16:00:39 <adrian_otto> #startmeeting containers
16:00:40 <openstack> Meeting started Tue Apr 28 16:00:39 2015 UTC and is due to finish in 60 minutes.  The chair is adrian_otto. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:41 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:43 <openstack> The meeting name has been set to 'containers'
16:00:46 <adrian_otto> #link https://wiki.openstack.org/wiki/Meetings/Containers#Agenda_for_2015-04-28_1600_UTC Our Agenda
16:00:50 <adrian_otto> #topic Roll Call
16:00:53 <sdake_> o/
16:00:54 <apmelton> o/
16:00:55 <adrian_otto> Adrian Otto
16:01:01 <Tango> Ton Ngo
16:01:08 <rpothier> o/
16:01:42 <juggler> o/
16:01:51 <dims> o/
16:01:55 <tcammann_> o/
16:01:59 <thomasem> o/
16:02:17 <adrian_otto> hello sdake_, apmelton, Tango, rpothier, juggler, dims, tcammann_, and thomasem
16:02:25 <adrian_otto> we will begin in a moment
16:02:27 <jjlehr> o/
16:02:45 <diga_> o/
16:03:40 <adrian_otto> hello jjlehr, and diga_
16:03:59 <diga_> hello adrian_otto
16:04:18 <adrian_otto> ok, let's begin
16:04:26 <adrian_otto> #topic Announcements
16:04:34 <adrian_otto> 1) Reminder: Our IRC Meeting will be skipped on 2015-05-19 because we will be at the Vancouver Design Summit
16:04:59 <adrian_otto> 2) We did not release on schedule on 2015-04-25. I expect to complete the release work today.
16:05:16 <adrian_otto> after the release we will have two branches: master, and stable/kilo
16:05:40 <adrian_otto> once we do that, we fix bugs in master, and backport defect fixes to stable/kilo
16:05:51 <adrian_otto> not all bug tickets are defects
16:06:09 <adrian_otto> so please use the tag "defect" to identify a bug ticket as a defect
16:06:21 <sdake> which, in the git log?
16:06:24 <adrian_otto> and be sure to indicate which branch(es) that defect impacts.
16:06:43 <adrian_otto> no, the tag I am talking about is actually in LP (in the ticket system)
16:06:52 <sdake> got it
16:07:04 <adrian_otto> but being clear in the commit log is also wise, so plan to do both
16:07:16 <adrian_otto> any questions on that?
16:07:22 <sdake> in the git log add
16:07:26 <sdake> tag: defect
16:07:27 <sdake> ?
16:08:03 <adrian_otto> no, because the git log are the fixes to the bugs
16:08:14 <adrian_otto> Closes-Bug: NNNN is enough for that
16:08:18 <sdake> ok
16:08:34 <adrian_otto> but you can mention in the narrative of the commit message that this is a fix for master, or a fix for stable/kilo
16:08:57 <adrian_otto> that is visible in the patch metadata as well, but because this is something new, this should help reviewers get used to it
16:08:59 <sdake> that should be obvious in gerrit becasue of the repo it is applied against :)
16:09:12 <adrian_otto> and if for any reason something is not clear, or you think it would work better another way, let's discuss it
16:09:32 <adrian_otto> ok, I have a special update to share with you
16:09:47 <adrian_otto> I'd like to keep this to the team, but I think it's important to mention so that we are ready
16:09:54 <adrian_otto> 3) Special Update
16:10:14 <adrian_otto> as you know, we will be showing a demo of Magnum at the OpenStack Design Summit in Vancouver in a few weeks
16:10:31 <adrian_otto> that demo will happen in a conference track session
16:10:40 <adrian_otto> the link to that is on our wiki page
16:11:00 <adrian_otto> the new info is that we will also be giving the demo in one of the keynotes
16:11:08 <dims> nice!
