16:00:23 <sdake> #startmeeting kolla 16:00:29 <openstack> Meeting started Wed Jul 8 16:00:23 2015 UTC and is due to finish in 60 minutes. The chair is sdake. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:29 <mfalatic> o/ 16:00:30 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:33 <openstack> The meeting name has been set to 'kolla' 16:00:35 <sdake> #topic rollcall 16:00:38 <mfalatic> oops 16:00:40 <mfalatic> o/ 16:00:46 <rhallisey> hi 16:00:48 <inc0> o/ 16:00:51 <mstachow> o/ all 16:00:53 <sdake> o/ folks - lots of progress in the code base yay ;) 16:00:58 <SamYaple> \o 16:01:42 <kevsi_> Hi. 16:01:48 <pbourke> o/ 16:02:00 <sdake> pretty sure jpeeler is on pto because of impending loss of freedom ;) 16:02:13 <jpeeler> no i'm here still! 16:02:18 <sdake> oh yay :) 16:02:20 <SamYaple> /kick jpeeler 16:02:21 <jpeeler> thanks for the ping 16:02:24 <jpeeler> noooooooo 16:02:26 <SamYaple> ;) 16:02:35 <sdake> #topic announcements 16:02:52 <sdake> this timeslot hs historically been at 16:00 UTC 16:03:02 <sdake> we are goingto try movin it 30 minutes later to make more people happy 16:03:07 <sdake> that will happen in 2 weeks 16:03:16 <sdake> so not next meeting but the meeting after 16:03:22 <sdake> i'll send an announcement to the ml 16:03:30 <sdake> if the time is problematic now is your chance to speak up :) 16:03:47 <sdake> the 2200 timeslot remains the same 16:03:53 <inc0> well, this one is late anyway 16:04:02 <SamYaple> 2200 that one is problematic 16:04:05 <SamYaple> :) 16:04:20 <sdake> samyaple lets fine tuen that after we get 1630 solid ;) 16:04:26 <SamYaple> i wont derail further 16:04:51 <sdake> #2 - please register for kolla-palooza midcycle - link in #kolla topic 16:05:07 <sdake> need an accurate count for dinner , hopefully by 9th of july ;) 16:05:15 <sdake> (as in tomorrow) 16:05:21 <sdake> any other announcements from folks? 16:05:30 <pbourke> sdake: do remote attendees need to register? 16:05:46 <sdake> pbourke please do so I can sort out logistics 16:05:53 <sdake> but its not mandatory 16:05:57 <pbourke> ok 16:06:09 <sdake> that way I can send you a personal email with login info etc 16:06:22 <sdake> we will probably use google groups or webex 16:06:25 <sdake> webex scales better 16:06:35 <sdake> but runs on linux poorly 16:06:58 <sdake> i'm going to try to find some usb handheld microphones prior to the event 16:07:17 <sdake> if anyone has a suggestion on mic hardware, send me a private note at stdake@cisco.com pls 16:07:38 <sdake> #topic CI Status 16:08:18 <sdake> so still blocked on internal config for openvswitch 16:08:24 <diga> O/ 16:08:26 <sdake> anyone want to tackle this problem? 16:08:27 <sdake> hey diga 16:08:40 <diga> hi sdake 16:08:50 <diga> I am working on that 16:08:59 <diga> Will resolve that stuff by tomorrow 16:09:04 <sdake> diga nnice, let me set arbitrary deadline of 15th for that ;) 16:09:14 <diga> +1 16:09:16 <diga> sorry was too much busy from last two days 16:09:23 <sdake> welcome to life ;) 16:09:38 <diga> :) 16:09:38 <sdake> cool so once that is unblocked samyaple has the multi-host setup for it ready to go I think 16:09:46 <sdake> and we can put it in the gate 16:09:52 <sdake> and be ready to functional test via smoke testing 16:09:54 <jpeeler> yay 16:10:05 <diga> yes 16:10:09 <jpeeler> on that topic, is the new build script ready to be put in the gate? 16:10:22 <sdake> jpeeler ya i think samyaple was going to hack on project-config to do it 16:10:26 <SamYaple> jpeeler: i was thinking a test non-voting gate 16:10:29 <sdake> but if someone else wants to do it that would be fantastic 16:10:33 <SamYaple> but i dont know how to do that 16:10:34 <sdake> but we dont want to edit the existing gate 16:10:37 <sdake> we want a new one 16:10:40 <SamYaple> yes 16:10:52 <jpeeler> i can help SamYaple 16:10:55 <sdake> jpeeler interested in that? 16:11:05 <SamYaple> yay 16:11:14 <SamYaple> i get to learn 16:11:16 <sdake> #action jpeeler to help sam sot out new python script as a nonvoting