16:02:16 <adrian_otto> #startmeeting containers
16:02:16 <openstack> Meeting started Tue Jul 21 16:02:16 2015 UTC and is due to finish in 60 minutes.  The chair is adrian_otto. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:02:18 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:02:20 <openstack> The meeting name has been set to 'containers'
16:02:27 <adrian_otto> #link https://wiki.openstack.org/wiki/Meetings/Containers#Agenda_for_2015-07-21_1600_UTC Our Agenda
16:02:27 <mfalatic> o/
16:02:33 <adrian_otto> #topic Roll Call
16:02:36 <adrian_otto> Adrian Otto
16:02:37 <tcammann> \o/
16:02:41 <chirag> chirag arora
16:02:42 <dane_leblanc> o/
16:02:42 <diga_> Digambar Patil
16:02:43 <suro-patz> o/
16:02:43 <Tango> Ton Ngo
16:02:44 <mfalatic> Marty Falatic
16:02:47 <rods> o/
16:02:49 <daneyon> hello
16:02:49 <hongbin> o/
16:02:51 <jjfreric> Jacob Frericks
16:03:27 <brendenblanco> hello
16:03:48 <adrian_otto> hello mfalatic, tcammann, chirag, dane_leblanc, diga_, suro-patz, Tango, rods, daneyon, hongbin, jjfreric, brendenblanco
16:05:33 <adrian_otto> ok, let's begin
16:05:39 <adrian_otto> #topic Announcements
16:06:01 <adrian_otto> #link https://blueprints.launchpad.net/magnum/+spec/mesos-bay-type Our Mesos bay type is complete!
16:06:07 <adrian_otto> 1) WHOOT, we have Mesos
16:06:18 <adrian_otto> Feature is complete, with ongoing work for additional enhancements
16:06:40 <adrian_otto> I'll pause here from comments from the contributors who worked on this.
16:06:47 <daneyon> congrats!!!
16:07:12 <diga_> good work hongbin
16:07:13 <tcammann> nice one
16:07:14 <adrian_otto> this is a good example of our community working together to shape the future of OpenStack
16:07:28 <hongbin> :)
16:07:28 <diga_> yes
16:07:52 <tcammann> We said we would do this at the summit, and now its done :)
16:08:13 <mfalatic> :-)
16:08:19 <adrian_otto> yes, this is really worth recognition.
16:08:52 <adrian_otto> ok,  this was the only prepared announcement I had on our agenda
16:09:10 <adrian_otto> but I just thought of another
16:09:18 <adrian_otto> 2) sdake is on vacation this week
16:09:30 <diga_> adrian_otto: I am targeting to add support for mesos bay with coreos in this week https://blueprints.launchpad.net/magnum/+spec/mesos-bay-with-coreos
16:09:48 <adrian_otto> diga_: thanks for submitting that
16:10:08 <diga_> welcome adrian_otto
16:10:19 <adrian_otto> diga_: we'll take a look at that one. Have you submitted any reviews against that yet, or is that still in ideation phase?
16:10:40 <adrian_otto> any other announcements from team members?
16:10:43 <diga_> no, I am testing some patches at my end
16:10:43 <juggler_> o/
16:10:51 <adrian_otto> thanks diga_
16:11:07 <juggler_> (some minor technical difficulties...hello)
16:13:17 <adrian_otto> ok, let's proceed
16:13:32 <adrian_otto> #topic Container Networking Subteam Update
16:13:40 <daneyon> The networking subteam had a meeting last Thursday. The meeting summary is here:
16:13:44 <daneyon> #link http://eavesdrop.openstack.org/meetings/container_networking/2015/container_networking.2015-07-16-18.03.html
16:13:52 <daneyon> Based on feedback from the subteam, the language of the networking blueprint has been modified:
16:13:58 <daneyon> #link https://blueprints.launchpad.net/magnum/+spec/extensible-network-model
16:14:11 <daneyon> ^ pls review and let me know if you have any questions.
16:14:17 <daneyon> I am working through the networking spec and it will be submitted for review tomorrow. I appreciate everyone's input over the past few weeks and look forward to upcoming feedback from the community.