16:11:10 <apmelton> \o/
16:11:28 <thomasem> noice
16:11:30 <juggler> +1
16:11:32 <adrian_otto> so we will be laser focused on making that demo really smooth, and compelling
16:11:33 <Tango> yea
16:11:49 <adrian_otto> this is obviously good news because we will gain new users and new contributors
16:11:52 <sdake> by we, he means everyone ;)
16:12:20 <adrian_otto> yes, I'm mentioning this to you because I'm seeking all of your help to get this go play really smooth
16:12:34 <adrian_otto> rehearsals happen on May 7th, so we need to be done basically in a week
16:12:51 <adrian_otto> and if anything is glitchy, we will have a little time to iron it out.
16:13:04 <dims> adrian_otto: so "stable/kilo" should be really stable :)
16:13:11 <adrian_otto> yes, dims!
16:13:21 <dims> ++
16:13:27 <Tango> What's planned for the demo?
16:13:45 <sdake> wordpres
16:13:47 * sdake runs
16:13:49 <diga_> https://libertydesignsummit.sched.org/event/44475e888cc90b5e7b14ac61a0b8de0e#.VT-xrIVxMxA
16:13:57 <diga_> talk link :)
16:14:00 <adrian_otto> Tango, I'm still open to suggestions, but Bay creation is the thrust of our value right now, so we'll want to do that
16:14:11 <adrian_otto> #link https://libertydesignsummit.sched.org/event/44475e888cc90b5e7b14ac61a0b8de0e#.VT-xrIVxMxA Magnum Session
16:14:21 <dims> adrian_otto: the main thing that i can think of is the time taken for things to come up (which i am facing on the CI job)
16:14:38 <sdake> we are not doing virt on virt dims
16:14:43 <sdake> we are doing bare metal bay creation
16:14:44 <adrian_otto> yes, that's something that we will want to apply some creativity to
16:14:46 <tcammann_> yeah, don't do virt nesting..
16:14:49 <dims> cool
16:15:09 <dims> just making sure :)
16:15:19 <adrian_otto> We might start with an existing bay, and show how we can scale it, interact with it, and explain how it works
16:15:56 <sdake> bay creation takes about 1 minute on my hardwqare
16:16:02 <adrian_otto> and with any live demo, we will want to show an actual app running, and some way to demonstrate that it's changing while we give Magnum instructions
16:16:04 <sdake> so doing bay creation is probably feasible in a demo
16:16:11 <adrian_otto> I know I'll ahve a dedicated wired network connection
16:16:38 <adrian_otto> so I can create a really fast demo environment on public cloud (bare metal)
16:16:49 <adrian_otto> I can also carry in a server, and have something local on the laptop
16:17:02 <adrian_otto> so there are plenty of options for making sure e have fast hardware
16:17:19 <dims> awesome
16:17:26 <adrian_otto> so we can come back to this in Open Discussion. Let that simmer for now.
16:17:52 <adrian_otto> #topic Blueprint/Task Review
16:18:08 <adrian_otto> do any current work items require team discussion today?
16:18:22 <sdake> lots of our blueprints are not going to make the cutoff
16:18:54 <adrian_otto> yes, I'll be retargeting those that are incomplete
16:19:22 <jjlehr> I was wondering if the blueprint for creating swarm bays will make this cutoff.
16:19:25 <adrian_otto> if you have a BP that you know is complete, be sure to mark it as "Implemented" now, so we record that in our release notes.
16:19:29 <jjlehr> https://blueprints.launchpad.net/magnum/+spec/multiple-bay-templates
16:19:47 <adrian_otto> apmelton: can you give jjlehr an update on that?
16:20:00 <apmelton> I believe at this point that bp is complete
16:20:20 <apmelton> rather that entire flow of bps
16:20:24 <adrian_otto> there were a couple of nice-to-have reviews in process on Friday
16:20:31 <adrian_otto> I expect those have also merged?
16:20:46 <apmelton> jjlehr: baymodel.coe is complete, right?
