16:00:29 <kozhukalov> #startmeeting Fuel
16:00:31 <openstack> Meeting started Thu Dec 18 16:00:29 2014 UTC and is due to finish in 60 minutes.  The chair is kozhukalov. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:32 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:35 <openstack> The meeting name has been set to 'fuel'
16:00:38 <kozhukalov> hi everyone
16:00:45 <kozhukalov> #chair kozhukalov
16:00:45 <meow-nofer> hi
16:00:45 <openstack> Current chairs: kozhukalov
16:00:56 <kozhukalov> agenda as usual
16:01:03 <tdubyk> hi
16:01:06 <docaedo> hey
16:01:06 <kozhukalov> #link https://etherpad.openstack.org/p/fuel-weekly-meeting-agenda
16:01:13 <evgeniyl__> hey
16:01:17 <dpyzhov> hi
16:01:28 <kozhukalov> #topic Announcements (dpyzhov)
16:02:51 <xarses> hi
16:03:01 <kozhukalov> dpyzhov: around?
16:03:08 <dpyzhov> Hi. I’m pretty busy with agile trainings these days. As I see, we are pretty close to finally release 6.0
16:03:33 <dpyzhov> There are 3 High bugs in progress
16:03:47 <dpyzhov> #link https://bugs.launchpad.net/fuel/+bug/1403798
16:03:53 <uvirtbot> Launchpad bug 1403798 in fuel "[Upgrade] Upgrade of fuel failed on upgrading openstack stage withHTTPError: 401 Client Error: Unauthorized " [High,In progress]
16:03:59 <dpyzhov> This one should be moved to 6.1
16:04:13 <dpyzhov> #link https://bugs.launchpad.net/fuel/6.0.x/+bug/1402641
16:04:14 <uvirtbot> Launchpad bug 1402641 in fuel/6.0.x "No logs from neutron openvswitch agent" [High,In progress]
16:04:19 <dpyzhov> #link https://bugs.launchpad.net/fuel/6.0.x/+bug/1393557
16:04:22 <uvirtbot> Launchpad bug 1393557 in fuel/6.0.x "Quick Start scripts cannot locate Fuel ISO in 'iso' directory when executed inside Cygwin" [High,In progress]
16:04:29 <dpyzhov> I’m not sure about this bugs
16:04:42 <dpyzhov> But we have a rule to merge only Critical bugfixes
16:05:21 <dpyzhov> Does anyone has info about this bugs?
16:05:58 <kozhukalov> looks like there are no criticals
16:05:59 <angdraug> virtualbox scripts are not shipped with the iso, so we can fix them on master and publish that
16:06:50 <dpyzhov> AFAIK, bug with logs should be mentioned in release notes and moved to 6.1
16:06:59 <angdraug> +1
16:07:03 <dpyzhov> So for me it looks like we are ready for new RC
16:07:28 <angdraug> do you mean we keep #56?
16:07:42 <kozhukalov> yep it is rc3
16:08:01 <dpyzhov> do we have any merged fixes after #56?
16:08:32 <xarses> no logs from neutron agent? its that in master or locally or at all?
16:08:57 <dpyzhov> xarses: on master. Actually logs are on master node. But in wrong place
16:09:15 <xarses> ok, but they are somewhere
16:09:41 <dpyzhov> xarses: yes
16:09:48 <evgeniyl__> also we've got similar bug with sahara https://bugs.launchpad.net/fuel/+bug/1403877
16:09:51 <uvirtbot> Launchpad bug 1403877 in fuel/6.1.x "Sahara logs don't present in UI because it was renamed to "sahara-all"" [Medium,In progress]
16:09:56 <xarses> ok  thx, editing name to reflect that
16:09:57 <docaedo> any risk of those logs, due to being in the wrong place not getting rotated leading to out of disk space eventually?
16:10:02 <dpyzhov> User have to log in to the master node and read the file locally
16:10:15 <dpyzhov> It is a really annoying bug, but not Critical
16:10:37 <angdraug> the only changes we have merged since #56 are two commits in fuel-main
16:11:10 <dpyzhov> angdraug: anything critical?