16:14:25 <daneyon> We will meet again this Thursday. Meeting details are here:
16:14:26 <adrian_otto> also note that the URL of the BP also changed
16:14:31 <daneyon> #link https://wiki.openstack.org/wiki/Meetings/Containers#Container_Networking_Subteam_Meeting
16:14:42 <daneyon> Additional supporting blueprints have also been filed. One for expanding the scope of labels and the other for a common pluggable framework:
16:14:46 <daneyon> #link https://blueprints.launchpad.net/magnum/+spec/expand-labels-scope
16:14:51 <daneyon> #link https://blueprints.launchpad.net/magnum/+spec/common-plugin-framework
16:15:01 <daneyon> I look forward to discussing these bp's
16:15:21 <adrian_otto> daneyon: it looks like the participation level there is healthy
16:15:22 <daneyon> I believe that's it for announcements. Any questions?
16:15:33 <adrian_otto> do you feel like you have the support needed to succeed?
16:15:33 <daneyon> adrian_otto agreed
16:15:53 <daneyon> and a lot of good dialogue during the meetings and in the etherpad.
16:16:04 <daneyon> adrian_otto I do
16:16:10 <adrian_otto> ok, great.
16:16:15 <adrian_otto> any questions about this?
16:16:24 <daneyon> I think it will be important to discuss the above blueprints
16:16:36 <daneyon> neither of them are show stoppers for magnum networking
16:16:55 <daneyon> but they could be beneficial to networking and the magnum project as a whole
16:16:58 <adrian_otto> daneyon: did you want to give a quick update on the BP as well?
16:17:05 <daneyon> sure
16:17:16 <adrian_otto> we can do that now
16:17:22 <daneyon> I am half way through writing the bp
16:17:44 <daneyon> i have incorporated most o the feedback in the etherpad:
16:17:50 <daneyon> #link https://etherpad.openstack.org/p/magnum-native-docker-network
16:18:19 <daneyon> if your feedback was not incorporated, then i should have provided feedback in the ep directly
16:18:38 * tcammann adds to reading list
16:18:49 <daneyon> most of the details of the spec can be viewed in the ep
16:19:30 <adrian_otto> ok, thanks daneyon
16:19:36 <daneyon> their is the idea of loading plugins through magnum.conf as a near-term step until labels are examined further
16:20:03 <daneyon> pls let me know if you have any concerns managing plugins through magnum.conf
16:20:16 <daneyon> this would be an approach similar to how Neutron manages plugins
16:20:27 <brendenblanco> it seems that the third link in the Common Plugin Framework BP is broken, is that expected?
16:20:32 <daneyon> if no questions, then that's it
16:20:42 <daneyon> let me check
16:21:01 <adrian_otto> #topic Review Action Items
16:21:04 <daneyon> brendenblanco good catch
16:21:08 <daneyon> let me update the link
16:21:09 <mfalatic> fixed
16:21:14 <adrian_otto> 1) adrian_otto to follow up with Barbican team to arrange assistance for integration with Magnum
16:21:23 <mfalatic> the parentheses confuses etherpad.
16:21:33 <mfalatic> suspect that wouldn't be a problem in normal rest docs
16:21:47 <adrian_otto> I did follow up, but I'm not sure if madhuri got enough support. I did see a new introduction yesterday on this.
16:22:03 <adrian_otto> so although the item is complete, I think this needs continued attention
16:22:14 <daneyon> brendenblanco updated... go ahead and refresh
16:22:33 <brendenblanco> daneyon: works, thanks!
16:22:40 <adrian_otto> #action adrian_otto to verify Magnum team has enough support to integrate with Magnum
16:22:46 <daneyon> brendenblanco yw
16:22:58 <adrian_otto> 2) All core reviewers to review https://review.openstack.org/194905
16:23:16 <tcammann> adrian_otto: s/Magnum/Barbican/ ^^?
16:23:17 <adrian_otto> it looks like this review needs a revision
16:23:24 <adrian_otto> #undo
16:23:26 <openstack> Removing item from minutes: <ircmeeting.items.Action object at 0xacb4e90>
16:23:39 <adrian_otto> irc://chat.freenode.net:6667/#action adrian_otto to verify Magnum team has enough support to integrate with Barbican
16:23:50 <adrian_otto> #action adrian_otto to verify Magnum team has enough support to integrate with Magnum
16:23:54 <adrian_otto> ok
16:24:26 <adrian_otto> any remarks on the "Add TLS support in Magnum" spec?
