:00 <aglarendil> kozhukalov: well, switching power management between cobbler and ironic should not be a rocket science
16:19:29 <aglarendil> next feature which is also important
16:19:31 <kozhukalov> aglarendil: it is still in question whether ironic is ready for widening their scope
16:19:48 <aglarendil> is improvement of our HA setup
16:19:51 <aglarendil> #link https://blueprints.launchpad.net/fuel/+spec/pacemaker-improvements
16:19:58 <aglarendil> we are going to switch to corosync 2.x
16:20:05 <aglarendil> improve galera pacemaker script
16:20:33 <aglarendil> make rabbitmq cluster HA more resilient
16:20:42 <aglarendil> refactor our puppet manifests for pacemaker and corosync
16:21:05 <aglarendil> along with working on upstreaming this code to puppetlabs-openstack community
16:21:32 <aglarendil> and also, if power management feature goes into release, we are going to enable fencing out of box
16:21:51 <aglarendil> so that our setup become rock solid and invulnerable
16:22:11 <aglarendil> also, we are eager to support SSL deployment
16:22:22 <aglarendil> in the simplest mode with self-signed certs
16:22:29 <aglarendil> both for FUEL and OpenStack endpoints
16:22:38 <aglarendil> #link https://blueprints.launchpad.net/fuel/+spec/ssl-endpoints
16:22:48 <aglarendil> #link https://blueprints.launchpad.net/fuel/+spec/fuel-ssl-endpoints
16:23:09 <kozhukalov> ssl support gonna be very important for some users
16:23:28 <aglarendil> yep, and that's why we are targeting it to 6.1
16:23:50 <aglarendil> we've been a little bit slow with it, but we want to deliver it with a high level of quality
16:24:20 <aglarendil> and there are also many other important blueprints, but I  have no time
16:24:25 <aglarendil> to enumerate them all
16:24:41 <aglarendil> so, that's all so far
16:24:41 <kozhukalov> aglarendil: thanx a lot
16:24:58 <kozhukalov> about fuel-web
16:25:11 <mattymo> I hope to help us tackle the SSL approach for 6.1 and make significant contributions there
16:25:49 <kozhukalov> we are working hard on integrating Ironic into Fuel but it is still in question
16:26:13 <kozhukalov> there is a spec
16:26:19 <kozhukalov> #link https://review.openstack.org/#/c/138115/
16:26:28 <kozhukalov> for implementing Fuel Agent driver
16:27:03 <kozhukalov> anyway by 6.1 we are planning to make image based scheme the only provisioning way available in Fuel
16:27:37 <kozhukalov> that gonna significantly decrease provisioning time
16:27:51 <kozhukalov> currently it take around 20 minutes
16:28:11 <kozhukalov> it is supposed image based gonna take couple minutes
16:28:34 <xarses> image does take about 3-5 min for me
16:28:49 <mattymo> I imagine that time is based on hardware
16:29:04 <izinovik> it is a tremendous speedup
16:29:19 <kozhukalov> another great advantage is that we will be able to  increase flexibility level (really adavanced partition schemes including software raids)
16:29:54 <kozhukalov> ok, next feature is refactoring nailgun  volume manager
16:30:06 <kozhukalov> #link https://blueprints.launchpad.net/fuel/+spec/volume-manager-refactoring
16:30:40 <kozhukalov> this gonna address many bugs which are quite difficult to address using current scheme
16:31:08 <kozhukalov> and again it is also necessary for providing additional flexibility for end user
16:31:39 <kozhukalov> anther one is artifact based build system
16:31:51 <kozhukalov> #link https://blueprints.launchpad.net/fuel/+spec/artifact-based-build-process
16:32:09 <kozhukalov> we gonna make it much more scalable
16:32:29 <kozhukalov> and significantly decrease iso build time
16:32:55 <kozhukalov> it also gonna help to improve our functional tests
16:33:12 <kozhukalov> because it opens a way for granular testing
16:33:47 <kozhukalov> another extremely important activity is about pluggable architecture
16:34:10 <kozhukalov> unfortunately, i know nothing about its status
16:34:27 <kozhukalov> evgeniyl__: can you share some info about that?