16:20:57 <jjlehr> Yes it is. It was merged into master last night.
16:21:42 <adrian_otto> that's awsome
16:21:54 <adrian_otto> (stop a minute to recognize this milestone)
16:22:03 <dims> yay :)
16:22:06 <adrian_otto> thanks to everyone who worked on this feature
16:22:11 <apmelton> thanks a bunch to jjlehr and diga_ for helping out on this!
16:22:22 <adrian_otto> this is one of the key reasons that Magnum even makes sense for OpenStack operators
16:22:27 <juggler> thx jj/diga
16:22:29 <diga_> apmelton: welcome!
16:22:50 <jjlehr> You're welcome! I enjoyed working on it as my first contribution.
16:22:52 <adrian_otto> because it can give them a framework and an API that the can plug in whatever the prevailing technology is as that evolves
16:23:23 <adrian_otto> that's really important because large scale operators are reluctant to invest in emerging technologies that are at high risk of changing (quickly becoming obsolete)
16:23:27 <apmelton> trial by fire jjlehr :)
16:23:54 <adrian_otto> but with a feature like multi-bay-type they can bet on Magnum, and have confidence that they can fit in whatever COE users like best over time
16:24:19 <adrian_otto> I do want to make a special call out to jjlehr and apmelton
16:24:37 <adrian_otto> jjlehr: because you handled the learning curve with such grace
16:24:47 <adrian_otto> apmelton: for just kicking butt
16:24:54 <apmelton> thanks adrian_otto!
16:25:00 <adrian_otto> who else worked on this feature?
16:25:11 <jjlehr> Thank you to everyone who was so patient with me :_
16:25:25 <apmelton> diga_ worked on the docker conductor
16:25:28 <jjlehr> joffter helped me get started
16:25:36 <adrian_otto> jjlehr: do you feel like we are doing a good job welcoming new contributors?
16:25:49 <adrian_otto> is there anything we should plan to do better?
16:25:53 <jjlehr> Definitely! This seems like a very friendly community.
16:26:24 <adrian_otto> if you think of anything, just let us know, and we;ll find a way to work it in
16:26:31 <jjlehr> Sounds good.
16:26:48 <apmelton> diga_ also worked on bay_distro, which is extremely important for really supporting images other than atomic
16:27:00 <apmelton> cluster_distro*
16:27:03 <adrian_otto> that's right!!
16:27:15 * adrian_otto offers a high five to diga_
16:27:32 <diga_> thanks :)
16:27:43 <adrian_otto> diga_ has worked on a really wide range of stuff
16:27:49 <adrian_otto> thanks for being so versitile
16:28:24 <adrian_otto> so our current topic is Blueprint/Task Review
16:28:26 <diga_> adrian_otto:  I enjoy working on magnum
16:28:31 <adrian_otto> ;-)
16:28:39 <juggler> I agree. diga's very helpful!
16:29:01 <adrian_otto> any other work items we should review together today?
16:29:23 <diga_> juggler: thanks :)
16:30:02 <adrian_otto> ok, one thing I did not add to the agenda yet is an etherpad to name our Design Summit session topics
16:30:12 <adrian_otto> would you like to make one now?
16:30:45 <apmelton> adrian_otto: how many sessions do we have?
16:31:06 <adrian_otto> 6+4 I think… I will confirm… one moment
16:31:17 <dims> 6 fishbowls?
16:32:04 <sdake> 4 fishbowl
16:33:53 <adrian_otto> We have 4 Fishbowl (large audience) and 6 Workrooms
16:34:51 <adrian_otto> and 1 contributors meetup on Friday afternoon
16:35:04 <adrian_otto> #link https://libertydesignsummit.sched.org/ Summit Schedule
16:35:17 <adrian_otto> search for "Magnum" on that page and check all the Magnum ones
16:35:37 <diga_> I wish you all the best & have a fun at summit :)
16:36:30 <apmelton> adrian_otto: I guess they didn't tag each event with the project?