16:11:15 <xarses> i fixed name to make it not sound critical
16:11:20 <angdraug> #link https://review.openstack.org/142511
16:11:22 <bogdando> docaedo, unmatched logs are still at /var/log/remote of master node and this dir rotated as always
16:11:28 <xarses> #link https://bugs.launchpad.net/fuel/6.0.x/+bug/1402641
16:11:33 <uvirtbot> Launchpad bug 1402641 in fuel/6.0.x "No logs from neutron openvswitch agent" [High,In progress]
16:11:36 <angdraug> #link https://review.openstack.org/142483
16:12:03 <dpyzhov> angdraug: I see only test fixes
16:12:11 <angdraug> yes, Medium and High priority
16:12:30 <angdraug> so no, nothing Critical
16:12:47 <dpyzhov> excellent, we don’t need to respin RC =)
16:13:14 <dpyzhov> I don’t see nurla in the meeting. Does anyone know status of our acceptance tests?
16:14:44 <dpyzhov> Looks like raw estimate is about a week for testing
16:14:58 <dpyzhov> Unfortunately there is noone here to confirm
16:15:47 <kozhukalov> dpyzhov, thanks
16:15:53 <kozhukalov> are you done?
16:16:07 <dpyzhov> thats all from my side
16:16:20 <kozhukalov> ok, moving on then
16:16:25 <kozhukalov> #topic Ubuntu 14.04 update (msemenov)
16:16:32 <msemenov> hi all
16:16:42 <msemenov> So, it's official - we are moving to Ubuntu 14.04 (trusty) in 6.1 scope
16:16:50 <msemenov> Now we can build ISO with trusty based on 6.0 branch
16:16:57 <msemenov> Alexei Sheplyakov is working on it (asheplyakov)
16:17:04 <msemenov> There is a heap of requests related to trusty
16:17:12 <msemenov> #link https://review.openstack.org/#/c/138010
16:17:12 <msemenov> #link https://review.openstack.org/#/c/138110
16:17:12 <msemenov> #link https://review.openstack.org/#/c/138298
16:17:12 <msemenov> #link https://review.openstack.org/#/c/142404
16:17:12 <msemenov> #link https://review.openstack.org/#/c/142405
16:17:13 <msemenov> #link https://review.openstack.org/#/c/142406
16:17:14 <msemenov> #link https://review.openstack.org/#/c/142407
16:17:16 <msemenov> #link https://review.openstack.org/#/c/142408
16:17:20 <msemenov> For these and further requests we need assistance of fuel-main and fuel-library teams
16:17:28 <msemenov> I mean review and some help if it's needed
16:17:43 <msemenov> Also, today we scheduled a weekly cross-team sync up about Ubuntu 14.04 in MOS 6.0
16:17:52 <kozhukalov> msemenov: unfortunately, i did not have time today to review them
16:18:05 <dpyzhov> msemenov: were you able to deploy on 14.04 with these fixes?
16:18:06 <kozhukalov> i hope tomorrow i will
16:18:51 <msemenov> dpyzhov: I think no, today we can just build iso and boot the master node. I'm not sure about the whole deploy
16:19:22 <dpyzhov> msemenov: Ok, looking forward for it
16:19:26 <msemenov> ok
16:20:17 <msemenov> any other questions on this topic?
16:20:54 <mattymo> wait wait wait
16:20:55 <dpyzhov> msemenov: we’ll review patches
16:20:59 <mattymo> fuel master node on Ubuntu?
16:21:45 <msemenov> no, ISO containing trusty
16:22:02 <mattymo> oh just meeting requirements based on Trusty. got it
16:22:04 <msemenov> deployment on the nodes doesn't work well
16:22:29 <msemenov> because of astute problems for example
16:23:01 <msemenov> https://bugs.launchpad.net/mos/+bug/1385079
16:23:03 <uvirtbot> Launchpad bug 1385079 in mos "astute erroneously decides that puppet on nodes has hung and aborts the deployment" [High,Triaged]
16:24:22 <kozhukalov> msemenov: ok, looks like it is a wise decision to move this feature to 6.1
16:24:32 <kozhukalov> moving on?
