16:00:27 <vkozhukalov> #startmeeting Fuel
16:00:28 <openstack> Meeting started Thu Jul 17 16:00:27 2014 UTC and is due to finish in 60 minutes.  The chair is vkozhukalov. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:29 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:31 <openstack> The meeting name has been set to 'fuel'
16:00:37 <vkozhukalov> #chair vkozhukalov
16:00:38 <openstack> Current chairs: vkozhukalov
16:00:39 <evgeniyl> Hi
16:00:48 <vkramskikh> hi
16:00:53 <xarses> Hi Dr. Nick
16:00:55 <ikalnitsky> o/
16:00:55 <vkozhukalov> #topic Greetings, announcements
16:01:25 <vkozhukalov> #topic 5.1 feature freeze status
16:01:26 <meow-nofer> hi
16:01:35 <mattymo> hi
16:01:46 <aviramb_> hi
16:01:52 <vkozhukalov> dpyzhov: do you have any details on this topic
16:02:55 <vkozhukalov> afaik we merged a lot of new code and found a lot of new issues
16:03:05 <angdraug> doesn't look like we have any changes WRT features since last week
16:03:17 <angdraug> bug numbers from the Tuesday's bug squashing are here:
16:03:17 <angdraug> #link https://docs.google.com/a/mirantis.com/spreadsheets/d/10mUeRwOplnmoe_RFkrUSeVEw-__ZMU2nq23BOY-gzYs/edit#gid=1298115632
16:03:39 <vkozhukalov> thanx angdraug
16:03:47 <angdraug> we should be making this much progress everyday to fix enough high priority bugs by 5.1 release
16:04:12 <angdraug> mihgen proposed to have soft code freeze next thursday
16:04:33 <vkozhukalov> current topic is feature freeze for 5.1
16:05:27 <aglarendil> so, I will replace mihgen
16:05:59 <aglarendil> we have almost all features merged into the master, except some of mlnx code
16:06:08 <aglarendil> we are waiting for CI to pass for this code
16:06:30 <aglarendil> there is a floating bug related to swift that breaks ubuntu CI pass
16:06:33 <aglarendil> #link https://review.openstack.org/#/c/107749/
16:06:58 <aglarendil> as soon as it is merged we can retrigger jobs and if they pass - merge mlnx code into fuel-library
16:07:10 <aglarendil> also, there is some code in merge queue for fuel-web
16:07:22 <aglarendil> dpyzhov can provide update on it
16:07:36 <angdraug> aglarendil: what about NSX? did it land?
16:07:42 <aglarendil> angdraug: yep
16:07:55 <aglarendil> angdraug: also there is galera improvements code that we need to merge into fuel-library
16:08:19 <vkozhukalov> yes a lot of requests in fuel-web are waiting to be merged
16:08:33 <aglarendil> #link https://review.openstack.org/#/c/106516/
16:08:42 <aglarendil> it is also blocked by swift failure
16:08:52 <aglarendil> xarses: can provide update on ML2 support
16:09:05 <dpyzhov> I see two unmerged mellanox reviews: https://review.openstack.org/#/c/106007/ and https://review.openstack.org/#/c/104501/
16:09:37 <dpyzhov> All other mellanox code is in master
16:10:08 <xarses> aglarendil: the last update in the ML is still correct. HA is still a problem because of OCF scripts and upstart are fighting to run the services
16:10:31 <aglarendil> xarses: we do not know about this issue. you need to land override scripts for upstart
16:10:35 <vkozhukalov> aglarendil wrote ^^^ that they are waiting for mlnx code to pass ci tests
16:10:53 <dpyzhov> vkozhukalov: I’m talking about fuel-web repo only
16:10:55 <aglarendil> xarses: as we did in our code for other neutron agents
16:10:55 <xarses> aglarendil: correct, that is the next step
16:11:07 <aglarendil> xarses: in our current code it is already done
16:11:26 <aglarendil> xarses: regarding swift bug, you can apply https://review.openstack.org/#/c/107749/
16:11:31 <vkozhukalov> dpyzhov: what about fuel_agent stuff? there are 3 requests
16:11:47 <vkozhukalov> 2 for fuel-main and 1 for fuel-web
16:12:01 <xarses> aglarendil: the code in our puppet-neutron is very inconsistent with how each service is stopped, so I'm trying to work out how much beating each service needs
16:12:45 <aglarendil> xarses: we are using two approaches: 1) disable default provider's management of services 2) add override files for upstart
16:13:14 <xarses> 3) use an exec to ensure that it's killed
16:13:35 <aglarendil> xarses: this is pretty harsh way. I would not use it
16:13:47 <xarses> We already do
16:14:01 <xarses> which is why im iterating through how much 'beating' the services need
16:14:20 <aglarendil> xarses: let's move it to follow up and discuss separately in fuel-dev after the meeting
16:14:25 <xarses> yes
16:14:31 <angdraug> doesn't sound like ML2 is making it to 5.1
16:14:52 <angdraug> it's already past its FF exception deadline
16:14:57 <angdraug> I mean upstream ML2
16:15:31 <aglarendil> angdraug: unless we do very thorough testing and merge it carefully
16:15:45 <aglarendil> angdraug: but I would move it to 6.0 along with all the other upstream manifests merge
16:15:52 <angdraug> +1
16:16:08 <angdraug> moving on?