gate 16:11:34 <sdake> #topic Ansible Modules License Status 16:11:53 <sdake> so in essence we need modules that are licensed pglv3 but we can't use them 16:12:01 <sdake> because it affects our ability to get into big tent 16:12:07 <sdake> so samyaple wrote his own ;) 16:12:16 <sdake> samyaple can you give us a summary pls 16:12:22 <SamYaple> sure 16:12:49 <SamYaple> openstack-infra/shade is a library that allows generic wrappings for openstack environemnts 16:12:54 <SamYaple> the ansible modules use shade 16:13:06 <SamYaple> the modules are gplv3 licensed and cant be relicensed 16:13:22 <SamYaple> i have built some "they do the job" modules in the meantime 16:13:36 <SamYaple> they only add, no updates or deletes, but it should get us by 16:13:52 <SamYaple> ansible 2.0 drops and we will be able to switch to the upstream modules 16:14:03 <SamYaple> thats about everything 16:14:14 <SamYaple> we are unblocked, but not 100% strength 16:14:27 <sdake> re ansible 2.0 drops and we will be able to use upstream modules - I need to run that by legal apparently samyaple 16:14:30 <sdake> which is in progress 16:14:38 <SamYaple> why they accepted OSAD 16:14:43 <SamYaple> same deal 16:15:10 <sdake> agree we are using same model as osad here with the modules - make our own 16:15:27 <sdake> but both of our projects need to work together to make sure we can use these new moduels with the legal team at the foundation 16:15:46 <sdake> or it could affect our ability to use the trademark 16:15:52 <SamYaple> what i am saying is the "new modules" will be just like any other ansible modules. 16:15:55 <sdake> (which we want to be able to use) 16:16:00 <SamYaple> and we use lots of baked in ansible modules 16:16:07 <SamYaple> thats basically all ansible is, gplv3 modules 16:16:31 <sdake> ya best not to thin about it until we are presented wit hthe problem ;) 16:16:34 <pbourke> was going to say the same. they will essentially be like using 'copy' 16:16:36 <SamYaple> fair 16:16:43 <pbourke> but best be safe 16:16:45 <SamYaple> well move on 16:17:02 <sdake> #tpoic Liberty-2 planning (#link https://launchpad.net/kolla/+milestone/liberty-2) 16:17:59 <sdake> #topic Liberty-2 planning 16:18:10 <sdake> #link https://launchpad.net/kolla/+milestone/liberty-2 16:18:16 <sdake> pls open 16:18:36 <sdake> lots of blue - yay :) 16:18:42 <SamYaple> pls give me more blueprints 16:18:48 <sdake> some purple in there - that is fantastic for only the first week :) 16:19:02 <sdake> you definately need m0ar sam :) 16:19:16 <sdake> so just a reminder our deadline is july 31st 16:19:24 <sdake> can we get a quick update of two blueprints: 16:19:34 <diga> sdake: Need to talk about this BP - https://blueprints.launchpad.net/kolla/+spec/kolla-compose-script 16:19:46 <sdake> #link https://blueprints.launchpad.net/kolla/+spec/ansible-service 16:19:48 <sdake> diga ok hang tight bro 16:19:58 <sdake> samyaple can you give us quick update on that 16:20:02 <diga> ok 16:20:25 <SamYaple> we have 4 services done, mariadb,rabbitmq,keystone,glance (last two in code review) 16:20:38 <SamYaple> I will be handling nova and neutron 16:20:53 <SamYaple> then it will be funcitonal and the others can be handled in no particular order 16:20:53 <sdake> who is tackling heat? 16:21:05 <sdake> heat needs to be in our core list as well 16:21:11 <SamYaple> we have no assignes for these others yet. they would need constant rebasing 16:21:18 <sdake> ok 16:21:29 <sdake> we willw ait for you to wrap up then on nova and neutron 16:21:29 <SamYaple> it will probably be a week before we can do this async 16:21:34 <sdake> and then we will ask people to volunteer for the remainder 16:21:40 <SamYaple> agreed 16:22:01 <sdake> #link https://blueprints.launchpad.net/kolla/+spec/standard-start 16:22:42 <sdake> the progress on this is fantastic 16:22:45 <sdake> I think we are almost done here 16:22:57 <sdake> everything is up for code review afaik 16:23:26 <sdake> there may be a few lagging, so if you have implementation for this blueprint pls get her done this week ;) 16:23:46 <sdake> diga shoot with your blueprint 16:24:13 <diga> just want to discuss on - https://blueprints.launchpad.net/kolla/+spec/kolla-compose-script 16:24:27 <sdake> that should bascially be a git mv 16:24:30 <sdake> not much to that one 16:24:47 <diga> okay 16:24:54 <sdake> did that answer your q? 16:24:55 <SamYaple> sdake: it also includes writing the kolla-ansible script 16:25:01 <SamYaple> "Rename the kolla script to kolla-compose and create a new kolla-ansible script to manage playbook operation." 