16:24:38 <msemenov> +
16:24:52 <kozhukalov> #topic Artifact based build process (nmarkov)
16:25:19 <meow-nofer> so, I'm already working on POC of build system for several days
16:25:58 <meow-nofer> it's not really ready to show right now, but kozhukalov is also monitoring progress and we made multiple important architecture decisions
16:26:15 <angdraug> #link https://review.openstack.org/#/c/138994/2/specs/6.1/artifact-based-build-system.rst
16:26:31 <kozhukalov> good to hear we are on the same page about its design now
16:26:38 <meow-nofer> also, there were a couple of meetings with OSCI team on how their system is building packages and how we can integrate together in the future
16:26:54 <meow-nofer> kozhukalov, mostly, yes
16:26:56 <kozhukalov> angdraug: yes, thanks for link
16:27:34 <meow-nofer> so, our current target is to provide somewhat working POC next week and maybe some beautiful scheme of how it will work
16:27:42 <meow-nofer> apart from spec itself
16:28:25 <kozhukalov> meow-nofer: ok, when ready, please ping me earlier to assist with assemblingPOC infrastructure
16:28:28 <angdraug> is there a transition plan?
16:28:30 <meow-nofer> as for me, current code looks promising, but I'm its author and may be not objective
16:28:42 <meow-nofer> kozhukalov have a side look
16:29:00 <angdraug> is it going to be a gradual transition or a single point where we introduce all artefacts?
16:29:09 <meow-nofer> angdraug, transition from make, you mean?
16:29:13 <angdraug> yes
16:29:24 <kozhukalov> angdraug: gradual
16:29:30 <meow-nofer> angdraug, first of all, we'll be able to run our system on top of make
16:29:42 <meow-nofer> and then move on step by step with no hurry
16:29:51 <kozhukalov> at first we are going to split make system into a set of independent parts
16:30:07 <kozhukalov> and create a set of jenkins jobs for them
16:30:53 <kozhukalov> next step is to run make scripts from inside new build system
16:30:56 <meow-nofer> the first achievement should be, we'll be able to build ISO much faster than we do it now
16:31:52 <kozhukalov> meow-nofer: i think there will be many many questions during POC presentation
16:32:20 <angdraug> sounds good, looking forward to see the POC
16:32:22 <meow-nofer> kozhukalov, sure, but as for me, our current implementation is as simple as an axe
16:32:44 <kozhukalov> meow-nofer: we need to start documenting architecture as well as cases
16:32:51 <meow-nofer> kozhukalov, and test it
16:33:30 <meow-nofer> unit testing is strongly required before we'll run something "in production"
16:33:38 <meow-nofer> that's all from me on topic
16:33:44 <kozhukalov> ok
16:33:47 <kozhukalov> thanks
16:33:52 <kozhukalov> moving
16:34:06 <kozhukalov> #topic Open discussion
16:34:29 <kozhukalov> does anyone has any q?
16:34:34 <angdraug> #link https://blueprints.launchpad.net/fuel/+spec/merge-openstack-puppet-modules
16:34:50 <angdraug> can we talk about breaking up this bp?
16:35:29 <angdraug> I think it's an ongoing effort, and since it's not something that requires much design discussions, we can track it as bugs
16:35:49 <angdraug> one bug per module, e.g. "update puppet-haproxy to 0.9.0"
16:35:53 <kozhukalov> aglarendil: around?
16:36:29 <angdraug> aglarendil raised a point over email that we agreed not to track non-defect work in bugs
16:37:03 <angdraug> I think we should be pragmatic about this case
16:37:40 <angdraug> (just to be clear, I think pacemaker-improvements should still be tracked in blueprints, not bugs)
16:37:53 <angdraug> thoughts anyone?
16:38:17 <adanin> There are 2 types of blueprints
16:38:24 <adanin> simple and complex
16:38:44 <adanin> a complex blueprint must have a specification
16:38:52 <adanin> but a simple one doesn't
16:39:23 <adanin> we can track little improvements as simple blueprints
16:39:36 <angdraug> good point
16:39:51 <adanin> and let developer to file and implement them after FF but before SCF
16:40:22 <angdraug> that far, I wouldn't go
16:40:36 <kozhukalov> is there special field in BP (simple or complex) or it is on BP reporter? how can i know which of them are simple?