16:16:15 <vkozhukalov> moving
16:16:17 <aglarendil> also, upstream manifests merge is half completed
16:16:22 <aglarendil> just a moment, guys
16:16:25 <aglarendil> #link https://blueprints.launchpad.net/fuel/+spec/merge-openstack-puppet-modules
16:16:27 <vkozhukalov> #topic 5.0.1 release status
16:17:00 <angdraug> we still have 2 bugs blocking 5.0.1 afaiu:
16:17:11 <angdraug> #link https://bugs.launchpad.net/fuel/+bug/1342719
16:17:13 <uvirtbot> Launchpad bug 1342719 in fuel "Rabbit is unavailable" [Critical,Confirmed]
16:17:16 <angdraug> #link https://bugs.launchpad.net/fuel/+bug/1305826
16:17:18 <uvirtbot> Launchpad bug 1305826 in fuel/5.1.x "Swift Ringbuilder rebalance fails" [Critical,In progress]
16:17:30 <angdraug> anyone got update on those 2?
16:17:58 <aglarendil> I have
16:18:04 <aglarendil> the first one should be addressed by oslo.messaging patch
16:18:12 <aglarendil> which is under development and testing now
16:18:22 <angdraug> it should have been by previous 2 versions of that patch
16:18:43 <aglarendil> it requires more work as it seemed
16:19:04 <angdraug> #link https://bugs.launchpad.net/fuel/+bug/1340711
16:19:08 <uvirtbot> Launchpad bug 1340711 in fuel/5.0.x "[OSCI] building oslo-messaging for rpm and deb" [Critical,In progress]
16:19:08 <angdraug> is this the same thing?
16:19:33 <aglarendil> looks like
16:19:44 <angdraug> also, why is #1305826 different priority for 5.0.1 and 5.1? is it critical in 5.0.1 or not?
16:19:59 <aglarendil> angdraug: it is not breaking current deployments
16:20:08 <aglarendil> angdraug: but may if puppet changes ordering of resources
16:20:11 <vkozhukalov> do we need to sqash those bugs?
16:20:14 <aglarendil> angdraug: and it is blocking CI now
16:20:26 <aglarendil> there is fix available already
16:20:40 <aglarendil> #link https://review.openstack.org/#/c/107767/
16:20:43 <angdraug> anyone from QA around?
16:21:24 <nurla> yes
16:21:28 <angdraug> is ISO #135 our RC2?
16:21:42 <angdraug> latest comment on the rabbitmq bug is that it's fixed there
16:21:55 <angdraug> if it is, I think it should be the RC2
16:22:02 <nurla> looks like no. because we wait new iso with oslo path
16:22:22 <angdraug> if it's not, then we need to make another ISO and that means we can as well merge the fix for swift ringbuilder bug
16:22:23 <nurla> path only in custom repo
16:22:49 <nurla> yes, fix for swift we can merge
16:22:56 <angdraug> thanks!
16:23:16 <nurla> uw)
16:23:23 <angdraug> summary for 5.0.1 status: 2 more fixes to merge before we can build an RC, links above
16:23:49 <vkozhukalov> ok, moving
16:24:05 <vkozhukalov> #topic LP tags vs [tags] in bug summary vs LP karma
16:24:29 <vkozhukalov> rmoe: around?
16:24:34 <xarses> he's typing
16:24:34 <rmoe> yep
16:25:01 <rmoe> starting tuesday (maybe?) I noticed a whole bunch of title updates adding free-form tags like [nailgun]
16:25:03 <mattymo> I like duplicate of [tag] and tag
16:25:05 <rmoe> 1. why are we doing this?