16:24:59 <adrian_otto> ok, next action item
16:25:00 <adrian_otto> 3) tcammann to move heat-coe-templates repo to project attic, and to relay our resolution on the ML thread: http://lists.openstack.org/pipermail/openstack-dev/2015-July/068381.html
16:25:05 <adrian_otto> Status?
16:25:09 <tcammann> Its on my list todo
16:25:12 <tcammann> busy week
16:25:24 <adrian_otto> #action tcammann to move heat-coe-templates repo to project attic, and to relay our resolution on the ML thread: http://lists.openstack.org/pipermail/openstack-dev/2015-July/068381.html
16:25:39 <daneyon> adrian_otto you may want to double-check your action item above
16:25:39 <adrian_otto> #topic Blueprint/Bug Review
16:25:50 <adrian_otto> daneyon: ?
16:25:57 <daneyon> verify Magnum team has enough support to integrate with Magnum
16:26:03 <daneyon> Should that be Barb not M?
16:26:05 <adrian_otto> I did an undo on that
16:26:14 <daneyon> oops, didn;t catch that
16:26:16 <daneyon> sorry
16:26:20 <adrian_otto> and then I made the same error
16:26:24 <adrian_otto> whatever. I know what it means
16:26:46 <adrian_otto> New Blueprints for Discussion
16:27:02 <adrian_otto> #link https://blueprints.launchpad.net/magnum/+spec/hyperstack Power Magnum to run on metal with Hyper
16:27:16 <adrian_otto> we have had an ongoing discussion about this on our ML
16:27:45 <adrian_otto> I've formed an objection to using Magnum as a way to re-implement aspects of Nova
16:28:09 <adrian_otto> so I've offered guidance to explore adding Hyper as a virt-driver for nova
16:28:29 <adrian_otto> if you have opinions on this, please join the ML discussion. Let me find a link to that.
16:29:10 <hongbin> adrian_otto: If there is a hyper virt-driver for nova, it will be something similar as nova-docker?
16:29:14 <adrian_otto> #link http://lists.openstack.org/pipermail/openstack-dev/2015-July/068574.html Hyper/Magnum ML Discussion
16:29:34 <adrian_otto> hongbin: yes, nova-docker is also a virt driver
16:29:55 <adrian_otto> we never anticipated using Magnum as a tool for creating baremetal hosts to bypass Nova
16:30:12 <adrian_otto> so I'm concerned about that aspect of the proposal
16:30:39 <hongbin> I think we bypass Nova to create a k8s ironic bay?
16:31:06 <adrian_otto> hongbin: no, we use a Heat template that calls for an OS::Server resource. That uses Nova.
16:31:15 <adrian_otto> through ironic, correct?
16:31:22 <hongbin> I think so
16:31:45 <adrian_otto> so I'd be happy to endorse an approach that would closely track that implementation
16:32:01 <adrian_otto> but that works because there is an ironic virt driver for nova
16:32:29 <tcammann> +1
16:32:44 <hongbin> k
16:32:51 <daneyon> and if/when hyper is added to nova as a virt driver, magnum can take the same approach to support the hyper virt driver
16:33:00 <adrian_otto> ok, please follow up on the ML for this
16:33:14 <Tango> Sounds like a consistent approach
16:33:56 <adrian_otto> I'll keep it on the agenda for next week, to see if we can tweak the BP so we can all feel comfortable with it
16:34:06 <adrian_otto> Essential Blueprint Updates
16:34:12 <adrian_otto> #link https://blueprints.launchpad.net/magnum/+spec/objects-from-bay Obtain the objects from the bay endpoint (sdake)
16:34:30 <adrian_otto> sdake is not present. Did he submit status for this for anyone to share?
16:34:35 <adrian_otto> my guess is that it's pending.
16:35:55 <adrian_otto> #link https://blueprints.launchpad.net/magnum/+spec/secure-kubernetes Secure the client/server communication between ReST client and ReST server (madhuri)
16:36:26 <adrian_otto> my understanding on this one is that we are making progress on our Barbican integration
16:36:35 <tcammann> Patch up for cert controller https://review.openstack.org/#/c/203901/
16:36:45 <adrian_otto> thanks tcammann
16:36:45 <tcammann> I would like to see some more work on the spec though
16:36:53 <adrian_otto> agreed
16:37:21 <adrian_otto> tcammann: are you ever around when madhuri is? I'm usually leaving for the day as she is arriving
16:37:40 <tcammann> In the mornings I believe so
16:38:35 <adrian_otto> ok, can you please check with her to see if she needs more help?