16:25:06 <sdake> samyaple there is a separate blueprint for kolla-ansible 16:25:06 <diga> yepm yep 16:25:15 <SamYaple> ill edit the BP then 16:25:20 <sdake> sounds good 16:25:38 <sdake> ya guys when i write these blueprints i dont always do a perfect job 16:25:46 <sdake> feel free to edit them to your liking ;) 16:25:49 <diga> :) 16:26:01 <sdake> liberty 2 is off to a great start 16:26:10 <sdake> anyone know status of cinder or horizon? 16:26:16 <sdake> rhallisey on cinder ? 16:26:20 <sdake> since these were broken in liberty 1 release 16:26:47 <pbourke> sdake: a recent joiner to our team has a fix that apparently fixes horizon 16:26:54 <sdake> nice! 16:26:55 <rhallisey> sdake, nothing has changed, I just added oslo in so it won't sit in restarting 16:27:05 <sdake> rhallisey does that fix cinder? 16:27:07 <pbourke> sdake: he's in the process of filing a bug and creaing a review 16:27:16 <sdake> pbourke there is a bug already 16:27:20 <sdake> pbourke what was his name do you recall? 16:27:20 <pbourke> aha 16:27:28 <sdake> or her name 16:27:28 <rhallisey> I'm assuming the iscsi issue is still hanging around, so no not yet 16:27:48 <rhallisey> I've been cleaning the 200 docker images off my machine for the last 2 hours 16:27:54 <sdake> rhallisey ack - if you need some kind of help feel free to ask in #olla ;) 16:28:00 <rhallisey> sure thing 16:28:06 <sdake> fastest way to do that is rm -rf /var/lib/docker ;) 16:28:16 <sdake> and resstart docker 16:28:24 <rhallisey> that would do it 16:28:27 <rhallisey> lol 16:28:32 <sdake> thats what I use ;) 16:28:39 <rhallisey> drop a bomb on docker lol 16:28:43 * jpeeler has wondered if that would work 16:29:10 <sdake> #topic Midcycle sesion planning 16:29:16 <sdake> jpeeler it works fantastically 16:29:39 <pbourke> sdake: his name is mark I'll get him to give you a shout about it 16:29:40 <sdake> #link ttps://etherpad.openstack.org/p/liberty-kolla-midcycle-planning 16:29:53 <sdake> pbourke just point him at the correct existing bug 16:30:01 <pbourke> will do 16:30:05 <sdake> if you woudln't mind :) 16:30:12 <rhallisey> sdake, missing an h on the paste 16:30:15 <sdake> if folks can open that midcycle planning page 16:30:19 <rhallisey> 'ttps' 16:30:34 <sdake> #link https://etherpad.openstack.org/p/liberty-kolla-midcycle-planning 16:30:39 <SamYaple> you have to break the code rhallisey 16:32:58 <markd_> sdake: just creating the horizon bug that Pbourke mentioned. is there a bug already on the Horizon login ? 16:33:01 <sdake> ok line 9 guys 16:33:14 <pbourke> markd_: check your pidgin ;) 16:33:16 <sdake> markd_ yes let me get with you after meeting ok? 16:33:27 <markd_> Ok 16:33:29 <sdake> or pbourke can point the bug out to you 16:33:46 <sdake> so add design sessiosn you would like to discuss at line 9 16:33:48 <sdake> each session is 50 mins in length 16:33:50 <sdake> with a 10 min break 16:34:00 <sdake> maybe 45/15 , have to see how thigns are going ;) 16:40:54 <SamYaple> diga: for the checks, what do you mean? 16:42:34 <diga> I mean, when we create containerfor services - Need to have check whether service is running properly & when it goes down kolla should know 16:43:10 <SamYaple> diga: "kolla should know" hmmm well have to talk about what kolla is then 16:43:31 <diga> SamYaple: I dont know but this is something important for multinode depoyment model 16:43:46 <SamYaple> agreed, im not trying to kick it away :) 16:43:52 <sdake> we