16:25:09 <rmoe> 2. why was there no discussion of this?
16:25:37 <rmoe> launchpad already has tags so adding [tag] to everything seems dumb
16:25:53 <xarses> it also breaks email update threads
16:26:07 <rmoe> yes
16:26:09 <vkozhukalov> yes, it is a little bit strange
16:26:22 <angdraug> and takes away horizontal space from bugs table that could be used for information that's actually useful
16:26:27 <rmoe> and all free-form tagging systems devolve into messes of 20 variations of each tag
16:26:34 <angdraug> I'd rather have a short and descriptive title than a mix of tags
16:26:41 <angdraug> especially if there's more than 2 of them on the bug
16:27:19 <angdraug> proposal: let's stop doing this for now and have a discussion on ML
16:27:32 <xarses> +1
16:27:33 <AndreyDanin> I agree.
16:27:34 <vkozhukalov> does launchpad allow to delete tags we think are not needed?
16:27:39 <angdraug> of course
16:28:14 <vkozhukalov> automatically? api?
16:29:02 <vkozhukalov> or we need to go through all bugs and remove [tags] manually?
16:29:30 <angdraug> I think we should do it manually
16:29:35 <xarses> in titles? lets just leave them alone
16:29:41 <AndreyDanin> You can move “oficial” tag to “unofficial” tag list
16:29:41 <angdraug> and while at it, improve bug titles
16:29:50 <angdraug> xarses: actually yes
16:30:09 <angdraug> if you can improve a bug title, just remove [tags] from it while you're at it
16:30:24 <angdraug> otherwise, let's just avoid even more spam and email thread breakage
16:30:59 <vkozhukalov> ok, moving
16:31:18 <vkozhukalov> #topic Fuel public builds
16:31:27 * xarses cheers
16:31:30 <teran> Recently we had updated the status of our task to publish nightly builds
16:31:30 <teran> #link https://bugs.launchpad.net/fuel/+bug/1308502
16:31:30 <teran> Currently it’s publicly available to download with torrents
16:31:32 <uvirtbot> Launchpad bug 1308502 in fuel "[devops] publish nightly builds for seeding" [High,Confirmed]
16:31:32 <teran> #link https://fuel-jenkins.mirantis.com/view/ISO/job/publish_fuel_community_iso/
16:31:34 <teran> Please feel free to use :)
16:31:42 <xarses> +100500
16:32:06 <teran> The only left item in that case is to set up smoke tests for them, and we'll done it soon
16:32:26 <angdraug> that's icehouse based builds?
16:32:37 <angdraug> or juno/master?
16:32:39 <teran> yeah, it's master
16:32:46 <angdraug> woot!
16:32:47 <teran> icehouse
16:32:51 <angdraug> boo :(
16:32:51 <nurla> icehouse
16:32:57 <vkozhukalov> also there are some ideas how to make it faster building iso
16:33:53 <vkozhukalov> briefly the idea is to move from file based build system to artifact build system
16:33:58 <angdraug> what about juno packages?
16:34:00 <angdraug> #link https://blueprints.launchpad.net/fuel/+spec/build-packages-for-openstack-master-rpm
16:34:18 <angdraug> combined with above, would be really really awesome
16:34:25 <nurla> this feature will be available on osci-ci)
16:34:56 <angdraug> my ultimate dream here is to use fuel ci for commits in openstack master branches
16:34:58 <xarses> vkozhukalov: as long as we can still build ISO's with out needing lots of tools that are hard to build on your own, like your on Jenkins server
16:35:13 <dilyin> hello
16:35:22 <dilyin> about tagging bugs.
16:35:50 <angdraug> dilyin: we agreed to have further discussion on ML
16:35:55 <xarses> We moved on, we asked to stop doing it in titles, and will discuss further in ML
16:36:52 <vkozhukalov> xarses: it is not supposed to make harder to build iso on one's own
16:37:05 <vkozhukalov> let's discuss it next time
16:37:15 <xarses> vkozhukalov: sounds good!
16:37:33 <vkozhukalov> i am going to recall my experience in this field from my previous job
16:38:23 <vkozhukalov> ok, sounds like we have more than 20 minutes for open discussion
16:38:50 <aglarendil> is anyone from mlnx here ?