16:38:48 <adrian_otto> we'll pull in help if she needs it
16:39:06 <tcammann> Ok, I'll try and catch up with her
16:39:15 <adrian_otto> for a while there we were stuck on Barbican, but I expect we have started moving again
16:39:20 <adrian_otto> thanks tcammann
16:39:27 <adrian_otto> #link https://blueprints.launchpad.net/magnum/+spec/external-lb Support the kubernetes service external-load-balancer feature (Tango)
16:39:52 <Tango> I moved to Kubernetes V1
16:40:07 <Tango> It did require further tweak in the templates
16:40:56 <adrian_otto> Tango: are those changes up for review, or might we expect those to come soon?
16:41:42 <Tango> I have 2 issues right now:  resolve the bug in the atomic build, and debugging the config for Kubernetes to talk to OpenStack (as it has changed)
16:42:27 <Tango> sdake and I were talking to the Atomic folks to get some help.  They confirmed the bug and suggested we move to Fedora 22.
16:42:44 <adrian_otto> so we have a Fedora problem
16:43:00 <Tango> But so far I have not been able to build with F22.  I will need to bug them again
16:43:06 <adrian_otto> we have tooling (that I think you contributed?) for building new images
16:43:23 <adrian_otto> ok, so I look forward to an update next time
16:43:31 <adrian_otto> if you want help before then, let's fire something up on the ML
16:43:50 <Tango> Right, the build process is ok for f21, but for f22 it's still not fully developed yet from the Atomic side
16:43:57 <adrian_otto> We might want to start getting creative about where to find help if you feel like you need it
16:44:22 <Tango> I can certainly use help with the Atomic build problem
16:44:42 <adrian_otto> another option would be to pick an older version of k8s
16:44:47 <juggler> Tango so the present recommendation then is to maintain f21 for build setup then?
16:45:11 <Tango> I can only build with F21 and use a work around for the bug
16:45:29 <juggler> ok
16:45:32 <Tango> So with this, I can get Kubernetes V1 running
16:46:14 <Tango> I did get confirmation from someone at eBay that they did get V1 load balancer working with OpenStack, so that's good to know
16:46:40 <Tango> We just need to get the configuration correct
16:46:42 <daneyon> adrian_otto I do not see anything in the tools dir related to building new images. Can you provide a link to the build tool you are referencing?
16:46:59 <Tango> daneyon: It's still a patch from me
16:47:14 <Tango> I need to submit another patch based on sdake comment
16:47:18 <hongbin> #link https://review.openstack.org/#/c/196145/
16:47:30 <adrian_otto> thanks Hongbin
16:47:40 <Tango> I haven't heard from anyone else, so I will do that update now
16:47:46 <adrian_otto> my memory was flawed. I's a guide, not a tool
16:48:05 <Tango> right, just instruction on how to build
16:48:17 <adrian_otto> ok, I have a couple of more status'es to request
16:48:24 <adrian_otto> thanks for the update Tango
16:48:44 <adrian_otto> #link https://blueprints.launchpad.net/magnum/+spec/magnum-swarm-scheduling Provide container anti-affinity through swarm constraint filtering (diga)
16:48:58 <adrian_otto> diga: I downgraded this from Essential to Medium last week
16:49:08 <adrian_otto> is that something you want us to revisit?
16:49:08 <diga_> yes adrian_otto
16:49:17 <diga_> yes
16:49:51 <diga_> I will submit a patch on this
16:50:08 <adrian_otto> because I adjusted the priority I will not ask for weekly reports back on this
16:50:28 <adrian_otto> but we will look forward to your patches in gerrit
16:50:29 <diga_> yes, I understand
16:50:35 <adrian_otto> ok, next one here...
16:50:37 <adrian_otto> #link https://blueprints.launchpad.net/magnum/+spec/secure-docker Secure client/server communication using TLS (apmelton)
16:50:50 <adrian_otto> apmelton is not present today
16:50:54 <tcammann> Still dependent on tls kube work
16:51:04 <adrian_otto> ok, I'll advance to the next
16:51:13 <adrian_otto> this one is actually a new one
16:51:33 <adrian_otto> we should have looked at it in the previous sub-topic, but let's check it now….