16:40:41 <angdraug> I'm quite ok with not doing upstream manifest merges after FF
16:41:08 <angdraug> kozhukalov: I guess it just wouldn't have a link to specification
16:41:57 <kozhukalov> angdraug: or maybe it is better to put that into description template we use now for all our BPs
16:42:07 <adanin> angdraug, kozhukalov: we can explicitly specify that it’s a simple BP w/o spec
16:42:28 <xarses> i think this is not angdraug's point
16:42:32 <angdraug> yes, explicitly specifying is better, would help to tell if we don't need a spec or simply haven't created a spec yet
16:42:41 <xarses> we run massive bp cycle over cycle
16:42:43 <xarses> its bad
16:42:50 <xarses> its hard to trak
16:43:05 <angdraug> xarses: I think even aglarendil has agreed that it's bad, based on his email from this morning :)
16:43:15 <xarses> =)
16:43:46 <angdraug> the only downside with tracking upstream merges as simple bp's instead of bugs is that there'll be a lot of them
16:44:06 <angdraug> (hopfully :)
16:44:29 <kozhukalov> adanin:  could you please put your idea about simple/complex BPs to ML thread
16:44:32 <kozhukalov> ?
16:44:46 <xarses> aren't we going to eventually stop merging upstream altogether though
16:44:47 <adanin> kozhukalov: yes, I will
16:45:17 <angdraug> as I mentioned on that thread, we can only stop merging upstream when we continuously use latest upstream
16:46:00 <angdraug> and that's definitely not in scope for 6.1
16:46:30 <kozhukalov> angdraug: is it possible? isn't that going to break anything? i mean using latest upstream
16:46:47 <angdraug> it is possible in theory, but a lot of work to get there in practice
16:47:15 <angdraug> most of the changes we do to upstream manifests can either be pushed to upstream, or moved to the openstack module
16:47:53 <alex_didenko> +1
16:47:54 <angdraug> not to mention that we'll need a much better ci for puppet modules
16:48:13 <angdraug> so that we can quickly catch and fix breakage introduced by upstream changes
16:48:28 <angdraug> same as we currently beginning to do with openstack master in mos
16:48:38 <kozhukalov> angdraug: and if you are going to push your changes to upstream it means you are going to slow down your development process, am i right?
16:49:11 <angdraug> kozhukalov: if you move most of your customizations to openstack module, you won't have to push to upstream that often
16:49:19 <alex_didenko> nope, we can push change upstream and apply it on our copy of module
16:49:39 <angdraug> you can implement an external fix first, then propose native fix to upstream, then remove external fix when upstream version having your fix lands
16:49:50 <angdraug> and what alex_didenko said, too
16:50:20 <kozhukalov> ok, sounds like a great plan
16:50:25 <angdraug> before we consider that we need to meet 2 large goals
16:50:32 <angdraug> 1) finish catching up to upstream
16:50:41 <angdraug> 2) expand scope of ci for fuel-library
16:50:58 <angdraug> that's why I don't think we can even consider this in 6.1
16:51:19 <kozhukalov> yes, it is not so easy
16:51:36 <angdraug> I think we're done with this topic, anyone else for open discussion?
16:51:42 <OKosse> o/
16:52:01 <kozhukalov> OKosse: hi
16:52:05 <OKosse> We've already finished acceptance tests of vcenter , vcenter+glance, nsx, vcenter+vlan on rc2. We are in progress of testing our features on rc3. we have 3 bugs in progress and moved to 6.1, but they are not critical
16:52:05 <adanin> OKosse: please
16:52:29 <angdraug> that's great news
16:53:46 <kozhukalov> OKosse: good to hear, thanks
16:53:56 <kozhukalov> anything else?
16:54:18 <kozhukalov> looks like we are done
16:54:34 <kozhukalov> thanks everyone
16:54:36 <kozhukalov> closing
16:54:42 <kozhukalov> #endmeeting