16:38:54 <vkozhukalov> it does not make much sense to discuss statuses of particular features
16:39:06 <aviramb_> yes
16:39:06 <vkozhukalov> #topic open discsussion
16:39:10 <nuritv> i'm from mellanox
16:39:27 <aglarendil> aviramb_: you can discuss partiuclar questions with fuel-web guys if you want
16:40:21 <vkozhukalov> nuritv: aviramb_ please feel free to add topics in agenda https://etherpad.openstack.org/p/fuel-weekly-meeting-agenda if you have something to discuss
16:40:40 <aviramb_> we still need some reviews for https://review.openstack.org/#/c/104501/11 and https://review.openstack.org/#/c/106007/
16:41:13 <angdraug> dpyzhov: can anyone from fuel-web go through these?
16:41:31 <angdraug> I'm plan to have a look later today but I'd prefer to have more than one pair of eyeballs on it
16:41:48 <vkozhukalov> AndreyDanin: is going to review some of your patches
16:42:10 <aviramb_> already fixed some comments of angdraug
16:42:11 <AndreyDanin> aviramb_: I’ll review these PR today
16:42:30 <aviramb_> centos packages checked and will be merge soon
16:42:40 <aviramb_> then https://review.openstack.org/#/c/101126/ can be merged
16:42:50 <aviramb_> about bootstrap
16:42:58 <dpyzhov> angdraug: I’ve added meow-nofer to reviews in fuel-web
16:43:04 <aviramb_> iSER rename+manifests(library) https://review.openstack.org/#/c/104323/ , https://review.openstack.org/#/c/104621/ waits to pass CI
16:43:51 <aviramb_> library commits already passed many review cycles
16:43:53 <dpyzhov> nuritv aviramb_:  and we are looking forward for unit tests
16:44:17 <nuritv> dpyzhov: will be ready early next week!
16:44:27 <dpyzhov> nuritv: great
16:45:10 <aviramb_> we need some POC to work with on that, in order to make this efficient, talked about it with Evgeniya
16:45:22 <aviramb_> (UT)
16:46:52 <nuritv> also, we still have this bug that affects us (https://bugs.launchpad.net/fuel/+bug/1341486).
16:46:53 <vkozhukalov> dpyzhov: what about https://review.openstack.org/#/c/101126/ ?
16:46:55 <uvirtbot> Launchpad bug 1341486 in fuel "[provision] additional kernel_params not reflected in  ubunut /boot/grub/grub.cfg" [High,Confirmed]
16:47:19 <dpyzhov> vkozhukalov: no
16:47:42 <dpyzhov> We will have separate bootstrap for now
16:48:03 <dpyzhov> Because we need to be really sure that extra modules do not affect anything
16:48:32 <vkozhukalov> dpyzhov: what does it mean separate?
16:48:39 <nuritv> dpyzhov: we only load it on mlnx HW  and we also did some test in our lab
16:48:42 <dpyzhov> it will be another bootstrap image
16:48:47 <nuritv> do you need us to do anything else?
16:49:00 <eshumakher> dpyzhov MLNX team changed the implementation
16:49:01 <dpyzhov> nuritv: did you test it on other hardware?
16:49:04 <aviramb_> this is the way centos loads mlnx in case of mlnx hw
16:49:36 <eshumakher> the MLNX kernel modules uploads only if MLNX HW is in place
16:49:37 <nuritv> dpyzhov: we tested it on VM with no MLNX driver
16:49:51 <dpyzhov> nuritv: nice
16:50:04 <vkozhukalov> dpyzhov: afaiu we gonna have saparate bootstrap for every extra kernel module? right?
16:50:05 <eshumakher> i have a test report btw
16:50:11 <angdraug> dpyzhov: confirmed, the way it's implemented now won't impact non mlnx hw configurations
16:50:14 <eshumakher> will share it with the QA team
16:51:03 <vkozhukalov> dpyzhov: if so then it is a bit strange. isn't it?
16:51:44 <dpyzhov> vkozhukalov: plan was to use separate bootstrap as a temporary solution
16:51:55 <dpyzhov> good that we do not need it now
16:52:18 <dpyzhov> +1 from me, and our osci team will merge it after update of our mirrors
16:52:40 <vkozhukalov> dpyzhov: great
16:54:29 <vkozhukalov> ok, looks like we've managed to discuss everything we had
16:54:35 <vkozhukalov> closing?
16:54:48 <vkozhukalov> thanx everyone
16:55:01 <vkozhukalov> #endmeeting