16:37:06 <juggler> regarding adrian's earlier question, an etherpad sounds like a great idea.
16:37:16 <adrian_otto> apmelton: correct. not yet
16:37:25 <apmelton> gotcha
16:37:26 <adrian_otto> we can get these placeholders updated with actual topics
16:38:19 <adrian_otto> #link https://etherpad.openstack.org/p/liberty-magnum-topics OpenStack Design Summit Topic Etherpad
16:38:32 <juggler> Everything You Wanted to Know about Magnum, but Were Afraid to Ask :)
16:40:03 <adrian_otto> juggler: oh, my
16:40:16 <adrian_otto> that would be a big Fishbowl
16:40:25 <juggler> prolly
16:40:59 <adrian_otto> let's just get all the ideas up first
16:41:06 <adrian_otto> and then we can slot them after
16:42:42 <adrian_otto> it might make sense to have networking in a fishbowl as well
16:43:07 <adrian_otto> ideally one that happens after a workroom… not sure if we have any in that order… looking at that now
16:44:35 <thomasem> adrian_otto: that's fair
16:44:47 <thomasem> I wasn't sure about where to put that
16:45:17 <thomasem> Yay, etherpad formatting!
16:48:44 <thomasem> What's the 'c' or 'd' for?
16:52:45 <adrian_otto> C = Conference Settion
16:52:49 <adrian_otto> Session
16:52:56 <adrian_otto> D = Design Session
16:52:59 <thomasem> Ahhhh, I see
16:53:00 <thomasem> thanks, adrian_otto
16:53:03 <adrian_otto> np
16:54:09 <apmelton> adrian_otto: so we have fishbowl early on wednesday, then design?
16:54:33 <adrian_otto> apmelton: I pasted the schedule in the bottom of the etherpad for your reference
16:54:52 <adrian_otto> so the fishbowls will be before all the work sessions
16:54:53 <apmelton> adrian_otto: I'm just making sure I understand what's pasted
16:54:59 <adrian_otto> yes
16:55:16 <adrian_otto> the Room 306 ones are the Fishbowls
16:55:53 <diga_> adrian_otto: Can we add volume topic for containers "add volume to the container from cinder" ? I never tried yet, hope someone tried that
16:55:59 <adrian_otto> and I think Room 121/122 is a giant combined room
16:56:56 <adrian_otto> diga_: well as long as you mount the volume in the nova instance and put a filesystem on it before you do a bind mount of the filesystem, that will work
16:57:20 <adrian_otto> but you can not "mount" a volume in an ordinary Docker container because it drops the CAP_SYS_ADMIN capability
16:57:23 <diga_> okay
16:57:32 <adrian_otto> which is needed to do the mount of the filesystem on the volume
16:57:40 <diga_> got it
16:57:56 <adrian_otto> you could actually expose a raw volume to a container, as long as you did not want to put a filesystem on it
16:58:17 <diga_> okay
16:58:20 <adrian_otto> there are a few use cases for that, but they are much less common than a filesystem use case
16:58:25 <adrian_otto> that's actually a good topic
16:58:30 <diga_> yep
16:58:41 <adrian_otto> oh, wow, where did the time go!
16:58:47 <adrian_otto> #topic Open Discussion
16:58:47 <diga_> I will study on that
16:58:49 <diga_> yes
16:59:18 <juggler> for the benefit of new and existing contributors, may i recommend that a reference to https://github.com/openstack/magnum/blob/master/doc/source/dev/dev-manual-devstack.rst be in the quickstart or the contributing page or both.
16:59:24 <adrian_otto> ok, so I used up all the time on that, sorry
16:59:47 <adrian_otto> our next meeting will be 2015-05-05 at 2200 UTC
16:59:48 <juggler> no prob
16:59:53 <adrian_otto> see you all then!!
17:00:08 <adrian_otto> #endmeeting