will wrap this up at 50 after 16:44:00 <diga> I mean it should return the status of conainers 16:44:16 <sdake> then have anothe collaborative editing session at our next meeting to get wit hteh folks that can only make our 2200 slot 16:44:40 <SamYaple> after it deploys, yes. actively check containers as if kolla were a service? i would argue no 16:44:46 <sdake> please add more detail for lines 23/24/25/26 etc ;) 16:50:04 <SamYaple> Session Topic: How to pronouce Kolla 16:50:12 <sdake> haha 16:50:12 <rhallisey> nice 16:50:14 <sdake> who nows ;) 16:50:18 <sdake> #topic open discussion 16:50:22 <sdake> ok thanks folks that was really good 16:50:25 <SamYaple> were all going to show up and say it differently 16:50:27 <sdake> we will finish the job on that next week 16:50:31 <jpeeler> hey, so i have a pending blueprint i need to write 16:50:34 <sdake> and have a final schedule published friday of next week 16:50:58 <jpeeler> it's a container that uses heat to deploy our kolla containers 16:51:10 <jpeeler> i just wanted to mention it here... since i've been mostly silent about it. apologies 16:51:18 <jpeeler> plan on writing it up today 16:52:04 <SamYaple> I will use this as an oppurunity to ask for reviews on two important patches I would like to see merged ASAP so i dont have 4 layer deep patchsets to rebase again :) https://review.openstack.org/#/c/198494/ https://review.openstack.org/#/c/199406/ 16:52:23 <sdake> samyaple will review after meeting 16:52:39 <SamYaple> jpeeler: i like that idea! i was thinking of a container to deploy to keep the host clean as well 16:52:43 <sdake> jpeeler osunds good 16:52:43 <SamYaple> looking forward to it! 16:53:17 <rhallisey> ya it's going to be pretty cool 16:53:48 <SamYaple> i want to get the nova-docker driver running in dokcer ;) 16:53:55 <sdake> groan 16:54:02 <SamYaple> ok fine, just running at all 16:54:08 <sdake> lol 16:54:31 <sdake> jpeeler go ahead file your blueprint i'll put in discussion state 16:54:37 <sdake> and add to next week agenda for discussion 16:54:46 <sdake> new blueprints go through discussion->approved 16:54:46 <jpeeler> ok will do 16:54:47 <sdake> unless they com with code 16:54:51 <sdake> then they generally go straight to approved :) 16:54:52 <SamYaple> hey real quick poll. Are we going to be including ceph in Kolla? or just supporting it as a backend? +1 or -1 16:55:02 <sdake> ceph yes 16:55:03 <jpeeler> sdake: oh well then i'll go the code route heh 16:55:05 <sdake> if we can make it happen 16:55:05 <inc0> +1 16:55:07 <mstachow> +1 16:55:37 <pbourke> +1 16:55:50 <rhallisey> let me fix cinder then I'll move on to that +1 16:55:58 <SamYaple> if you guys want ceph containers, I can make that happen real quick like. 16:56:03 <SamYaple> rhallisey: i already have ceph containers 16:56:17 <inc0> please do Sam 16:56:24 <rhallisey> that would be cool 16:56:31 <inc0> and Ryan can move to ceph as default 16:56:36 <rhallisey> yup 16:56:41 <inc0> and ditch whole iScsi thing alltogether 16:56:42 <SamYaple> fair enough. ill add an implementation of it and you just kick it around 16:56:46 <sdake> keep in mind folks liberty 2 needs to work out of the box ;) 16:56:55 <SamYaple> yea this is L3 16:57:08 <sdake> wfm 16:57:13 <SamYaple> sdake: what needs to work in L2....? 16:57:19 <SamYaple> just core or ALL containers? 16:57:26 <SamYaple> define core 16:57:28 <sdake> bascailly all our containers in ansible would be my goal 16:57:32 <sdake> what we deploy today 16:57:37 <sdake> in compose 16:58:05 <sdake> i think gaps are heat, horizon, cinder 16:58:05 <sdake> that we dont have ansible code for 16:58:12 <SamYaple> yea cant promise cinder since it deals with the disks 16:58:15 <SamYaple> the rest yea 16:58:16 <sdake> and then we need a swift and ceilometer implemtnation at some point down the road 16:58:28 <sdake> ya cinder is busted container wise 16:58:33 <sdake> we need it working :) 16:58:46 <sdake> ok ending meeting we can overflow in #kolla 16:58:47 <sdake> #endmeeting