16:34:52 <evgeniyl__> kozhukalov: we discuss details in ml
16:35:02 <evgeniyl__> kozhukalov: there is no bp or spec
16:35:09 <evgeniyl__> kozhukalov: we are trying to define the scope
16:35:39 <evgeniyl__> kozhukalov: we think it will be role as a plugin
16:36:42 <kozhukalov> evgeniyl__: thanx, waiting for spec PR to be available
16:37:20 <kozhukalov> ok, is there any other info about fuel-web activities?
16:37:27 <kozhukalov> dpyzhov: around?
16:37:33 <akasatkin> https://blueprints.launchpad.net/fuel/+spec/advanced-networking
16:37:48 <akasatkin> seems it 'should' be in 6.1
16:38:22 <angdraug> personally I think it should be a 'must'
16:38:26 <akasatkin> it is in progress now
16:39:01 <kozhukalov> akasatkin: thanx
16:39:17 <dpyzhov> kozhukalov: we are working on unit tests restructure, json validation, granular deployment
16:39:38 <kozhukalov> is anyone from polish team around?
16:40:04 <dpyzhov> also we are going to improve our system tests
16:40:34 <kozhukalov> dpyzhov: great, more tests better
16:41:14 <dpyzhov> advanced networking and granular deployment seems to be must have in 6.1. Too many stuff rely on it
16:41:41 <kozhukalov> yes
16:42:04 <sambork> Yes, I'm here from poznan team
16:42:21 <dpyzhov> And we are working on tests not because we have nothing important to do. Our tests became high priority task.
16:42:37 <kozhukalov> sambork: could you please share your team's plans for 6.1?
16:43:52 <sambork> Our team will mostly focuse on 100-200 nodes deployment, also will help with ssl and re-thinking fuel client
16:44:41 <kozhukalov> sambork: great, scalability is also extremely important
16:44:59 <dpyzhov> sambork: will it involve all of you or we can share some extra tasks?
16:45:06 <kozhukalov> hope all those features will be delivered by 6.1
16:45:10 <kozhukalov> many many plans
16:45:45 <sambork> I can tell only about python team: yes it will involve all of us
16:45:55 <dpyzhov> sambork: ok, thanks
16:46:45 <kozhukalov> ok, looks like we are done here
16:46:56 <kozhukalov> #topic open discussion
16:47:25 <dpyzhov> Sorry, I’ve missed the begining of the meeting. Have we discussed 6.0 status?
16:47:30 <kozhukalov> if anyone has any q please you are welcome
16:47:30 <angdraug> yes
16:47:36 <kozhukalov> yes
16:47:46 <dpyzhov> Ok, thanks. I’ll read it in summary
16:47:56 <kozhukalov> there are still some bugs, but looks like we are in time
16:48:03 <angdraug> everyone, I'd like to duplicate my recent update about fuel architecture review
16:48:06 <kozhukalov> by December 18
16:48:34 <angdraug> a lot of people are unhappy with the current format, and today's google calendar glitch was the straw that broke camel's back
16:48:51 <angdraug> the current weekly meeting is cancelled, and will be reborn in a different format very soon
16:49:45 <kozhukalov> angdraug: yes, i have some topics to discuss but previous format is not very suitable
16:50:01 <kozhukalov> because topic are quite technical
16:50:18 <kozhukalov> angdraug: thanx
16:50:28 <kozhukalov> are there any other q?
16:50:38 <kozhukalov> announcements?
16:50:46 <kozhukalov> wishes?
16:50:50 <kozhukalov> claims?
16:50:53 <ntrueblood> I suggest for future Fuel architecture review meetings, we make the Agenda and meeting objectives more clear, so that people can determine about participation.
16:52:11 <kozhukalov> ntrueblood: ok, looking forward for first reborn meeting
16:52:25 <angdraug> ntrueblood: it
16:52:28 <angdraug> oops
16:52:42 <xarses> next one will be about plugins
16:53:12 <angdraug> I think we should fork architecture review into parallel feature specific working groups
16:53:46 <angdraug> I'm still thinking about specific and I welcome ideas
16:54:17 <angdraug> having targeted working groups should help address ntrueblood's concern about clear meeting objectives
16:54:22 <xarses> angdraug: +1
16:55:19 <kozhukalov> ok, looks like we are done
16:55:23 <kozhukalov> closing?
16:55:43 <kozhukalov> thanx everyone, great meeting, many plans
16:55:52 <kozhukalov> let's do our best
16:55:55 <dpyzhov> thank you
16:56:04 <kozhukalov> #endmeeting