16:51:34 <adrian_otto> #link https://blueprints.launchpad.net/magnum/+spec/magnum-service-list Add service list to magnum (suro-patz)
16:51:41 <adrian_otto> suro-patz: proceed
16:51:55 <suro-patz> I am looking for some clarification on the ask of this BP
16:52:27 <adrian_otto> so this is a previously triaged BP, set to priority Medium
16:52:41 <suro-patz> it seems jay-lau-513 wanted that we should implement 'magnum service-list' as similar to 'nova service-list'
16:52:46 <adrian_otto> so we are looking at this for team discussion today, not administrative processing?
16:53:01 <suro-patz> yes
16:53:05 <adrian_otto> ok, thanks suro-patz
16:53:16 <tcammann> Can we move this to the ML?
16:53:28 <suro-patz> sure tcammann:
16:53:41 <suro-patz> I will initiate a thread there, then
16:53:42 <adrian_otto> tcammann: I did agree to taking a look today, but we are running a little short on time
16:54:16 <adrian_otto> so let's take this one to the ML, and we can pull in stakeholders to offer feedback
16:54:37 <adrian_otto> please add yourself as a subscriber on this BP if you would like to discuss it
16:54:50 <adrian_otto> suro-patz: is that ok with you?
16:55:03 <suro-patz> adiran_otto: absolutely
16:55:18 <adrian_otto> thanks so much. would you like to be assigned to this?
16:55:18 <suro-patz> let's proceed
16:55:23 <suro-patz> yes, please
16:55:42 <adrian_otto> ok, you got it.
16:55:48 <adrian_otto> #topic Open Discussion
16:56:07 <daneyon> A lot of talk around google joining openstack. Have their been any discussions on Google joining the M community?
16:56:23 <adrian_otto> daneyon: yes. I have been in touch with them
16:56:43 <daneyon> any details that can be shared with the community?
16:56:45 <adrian_otto> we expect to regroup and plan participation ofter OSCON dust has settled
16:56:58 <daneyon> OK
16:57:04 <daneyon> I would like to discuss our heat templates, template definitions and definition entry points.
16:57:20 <adrian_otto> let me find the announcement for reference, one moment
16:57:27 <daneyon> seems like their is considerable amount of code dup among templates of the same coe ttyoe
16:57:35 <daneyon> do people agree?
16:57:40 <adrian_otto> #link http://techcrunch.com/2015/07/16/google-signs-on-as-openstack-sponsor-will-contribute-container-tech/ Google Sponsors OpenStack
16:57:53 <tcammann> daneyon: I agree, was thinking about this to the other day
16:57:56 <adrian_otto> they would have been a higher level of sponsorship but the number of seats are limited
16:58:07 <daneyon> if so, it seems like their is an opportunity to make temapltes as the associated def's/entry points more composible
16:59:02 <daneyon> tcammann have you put thought into refactoring the templates/defs/ep's?
16:59:10 <adrian_otto> daneyon: yes, we talked in the past about applying more DRY principles to our templates
16:59:32 <tcammann> My wild idea was to make the templates programmatically, but I haven't thought too much about it
16:59:40 <daneyon> tcammann let me know if you would like to have a brainstorming session
16:59:53 <adrian_otto> we stumbled on the fact that HOT format does not allow for conditional logic, so we started thinking about maybe using the environment feature as a workaround
16:59:54 <daneyon> seems like a bp should be filed to refactor our template approach
17:00:06 <adrian_otto> I'm tempted to table that pursuit until we have a better way to solve it
17:00:19 <daneyon> adrian_otto OK
17:00:26 <adrian_otto> I am happy to have a BP for that
17:00:35 <adrian_otto> but I am concerned about distracting us
17:00:40 <adrian_otto> time's up for today
17:00:45 <adrian_otto> thanks everyoen for attending
17:00:55 <daneyon> tcammann let me know if you want to create the heat refactor bp or i will
17:01:03 <adrian_otto> our next meting is Tuesday 2015-07-28 at UTC 2200
17:01:05 <juggler> np!
17:01:14 <adrian_otto> #endmeeting